ye & V- /2Sf 7 DAYS Batholithic Rocks of Southern California—A Model for the Petrochemical Nature of their Source Materials U.S. .GEOLOGICAL SURVEY ^PROFESSIONAL PAPER 1284  Batholithic Rocks of Southern California—A Model for the Petrochemical Nature of their Source Materials By A. K. BAIRD and A. T. MIESCH U.S. GEOLOGICAL SURVEY PROFESSIONAL PAPER 1284 A mathematical model is used to remove the effects of magmatic differentiation from the chemical data on 480 samples UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1984UNITED STATES DEPARTMENT OF THE INTERIOR WILLIAM P. CLARK, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Baird, Alexander K. Bathoiithic rocks of southern California. (Geological Survey Professional Paper ; 1284) Bibliography: 42 p. 1. Batholiths—California, Southern. I. Miesch, A. T. (Alfred T.) II. Geological Survey (U.S.) III. Title. IV. Series. QE461.B24 1983 552'.3 82-21048 For sale by the Branch of Distribution U.S. Geological Survey 604 South Pickett Street Alexandria, VA 22304CONTENTS Page Abstract ................................................... 1 Introduction ............................................... 1 Regional geologic setting................................. 1 Previous geochemical studies................................ 4 Petrologic nomenclature..................................... 4 Acknowledgments............................................. 5 Development of the model.................................... 5 Number of end members ................................... 5 Compositional structure ................................. 8 Derivation of the differentiate end-members D-1 and D-2 ............................................. 10 Derivation of the magma end-members M-1 and M-2 . 11 Page Characteristics of the model................................... 12 Petrology of the magma end-members M-l and M-2 . 12 Mixing proportions.................................... 14 Regional variation in the compositions of the magmas and differentiates................................. 16 Anomalous samples.......................................... 18 Relations to other geologic features....................... 19 Discussion ................................................... 21 Summary and conclusions........................................ 22 References cited............................................... 23 Appendixes 1 and 2............................................. 35 ILLUSTRATIONS Page Figure 1. Map showing geologic provinces and major fault patterns of southern and central California and northern Baja California, Mexico............................................................................................................... 2 2. Generalized map of the Mesozoic plutonic rocks of the northern Peninsular and Transverse Ranges Provinces, Calif. . . 3 3. Map of sampling localities in Mesozoic plutonic rocks of the northern Peninsular and Transverse Ranges, California . . 6 4. Factor-variance diagram for the bathlolithic rocks of southern California ............................................ 9 5. Factor-variance diagram for the Santa Ana block ................................................................. 12 6. Factor-variance diagram for the San Bernardino block................................................................. 13 7. Stereogram for the Santa Ana block................................................................................... 14 8. Stereogram for the San Bernardino block............................................................................. 15 9-12. Maps showing: 9. Relative amounts of magma required for formation of batholithic rocks....................................... 18 10. Percentages of end-member M-2............................................................................. 19 11. Percentages of end-member D-2............................................................................. 20 12. Variability of Si02, A1203, FeO, MgO, CaO, Na20, K20, and Ti02 in the batholithic rocks of southern California and in the magmas and differentiates as interpreted from the model.................................... 25 TABLES Page Table 1. Average compositions of the batholithic rocks.............................................................................. 8 2. Chemical variances in the batholithic rocks............................................................................... 9 3. Biotite and hornblende compositions...................................................................................... 10 4. Compositions of two mineral assemblages and end members D-l and D-2.................................................. 10 5. Compositions represented by some points on figure 7 ................................................................. 14 6. Compositions represented by some points on figure 8 ................................................................. 15 7. Chemical and normative compositions of the end members................................................................. 16 8. Compositions of two samples of gabbros................................................................................... 16 9. Analyses from the U.S. Geological Survey RASS file ...................................................................... 17 10. Statistical summary of mixing proportions.............................................................................. 17 11. Mean values of parameters.............................................................................................. 17 12. Average compositions of the magmas..................................................................................... 21 13. Average compositions of the differentiates............................................................................. 21 illBATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA—A MODEL FOR THE PETROCHEMICAL NATURE OF THEIR SOURCE MATERIALS By A. K. Baird1 and A. T. Miesch ABSTRACT Major-element analyses of 497 composite samples of batholithic rocks (quartz diorites, granodiorites, and quartz monzonites) from the northern Peninsular Ranges and Transverse Ranges Provinces, southern California, form the basis for a mixing model that accounts for most of the compositional variation in the rocks. The compositional structure in the batholithic rocks as a group was found to be similar to that in the Sierra Nevada batholith, and indicates that four end members are sufficient to account for 85-97 percent of the variability in each of the eight major oxides. According to the model, the batholithic rocks formed from the mixing of basaltic and quartzo-feldspathic end-member magmas, and the removal of variable proportions of plagioclase and mafic minerals, principally hornblende. Gab-broic rocks, common only in the western part of the region, could have formed from nearly uncontaminated magmas of the basaltic end member. Variations in the mixtures of basaltic and quartzo-feldspathic magmas are presumed to reflect variations in their source materials at depth. According to the model, the compositions of the source materials do not vary smoothly over the region, but display a discontinuity along a line approximately coincident with the present San Jacinto fault zone. The discontinuity is roughly coincident with previously noted petrologic and isotopic discontinuities in the northern Peninsular Ranges and is interpreted as the western limit of significant contribution of continental materials to the batholithic rocks. The model gives no evidence of a discontinuity in the vicinity of the San Andreas fault zone. INTRODUCTION The batholithic rocks of late Mesozoic age in southern California are exposed over 18,000 km2 (square kilometers) in parts of two geologic provinces, the Peninsular Ranges Province and the Transverse Ranges Province (fig. 1), and range in composition from gabbro to quartz monzonite. Composite samples from 548 localities were analyzed for eight major elements (Baird and others, 1979) and these data form the basis for compositional modeling with an extended method of Q-mode factor analysis (Miesch, 1976b). The method was used to determine the number and nature of end members required to adequately explain the observed chemical variations. Areal variations in the corresponding mixing proportions and in the derived compositions of the end-member magmas, or magma-source materials, display a distinct discontinuity in the eastern part of the Peninsular Ranges Province. 'Present address: Pomona College, Claremont, CA 91711 REGIONAL GEOLOGIC SETTING The northern Peninsular Ranges lie at the northern end of a narrow (120 km) belt of batholithic rocks of Mesozoic age that can be traced southward for hundreds of kilometers into Baja California, Mexico. The eastern limit of this belt, over all its length, is in the flooded depression of the Gulf of California and its northward extension, the Salton Trough. The Gulf is interpreted to have formed in latest Tertiary time by oceanic spreading on the East Pacific Rise; the Salton Trough is believed to have formed by a series of rhom-bochasms developed along the San Andreas fault zone (fig. 2). The northern terminus of the Peninsular Ranges Province, and the southern boundary of the Transverse Ranges Province, is at the east-striking Malibu Coast-Cucamonga fault zone and its possible eastward continuation, the Banning fault (fig. 2). The San Andreas fault zone continues northwestward from the northern margin of the Peninsular Ranges diagonally across the Transverse Ranges Province. Other fault zones, especially the San Jacinto and Elsinore (both subparallel to the San Andreas) and the east-striking Pinto Mountain fault, provide boundaries for further subdivisions of the provinces. We recognize six structurally-bounded units or blocks (fig. 2): San Gabriel, San Bernardino, and Little San Bernardino (Transverse Ranges), and San Jacinto, Perris, and Santa Ana (Peninsular Ranges). To claim that this region is the most geologically enigmatic part of California is an understatement. Certain petrologic-structural aspects have been agreed to by most workers, but other aspects have received diverse interpretations. General, if not universal, agreement exists on the following aspects: 1. Plutonic igneous rocks are mostly late Mesozoic in age and are intrusive into metamorphic rocks apparently no older than Triassic in the Peninsular Ranges but as old as late Precambrian in the Transverse Ranges. 2. Plutonic igneous rocks vary compositionally from southwest to northeast, from mafic to felsic. Gabbroic rocks are found only in the Peninsular Ranges and most are southwest of the San Jacinto fault. 3. Plutonic igneous rocks are dominantly the product l2 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA Figure 1.—Geologic provinces of southern and central California and northern Baja California, Mexico (dotted lines). Heavy lines delineate major fault patterns, dashed where inferred. Shaded rectangle outlines area of figures 2 and 3. of magmatic activity; that is, they form numerous individual plutons that have internal structures characteristic of flow. 4. At least some, if not most, of the faults that bound structural blocks postdate the emplacements of the plutonic rocks, and movements on the faults have continued to Holocene times. The nature and magnitude of separation on the San Andreas fault in southern California remain subjects of major disagreement. (The total separations on the Elsinore and San Jacinto faults have been relatively small— 15 km or less—by most interpretations.) The San Andreas fault has been judged (1) a major strike-slip fault in California with hundreds of kilometers of movement dating back to (perhaps) early Mesozoic time (Hill and Dibblee, 1953), (2) a major continental transform and plate boundary also with hundreds of kilometers of movement (Atwater, 1970), (3) a boundary of minor importance in the Transverse Ranges (Hadley and Kanamori, 1977; Yeats, 1981), (4) an important strike-slip fault in southern California that has 250 km of separation (Crowell, 1979), (5) a relatively minor strike-slip fault in southern California that has 35-70 km of right separation at the Transverse Ranges (Baird, Morton, Woodford, and Baird, 1974), (6) a minor structure that does not significantly offset bedrock patterns in southern California (Woodford, 1960), and, finally, (7) a minor participant in large-scale rotational tectonics of southern California that does not cut the Transverse Ranges (Luyendyk and others, 1979). Thus, depending upon the interpretation, the San Andreas fault in this part of California may be young, may be old, or may0=3 SAN GABRIEL BLOCK SAN BERNARDINO BLOCK ^A°NGA~FAUrf LITTLE banning_\ fault v '''ty Los Angeles BLOCK EXPLANATION Quartz monzonite SAN JACINTO BLOCK Granodiorite Quartz diorite Gabbro MAJOR FAULT; dashed where inferred PERRIS BLOCK 50 KILOMETERS SANTA ANA BLOCK Figure 2.—Generalized map for the Mesozoic plutonic rocks of the northern Peninsular and Transverse Ranges Provinces, Calif. Fault-bounded structural blocks are indicated. Rock nomenclature follows that of Bateman and others (1963). Map modified from Baird, Baird, and Welday (1979). INTRODUCTION4 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA have only a few kilometers or more than 1000 km of separation. If the separation since Mesozoic time has been large then, obviously, it is fortuitous that the plutonic rocks east of the San Andreas (San Bernardino and Little San Bernardino blocks) are presently adjacent to similar rocks west of the fault, it is accidental that apparently continuous areal patterns of petrologic variation occur across the region, and a happenstance that distinct, internally consistent patterns of chemical variations in all these rocks suggest a common origin and evolution that are different from those of other Cor-dilleran batholithic masses (Baird and others, 1979). PREVIOUS GEOCHEMICAL STUDIES The first extensive and the definitive work on the batholithic rocks of the area was by E. S. Larsen, Jr. (1948) who described a number of plutonic units in the northern Peninsular Ranges and published dozens of chemical and modal analyses. Through his work, these rocks became known as the “southern California batholith,” although Larsen clearly recognized that the same batholithic mass extends great distances into Mexico. The area he studied and mapped is essentially that of the Perris and Santa Ana blocks of the present paper (fig. 2). The name “southern California batholith” has been restricted to the Peninsular Ranges Province and has not been extended to include the Transverse Ranges, perhaps mainly because of Larsen’s work. Within and southeast of the area described by Larsen, a number of more detailed studies bearing on the geochemistry of plutonic rocks have been made (for example, Morton and others, 1969; Miesch and Morton, 1977; Morton and Baird, 1976; Nishimori, 1976; Todd and Shaw, 1979; Taylor and Silver, 1978; Walawender, 1979). Reconnaissance work in Mexico by Gastil and others (1975) has established the general geology of the batholithic rocks there. Isotope, trace-element and radiometric-dating studies have been pursued by Silver and colleagues (for example, Silver and others, 1975; Taylor and Silver, 1978; Gromet and Silver, 1979), by Krummenacher and others (1975), and by DePaolo (1980). Much of this work, especially as it pertains to the southern Peninsular Ranges within California, has been carefully summarized by Abbott and Todd (1979). To the north, in the Transverse Ranges, considerably less work bearing on the geochemistry of batholithic rocks has been done. However, individual plutons have been studied (for example, Baird and others, 1967; Richmond, 1965), problems of the older Mesozoic syeni-tic rocks have been described (Miller, 1977), and radiometric dates for both wallrock and some intrusives have been reported (Silver, 1971). The basis for the present report is the group of 548 composite samples collected by Baird and his colleagues (fig. 3). The purpose of the sampling was to provide unbiased representative samples of the batholithic rocks that could be used to determime areal distributions of the major and minor elements. Details of the methods, analyses, and discussions of elemental distributions are given elsewhere (Baird, 1975; Baird, Morton, Woodford, and Baird, 1974; and Baird, Baird, and Welday, 1974, 1979). In summary, the principal findings from these prior studies are: 1. All elements, except aluminum and titanium, have statistically significant regional trends that increase or decrease to the northeast. 2. Silicon, potassium, and to a lesser extent sodium, increase markedly toward the northeast in a fashion directly predictable from areal distributions of rock types (gabbro to quartz monzonite) and from position with respect to the continental margin. 3. Within individual quartz-rich rock types (quartz monzonite, granodiorite and quartz diorite), however, silicon varies antipathetically with potassium. 4. Potassium, the element that exhibits the strongest trend, seems to vary independently of the other elements within the quartz-rich rock types, but dependency in gabbroic rocks. 5. Chemical variations are continuous, without the stepwise sequences thought to characterize the Sierra Nevada trends (Bateman and Dodge, 1970). 6. Monzo-syenitic rocks in the Transverse Ranges are genetically unrelated to the quartz plutonites. 7. Gabbroic rocks have patterns of geochemical behavior different from the quartz plutonites and probably had different sources. For the present investigation, we have eliminated from consideration 23 samples from localities underlain by monzo-syenites, and an additional 26 samples from localities at which the rocks are gabbroic, for the reasons cited and because attempts to develop a model with these analyses included in the data gave evidence of the petrogenetic differences. A final two samples were removed from the San Gabriel block because the rocks are probably Miocene in age and genetically are unrelated to the batholith (D. M. Morton, oral commun. 1980). Thus, a total of 497 composite samples from the studies summarized have been considered in this investigation. PETROLOGIC NOMENCLATURE This report uses the terminology of Bateman and others (1963, fig. 2), rather than that of the IUGS (International Union of Geological Sciences) Commission (Streckeisen, 1973) for several reasons: (1) for consistency with prior discussions of the same chemical data; (2) for consistency with studies in the Sierra Nevada; and (3) for flexibility because the Bateman classificationDEVELOPMENT OF THE MODEL 5 provides five divisions of quartz plutonites whereas the IUGS scheme uses only four (effectively three because no rocks in southern California fall in the alkali granite category). ACKNOWLEDGMENTS We are indebted to T. H. McCulloh and V. R. Todd of the U.S. Geological Survey and to W. B. Wadsworth of Whittier College for helpful criticism of the manuscript. DEVELOPMENT OF THE MODEL Chemical variation in nearly all rock bodies has resulted from processes of mixing and unmixing (for example, differentiation), and appropriate petrogenetic models are comprised of end-member compositions and estimated mixing proportions for representative rock samples. The observed compositions of the rock samples can be approximated by combining (forming linear combinations of) the end-member compositions according to the derived mixing proportions. The first task in the development of such a model is to determine the number of end members required; the next is to derive the end-member compositions. The first task can be accomplished rather objectively by mathematical analysis, but the second requires geologic reasoning and speculation as well as mathematics. Once the number of end members and their compositions have been determined, each individual sample composition can be examined to see if it can be approximated by reasonable combinations of the end members. If most or all the samples can be explained, the model is said to be mathematically adequate. Whether the model is valid or not depends on whether the end-member compositions are those of the materials that were actually involved in the mixing or unmixing processes that caused the compositional variations in the rocks. The methods used here serve to reject some selected end-member compositions as mathematically impossible, but the final selections are not unique. The modeling methods used here have been described elsewhere (Miesch, 1976a, b, c) and have been applied previously to a variety of petrogenetic problems (for example, Miesch and Morton, 1977; Miesch and Reed, 1979; Miesch, 1979; Stuckless and others, 1981; Stuck-less and Miesch, 1981). The mathematical development is not repeated here and the reader is referred to the cited papers for details. However, an easily visualized application of the modeling concepts involves the plagioclase feldspar system. If one were given a number of feldspar crystals as unknowns (when in reality they were all samples of oligo-clase, andesine, labradorite and bytownite), one could determine analytically that they were composed principally of four oxides: Si02, A1203, Na20 and CaO. But, the compositions of the plagioclases can be described at least approximately in terms of only two end members, albite and anorthite. Similarly, although the batholithic rocks of southern California are composed mostly of eight major-oxide constituents, we will show that the compositions of these rocks can be approximated in terms of only four end members. The end members chosen for the plagioclase series, albite and anorthite, are conventional, but they are not the only end members that could be used. In a study of island arcs, for example, An30 and An55, rather than Ab and An, might be appropriate. In fact, the only requirement for the plagioclase end members is that they be some combination of albite and anorthite. Otherwise, they would not be mathematically compatible with the plagioclase series. That is, the compositions of the plagioclase crystals could not be approximated as linear combinations of the end members. Just as there is a choice for the plagioclase end members, there is a choice of end members that can be used to describe the compositional variation within the batholithic rocks of southern California; the choices must be made on the basis of both mathematical evidence and geologic observation or speculation regarding the materials actually involved in the mixing processes. NUMBER OF END MEMBERS The data matrix (Appendix 1) consists of 497 rows that represent composite samples and eight columns that represent the oxide variables: Si02, A1203, FeO (total iron), MgO, CaO, Na20, K20, and Ti02. The 497 rows of the original data matrix were adjusted so that the eight variables in Appendix 1 sum to 100 percent for each sample. This constant row-sum is required in order to recalculate the results of conventional Q-mode factor analysis into the end-member compositions and mixing proportions that comprise the model. The original data are given in Baird and others (1979) where they are keyed with rock-type designations and the U-V geographic coordinates identified on figure 3 of this report. The average compositions of all samples and of the samples from each of the six structural blocks identified in figure 2 are given in table 1. In order to give each oxide equal weight in the modeling procedure, each column of the matrix was scaled to have the same mean and variance, according to a technique described previously (Miesch, 1980). Also, in order to treat the data for each sample as a vector of unit length, the scaled data for each sample were divided through by the square root of the sum of squares (that is, normalized). The scaled and normalized data were then regarded as the coordinates of 497 unit vectors with a common origin in eight-dimensional space.6 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA Figure 3.—Sampling localities in Mesozoic plutonic rocks of the nothem Peninsular and Transverse Ranges, Calif. Locality numbers are used in Baird and others (1979) forDEVELOPMENT OF THE MODEL 117° 116° 7 keyed to Appendixes 1 and 2 (prefix B has been deleted from locality numbers on map to save space). The U-V coordinate system was convenience in referring to sampling localities.8 BATHOLITHIC ROCKS OP SOUTHERN CALIFORNIA Table 1.—Average compositions (in percent) of the batholithic rocks ivithin structural blocks of southern California [All analyses were expressed as oxides and recomputed to 100 percent before computation of averages] Number of Oxide Structural block samples Si02 A1203 FeO MgO CaO Na20 Ka20 Ti02 San Gabriel block.................................. 47 67.10 16.41 3.81 1.58 3.47 3.90 3.15 0.57 San Bernardino block.............................. 125 70.22 15.45 2.22 0.88 2.86 3.76 4.06 0.55 Little San Bernardino block........................ 25 69.90 15.65 2.56 1.16 2.59 3.87 3.88 0.40 San Jacinto block................................. 101 67.54 16.41 3.56 1.45 3.89 3.65 2.76 0.73 Perris block...................................... 119 67.12 15.99 4.02 1.85 4.43 3.78 2.23 0.59 Santa Ana block.................................... 80 68.38 15.17 4.07 1.70 3.99 3.91 2.27 0.52 All batholithic rocks............................. 497 68.32 15.83 3.39 1.44 3.67 3.79 2.97 0.58 Methods of principal-components analysis were then used to project the vectors into two dimensions, then three, and so forth up to seven dimensions (one less than the number of oxides). After each projection, the compositions represented by the projected vectors were determined and were compared with the compositions represented by the vectors in the original eight-dimensional space. The correlation coefficients between the original and recomputed data were determined and squared to give coefficients of determination for each oxide. These latter values are measures of the variances in the original data that can be explained by mixing models with two to seven end members, and are summarized in the factor-variance diagram in figure 4. This figure clearly shows that the original data can be closely approximated as a four-dimensional vector system, or as a four-factor, four-component, or four-end-member compositional series. Figure 4 shows that a model with only three end members would explain only about 71 percent of the variance in A1203 and that a five-end-member model offers no substantial improvement over a four-end-member model. The values of the coefficients of determination for four factors range from 0.85 to 0.97 (fig. 4), indicating that a four-end-member model can account for 85-97 percent of the variance in each oxide constituent. Thus, 3-15 percent of the variance in each constituent is ascribed to analytical imprecision and to other factors, such as minor petrologic processes that will not be represented in the model. Also, at least some part of this unaccounted for variance may be attributed to the fact that the four end-member compositions to be derived, in actuality, rather than being fixed, varied somewhat in both space and time. The absolute variances accounted for and not accounted for by four factors are listed in table 2. COMPOSITIONAL STRUCTURE The concept of compositional structure in igneous bodies and rock series was described in a previous report (Miesch and Reed, 1979) and refers to the nature of the relations among the oxide constituents as represented by a factor-variance diagram. These relations determine the number of end members that will be required in a petrogenetic mixing model. The compositional structure in the southern California batholithic rocks is closely similar to that of the Sierra Nevada batholith (Miesch and Reed, 1979, fig. 7). The fact that the variances in the eight oxides can be closely accounted for by the mixing and unmixing of four end members is not accidental. It has arisen from the fact that the southern California batholithic rocks originated mostly as a result of a simple combination of processes that involved only four dominant compositions, or perhaps four compositional extremes. That is, the batholithic rocks might have formed mainly by the separation of three independent mafic phases from a magma, by the separation of two phases from mixtures of two magmas, or by some other process that involved four, and only four, dominant end members. A likely possibility is that the batholithic rocks formed primarily by the separation of differentiates (mineral assemblages) that varied in composition between two end members from a magma that also varied within a two-end-member system. We shall assume that the magma varied in composition between an end member (which we label M-l) derived from near the present continental margin and another end member (labeled M-2) derived from farther to the northeast. Composition M-l may represent a magma that has undergone relatively little or no contamination by continental materials, and M-2 may represent a magma that includes more continental material. We shall also assume that the differentiates varied between extremes which we shall label D-1 and D-2. The compositions of the differentiates, presumably, varied with the composition of the magma, temperature, and other local conditions. Now the task is to estimate the compositions M-l, M-2, D-1, and D-2 in the most reasonable way and then to describe the compositions of all 497 samples in terms of these four components. The only mathematical requirement for each of the four compositions is that they be representable as vectors in the four-dimensionalDEVELOPMENT OF THE MODEL 9 Cn O Li- CH Ll3 E-» O O O a= LlJ CJ z cn I—» cn az Li_ O O i—i E—< cn CD Q_ O cn cn NUMBER OF FRCTORS Figure 4.—Factor-variance diagram for the batholithic rocks of southern California. Table 2.—Chemical variances in the batholithic rocks of southern California that can and cannot be accounted for by a four-end-member mixing (factor) model Variance Oxide Accounted for Not accounted for Total Si02............................. 22.3617 0.8719 23.2336 A1203............................. 1.9365 .0656 2.0018 FeO............................... 2.7419 .2062 2.9481 MgO................................ .9547 .1404 1.0951 CaO..................... 2.7561 .1334 2.8895 NaaO............................... .1515 .0089 .1604 K20..................... 1.1326 .1138 1.2464 Ti02 .............................. .0493 .0089 .0582 space that contains the projected sample vectors. Any four compositions that can be represented in this manner can be mixed in varying proportions to approximate the original data to the degree indicated in figure 4 for a four-factor model and will lead to variances accounted for and not accounted for as listed in table 2.10 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA The need to select M-1, M-2, D-1, and D-2 so that they can be represented as vectors in the four-dimensional space of the sample vectors is analogous to the need to describe the plagioclase compositions in terms of combinations of albite and anorthite. DERIVATION OF THE DIFFERENTIATE END-MEMBERS D-l AND D-2 The principal minerals in the quartz diorite, granodiorite, and quartz monzonite are quartz, potassium feldspar, plagioclase, hornblende, biotite, and iron oxides, chiefly magnetite and ilmenite. Presumably, the differentiates that were precipitated from the magmas consisted mainly of the least-soluble phases, plagioclase, hornblende, biotite, and the iron oxides. Consequently, the compositions of end members D-l and D-2 were sought by testing mixtures of the compositions of these minerals. Plagioclase was represented by the ideal compositions of albite and anorthite, and ideal compositions were also used to represent magnetite and ilmenite. Biotite was represented by the average of eight analyses (table 3) from Larsen and Draisin (1950, p. 69). Hornblende was first represented by the average Table 3.—Biotite and hornblende compositions (in percent) used in determination of end-member compositions, D-l and D-2 Oxide Biotite1 Hornblende2 Si02.......................... 37.% 47.57 A1203........................ 17.70 8.64 FeO.......................... 21.20 17.46 MgO........................... 9.52 11.22 CaO........................... 0.39 11.82 Na20.......................... 0.22 1.18 K20........................... 9.49 0.72 Ti02.......................... 3.51 1.37 'Average of eight analyses from Larsen and Draisin (1950, p. 69). 2Analysis 7 from Larsen and Draisin (1950, p. 71). of eleven analyses given by Larsen and Draisin (1950, p. 71), then by the average analysis of the six hornblendes from nongabbroic rocks, and finally by the individual analyses. The best results were achieved using the single hornblende analysis given in table 3. A computer program similar to the EQSCAN program described previously (Miesch, 1976c) was used to form 53,130 systematic mixtures of the six selected compositions. The value 53,130 is the number of points at 5-percent increments within the six-dimensional space representing the six compositions. Each mixture was represented as a unit vector in the original eight-dimensional space of the 497 sample vectors, and then projected into the four-dimensional space containing the projected sample vectors. After each projection, the vector communality (square of vector length) was computed and taken as a direct measure of the nearness of the original unit vector to the four-dimensional subspace. The average communality of the 497 sample vectors in four-dimensional space is 0.9951. Of the vectors representing the 53,130 mathematical mixtures, only a few were represented by vectors with communalities this high. The two that led to the highest communalities were mixtures of (1) 33.9 percent anorthite, 63.3 percent hornblende, 0.9 percent magnetite, and 1.9 percent ilmenite (assemblage 1); and (2) 60.6 percent albite, 35.5 percent anorthite, 2.0 percent magnetite, and 1.9 percent ilmenite (assemblage 2). No biotite is included in either mixture. Thus, of all possible combinations of plagioclase, hornblende, biotite and the iron oxides, these two assemblages, and any combinations of them, come closest to satisfying the mathematical requirements for the differentiates, D-l and D-2. The requirements of the model are perfectly satisfied, however, by the two vectors representing these compositions after projection into the four-dimensional space. The compositions represented by the projected vectors are given in table 4 (D-l and D-2) along with the compositions rep- Table 4.—Compositions (in percent) of two mineral assemblages and end members D-l and D-2 Mineral assemblage1 End member Oxide Assemblage 1 Assemblage 2 D-l D-2 Si02.................... 44.75 57.00 44.68 56.31 A12Os................... 17.89 24.81 18.49 25.12 FeO..................... 12.87 2.87 13.01 3.55 MgO...................... 7.10 0.00 8.01 0.44 CaO..................... 14.31 7.16 12.81 5.91 NaaO..................... 0.75 7.16 0.72 7.39 KzO...................... 0.46 0.00 0.42 0.14 Ti02..................... 1.88 1.00 1.83 1.13 'Assemblage 1 contains the following percentage concentrations of minerals: anorthite, 33.89; hornblende (see table 3), 63.27; magnetite, 0.91; and ilmenite, 1.92. Assemblage 2 contains the following percentage concentrations of minerals: albite, 60.59; anorthite, 35.53; magnetite, 1.97; and ilmenite, 1.90.DEVELOPMENT OF THE MODEL 11 resented by the vectors before projection (assemblage 1 and assemblage 2). The bulk compositions of all materials precipitated from the magmas in the formation of the observed samples, therefore, are assumed to have ranged between composition D-1 and composition D-2. Materials rich in D-1 and sparse in D-2 contained a more calcic plagioclase and were more mafic. Those rich in D-2, on the other hand, contained a more sodic plagioclase and were more felsic. DERIVATION OF THE MAGMA END-MEMBERS M-I AND M-2 The remaining problem is to estimate the compositions of the extremes, M-1 and M-2, in the range of magmas from which the batholithic rocks were derived. Like compositions D-1 and D-2, compositions M-l and M-2 must also be representable as vectors in the fourdimensional space of the 497 sample vectors. But, an additional requirement in the selection of compositions M-l and M-2 pertains to the mixing proportions that will be required for the 497 samples. Acceptable mixing proportions will necessarily be positive for magma compositions M-l and M-2 and must be generally negative for differentiate compositions D-l and D-2, inasmuch as the differentiates should be separated from the magmas, not added to them, in the formation of most of the batholithic rocks. If compositions M-l and M-2 are selected arbitrarily, for example, many samples may require negative proportions of M-l and (or) M-2. Also, some samples may require opposite signs in the mixing proportions for D-l and D-2, suggesting that the differentiates separated from the magmas had bulk compositions outside the range for D-l and D-2 (table 4); possibly the required bulk composition of the differentiate may be partly negative. In addition, the mixing proportions must not be large in absolute value. The sum of the mixing proportions for M-l, M-2, D-l and D-2 will always equal unity, so if the sum of those for M-l and M-2 equals 20, for example, the sum of those for D-l and D-2 must equal minus 19. Mixing proportions such as these would indicate that the specific sample resulted from a differentiation process that went 95 percent of the way towards completion—possible for some individual samples, but not likely for any large mass of batholithic rocks. The two structural units within the region that are the most widely separated in a direction perpendicular to the continental margin are the Santa Ana and San Bernardino blocks (fig. 2). The chemical data for samples from these two blocks were extracted from the main data matrix (Appendix 1) and were examined by independent factor analyses of the same type used to examine the entire data set. Factor-variance diagrams for the two subsets of the data are given in figures 5 and 6. Both diagrams indicate that the chemical data from the corresponding structural blocks can be well represented in three-dimensional vector systems. Stereograms showing the configurations of the vector systems are given in figures 7 and 8; the compositions represented by selected vectors on the stereograms are given in tables 5 and 6. Tables 5 and 6 may be used to observe the nature of compositional changes in various directions across the stereograms. Note that both configurations of sample vectors (figures 7 and 8) are approximately triangular and that those points near the lower left corners represent compositions relatively low in Si02 and high in FeO and MgO, whereas those points to the right represent more siliceous compositions. The compositions of the parent magmas for each of these groups of samples seem to be at or near the point “N” on the corresponding stereogram; the compositions represented by these points are given in tables 5 and 6. These compositions were represented as vectors in the four-dimensional space of the 497 sample vectors. They were then taken tentatively as the M-l and M-2 vectors being sought and as representing two of the end members for the model. Vectors D-l and D-2, previously deduced as representative of the range of precipitates separated from the magmas to form the batholithic rocks, were taken to represent the other two end members. We found, however, that the compositions of many samples could be approximated in the mixing computations only by using negative proportions of composition M-l. Consequently, alternative compositions for M-l and M-2 were sought by trying vectors in the same plane as those representing M-l and M-2, but separated by a wider angle. This procedure led to the identification of the compositions M-l and M-2 listed in table 7. The four end-member compositions given in table 7 can be mixed in various proportions (Appendix 2) to approximate the compositions (to the degrees indicated for four factors in fig. 4 and table 2) of all 497 samples (Appendix 1). With the exception of one sample, all the proportions for end members M-l and M-2 are positive and in the range from: Minimum Maximum M-l 0.0 to 2.5641 M-2 0.0 to 1.8051 With the exception of 16 samples, the mixing proportions for D-l and D-2 are of the same sign and in the range from: Minimum Maximum D-l -1.4776 to 0.1996 D-2 -1.0498 to 0.1304 Only a few of the samples required positive mixing proportions for D-l and D-2, indicating that compositions D-l and D-2 must be subtracted from compositions M-l and M-2 in order to approximate the compo-12 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA QZ O Q_ □ LJ E—' 2! ZD CD CD CD QZ 1x2 CD QZ az L_ o t-l QZ CD CL. CD QZ Q_ NUMBER OF FRCTORS Figure 5.—Factor-variance diagram for the batholithic rocks of the Santa Ana block of southern California. sitions of most samples. The compositions of the 16 samples that require mixing proportions of opposite sign for D-1 and D-2 can be produced from compositions M-1 and M-2 only by adding or subtracting compositions that include negative values for one or more oxides. Therefore, these 16 samples and one (B429)2 that requires a negative amount of M-1 are not accounted for by the model proposed here. Of the 17 anomalous samples, 11 are from the San Jacinto block (B226, B229, B283, B286, B301, B302, B303, B304, B306, B308, and B321), 4 from the Perris block (B338, B392, B420, and B429), and 1 each from the San Bernardino (B99) and Santa Ana blocks (B513). The anomalous samples are discussed in a later section of this report. CHARACTERISTICS OF THE MODEL PETROLOGY OF THE MAGMA END-MEMBERS M—1 AND Af—2 The normative mineralogy of M-l (table 7) is that of a saturated basaltic or gabbroic rock. The 26 gabbro compositions (heretofore not considered in the modeling) were each tested for compatibility with the more silicic rocks of the Santa Ana block by computation of vector communalities. The two gabbros most compatible (highest communalities) are B545 and B511 (table 8). The compositions of these two samples need only be slightly modified to be perfectly representable as vectors (G-l and G-2) in the three-dimensional vector space formed by the Santa Ana block sample compositions (fig. 7). The compositions represented by vectors 2 See figure 3 for sampling localities and Appendix 1 for the analyses.CHARACTERISTICS OF THE MODEL 13 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 Figure 6.—Factor-variance diagram for the batholithic rocks of the San Bernardino block of southern California. G-1 and G-2 are given in table 8. The compositions are notably close to that of end member M-1 (table 7). Nockolds and Allen (1953) used variation diagrams to deduce the compositions of parental magmas for calc-alkaline rocks. Their average values, compared with M-1 (table 7), are only slightly higher in MgO and K20, and slightly lower in FeO and CaO. Because both compositions M-1 and M-2 (table 7) were derived by mathematical procedures, we were interested in determining how closely they correspond to actual rock compositions. Although composition M-1 is close to the compositions of some gabbros from the region, as pointed out in the preceding paragraph, none of the batholithic rocks have compositions close to composition M-2. Consequently, a search was made of the RASS computer-based file (Van Trump and Miesch, 1977) which contains about 130,000 rock analyses by laboratories of the U.S. Geological Survey since 1967. The analyses are of samples collected throughout the United States. The igneous rock analyses in the RASS file closest to compositions M-1 and M-2 are given in table 9. The measure of closeness used was cosine theta of Imbrie and Purdy (1962). The closest match to M-1 is a gabbro of Jurassic age from the Alaska-Aleutian Range batholith (Reed and Lanphere, 1973). The closest matches to M-2 are an andesite from southwest Nevada and tuffaceous rocks from southwest Colorado and northern Washington, all of Tertiary age. The average norm for these matches to M-2 includes 16 percent quartz, 23 percent orthoclase, and 53 percent sodic an-desine, consistent with a leucocratic granodiorite. Composition M-2 is incompatible with any magma or source material of an oceanic nature. One possible origin for end-member M-2 is by mixing14 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA Figure 7.—Sterogram showing the three-dimensional vector system for the batholithic rocks of the Santa Ana block of southern California. Open circles represent 80 sample vectors. Dashed line outlines area wherein all vectors represent compositions that are entirely nonnegative. All vector projections have been from an upper hemisphere vertically onto the plane of the stereogram. See table 5 for compositions represented by vectors A through N. Table 5.—Compositions represented by some points on figure 7 Oxide ABCDEFGHI J K L M N Si02...................... 71.46 74.67 77.79 80.45 81.20 65.97 51.52 37.21 32.12 48.31 56.70 62.04 68.47 54.30 A1203..................... 15.75 14.03 12.35 10.82 9.16 13.29 17.23 21.15 22.91 20.07 18.59 17.65 16.52 18.60 FeO......................... 3.29 2.27 1.29 0.44 . 011 4.62 8.90 13.13 14.67 9.99 7.57 6.03 4.17 8.23 MgO......................... 0.00 0.00 0.00 0.09 1.04 4.10 6.98 9.83 10.56 6.14 3.85 2.39 0.63 4.81 CaO......................... 2.63 1.69 0.77 0.00 0.05 5.38 10.43 15.44 17.14 11.15 8.04 6.07 3.69 9.02 NaaO........................ 6.08 4.78 3.52 2.29 0.05 0.00 0.00 0.00 0.59 3.01 4.25 5.05 6.01 3.27 K20 ........................ 0.40 2.29 4.14 5.88 8.39 6.01 3.72 1.43 0.00 0.00 0.00 0.00 0.00 0.67 Ti02 ....................... 0.39 0.26 0.14 0.03 0.00 0.62 1.21 1.79 2.00 1.34 0.99 0.78 0.51 1.09 of the gabbroic magma represented by M-1 with a partial melt from the continental crust. For example, composition M-2 of table 7 could have been produced by mixing 11.76 percent magma of composition M-1 with 88.24 percent melt with a composition as follows: Si02 A1203 FeO MgO CaO Na20 K20 Ti02 68.02 18.70 1.29 0.00 2.76 4.78 3.79 0.67 The equivalent normative composition is: Q C Or Ab An Fs II 19.2 1.7 22.4 40.4 13.7 1.7 1.3 MIXING PROPORTIONS A statistical summary of the mixing proportions for the four end members, as required for the 480 samples accounted for by the model, is given in table 10. Some parameters derived from these values are summarized in table 11. The first of these parameters is the sumCHARACTERISTICS OF THE MODEL 15 Figure 8.—Stereogram showing the three-dimensional vector system for the batholithic rocks of the San Bernardino block of southern California. Open circles represent 125 sample vectors. Dashed line outlines area wherein all vectors represent compositions that are entirely nonnegative. All vector projections have been from an upper hemisphere vertically onto the plane of the stereogram. See table 6 for compositions represented by vectors A through N. Table 6.—Compositions represented by some points on figure 8 Oxide ABCDEFGH I J K L M N Si02...................... 68.12 72.37 75.29 77.16 80.54 80.06 66.25 49.93 9.94 33.21 46.51 55.78 62.86 56.67 A1203 .................... 20.09 16.50 14.04 12.44 9.53 8.04 10.68 13.82 21.77 21.85 21.93 21.96 22.02 18.33 FeO......................... 0.02 0.28 0.46 0.59 0.84 2.18 7.83 14.49 30.64 17.85 10.52 5.43 1.51 7.55 MgO......................... 0.00 0.00 0.00 0.00 0.00 0.60 3.38 6.65 14.60 8.53 5.05 2.64 0.78 3.52 CaO......................... 3.51 2.31 1.49 0.96 0.01 0.11 3.85 8.27 19.10 12.88 9.32 6.84 4.95 6.54 NaaO........................ 6.63 4.75 3.46 2.62 1.08 0.00 0.00 0.00 0.23 3.30 5.08 6.30 7.26 3.96 K20........................ 1.20 3.41 4.94 5.92 7.72 8.64 6.97 5.01 0.00 0.00 0.00 0.00 0.00 2.25 Ti02 ....................... 0.42 0.36 0.33 0.30 0.26 0.37 1.03 1.82 3.72 2.36 1.58 1.04 0.62 1.18 of the mixing proportions for end members M-\ and M-2, which gives the amount of magma from which each unit amount of sample was derived. The mean value of this parameter for all 480 samples is 1.8744, indicating that, according to the model, the average sample was formed by removal of 0.8744 parts of the differentiates (D-1 and D-2) from 1.8744 parts of magma (Af—1 plus M-2); this corresponds to differentiation of about 47 percent. Inspection of the mean values for batholithic rocks from the individual structural blocks (table 11) shows that those for the San Jacinto block call for the least amount of differentiation and that, in general, the amount of differentiation required tends to increase from the northeast to the southwest. This regional variability is also apparent from the map of individual sample values for this parameter in figure 9. (Note: The class intervals in figures 9-12 are defined by the 20th, 40th, 60th, and 80th percentiles of the mapped variable.) The second parameter derived from the mixing pro-16 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA Table 7.—Chemical and normative compositions of the end members for the mixing model End member M-1 M-2 D-2 D-2 Chemical composition (in percent) Si02 54.41 66.42 44.68 56.31 A1203 18.37 18.66 18.49 25.12 FeO 8.86 2.18 13.01 3.55 MgO 4.76 0.56 8.01 0.44 CaO 8.89 3.48 12.81 5.91 NagO 3.18 4.59 0.73 7.39 K20 0.30 3.38 0.42 0.14 Ti02 1.24 0.74 1.83 1.13 Normative composition (in percent) Q 4.2 17.3 0.0 0.0 C 0.0 1.1 0.0 2.1 Or 1.7 20.0 2.5 0.8 Ab 26.9 38.8 6.2 59.8 An 35.0 17.2 45.9 29.3 Ne 0.0 0.0 0.0 1.5 Wo 3.8 0.0 7.4 0.0 En 11.9 1.4 8.3 0.0 Fs 14.2 2.8 8.7 0.0 Fo 0.0 0.0 8.2 0.8 Fa 0.0 0.0 9.4 3.6 11 2.4 1.4 3.5 2.1 Table 8.—Compositions (in percent) of two samples of gabbros and the compositions represented by vectors G-1 and G-2 on figure 7 Sample Vector Oxide B545 B611 G-l G-2 Si02................................ 53.96 53.02 53.89 52.65 A1203............................... 18.01 19.25 18.64 19.02 FeO.................................. 8.86 8.71 8.35 8.72 MgO.................................. 5.04 4.73 4.96 5.16 CaO.................................. 9.14 9.66 9.18 9.60 Na20................................. 3.17 3.19 3.15 3.23 K20 ................................. 0.71 0.44 0.74 0.46 Ti02 ................................ 1.13 1.00 1.11 1.16 portions is the percentage of end member M-2 in the total magma (Af-1 plus M-2) required for each sample. The mean values (table 11) and the map of individual values (fig. 10) show that this parameter increases from the southwest to the northeast, suggesting that the more silicic and potassic extreme, M-2, was only a minor component in the magmas that formed the rocks of the Santa Ana and Perris blocks, but a major component in the magmas farther to the northeast. A major discontinuity in the map pattern on figure 10 occurs near the San Jacinto fault, in the eastern part of the Perris block. This discontinuity is discussed further in the next section of this report. The third parameter is the percentage of end member D-2 in the total of the differentiates (D-l plus D-2) that separated from the magmas to yield the liquids that later crystallized to form each sample. The mean values (table 11) and the map of individual values (fig. 11) show that this parameter is somewhat higher for the easternmost blocks of batholithic rocks, indicating that, according to the model, the precipitates in the eastern part of the region included a more sodic plagio-clase and less hornblende than those farther west in the Santa Ana and Perris blocks. The discontinuity in the pattern of figure 10 is also present in figure 11, but is somewhat less distinct. REGIONAL VARIATION IN THE COMPOSITIONS OF THE MAGMAS AND DIFFERENTIATES The compositions of the magmas and, presumably, the source materials required for each of the 480 samples, according to the model, can be determined byCHARACTERISTICS OF THE MODEL 17 Table 9.—Analyses from the U. S. Geological Survey RASS data file that compare closely with end-member compositions M-l and M-2 [Analyses of end members M-1 and M-2 are from table 7. See NOTE for description of samples (from RASS data file) for analyses 1-7. All analyses were recomputed to sum to 100 percent] M-l 1 M-2 2 3 4 5 6 7 Si02...................................... 54.41 54.01 66.42 66.05 65.90 65.56 65.49 66.58 66.39 A1203..................................... 18.37 19.06 18.66 18.78 18.22 18.19 18.43 18.67 18.75 FeO........................................ 8.86 8.39 2.18 3.17 3.27 3.21 3.18 2.33 2.53 MgO..................................... 4.76 4.92 . 56 .38 .61 .74 . 76 .42 . 76 CaO........................................ 8.89 8.61 3.48 3.44 2.97 3.35 3.50 3.39 3.79 NazO....................................... 3.18 3.48 4.59 3.86 4.30 4.27 4.12 3.59 5.22 K20 ..........................................30 . 52 3.38 3.76 4.20 4.17 4.02 4.62 2.15 Ti02.................................... 1.24 1.00 . 74 .56 .53 .50 .49 .40 . 41 NOTE.—Description of samples for analyses 1-7: Analysis 1. Gabbro of Jurassic age from southern Alaska, collected by B. L. Reed, 1970. Analysis 2. Andesite of Eocene age from southwest Nevada, collected by D. F. Crowder, 1969. Analysis 3. Quartz-latite welded tuff of Oligocene age from the San Juan Mts., Colo., collected by R. G. Luedke, 1969. Analysis 4. Welded tuff of Oligocene age from the Sand Juan Mts., Colo., collected by R. G. Luedke, 1970. Analysis 5. Crystal-rich welded tuff of Oligocene age from the Sand Juan Mts., Colo., collected by R. G. Luedke, 1970. Analysis 6. Quartz-latite welded tuff of Oligocene age from the San Juan Mts., Colo., collected by P. W. Lipman, 1974. Analysis 7. Volcanic tuff of Eocene age from northern Washington, collected by R. C. Pearson, 1973. Table 10.—Statistical summary of mixing proportions required for the batholithic rocks within structural blocks of southern California Number End member of M-l M-2 D-l D-2 M-l M-2 D-l D-2 Structural block samples Mean Standard deviation San Gabriel block........................ 47 0.8805 0.8571 -0.3889 -0.3488 0.2284 0.2924 0.1809 0.1875 San Bernardino block.................... 124 0.6422 1.1517 -0.3376 -0.4563 0.1828 0.2756 0.1321 0.1760 Little San Bernardino block......... 25 0.8778 1.0627 -0.4678 -0.4727 0.2997 0.2691 0.1846 0.1528 San Jacinto block........................ 90 0.7294 0.8825 -0.2816 -0.3303 0.2681 0.2459 0.1958 0.1872 Perris block............................ 115 1.2651 0.6742 -0.5421 -0.3973 0.3547 0.2669 0.2711 0.2105 Santa Ana block.......................... 79 1.6256 0.6400 -0.7744 -0.4912 0.3828 0.3055 0.3581 0.2422 All batholithic rocks................... 480 1.0053 0.8691 -0.4598 -0.4146 0.4645 0.3394 0.2883 0.2062 Table 11.—Mean values of parameters computed from the mixing proportions summarized in table 10 Relative amount of Percentage of M-2 Percentage of D-2 crust from which in crustal material in materials (D-l Structural block batholithic rocks (M-l plus M-2) from plus D-2) separated were derived which the batholithic from the crust to form rocks were derived the batholithic rocks San Gabriel block........................... 1.7377 48.68 47.26 San Bernardino block........................ 1.7938 63.74 57.12 Little San Bernardino block................. 1.9405 55.04 51.04 San Jacinto block........................... 1.6119 55.08 56.19 Perris block................................ 1.9394 34.71 41.75 Santa Ana block............................. 2.2656 27.12 38.04 All batholithic rocks....................... 1.8744 47.21 48.84 mathematical mixing of end members M-l and M-2 (table 7) according to the mixing proportions derived for each sample (Appendix 2), followed by adjustment of each mixture so that the sum of the oxide values is 100 percent. The average compositions of the magmas required for each structural block, and for the batholithic rocks as a whole, are given in table 12. Note that the magma composition required for the batholithic rocks of the Little San Benardino and San Jacinto blocks, on opposite sides of the San Andreas fault zone,18 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 0 25 50 MILES Figure 9.—Relative amounts of magma required for the formation of the batholithic rocks. are almost identical even though the average compositions of the samples from these two blocks are somewhat different (table 1). Average compositions of the differentiates that were separated from the magmas to form the batholithic rocks are given in table 13; these averages were obtained from compositions D-1 and D-2 in table 7 and the mixing proportions for these end members in Appendix 2. Maps of chemical variation in the batholithic rocks of southern California, as determined from the original data (Appendix 1) on the 480 samples accounted for by the model, are given in figure 12. This figure also shows the areal compositional variations in the magmas and in the differentiates according to the model. As described previously, the original chemical data show continuous patterns of areal change from the southwest toward the continental interior. In marked contrast, most maps of model-derived parameters (figs. 9-11) and compositions of magmas and differentiates (fig. 12) show pronounced discontinuities near the boundary between the Perris and San Jacinto blocks, within the Peninsular Ranges Province. The exact locations of these discontinuities differ slightly from map to map. Nevertheless, we regard the discontinuity as a single feature, although of uncertain exact location. No other feature is as prominent in these map patterns as is this discontinuity. Certainly, the San Andreas fault, both east of the Peninsular Ranges and through the Transverse Ranges, does not correlate with any map pattern as significant as the discontinuity near the San Jacinto fault trace. In fact, if these maps, derived from the model, were used to establish petrologic provinces, two such provinces would emerge: one composed of the San Gabriel, San Jacinto, San Bernardino, and Little San Bernardino blocks and another composed of the Perris and Santa Ana blocks. ANOMALOUS SAMPLES The distribution of sampling localities whose chemical compositions are not accounted for by the model proposed here is far from random; 11 of 17 are in the SanCHARACTERISTICS OF THE MODEL 19 0 25 50 MILES Figure 10.—Percentages of end-member M-2 in the magmas. Jacinto block. Further, 8 of the 11 localities are tightly clustered (fig. 3) in the easternmost part of the block. On the basis of strontium and oxygen isotopes, Taylor and Silver (1978, p. 425) interpreted the batholithic rocks of the San Jacinto block to be anomalous and “derived from a distinctive source rock at depth.” Further, high values of 8180, exceeding + 10 in the easternmost San Jacinto block, show an anomalous circular low that coincides with the cluster of eight sampling localities that are anomalous with respect to our model (fig. 3). With two exceptions in the Perris block, all other anomalous localities are at or near the exposed margins of the batholithic rocks, or in the zones of faulting that mark the boundaries between blocks. RELATIONS TO OTHER GEOLOGIC FEATURES The discontinuity in the areal distributions of the model parameters and compositions derived from the model correlate well with a number of field relations, age relations, and petrologic and structural features of the batholithic and pre-batholithic rocks. However, except that the most pronounced breaks in the map patterns occur near the San Jacinto fault, there is a poor correlation with the pattern of major faults and especially no correlation with the San Andreas fault zone. This lack of correlation is surprising in view of the huge displacements widely accepted (for example, Crowell, 1979) for the San Andreas fault in southern California. Regional asymmetries that appear to correlate with the discontinuity in the model parameters have been described by many investigators working over a larger area than that considered here. In each of the asymmetries, regional gradients are interrupted by sharp changes in the eastern Peninsular Ranges, near the trace of the San Jacinto fault: 1. Pre-batholithic rock types.—Metavolcanic rocks in the west are succeeded eastward by metasedimentary rocks (Gastil and others, 1978) over much of the length of the Peninsular Ranges. A belt of carbonate (shelf?) rocks has been recognized in the Transverse Ranges (Woodford, 1960). The overall west-to-east sequence20 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 0 25 50 MILES Figure 11.—Percentages of end-member D-2 in the differentiates. seems to be blueschist-metavolcanic-metasedimentary (clastic)-metasedimentary(carbonate). Batholithic rocks are emplaced in all types, except in blueschist. The most mafic parts of the batholithic rocks are associated with the metavolcanics, and the most felsic with metasedimentary rocks; this change is at or close to the boundary of the Perris and San Jacinto blocks. 2. Wallrock metamorphism.—Metamorphic grade increases eastward, with an abrupt increase in grade to higher amphibolite facies across a line in the central to eastern Perris block (Schwarcz, 1969). 3. Plutonic rock types.—Gabbroic rocks are restricted to the Santa Ana block and the western part of the Perris block (Baird and others, 1979) and similar relations are noted farther south (Gastil and others, 1975; Todd and Shaw, 1979). Granodiorite dominates in the San Jacinto block and quartz diorite in the Perris block (Baird and others, 1979, fig. 4). This change may mark the “quartz diorite boundary” of Moore (1959) in this region and is coincident with the discontinuities ap- parent in figure 10 and in the maps in figure 12 that represent the magmas and, presumably, their source materials. 4. Plutonic rock mineralogy.—Rocks east of a generally north-south line through the Perris block tend to be distinctly richer in potassic feldspar, commonly present as large phenocrysts, and richer in sphene (D. M. Morton, oral commun., 1981; Gastil and others, 1975). 5. Plutonic rock structure and form.—Plutons of the western Peninsular Ranges tend to be small, contain internal structures that suggest flow and (or) deformation, and have contact relations that suggest syntec-tonic emplacement (D. M. Morton, oral commun., 1980; Todd and Shaw, 1979). The eastern Peninsular Ranges and eastern Transverse Ranges are dominated by larger plutons that have irregular outlines and massive interiors; they are characteristically post-tectonic (Todd and Shaw, 1979). 6. Radiometric ages.—The extensive uranium-lead zircon dating program conducted by Silver and col-.DISCUSSION 21 Table 12.—Average compositions (in percent) of the magmas for the batholithic rocks within structural blocks of southern California Number of Oxide Structural block samples Si02 A1203 FeO MgO CaO Na20 K20 Ti02 San Gabrial block........................ 47 60.26 18.51 5.61 2.72 6.26 3.86 1.80 0.99 San Bernardino block.................... 124 62.06 18.55 4.60 2.09 5.44 4.08 2.26 0.92 Little San Bernardino block.............. 25 61.02 18.53 5.18 2.45 5.91 3.95 1.99 0.96 San Jacinto block........................ 90 61.00 18.53 5.18 2.46 5.92 3.95 1.99 0.96 Perris block............................ 115 58.58 18.47 6.54 3.30 7.01 3.67 1.37 1.06 Santa Ana block.......................... 79 57.67 18.45 7.05 3.62 7.42 3.56 1.13 1.10 All batholithic rocks................... 480 60.08 18.51 5.71 2.78 6.34 3.84 1.75 1.00 Table 13.—Average compositions (in percent) of differentiates separated from the magmas to form the batholithic rocks within structural blocks of southern California Number of Oxide Structural block samples Si02 A1203 FeO MgO CaO Na20 K20 TiOz San Gabrial block........................ 47 50.18 21.63 8.54 4.44 9.55 3.88 0.29 1.50 San Bernardino block.................... 124 51.32 22.28 7.61 3.69 8.87 4.53 0.26 1.43 Little San Bernardino block......... 25 50.62 21.88 8.18 4.15 9.29 4.13 0.28 1.47 San Jacinto block........................ 90 51.22 22.22 7.70 3.76 8.94 4.47 0.26 1.44 Perris block............................ 115 49.54 21.26 9.06 4.85 9.93 3.51 0.30 1.54 Santa Ana block.......................... 79 49.10 21.02 9.41 5.13 10.19 3.26 0.31 1.56 All batholithic rocks................... 480 50.36 21.73 8.39 4.32 9.44 3.98 0.28 1.49 leagues (summarized in Silver and others, 1979) has demonstrated an age “step” at 105 m.y. (million years) along a boundary coincident with the change from syn-tectonic to post-tectonic plutons noted in paragraph 5— older plutonic dates to the west, younger to the east. (Potassium-argon dates are also available, but these present further problems of varying cooling histories and argon retentions not relevant to this paper.) 7. Isotopic and trace-element patterns.—Strongly correlated with the age “step” (paragraph 6) is an abrupt increase in 8180 (Taylor and Silver, 1978; further discussed in Silver and others, 1979) over normal igneous values of +6 to +8 in the west to +9 to +11 in the eastern part of the Peninsular Ranges. Contours of 8lsO values trend more northerly than do the strikes of major fault zones and do not appear significantly offset by the faults. The step crosses the Perris block coincident with the other changes just noted. Other isotopic and trace-element data (strontium, rare-earth elements) show close correlations to the patterns of oxygen data (Gromet and Silver, 1979; Silver and others, 1979). The choice of end members M-1 and M-2 was not totally objective, but the derived map patterns are not highly sensitive to these choices. M-1 and M-2 can be changed significantly without eradicating the marked discontinuity near the San Jacinto fault zone. On the other hand, the map patterns can be eradicated almost completely by radical changes in the choice of end-member compositions. The most common difficulty with the alternative models that result, however, is that they call for far greater degrees of differentiation than the model proposed here. In summary, the model calls for a regional discontinuity in the compositions of the magmas and in the compositions of the differentiates separated from them. The discontinuity in the magmas is presumed to represent a discontinuity in the compositions of the magma source materials. The discontinuity occurs mainly in the vicinity of the San Jacinto fault zone near the center of the Perris block, and corresponds generally to the location of other geochemical, mineralogical, petrological, radiometrical and structural discontinuities. In addition, most of the 17 samples found to be anomalous with respect to the other 480 are from a part of the region identified as anomalous in previous studies of strontium and oxygen isotopes. DISCUSSION Investigations cited and summarized in this paper are leading toward some firm conclusions about the origins of at least the Peninsular Ranges part of the southern22 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA California batholithic rocks. The batholithic rocks are divided longitudinally into a western belt made up of older syntectonic plutons of quartz diorite and abundant gabbro that were derived from upper mantle rocks of a primitive nature, and an eastern belt of chiefly younger post-tectonic plutons of mainly granodiorite and quartz monzonite of a nonprimitive nature (Taylor and Silver, 1978). The oxygen-isotope data indicate that the eastern source materials once occurred in a nearsurface environment. Baird, Baird, and Welday (1974) and Todd and Shaw (1979) concluded, on separate grounds, that the gabbroic magmas were distinct from those that supplied the quartz plutonites. Silver and colleagues proposed (see summary in Silver and others, 1979) that sources of the Peninsular Ranges rocks were fundamentally basaltic; rare-earth-element patterns show that variations in source rocks, not high-level crystal fractionation or differentiation, were responsible for zonation across the region. They believed that the simplest explanation for all observed trace-element and isotopic patterns is a two-end-member source for the batholithic rocks. Allegre and Othman (1980) showed, on the basis of neodymium-strontium isotopic relations, that one end member must have consisted of a large fraction of recycled older continental crust. The model we have presented here seems compatible with these conclusions. M-1 is basaltic and could have been derived from upper mantle sources. More or less pure M-l, modified or unmodified by the redistribution of crystals, could have supplied the gabbroic magmas of the western Peninsular Ranges. M-2, a quartzo-feldspathic type, represents the nonprimitive end member and may be composed, in some large part, of a partial melt from the continental crust. With but few exceptions, the magmas required for each of the 497 samples of the batholithic rocks range in composition between M-l and M-2. The model calls for areal patterns with a discontinuity in the eastern part of the Peninsular Ranges, probably representing both the time and place where significant amounts of continental materials became involved. Although we interpret M-l and M-2 as representing magmas and D-l and D-2 as representing differentiates, we recognize that M-l and M-2 could also be interpreted as the extremes in a range of crustal rocks from which partial melts were derived. End-members D-l and D-2 then could be interpreted as the extremes in a range of mineral assemblages separated from the crustal rocks, not by magmatic differentiation, but by residual concentration as refractory materials left behind on crustal melting. Also possible is that both differentiation and residual concentration occurred to varying degrees over the region, but the two processes are not distinguishable from our model or from the raw chemical data. However, the compositional discon- tinuity in the eastern part of the Peninsular Ranges is a necessary part of the model regardless of how the end members are interpreted. SUMMARY AND CONCLUSIONS The batholithic rocks of southern California were examined by a method of Q-mode factor analysis and were found to have a compositional structure similar to that of the Sierra Nevada batholith. The compositional structure indicates that a petrogenetic model with only four end members will account for 85-97 percent of the variation in each of the eight major oxide constituents. We assumed that two of the end members represent an assemblage of plagioclase and mafic minerals, largely hornblende, that ranges in composition between D-l and D-2 of table 7. We further assumed that the other two end members represent a range of magmas from which plagioclase and the mafic minerals separated; the range was determined by computations based on the most likely source-magma compositions for the rocks of the southwestemmost Santa Ana block and the northeastemmost San Bernardino block. The limits of the range are given as end members M-l and M-2 in table 7. Each of 480 sample compositions can be closely approximated as a mixture of the four derived end-member compositions. The 17 sample compositions that cannot be approximated in this manner are regarded as anomalous. The mixing proportions indicate that the batholithic rocks as a group required differentiation of about 47 percent; that is, about 47 percent of the magma was removed in the form of plagioclase, hornblende, and other mafic minerals. The model proposed here is comprised of the end-member compositions given in table 7 and of the mixing proportions in Appendix 2. The mathematical validity of the model can be verified by mathematically combining the magma compositions (M-l and M-2 in table 7) and separating the differentiate compositions (D-l and D-2 in table 7) according to the mixing proportions in Appendix 2. The results will give close approximations of the original data in Appendix 1. The goodness-of-fit of the model to the original data is given by the parameters in table 2. The loss in goodness-of-fit that would be obtained by using fewer than four end members, or the improvement by adding more end members, can be observed from the factor-variance diagram in figure 4. The principal advantage of the model is that it allows examination of its separate components, specifically the magmas and the mineral assemblages (differentiates) that were separated from them. Comparison of the compositional variation in the batholithic rocks with the compositional variation in the magmas, has been of particular interest. The variation in the magmas, or theREFERENCES CITED 23 magma source materials, shows a discontinuity that is not evident in the original compositional data. The discontinuity is mainly near the San Jacinto fault zone close to the center of the Perris block. The general correspondence of the discontinuity to similar apparent discontinuities in mineralogic, petrologic, isotopic, and structural properties of the batholithic rocks suggest that the discontinuity is real and that the derived model is valid in at least a general way. The discontinuity is interpreted as the western limit of significant contribution of quartzo-feldspathic materials in the continental crust to the magmas that formed the batholithic rocks. The compositional variations in the magmas, or their source materials, show no discontinuity at or near the San Andreas fault zone; this suggests that no large amount of displacement has occurred along this particular strand of the fault since the emplacement of the batholithic rocks. REFERENCES CITED Abbott, P. L., and Todd, V. R., eds., 1979, Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, 286 p. Allegre, C. J., and Othman, D. B., 1980, Nd-Sr isotopic relationship in granitoid rocks and continental crust development; a chemical approach to orogenesis: Nature, v. 286, p. 335-346. Atwater, T., 1970, Implications of plate tectonics for the Cenozoic tectonic evolution of western North America: Geological Society of America Bulletin, v. 81, p. 3515-3536. Baird, A. K., 1975, Plutonic zones in the Peninsular Ranges of southern California and Northern Baja California—Comment: Geology, v. 3, p. 676-677. Baird, A. K., Baird, K. W., and Welday, E. E., 1974, Chemical trends across Cretaceous batholithic rocks of southern California: Geology, v. 2, p. 493-496. ------1979, Batholithic rocks of the northern Peninsular and Transverse Ranges, southern California, in Abbott, P. L., and Todd, V. R., eds., Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, p. 111-132. Baird, A. K., McIntyre, D. B., and Welday, E. E., 1967, Geochemical and structural studies in batholithic rocks of southern California; Part II, sampling of the Rattlesnake Mountain pluton for chemical composition and variation: Geological Society of America Bulletin, v. 78, p. 191-122. Baird, A. K., Morton, D. M., Woodford, A. 0., and Baird, K. W., 1974, The Transverse Ranges Province; a unique structural-petrochemical belt across the San Andreas fault system: Geological Society of America Bulletin, v. 85, p. 163-174. Bateman, P. C., Clark, L. D., Huber, N. K., Moore, J. G., and Rinehart, C. D., 1963, The Sierra Nevada batholith—A synthesis of recent work across the central part: U.S. Geological Survey Professional Paper 414-D, p. D1-D46. Bateman, P. C., and Dodge, F. C. W., 1970, Variation of major chemical constituents across the central Sierra Nevada batholith: Geological Society of America Bulletin, v. 81, p. 409-420. Crowell, J. C., 1979, The San Andreas fault system through time: Quarterly of the Journal Geological Society of London, v. 136, p. 293-302. DePaolo, D. J., 1980, Sources of continental crust; neodymium isotope evidence from the Sierra Nevada and Peninsular Ranges: Science, v. 209, no. 4457, p. 684-687. Gastil, G., Morgan, G. J., and Krummenacher, D., 1978, Mesozoic history of Peninsular California and related areas east of the Gulf of California, in Howell, D. G., and McDougall, K. A., eds., Mesozoic paleogeography of the western United States, Pacific Coast paleogeography symposium 2: Pacific Section, Society of Economic Paleontologists and Mineralogists, p. 107-115. Gastil, R. G., Phillips, R. P., and Allison, E. C., 1975, Reconnaissance geology of the State of Baja California: Geological Society of America Memoir 140, 170 p. Gromet, P. L., and Silver, L. T., 1979, Profile of rare earth element characteristics across the Peninsular Ranges batholith near the international border, southern California, U.S.A., and Baja California, Mexico, in Abbott, P. L., and Todd, V. R., eds., Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, p. 133-142. Hadley, D., and Kanamori, H., 1977, Seismic structure of the Transverse Ranges, California: Geological Society of America Bulletin, v. 88, p. 1469-1478. Hill, M. L., and Dibblee, T. W., 1953, San Andreas, Garlock, and Big Pine faults, California—A study of the character, history, and tectonic significance of their displacements: Geological Society of America Bulletin, v. 64, p. 443-458. Imbrie, J., and Purdy, E. G., 1962, Classification of modern Bahamian carbonate sediments, in Ham, W. E., ed., Classification of carbonate rocks: American Association of Petroleum Geologists Memoir 1, p. 253-272. Krummenacher, D., Gastil, R. G., Bushee, J., and Doupont, J., 1975, K-Ar apparent ages, Peninsular Ranges batholith, southern California and Baja California: Geological Society of America Bulletin, v. 86, p. 760-768. Larsen, E. S., Jr., 1948, Batholith and associated rocks of Corona, Elsinore, and San Luis Rey quadrangles, southern California: Geological Society of America Memoir 29, 182 p. Larsen, E. S., Jr., and Draisin, W. M., 1950, Composition of the minerals in the rocks of the southern California batholith: International Geological Congress, Report of 18th Session, Great Britain, 1948, pt. II, p. 66-79. Luyendyk, B. P., Kamerling, M. J., and Terres, R., 1979, Geometric model for Neogene crustal rotations in southern California [abs.]: Geological Society of America Abstracts with Programs, v. 11, p. 470. Miesch, A. T., 1976a, Q-mode factor analysis of compositional data: Computers & Geosciences, v. 1, p. 147-159. ------1976b, Q-mode factor analysis of geochemical and petrologic data matrices with constant row-sums: U.S. Geological Survey Professional Paper 574-G, 47 p. ------1976c, Interactive computer programs for petrologic modeling with extended Q-mode factor analysis: Computers & Geosciences, v. 2, p. 439-492. ------1979, Vector analysis of chemical variation in the lavas of Paricutin volcano, Mexico: Mathematical Geology, v. 11, p. 345-371. ------1980, Scaling variables and interpretation of eigenvalues in principal component analysis of geologic data: Mathematical Geology, v. 12, p.523-538. Miesch, A. T., and Morton, D. M., 1977, Chemical variability in the Lake view Mountains pluton, southern California batholith—A comparison of the methods of correspondence analysis and extended Q-mode factor analysis; U.S. Geological Survey Journal of Research, v. 5, p. 103-116. Miesch, A. T., and Reed, B. L., 1979, Compositional structures in two batholiths of circum-Pacific North America: U.S. Geological Survey Professional Paper 574-H, 31 p. Miller, C. F., 1977, Early alkalic plutonism in the calc-alkaline batholithic belt of California: Geology, v. 5, p. 685-688. Moore, J. G., 1959, The quartz diorite boundary line in the western United States: Journal of Geology, v. 67, p. 198-210.24 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA Morton, D. M., and Baird, A. K., 1976, The Paloma Valley ring dike complex: U.S. Geological Survey Journal of Research, v. 4, p. 83-89. Morton, D. M., Baird, A. K., and Baird, K. W., 1969, The Lakeview Mountains pluton, southern California batholith, Part II; Chemical composition and variation: Geological Society of America Bulletin, v. 80, p. 1553-1564. Nishimori, R. K., 1976, The petrology and geochemistry of the gab-bros from the Peninsular Ranges batholith, California, and a model for their origin: Scripps Institute of Oceanography, Ph.D. thesis, 234 p. Nockolds, S. R., and Allen, R., 1953, Geochemistry of some igneous rock series, I: Geochimica et Cosmochimica Acta, v. 4, p. 105-142. Reed, B. L., and Lanphere, M. A., 1973, Alaska-Aleutian Range batholith; geochronology, chemistry, and relation to circum-Pacific plutonism: Geological Society of America Bulletin, v. 84, p. 2583-2610. Richmond, J. F., 1965, Chemical variation in quartz monzonite from Cactus Flat, San Bernardino Mountains, California: American Journal of Science, v. 263, p. 53-63. Schwarcz, H. P., 1969, Pre-Cretaceous sedimentation and metamorphism in the Winchester area, northern Peninsular Ranges, California: Geological Society of America Special Report 100, 63 p. Silver, L. T., 1971, Problems of crystalline rocks of the Transverse Ranges [abs.]: Geological Society of America Abstracts with Programs, v. 3, p. 193-194. Silver, L. T., Early, T. O., and Anderson, T. H., 1975, Petrological, geochemical and geochronological asymmetries of the Peninsular Ranges batholith [abs.]: Geological Society of America Abstracts with Programs, v. 7, p. 375-376. Silver, L. T., Taylor, H. P., Jr., and Chappell, B., 1979, Some petrological, geochemical and geochronological observations of the Peninsular Ranges batholith near the international border of the U.S.A. and Mexico, in Abbott, P. L., and Todd, V. R., eds., Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, p. 83-110. Streckeisen, A. L., 1973, Plutonic rocks, classification and nomenclature recommended by the I.U.G.S. Subcommision on the sys-tematics of igneous rocks: Geotimes, v. 18, p. 26-30. Stuckless, J. S., and Miesch, A. T., 1981, Petrogenetic modeling of a potential uranium source rock, Granite Mountains, Wyoming: U. S. Geological Survey Professional Paper 1225, 34 p. Stuckless, J. S., Miesch, A. T., Goldich, S. S., and Weiblen, P. W., 1981, A Q-mode factor model for the petrogenesis of the volcanic rocks from Ross Island and vicinity, Antarctica: American Geophysical Union, Dry Valley Drilling Project, Antarctica Research Series, v. 33, p. 257-280. Taylor, H. P., Jr., and Silver, L. T., 1978, Oxygen isotope relationships in plutonic igneous rocks of the Peninsular Ranges batholith, southern and Baja California: U.S. Geological Survey Open-File Report 78-701, p. 423-426. Todd, V. R., and Shaw, S. E., 1979, Structural, metamorphic and intrusive framework of the Peninsular Ranges batholith in southern San Diego County, California, in Abbott, P. L., and Todd, V. R., eds., Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, p. 177-231. Van Trump, G., Jr., and Miesch, A. T., 1977, The U.S. Geological Survey RASS-STATPAC system for management and statistical reduction of geochemical data: Computers & Geosciences, v. 3, p. 475-488. Walawender, M. J., 1979, Basic plutons of the Peninsular Ranges batholith, southern California, in Abbott, P. L., and Todd, V. R., eds., Mesozoic crystalline rocks: San Diego State University, Department of Geological Sciences, p. 151-162. Woodford, A. O., 1960, Bedrock patterns and strike-slip faulting in southwestern California: American Journal of Science, v. 258A, p. 400-417. Yeats, R. S., 1981, Quaternary flake tectonics of the California Transverse Ranges: Geology, v. 9, p. 16-20.FIGURE 1 Maps of the variability of Si02, AI2O3, FeO, MgO, CaO, Na20, K20, and Ti02 in the batholithic rocks of southern California and in the magmas and differentiates as interpreted from the model. Class intervals are defined by the 20th, 40th, 60th, and 80th percentiles of the mapped variables.26 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 119° 118° 117° 116° 0 25 50 KILOMETERS 0 25 50 MILES Si02 IN THE DIFFERENTIATESFIGURE 12 27 o ~r 25 1 50 MILES28 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 0 25 0 50 KILOMETERS FeO IN THE DIFFERENTIATES 25 50 MILESFIGURE 12 2930 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 119° 118° 117° 116° 0 25 50 KILOMETERS 0 25 50 MILES CaO IN THE DIFFERENTIATESFIGURE 12 31 o 25 50 MILES32 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA 119“ 118“ 117“ 116“ 34“ 33“ 0 25 50 KILOMETERS _L k2o in the differentiates 0 T 25 1 50 MILESFIGURE 12 33APPENDIXES 1 AND 2 Chemical data on batholithic rocks of southern California (Appendix 1) and mixing proportions for end-members Af-1, M-2, D-l, and D-2 (Appendix 2)36 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA APPENDIX 1 Chemical data on batholithic rocks of southern California, in percent Oxide Sample Number Si02 AI2O3 FeO MgO CaO Na20 k2o Ti02 San Gabriel block B 1 71.21 15.41 2.50 0.87 2.28 4.14 3.16 0.44 B 2 69.06 16.59 2.71 0.95 2.71 4.38 3.12 0.49 B 3 63.09 16.98 5.45 2.48 4.66 3.92 2.65 0.77 B 4 65.87 15.57 4.23 2.81 3.87 3.88 3.19 0.58 B 5 67.56 15.68 3.91 1.60 3.65 3.62 3.42 0.56 B 6 75.05 14.39 1.19 0.25 0.83 4.01 4.11 0.17 B 7 69.56 15.78 3.07 1.21 2.63 4.03 3.32 0.40 B 8 67.77 16.49 3.71 1.57 3.21 4.22 2.55 0.49 B 9 72.99 15.30 1.64 0.50 1.67 4.17 3.44 0.30 B 10 71.97 15.33 1.71 0.50 1.73 3.94 4.46 0.37 B 11 70.81 16.05 2.94 0.97 2.12 3.36 3.27 0.49 B 12 66.75 16.44 4.14 1.52 3.57 3.91 2.99 0.68 B 13 67.53 16.17 3.95 1.31 3.52 3.70 3.16 0.64 B 14 59.91 17.51 6.65 3.32 5.45 3.96 2.27 0.92 B 15 63.66 16.48 5.43 2.61 4.82 3.81 2.42 0.78 B 16 62.14 17.27 5.43 2.63 5.49 3.93 2.32 0.79 B 17 68.15 15.33 4.13 1.49 3.25 3.68 3.38 0.59 B 18 73.72 14.76 1.61 0.60 1.93 3.61 3.55 0.22 B 19 75.01 14.25 1.15 0.22 0.80 3.52 4.83 0.22 B 20 75.46 14.57 0.91 0.23 0.76 3.16 4.76 0.15 B 21 66.36 16.46 4.11 1.54 4.10 4.09 2.56 0.79 B 22 62.10 18.33 5.78 2.81 5.21 3.55 1.46 0.75 B 23 64.83 17.30 4.47 1.67 4.64 4.13 2.07 0.89 B 24 62.14 16.93 5.64 2.88 5.43 3.85 2.21 0.92 B 25 74.14 14.51 1.46 0.47 1.32 3.79 4.16 0.17 B 26 73.29 15.56 1.55 0.40 1.55 3.53 3.88 0.24 B 27 63.59 18.44 4.66 1.52 4.01 4.53 2.64 0.62 B 28 68.64 16.90 2.59 0.95 2.47 3.85 4.18 0.42 B 29 66.93 16.47 3.63 1.55 3.79 3.75 3.18 0.70 B 30 67.26 16.05 4.25 1.50 3.14 4.10 2.99 0.70 B 31 69.51 16.12 2.37 0.67 2.07 4.21 4.63 0.42 B 32 65.29 16.85 4.51 1.57 4.26 4.09 2.90 0.54 B 33 60.46 18.66 6.29 2.49 5.44 3.83 2.16 0.67 B 34 68.13 15.92 3.63 1.53 3.20 3.64 3.42 0.54 B 35 73.05 14.80 1.68 0.57 1.83 3.76 4.10 0.22 B 36 67.43 16.02 3.77 1.58 3.42 3.72 3.57 0.49 B 37 62.59 17.03 5.70 2.43 4.36 3.89 3.17 0.83 B 38 58.08 18.48 6.82 3.48 6.73 3.61 1.98 0.82 B 39 58.46 18.09 7.05 3.31 5.86 4.16 2.23 0.85 B 40 63.68 16.58 5.29 2.49 4.49 3.96 2.85 0.66 B 41 56.51 19.22 7.05 3.33 6.79 4.12 1.92 1.06 B 42 60.61 17.37 5.96 3.08 5.16 4.16 2.86 0.79 B 43 67.08 16.51 3.88 1.38 3.95 3.79 2.64 0.77 B 44 63.60 17.92 4.74 1.51 4.03 4.10 3.37 0.74 B 46 71.05 16.28 1.94 0.57 2.22 4.00 3.57 0.37 B 47 70.16 16.16 2.17 0.53 2.17 4.08 4.35 0.37 B 48 71.25 16.15 1.81 1.03 2.70 3.96 2.73 0.37 San Bernardino block B 54 72.80 15.45 0.74 0.27 2.43 3.97 3.96 0.37 B 55 70.24 15.38 2.45 0.88 3.32 3.60 3.53 0.60 B 56 64.65 17.02 4.01 1.56 4.39 4.27 3.16 0.93 B 57 63.99 17.55 4.05 2.17 4.35 4.01 2.98 0.89 B 58 64.85 16.92 4.11 1.62 4.23 4.10 3.27 0.90 B 59 64.96 17.07 3.89 1.57 4.32 4.15 3.15 0.88 B 60 67.37 16.26 3.29 1.11 3.64 3.91 3.72 0.71 B 61 62.06 17.13 5.48 2.50 5.31 3.91 2.64 0.96 B 62 64.31 17.15 4.19 1.61 4.48 4.21 3.08 0.97 Oxide Sample Number sio2 A12°3 FeO MgO CaO Na20 k2° Ti02 B 63 70.13 15.49 2.14 0.80 3.02 3.71 4.17 0.55 B 64 73.67 14.58 0.81 0.17 1.92 3.71 4.81 0.34 B 65 73.66 14.34 1.41 0.25 1.94 3.64 4.38 0.37 B 66 66.84 16.15 3.48 1.46 3.80 3.71 3.83 0.74 B 67 66.97 17.34 2.93 1.63 3.44 3.84 3.20 0.66 B 68 73.30 14.65 1.19 0.27 2.13 3.46 4.63 0.37 B 69 69.82 15.54 2.41 0.85 3.28 3.94 3.54 0.62 B 70 64.11 16.67 4.46 1.99 4.61 3.88 3.35 0.93 B 71 68.14 15.86 3.25 1.10 3.58 3.94 3.45 0.68 B 72 74.06 14.40 0.89 0.20 1.82 3.42 4.85 0.37 B 74 63.26 16.49 5.08 2.64 4.87 3.94 2.76 0.96 B 75 73.89 14.53 0.78 0.30 2.24 3.79 4.11 0.37 B 78 70.40 15.96 2.51 0.83 1.60 3.98 4.16 0.55 B 79 70.35 15.31 2.27 0.78 2.43 4.41 3.86 0.58 B 80 66.63 16.46 3.51 2.08 3.86 3.87 2.74 0.85 B 81 70.77 15.22 2.40 0.98 2.15 4.04 3.78 0.66 B 85 68.75 15.53 2.90 0.96 2.94 3.71 4.52 0.68 B 86 76.76 12.83 0.16 0.02 1.28 3.65 5.04 0.27 B 94 71.45 15.20 1.49 0.40 2.23 3.67 5.04 0.52 B 95 72.71 14.72 1.02 0.30 2.03 3.64 5.10 0.47 B 96 73.80 14.62 0.89 0.35 2.23 3.81 3.94 0.37 B 97 62.97 16.93 4.60 2.32 4.82 3.97 3.55 0.83 B 98 67.21 15.86 2.92 1.17 2.82 4.54 4.74 0.73 B 99 65.37 15.99 3.44 1.58 4.12 3.64 5.05 0.82 B100 73.31 15.55 0.13 0.10 2.45 4.03 4.16 0.27 B101 68.82 15.55 2.58 1.11 3.21 3.95 4.15 0.63 B102 67.58 16.13 2.77 1.22 3.45 4.05 4.12 0.68 B103 70.07 15.51 2.14 0.84 2.75 3.74 4.37 0.57 B104 68.46 16.26 2.50 1.03 3.06 4.19 3.95 0.55 B105 75.51 13.35 0.76 0.20 1.50 3.93 4.38 0.37 B106 71.97 14.80 1.64 0.60 2.43 3.81 4.25 0.50 B107 71.15 15.22 1.73 0.71 2.57 3.83 4.26 0.53 B108 63.94 16.15 5.20 2.48 4.62 4.01 2.91 0.68 B109 75.28 13.32 1.95 0.32 1.04 2.74 5.02 0.34 B110 75.96 12.85 1.05 0.20 1.47 2.98 5.14 0.35 Bill 76.19 13.02 0.14 0.08 1.34 2.78 6.13 0.32 B112 75.73 12.36 1.39 0.35 1.70 3.27 4.81 0.39 B113 75.85 12.70 1.13 0.22 1.44 3.02 5.26 0.39 B114 77.40 12.11 0.69 0.08 1.28 2.87 5.22 0.33 B115 73.67 15.37 0.40 0.20 1.81 4.01 4.23 0.32 B116 63.70 16.70 4.76 2.18 4.75 3.85 3.31 0.75 B117 70.91 16.05 1.54 0.56 2.84 4.17 3.50 0.43 B118 73.59 14.63 1.10 0.31 2.23 3.47 4.28 0.38 B120 71.82 14.39 2.16 0.56 2.01 3.45 5.10 0.51 B121 68.79 15.48 3.02 1.25 2.61 3.82 4.24 0.77 B124 72.66 14.36 1.67 0.55 2.39 3.60 4.32 0.45 B125 73.18 14.67 1.12 0.37 2.08 3.80 4.31 0.47 B127 72.57 15.13 1.38 0.37 1.97 3.62 4.49 0.47 B128 63.78 16.91 4.50 2.07 4.89 3.81 3.28 0.77 B129 73.39 14.42 1.22 0.33 1.88 3.36 5.01 0.40 B130 72.84 15.25 1.30 0.49 1.78 3.48 4.43 0.44 B131 73.61 14.59 1.04 0.25 2.13 3.58 4.44 0.37 B132 73.20 14.95 1.25 0.30 2.24 3.62 4.03 0.40 B133 74.33 14.48 0.79 0.08 1.63 3.60 4.76 0.33 B134 73.95 14.64 0.90 0.18 1.85 3.79 4.35 0.34 B135 72.37 15.73 0.68 0.20 2.21 4.04 4.44 0.33 B136 71.72 15.20 1.51 0.55 2.52 3.81 4.18 0.52 B137 71.62 15.23 1.76 0.59 2.44 3.71 4.14 0.52 B138 72.45 14.96 1.42 0.40 2.27 3.70 4.34 0.46 B139 72.86 14.88 1.29 0.38 2.18 3.68 4.28 0.45 B140 72.44 14.99 1.42 0.38 2.24 3.58 4.50 0.45 B141 70.93 15.40 1.77 0.57 2.58 3.88 4.35 0.52 B142 71.81 15.15 1.54 0.51 2.34 3.62 4.54 0.48 B143 71.96 15.07 1.41 0.45 2.44 3.72 4.50 0.45 B144 72.13 14.72 1.69 0.51 2.35 3.62 4.49 0.48APPENDIX 1 37 APPENDIX 1.—Continued Oxide Sample Number Si02 A12°3 FeO MgO CaO Na20 K2° Ti02 San Bernardino block— -Continued B145 72.12 14.98 1.41 0.50 2.36 3.69 4.45 0.50 B146 73.32 14.66 1.22 0.33 2.07 3.71 4.27 0.42 B147 71.64 15.20 1.46 0.50 2.49 3.77 4.43 0.50 B148 71.87 15.27 1.40 0.50 2.50 3.81 4.16 0.49 B149 72.84 14.88 1.07 0.35 2.22 3.78 4.42 0.44 B150 72.08 15.18 1.36 0.42 2.37 3.78 4.34 0.47 B151 70.67 15.56 1.84 0.62 2.78 3.84 4.17 0.54 B152 65.95 16.37 3.90 1.67 4.31 3.79 3.33 0.70 B153 67.49 16.54 3.16 0.98 3.30 4.09 3.91 0.53 B155 61.93 16.89 5.10 2.42 5.18 3.74 3.90 0.84 B156 63.74 16.88 4.36 1.59 4.63 3.96 4.53 0.31 B157 71.88 15.31 1.33 0.33 2.15 3.98 4.61 0.42 B158 60.80 17.84 5.88 2.40 5.28 4.04 2.82 0.93 B159 64.43 16.55 4.58 2.25 4.76 3.54 3.16 0.74 B160 65.27 16.48 4.22 1.96 4.61 3.79 2.97 0.70 B161 68.35 16.39 3.86 1.66 1.84 3.74 3.48 0.68 B162 69.21 15.78 3.03 0.81 2.96 3.89 3.79 0.52 B163 74.59 15.17 0.86 0.22 1.83 3.81 2.79 0.73 B164 59.57 17.93 6.45 3.15 5.19 3.95 2.82 0.94 B165 74.49 14.43 0.86 0.20 2.18 3.66 3.86 0.32 B166 59.03 17.69 6.37 3.34 6.37 3.95 2.26 1.00 B168 59.70 17.36 6.36 3.14 6.11 3.76 2.60 0.97 B169 59.12 17.29 6.65 3.26 6.46 3.80 2.38 1.03 B170 70.61 16.45 1.53 0.49 3.19 4.33 2.94 0.46 B171 69.21 16.26 2.41 0.95 3.29 4.02 3.30 0.55 B172 61.21 16.90 6.02 2.80 5.72 4.12 2.30 0.94 B173 73.61 14.79 0.44 0.20 1.90 3.57 5.14 0.35 B174 72.10 14.97 1.56 0.57 2.51 3.44 4.39 0.45 B175 71.54 15.18 1.60 0.62 2.64 3.58 4.40 0.45 B176 71.18 16.15 1.40 0.53 3.17 4.03 3.09 0.45 B177 72.87 14.97 0.79 0.22 1.95 3.83 5.02 0.35 B178 74.89 14.44 0.21 0.08 1.59 3.42 5.08 0.29 B179 63.87 17.64 4.77 1.78 4.74 4.26 2.17 0.76 B180 72.32 15.27 0.92 0.52 2.33 4.11 4.11 0.42 B181 74.03 14.86 0.62 0.22 2.00 3.73 4.24 0.32 B182 72.67 14.49 1.45 0.53 2.37 3.31 4.73 0.45 B183 71.52 14.95 2.27 0.93 2.47 3.35 3.88 0.62 B184 75.38 13.80 0.91 0.18 1.06 2.54 5.73 0.40 B185 68.99 15.80 2.46 0.99 3.10 3.81 4.20 0.65 B186 71.38 15.73 1.73 0.85 2.22 3.84 3.73 0.50 B187 71.46 15.60 1.31 0.56 2.59 3.78 4.25 0.43 B188 71.10 15.23 1.93 0.72 2.57 3.52 4.45 0.49 B189 74.77 14.73 0.21 0.07 1.58 3.74 4.63 0.28 B190 73.35 15.24 0.49 0.20 2.26 3.81 4.26 0.39 B191 73.04 15.18 0.59 0.12 2.19 3.69 4.87 0.32 B192 72.48 15.53 1.01 0.41 2.28 3.76 4.18 0.36 B193 73.60 15.16 0.64 0.20 1.96 3.83 4.26 0.35 B194 69.38 16.08 2.29 0.95 3.08 3.99 3.70 0.54 B195 74.59 14.61 0.26 0.15 1.98 3.64 4.46 0.32 B196 68.12 16.35 2.63 1.07 3.23 3.96 4.03 0.61 B197 73.91 14.92 0.41 0.21 2.04 3.55 4.62 0.33 Little San Bernardino block B198 72.14 15.80 1.13 0.50 1.83 3.77 4.49 0.34 B199 72.49 15.06 1.79 0.45 1.80 3.80 4.37 0.24 B200 69.47 16.50 1.76 0.50 2.46 4.82 4.30 0.19 B201 70.43 15.90 2.34 0.76 2.33 3.88 3.94 0.41 B202 72.69 16.29 1.23 0.42 1.57 3.35 4.27 0.18 B203 72.40 16.64 1.17 0.29 1.28 3.75 4.33 0.15 B204 73.58 14.47 1.49 0.55 1.87 3.69 4.13 0.22 B205 71.21 14.97 2.15 0.69 2.02 3.33 5.19 0.43 B206 74.73 13.64 1.59 0.22 0.84 3.52 5.24 0.23 B207 73.43 15.04 1.25 0.43 1.58 3.91 4.17 0.20 Oxide Sample Number Si02 A12°3 FeO MgO CaO Na20 K2° tio2 B208 73.92 13.84 2.02 0.48 1.21 4.06 4.11 0.35 B209 68.91 14.62 2.80 3.02 3.16 3.47 3.65 0.38 B210 66.99 16.49 3.71 1.49 3.70 3.89 3.23 0.50 B211 71.33 16.04 1.96 0.61 2.51 4.15 3.09 0.31 B212 70.36 16.07 2.33 0.78 2.93 4.14 3.02 0.37 B213 62.87 17.35 5.70 2.25 4.84 3.62 2.59 0.77 B214 61.34 16.75 5.98 3.10 5.84 3.90 2.19 0.89 B215 75.13 14.17 0.99 0.25 1.10 3.82 4.46 0.08 B216 67.02 16.90 1.47 2.33 4.67 3.88 3.09 0.64 B217 72.26 14.74 2.19 0.30 1.58 3.76 4.89 0.27 B218 65.78 16.52 3.61 2.16 2.94 4.89 3.50 0.61 B219 68.12 16.09 3.19 1.81 2.48 4.01 3.80 0.51 B220 71.61 15.45 1.90 0.64 1.96 3.88 4.32 0.25 B221 64.72 16.34 4.80 2.67 3.41 3.93 3.48 0.64 B222 64.45 15.64 5.40 2.20 4.75 3.61 3.15 0.80 San Jacinto block B223 66.88 17.22 3.39 1.29 4.59 4.20 1.72 0.71 B224 71.27 15.72 2.14 0.72 2.73 3.39 3.50 0.53 B225 67.79 17.35 2.89 1.05 4.17 4.13 1.98 0.64 B226 66.49 16.78 4.21 1.64 4.40 3.22 2.28 0.99 B227 64.61 16.49 4.14 2.32 4.97 3.36 3.19 0.92 B228 68.69 16.05 3.20 1.12 3.78 3.68 2.79 0.69 B229 63.94 18.34 3.80 1.56 5.42 4.11 1.82 1.00 B230 67.61 16.21 3.67 1.35 4.01 3.62 2.77 0.76 B231 69.36 16.47 2.79 1.05 3.28 3.79 2.66 0.59 B232 64.29 17.64 4.36 1.75 4.96 4.06 1.93 1.01 B233 64.86 15.92 5.14 2.43 4.74 3.40 2.70 0.81 B234 67.55 16.88 3.46 1.32 4.11 3.91 2.01 0.76 B235 68.60 16.90 3.00 1.10 3.90 4.01 1.87 0.62 B236 68.00 15.78 3.68 1.54 3.82 3.37 3.19 0.64 B237 65.74 17.30 3.84 1.36 4.76 4.03 2.12 0.84 B238 68.19 16.24 3.21 1.14 4.00 3.84 2.66 0.72 B239 66.00 17.11 3.86 1.39 4.65 3.98 2.17 0.84 B240 67.46 17.22 3.09 1.13 4.09 4.02 2.31 0.68 B241 67.83 17.04 3.17 1.18 3.74 4.03 2.34 0.66 B242 67.17 16.90 3.65 1.36 3.97 3.89 2.28 0.78 B243 66.09 17.31 3.69 1.44 4.68 3.94 2.05 0.81 B244 64.96 16.50 4.36 2.52 5.14 3.59 2.12 0.81 B245 66.34 16.29 4.64 1.29 4.09 3.81 2.74 0.80 B246 69.67 16.33 2.73 0.93 3.07 3.87 2.87 0.53 B247 67.70 16.92 3.37 1.24 3.83 4.11 2.16 0.68 B248 70.43 15.89 2.78 0.92 3.04 3.82 2.59 0.54 B249 68.30 16.44 2.94 1.16 3.94 3.73 2.80 0.69 B250 69.17 16.34 2.64 1.00 3.62 3.54 3.06 0.62 B251 63.83 16.86 5.10 2.10 5.40 3.74 2.00 0.96 B252 67.92 17.00 3.18 1.17 3.63 4.06 2.36 0.68 B253 66.45 16.62 4.03 1.59 4.30 3.82 2.30 0.89 B254 67.45 17.03 3.45 1.33 3.87 3.97 2.21 0.71 B255 75.57 13.68 1.13 0.31 1.65 2.88 4.51 0.27 B256 66.56 16.45 4.04 1.51 4.31 3.60 2.68 0.85 B257 65.81 17.22 3.93 1.43 4.62 3.87 2.25 0.86 B258 66.73 16.72 3.98 1.45 4.10 3.91 2.27 0.84 B259 73.00 14.61 2.26 0.45 1.90 3.38 4.09 0.32 B260 65.86 16.74 4.58 1.50 3.89 4.23 2.42 0.78 B261 67.01 16.94 3.70 1.34 4.01 3.83 2.39 0.78 B262 71.26 15.94 2.07 0.99 2.78 3.46 3.02 0.49 B263 70.58 15.10 3.53 0.78 2.59 3.73 3.13 0.57 B264 72.61 14.98 2.27 0.55 2.21 3.45 3.59 0.34 B265 64.74 16.80 4.85 2.07 4.48 3.83 2.35 0.90 B266 75.41 14.29 0.46 0.15 1.18 2.75 5.62 0.15 B267 75.19 13.61 1.65 0.18 1.12 2.62 5.48 0.16 B268 69.90 15.43 3.58 1.09 3.00 3.52 2.98 0.50 B269 63.96 17.57 4.74 1.94 5.03 4.06 1.69 1.01 B270 75.26 13.73 1.64 0.32 1.31 3.84 3.70 0.22 B271 75.13 14.17 1.36 0.27 1.24 3.77 3.87 0.20 B272 72.69 14.56 3.30 0.58 2.20 3.95 2.36 0.3738 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA APPENDIX 1.—Continued Oxide Sample Number Si°2 A12°3 FeO MgO CaO Na20 k2° Ii02 San Jacinto block—Continued B273 63.62 16.97 4.63 1.79 4.51 3.98 1.57 0.94 B274 66.32 16.88 3.67 1.82 4.24 3.61 2.59 0.87 B275 70.18 16.01 2.20 0.98 3.16 3.44 3.39 0.64 B276 66.77 16.46 3.97 1.79 3.83 3.54 2.85 0.79 B277 64.05 16.69 5.01 2.70 4.29 3.46 2.87 0.92 B278 65.89 17.12 4.29 1.81 3.95 3.60 2.54 0.80 B279 70.28 15.77 3.24 0.65 2.45 3.80 3.36 0.44 B280 68.94 16.25 3.76 0.68 2.60 4.08 3.22 0.47 B281 70.36 15.70 2.94 0.89 3.16 3.43 2.90 0.63 B282 73.79 14.41 1.75 0.47 1.75 2.86 4.55 0.41 B283 66.23 16.08 4.53 2.10 3.25 2.99 3.96 0.86 B284 68.84 16.20 2.76 1.13 3.25 3.35 3.77 0.71 B285 66.83 16.39 3.41 1.72 4.25 3.39 3.21 0.80 B286 63.47 17.29 4.54 2.55 5.24 3.50 2.43 0.98 B287 69.17 15.63 2.96 1.22 3.71 3.36 3.22 0.73 B288 65.22 16.62 4.29 2.24 4.99 3.46 2.21 0.98 B289 68.85 15.99 2.65 1.34 3.59 3.67 3.24 0.67 B290 64.38 16.87 4.53 1.97 5.05 3.79 2.43 0.97 B291 69.21 15.43 3.00 1.14 3.54 3.35 3.67 0.67 B292 69.67 15.22 3.09 1.23 3.72 3.37 2.98 0.71 B293 68.89 15.53 3.22 1.37 3.52 3.49 3.24 0.74 B294 69.56 15.61 2.98 1.23 3.42 3.49 3.01 0.71 B295 62.78 17.27 4.90 2.65 5.47 3.55 2.34 1.03 B296 68.72 16.09 3.04 1.40 3.41 3.32 3.23 0.78 B297 65.50 16.85 3.86 2.16 4.40 3.39 2.97 0.87 B298 64.97 16.71 4.19 2.29 4.74 3.35 2.81 0.93 B299 65.85 17.13 3.49 1.69 4.46 3.68 2.78 0.92 B300 61.79 17.89 4.89 2.76 5.71 3.74 2.25 0.98 B301 65.68 16.69 3.90 2.11 4.56 3.29 2.88 0.89 B302 64.92 17.00 4.22 2.11 4.43 3.32 3.12 0.87 B303 67.89 16.87 2.87 1.28 3.93 3.29 3.14 0.73 B304 62.41 18.06 4.68 2.52 5.52 3.63 2.23 0.96 B305 62.25 17.73 4.97 2.56 5.22 4.03 2.22 1.02 B306 65.08 17.24 3.91 2.08 4.50 3.23 3.08 0.88 B307 65.56 16.66 3.82 2.16 4.64 3.30 3.02 0.86 B308 63.18 17.76 4.53 2.44 5.14 3.60 2.35 1.00 B309 68.70 16.08 3.04 1.40 3.42 3.32 3.23 0.80 B310 63.96 16.46 5.33 2.51 4.38 3.35 3.07 0.93 B311 64.18 17.31 4.83 2.11 4.60 3.64 2.47 0.84 B312 77.23 12.84 0.90 0.15 0.67 3.50 4.50 0.21 B313 75.87 13.66 1.29 0.19 1.34 3.39 3.97 0.29 B314 73.09 14.63 2.51 0.39 1.72 3.87 3.43 0.36 B315 64.60 17.43 4.80 1.86 4.77 3.78 1.92 0.84 B317 65.13 17.56 4.25 1.52 5.02 3.97 1.74 0.82 B318 63.96 17.52 4.57 1.80 5.34 3.85 1.97 1.00 B319 65.26 17.46 4.16 1.52 4.97 3.96 1.80 0.86 B320 63.77 17.74 4.58 1.83 5.43 3.78 1.83 1.03 B321 63.53 18.31 4.59 1.69 5.28 3.77 1.86 0.98 B322 65.58 16.81 4.33 1.76 4.69 3.90 1.96 0.98 B323 63.62 18.08 4.56 1.93 5.07 4.12 1.65 0.98 B324 66.09 17.16 4.53 1.50 4.41 4.08 1.46 0.77 Perris block B325 66.27 16.03 4.57 1.94 4.48 3.63 2.29 0.80 B326 64.42 15.81 5.16 2.88 5.14 3.48 2.38 0.73 B327 66.35 15.55 4.83 1.85 4.62 3.90 2.24 0.66 B328 68.65 15.28 4.06 1.35 3.84 3.79 2.48 0.55 B330 72.16 12.94 1.75 3.50 1.45 3.66 4.30 0.23 B331 72.79 14.29 2.51 0.75 2.21 3.91 3.24 0.30 B332 74.07 14.00 1.92 0.60 1.85 3.75 3.56 0.27 B333 73.74 14.19 1.97 0.61 2.17 3.86 3.18 0.27 B334 75.29 13.58 1.60 0.45 1.41 3.72 3.73 0.22 B336 60.86 16.99 6.44 3.38 6.71 3.42 1.42 0.78 Oxide Sample Number S102 A12°3 FeO MgO CaO Na20 K20 1i02 B337 69.53 14.58 3.66 1.64 3.84 3.35 2.87 0.53 B338 63.95 19.74 3.97 0.89 4.35 5.36 1.38 0.37 B339 63.45 16.86 5.04 2.65 6.29 3.81 1.20 0.70 B340 63.34 16.31 5.56 2.81 5.92 3.81 1.54 0.71 B341 54.16 19.25 7.65 4.80 9.01 3.06 1.08 0.98 B342 60.27 16.86 6.45 3.51 7.04 3.63 1.44 0.80 B343 64.76 15.94 5.18 2.60 5.49 3.54 1.81 0.67 B344 63.46 16.10 5.73 2.94 5.57 3.64 1.81 0.76 B345 65.94 15.13 5.35 2.48 5.31 3.44 1.70 0.64 B346 74.21 13.37 2.41 0.50 2.01 4.21 3.03 0.25 B347 62.31 16.75 7.11 1.31 6.90 3.54 1.25 0.82 B348 74.94 13.45 2.03 0.42 1.68 4.28 2.97 0.24 B350 68.91 15.74 3.99 1.24 3.45 3.92 2.27 0.49 B351 68.83 15.77 3.82 1.37 3.63 4.11 2.02 0.45 B352 63.36 16.31 5.66 2.74 6.04 3.34 1.89 0.67 B353 63.03 17.21 5.36 2.63 6.45 3.49 1.20 0.64 B354 67.46 15.64 4.45 1.76 4.31 3.50 2.27 0.61 B355 68.21 15.09 4.01 1.67 3.95 3.63 2.88 0.56 B356 68.71 15.91 4.03 1.14 3.31 4.06 2.35 0.49 B357 61.38 16.34 6.38 3.58 6.89 3.28 1.42 0.74 B358 64.19 16.06 5.44 2.74 5.72 3.65 1.52 0.67 B359 61.53 16.61 6.28 3.30 6.72 3.15 1.67 0.74 B360 68.74 15.54 4.08 1.29 3.72 4.39 1.76 0.48 B361 68.55 16.29 3.65 1.38 3.49 4.05 2.15 0.44 B363 64.47 14.07 6.32 3.11 5.43 3.79 2.09 0.72 B364 74.69 13.51 2.14 0.61 2.16 4.11 2.51 0.27 B366 69.53 15.68 3.63 1.20 3.51 4.33 1.69 0.42 B367 75.02 12.81 1.06 2.49 0.91 3.85 3.75 0.10 B368 74.22 14.10 2.06 0.56 2.14 3.92 2.72 0.27 B369 75.92 12.82 1.86 0.45 1*39 3.75 3.58 0.22 B370 65.54 15.12 5.12 2.59 4.85 3.45 2.51 0.82 B371 65.44 17.12 4.11 1.81 5.43 4.03 1.29 0.77 B372 65.43 16.48 4.34 1.80 5.07 3.82 2.17 0.90 B373 64.18 18.02 4.03 1.90 5.35 4.26 1.49 0.78 B374 64.46 16.73 5.04 1.88 5.71 3.34 2.15 0.69 B375 62.31 18.36 4.26 3.14 5.58 4.20 1.39 0.77 B376 62.99 17.20 5.60 2.63 5.43 3.33 2.06 0.77 B377 65.19 17.19 4.54 1.95 5.75 3.97 0.81 0.60 B378 67.17 16.40 4.20 1.60 4.97 3.89 1.25 0.52 B380 65.15 17.77 4.20 1.86 5.44 3.80 1.19 0.59 B381 62.76 18.45 4.41 2.14 6.08 4.20 1.12 0.84 B382 64.47 17.64 4.25 1.82 5.51 4.13 1.38 0.81 B383 64.38 17.84 4.00 1.90 5.85 4.01 1.20 0.82 B384 65.79 17.82 3.64 1.43 4.91 4.19 1.58 0.64 B385 64.08 17.32 4.54 2.15 5.69 4.00 1.36 0.87 B386 64.13 15.30 4.44 4.08 6.51 3.27 1.77 0.50 B387 73.77 14.91 1.79 0.45 1.48 3.64 3.72 0.24 B388 75.54 13.20 1.59 0.51 1.30 4.19 3.47 0.19 B389 74.26 13.97 1.92 0.49 1.40 4.36 3.39 0.22 B390 69.09 16.81 2.61 0.94 4.12 4.51 1.50 0.42 B391 70.98 16.02 2.09 0.81 3.53 4.33 1.87 0.36 B392 61.90 19.52 3.93 1.69 5.09 4.63 2.48 0.75 B394 66.46 16.65 3.94 1.72 4.71 3.91 1.89 0.72 B395 65.11 16.82 4.02 2.03 5.45 3.95 1.75 0.87 B396 72.41 15.00 1.87 0.71 2.84 3.57 3.28 0.32 B397 67.15 17.17 3.25 1.30 4.06 4.08 2.40 0.58 B398 64.77 17.41 4.19 1.99 5.06 3.91 1.84 0.83 B399 67.67 16.58 3.45 1.40 4.22 3.85 2.19 0.63 B400 68.40 16.57 3.01 1.17 4.25 4.65 1.47 0.49 B401 61.08 16.88 6.39 3.19 6.55 3.50 1.62 0.79 B402 69.23 14.42 4.16 1.98 2.84 2.86 3.58 0.93 B403 61.11 16.54 6.93 2.86 6.34 3.78 1.55 0.90 B404 66.67 16.13 5.20 0.38 5.08 3.61 2.23 0.69 B405 63.25 17.39 4.94 2.75 5.41 3.35 2.14 0.77 B406 71.33 13.91 3.20 2.05 2.66 2.46 3.76 0.63APPENDIX 1 39 APPENDIX 1.—Continued Oxide Sample Number Si02 A12°3 FeO MgO CaO Na20 k2° tio2 Perris block—Continued B407 71.97 15.41 2.02 0.62 2.61 4.09 2.93 0.34 B408 64.37 16.61 4.85 2.44 5.60 3.68 1.69 0.75 B409 66.48 16.31 3.92 1.81 5.04 3.87 1.92 0.65 B410 66.65 16.19 3.84 2.13 4.67 3.73 2.25 0.54 B411 65.14 16.76 4.93 0.20 9.30 2.59 0.61 0.47 B412 57.13 20.18 6.42 3.58 7.17 3.66 1.06 0.80 B413 67.00 15.46 4.51 2.15 4.40 3.48 2.36 0.62 B415 67.93 15.11 4.08 2.05 3.85 3.29 3.22 0.49 B416 74.79 14.39 1.22 0.37 1.75 3.62 3.73 0.13 B417 62.39 15.83 5.85 3.78 6.13 3.11 2.09 0.81 B418 74.14 13.87 1.64 0.41 1.71 3.52 4.45 0.26 B420 61.13 17.45 6.66 2.73 6.29 3.06 1.73 0.95 B421 76.29 12.85 1.64 0.85 1.15 3.01 3.94 0.27 B422 72.74 14.53 2.29 0.64 2.19 3.42 3.90 0.29 B423 63.51 15.79 5.91 3.06 5.44 3.08 2.47 0.73 B424 74.15 15.09 1.17 0.42 2.04 4.21 2.73 0.17 B425 72.13 14.60 2.12 0.86 2.62 3.59 3.79 0.30 B426 68.22 15.38 4.27 1.58 4.19 4.00 1.79 0.57 B427 62.88 16.01 6.01 3.13 6.51 3.47 1.13 0.85 B428 77.04 12.35 0.99 0.20 0.89 3.47 4.95 0.10 B429 59.97 18.83 5.67 1.99 5.70 7.18 0.10 0.57 B430 65.65 16.57 4.38 2.10 4.79 3.11 2.82 0.59 B431 67.18 15.35 4.73 1.99 4.16 3.50 2.47 0.63 B432 64.33 17.09 4.81 2.35 5.40 3.30 1.97 0.75 B433 65.94 16.24 4.31 2.27 4.93 3.63 2.06 0.62 B434 70.65 15.59 2.64 1.08 3.46 4.04 2.19 0.34 B435 64.42 17.04 4.52 2.18 5.58 3.97 1.51 0.77 B436 62.94 17.50 3.05 5.69 4.18 4.19 1.86 0.57 B437 65.61 17.31 3.76 1.78 4.97 4.19 1.72 0.66 B438 64.17 17.32 4.50 2.10 5.25 3.79 2.09 0.78 B439 65.64 16.80 3.85 1.91 5.10 3.98 2.04 0.67 B440 69.98 15.40 3.59 1.00 3.58 4.19 1.82 0.44 B441 66.77 16.47 4.01 1.78 4.61 3.71 1.90 0.74 B442 66.11 15.96 4.54 1.98 4.85 3.77 1.98 0.81 B447 70.07 16.05 2.69 0.90 3.38 3.95 2.50 0.47 B448 62.42 17.82 4.49 2.21 6.18 4.90 1.09 0.89 B449 66.36 16.12 4.55 2.15 4.54 3.64 2.00 0.63 B450 74.09 14.15 1.48 0.47 1.59 3.19 4.81 0.22 B451 69.88 15.54 3.37 1.29 3.77 3.79 1.92 0.44 B452 67.00 15.78 3.96 1.82 4.73 3.48 2.45 0.79 B453 66.51 16.59 4.09 1.76 4.39 3.64 2.29 0.75 B454 66.04 16.62 3.88 1.87 4.57 3.63 2.63 0.75 B455 67.83 15.76 4.00 1.80 3.91 3.50 2.50 0.69 B456 66.25 15.98 4.49 1.93 4.90 3.77 1.88 0.81 Santa Ana block B458 74.36 13.84 2.27 0.35 1.78 4.25 2.88 0.27 B459 76.20 13.31 1.46 0.17 1.11 4.02 3.57 0.17 B460 74.83 13.70 1.97 0.40 1.97 4.27 2.61 0.25 B461 75.88 12.09 2.54 0.38 1.48 3.91 3.50 0.22 B462 70.01 14.12 4.19 1.36 3.27 3.49 2.97 0.59 B464 72.75 13.77 3.02 0.89 2.50 3.58 3.09 0.41 B465 74.85 13.50 2.12 0.28 1.41 3.96 3.66 0.22 B466 73.21 13.99 3.02 0.31 1.79 4.42 2.97 0.29 B469 74.91 13.46 2.06 0.48 2.03 3.88 2.92 0.25 B470 75.04 13.41 2.33 0.35 1.76 3.85 2.97 0.28 B471 79.10 10.92 1.30 0.05 0.73 3.52 4.23 0.15 B472 72.51 14.55 2.71 0.54 2.35 4.29 2.70 0.34 B473 67.36 15.86 4.86 1.44 4.07 4.09 1.68 0.65 B474 65.86 15.06 5.09 2.74 5.17 3.30 2.13 0.65 B475 75.82 12.99 1.98 0.38 1.69 3.56 3.35 0.23 Oxide Sample Number Si02 A12°3 FeO MgO CaO Na20 K2° tio2 B476 70.82 13.92 3.71 1.45 3.33 3.66 2.67 0.44 B478 69.63 14.86 3.76 1.64 3.72 3.68 2.26 0.47 B479 74.84 13.70 2.37 0.35 1.81 3.54 3.14 0.25 B480 75.24 13.34 2.04 0.45 1.93 3.79 2.97 0.25 B482 77.21 12.75 1.32 0.05 0.77 4.12 3.68 0.10 B484 59.55 18.10 6.58 3.25 6.76 3.80 1.12 0.85 B485 73.26 14.53 2.26 0.55 2.17 3.70 3.24 0.29 B486 65.71 15.93 4.70 2.43 5.19 3.89 1.53 0.62 B487 58.14 18.37 6.79 3.65 7.72 3.62 0.86 0.85 B489 74.18 14.45 1.54 0.13 1.18 4.64 3.73 0.13 B490 75.99 13.35 1.49 0.23 1.26 3.87 3.64 0.18 B491 75.16 13.65 2.01 0.22 1.62 4.03 3.09 0.22 B492 67.26 15.67 4.42 1.88 4.42 3.47 2.27 0.61 B493 71.19 15.44 2.47 0.39 2.48 5.51 2.24 0.29 B494 64.42 16.05 5.21 2.75 5.39 3.59 1.92 0.67 B495 64.55 15.80 5.10 2.95 5.52 3.67 1.73 0.68 B496 72.59 14.03 3.62 0.74 2.75 4.22 1.57 0.48 B497 67.14 14.72 4.88 2.23 4.70 3.78 1.79 0.76 B498 61.14 16.05 6.64 3.69 6.50 3.49 1.59 0.90 B499 66.62 15.12 6.10 1.74 4.39 3.79 1.50 0.75 B500 61.79 16.72 6.06 3.22 6.13 3.55 1.70 0.83 B502 65.93 16.87 3.35 1.67 6.34 3.70 1.82 0.34 B503 60.08 17.45 6.19 3.43 7.25 3.47 1.33 0.79 B504 74.79 13.91 1.79 0.37 1.91 3.77 3.23 0.23 B506 71.54 14.66 3.21 0.74 2.73 4.46 2.28 0.39 B507 66.05 15.78 4.78 2.00 4.96 4.39 1.44 0.59 B508 72.63 14.55 2.70 0.54 2.45 4.47 2.33 0.32 B510 74.43 13.80 2.14 0.44 2.17 4.15 2.60 0.27 B513 53.86 18.93 8.14 4.64 9.16 3.61 0.71 0.95 B514 56.68 18.26 7.03 4.29 8.27 3.48 1.13 0.86 B515 77.17 12.41 1.74 0.02 0.93 4.19 3.41 0.13 B516 76.96 12.82 1.79 0.15 1.71 4.35 2.04 0.18 B517 69.05 14.43 4.54 1.68 4.11 3.67 1.96 0.55 B518 60.46 17.06 6.14 3.82 6.88 3.57 1.29 0.78 B519 63.48 15.92 5.67 3.12 5.59 3.78 1.71 0.74 B520 73.16 13.92 3.08 0.40 2.07 4.69 2.38 0.30 B521 73.95 14.32 2.22 0.37 2.00 4.40 2.49 0.23 B522 60.40 17.46 6.20 3.24 6.65 3.84 1.42 0.81 B523 58.70 17.73 6.83 3.71 7.37 3.62 1.19 0.86 B524 64.04 15.29 6.01 2.71 5.65 4.05 1.43 0.81 B525 65.91 14.92 5.22 2.32 5.17 3.96 1.75 0.76 B528 73.17 14.08 2.97 0.37 2.25 4.78 2.08 0.30 B529 61.20 17.19 6.03 3.14 6.20 3.57 1.93 0.74 B530 59.81 17.49 6.14 3.82 7.12 3.42 1.47 0.73 B531 61.53 16.46 6.36 3.24 6.10 3.34 2.11 0.86 B533 54.76 18.75 7.82 4.96 8.48 3.40 0.85 0.96 B536 55.28 17.66 8.06 4.93 8.75 3.39 1.00 0.93 B537 59.88 16.86 6.88 3.67 6.79 3.43 1.67 0.83 B538 73.04 14.10 2.67 0.62 2.33 4.15 2.75 0.34 B539 72.52 14.36 3.06 0.65 2.75 4.52 1.75 0.39 B540 72.49 14.31 2.71 0.60 2.34 4.20 3.00 0.34 B541 72.78 13.91 3.07 0.65 2.35 4.11 2.77 0.37 B542 70.34 14.97 3.96 0.82 3.24 4.91 1.23 0.52 B543 56.32 17.46 7.56 4.94 8.43 3.40 1.07 0.81 B544 64.55 16.75 5.30 2.33 5.63 3.57 1.33 0.54 B546 65.75 15.90 5.04 1.81 4.56 3.66 2.60 0.67 B547 59.48 17.81 6.10 3.54 6.55 3.94 1.60 0.98 B548 64.05 16.90 4.74 2.43 5.67 3.91 1.51 0.79 B549 70.45 14.68 3.91 0.84 2.87 4.23 2.50 0.52 B550 76.45 12.97 1.35 0.02 0.51 3.89 4.72 0.10 B551 72.03 15.33 3.03 0.60 2.44 4.25 1.98 0.34 B553 77.36 12.96 0.45 0.02 0.70 3.69 4.74 0.08 B555 63.72 16.68 5.19 2.51 5.80 3.77 1.48 0.85 B556 65.54 15.91 4.68 2.34 4.80 3.83 2.14 0.76 B557 59.80 18.24 5.95 3.15 6.48 3.85 1.57 0.9640 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA APPENDIX 2 Mixing proportions End member Sample Number M-1 M-2 D-1 D-2 San Gabriel block B 1 1.2174 0.9289 -0.6709 -0.4755 B 2 0.9619 0.8786 -0.5231 -0.3174 B 3 0.9302 0.6186 -0.3118 -0.2369 B 4 1.2767 0.7197 -0.5559 -0.4404 B 5 0.9208 0.9135 -0.3952 -0.4391 B 6 1.2507 1.2092 -0.7926 -0.6673 B 7 1.1276 0.9123 -0.5895 -0.4504 B 8 1.2485 0.7044 -0.6100 -0.3429 B 9 1.1917 1.0451 -0.7145 -0.5224 B 10 0.7177 1.2399 -0.4496 -0.5080 B 11 0.6233 1.1228 -0.2827 -0.4634 B 12 0.8445 0.8223 -0.3519 -0.3150 B 13 0.7706 0.9073 -0.3155 -0.3624 B 14 0.9792 0.4384 -0.2641 -0.1535 B 15 1.1089 0.5788 -0.3939 -0.2937 B 16 0.9389 0.5381 -0.2886 -0.1884 B 17 1.0423 0.8977 -0.4700 -0.4700 B 18 1.1551 1.1495 -0.6630 -0.6416 B 19 0.8019 1.4285 -0.5291 -0.7014 B 20 0.5837 1.5388 -0.3978 -0.7248 B 21 0.9391 0.7171 -0.3907 -0.2654 B 22 0.7831 0.4762 -0.1414 -0.1178 B 23 0.8008 0.6209 -0.2773 -0.1444 B 24 0.9681 0.5208 -0.2828 -0.2061 B 25 1.1512 1.2196 -0.7038 -0.6670 B 26 0.7134 1.2828 -0.4335 -0.5626 B 27 0.7229 0.6464 -0.2956 -0.0737 B 28 0.3545 1.1692 -0.1861 -0.3376 B 29 0.6434 0.9085 -0.2393 -0.3126 B 30 1.0392 0.7829 -0.4777 -0.3444 B 31 0.5875 1.1692 -0.3648 -0.3919 B 32 0.9966 0.7151 -0.4309 -0.2808 B 33 0.6869 0.5134 -0.1251 -0.0752 B 34 0.8412 0.9545 -0.3684 -0.4272 B 35 1.0494 1.1927 -0.6240 -0.6180 B 36 0.8666 0.9352 -0.3847 -0.4171 B 37 0.7273 0.7084 -0.2085 -0.2272 B 38 0.7242 0.4060 -0.0590 -0.0712 B 39 0.9959 0.3676 -0.2722 -0.0913 B 40 1.1020 0.6300 -0.4273 -0.3047 B 41 0.5382 0.3488 0.0236 0.0894 B 42 0.9570 0.5350 -0.3101 -0.1818 B 43 0.7424 0.8197 -0.2780 -0.2841 B 44 0.3696 0.8393 -0.0817 -0.1272 B 46 0.7325 1.1015 -0.4289 -0.4051 B 47 0.6038 1.1752 -0.3727 -0.4064 B 48 1.0573 0.9368 -0.5685 -0.4256 San Bernardino block B 54 0.7080 1.2247 -0.4546 -0.4781 B 55 0.7045 1.0839 -0.3346 -0.4538 B 56 0.5472 0.7952 -0.1855 -0.1568 B 57 0.4363 0.7923 -0.0867 -0.1418 B 58 0.5181 0.8350 -0.1629 -0.1903 B 59 0.5160 0.8225 -0.1667 -0.1718 B 60 0.5270 1.0037 -0.2191 -0.3116 B 61 0.7139 0.6223 -0.1693 -0.1669 B 62 0.4926 0.7891 -0.1410 -0.1407 B 63 0.5426 1.1837 -0.2808 -0.4455 End member End member Sample Number M-1 M-2 D-l D-2 Sample Number M-1 M-2 D-l D-2 B 64 0.6369 1.3884 -0.4281 -0.5972 B146 0.7554 1.2815 -0.4623 -0.5746 B 65 0.8435 1.2957 -0.5139 -0.6253 B147 0.5304 1.2744 -0.3194 -0.4854 B 66 0.4837 1.0171 -0.1670 -0.3337 B148 0.6137 1.2324 -0.3650 -0.4811 B 67 0.3876 0.9667 -0.1231 -0.2312 B149 0.6469 1.2973 -0.4073 -0.5369 B 68 0.5702 1.3892 -0.3572 -0.6022 B150 0.5771 1.2738 -0.3536 -0.4973 B 69 0.7879 1.0210 -0.4022 -0.4067 B151 0.5503 1.1946 -0.3100 -0.4349 B 70 0.5139 0.8303 -0.1225 -0.2218 B152 0.6898 0.8747 -0.2527 -0.3118 B 71 0.7682 0.9463 -0.3541 -0.3604 B153 0.6108 1.0084 -0.3028 -0.3163 B 72 0.5272 1.4569 -0.3500 -0.6341 B155 0.4314 0.8466 -0.0503 -0.2276 B 74 0.8694 0.6447 -0.2744 -0.2397 B156 0.6582 0.9458 -0.2783 -0.3258 B 75 0.8799 1.2560 -0.5439 -0.5920 B157 0.5937 1.2770 -0.3885 -0.4822 B 78 0.5754 1.1626 -0.3299 -0.4082 B158 0.5699 0.6217 -0.0971 -0.0945 B 79 0.9923 0.9954 -0.5698 -0.4178 B159 0.6619 0.8163 -0.1758 -0.3024 B 80 0.7478 0.8041 -0.2722 -0.2797 B160 0.8222 0.7752 -0.2971 -0.3003 B 81 0.8413 1.0596 -0.4558 -0.4451 B161 0.6458 0.9970 -0.2746 -0.3683 B 85 0.4104 1.1881 -0.1866 -0.4119 B162 0.7669 1.0411 -0.3915 -0.4165 B 86 1.0616 1.4356 -0.6989 -0.7983 B163 0.7325 1.1465 -0.4256 -0.4535 B 94 0.2961 1.3932 -0.2011 -0.4882 B164 0.6338 0.5702 -0.0955 -0.1084 B 95 0.3964 1.4304 -0.2761 -0.5507 B165 0.9658 1.2500 -0.5862 -0.6295 B 96 0.9281 1.2204 -0.5649 -0.5835 B166 0.8182 0.4391 -0.1585 -0.0987 B 97 0.5656 0.7990 -0.1500 -0.2146 B168 0.7335 0.5355 -0.1201 -0.1489 B 98 0.6223 1.0403 -0.3450 -0.3177 B169 0.8207 0.4652 -0.1485 -0.1373 B 99 0.1145 1.1810 0.0293 -0.3249 B170 0.8835 0.9454 -0.5081 -0.3209 B100 0.6653 1.2826 -0.4638 -0.4840 B171 0.7430 0.9793 -0.3765 -0.3458 B101 0.6535 1.0780 -0.3283 -0.4032 B172 1.0757 0.4596 -0.3519 -0.1834 B102 0.5175 1.0506 -0.2423 -0.3257 B173 0.3673 1.4962 -0.2835 -0.5801 B103 0.4882 1.2123 -0.2585 -0.4420 B174 0.5195 1.3222 -0.2926 -0.5492 B104 0.6614 1.0296 -0.3529 -0.3381 B175 0.5281 1.2866 -0.3003 -0.5144 B105 1.2093 1.2510 -0.7557 -0.7046 B176 0.7910 1.0308 -0.4448 -0.3771 B106 0.7563 1.2086 -0.4381 -0.5268 B177 0.5045 1.3977 -0.3611 -0.5411 B107 0.6321 1.2025 -0.3601 -0.4746 B178 0.4426 1.5445 -0.3357 -0.6514 B108 1.2012 0.6206 -0.4870 -0.3349 B179 0.8881 0.5737 -0.3306 -0.1312 B109 0.6404 1.5559 -0.3669 -0.8295 B180 0.7911 1.1830 -0.4966 -0.4775 B110 0.7476 1.5435 -0.4569 -0.8342 B181 0.7332 1.3216 -0.4761 -0.5787 Bill 0.2485 1.7961 -0.2146 -0.8300 B182 0.5015 1.3937 -0.2884 -0.6068 B112 1.1397 1.3705 -0.6659 -0.8443 B183 0.5930 1.2168 -0.2818 -0.5280 B113 0.7591 1.5381 -0.4642 -0.8331 B184 0.0000 1.8051 -0.0365 -0.7686 B114 0.8734 1.5886 -0.5394 -0.9226 B185 0.4551 1.1473 -0.2176 -0.3848 B115 0.6849 1.2952 -0.4731 -0.5070 B186 0.6726 1.1476 -0.3764 -0.4438 B116 0.7195 0.7742 -0.2294 -0.2643 B187 0.5206 1.2551 -0.3147 -0.4610 B117 0.8083 1.0521 -0.4734 -0.3870 B188 0.4868 1.2859 -0.2640 -0.5087 B118 0.6657 1.3399 -0.4026 -0.6030 B189 0.6386 1.4182 -0.4551 -0.6017 B120 0.5096 1.3764 -0.2935 -0.5925 B190 0.5585 1.3280 -0.3777 -0.5087 B121 0.5248 1.1163 -0.2401 -0.4011 B191 0.4142 1.4301 -0.3084 -0.5359 B124 0.8158 1.2503 -0.4657 -0.6004 B192 0.5742 1.2855 -0.3648 -0.4950 B125 0.7314 1.2730 -0.4523 -0.5521 B193 0.6465 1.3172 -0.4323 -0.5313 B127 0.4719 1.3466 -0.2979 -0.5205 B194 0.6755 1.0501 -0.3523 -0.3733 B128 0.5882 0.8055 -0.1571 -0.2365 B195 0.6606 1.3934 -0.4458 -0.6082 B129 0.4652 1.4625 -0.2998 -0.6279 B196 0.4675 1.0851 -0.2234 -0.3292 B130 0.4224 1.3774 -0.2644 -0.5354 B197 0.4803 1.4336 -0.3356 -0.5783 B131 0.6876 1.3436 -0.4301 -0.6011 B132 0.7061 1.2742 -0.4245 -0.5557 B133 0.6287 1.4218 -0.4263 -0.6241 B134 0.8028 1.3048 -0.5161 -0.5915 B135 0.5352 1.2980 -0.3801 -0.4530 B136 0.6184 1.2242 -0.3616 -0.4811 B198 0.4478 1.3347 -0.3044 -0.4782 B137 0.6047 1.2280 -0.3425 -0.4901 B199 0.8866 1.2342 -0.5413 -0.5795 B138 0.6140 1.2863 -0.3714 -0.5289 B200 0.9771 1.0126 -0.6330 -0.3567 B139 0.6512 1.2892 -0.3958 -0.5446 B201 0.7056 1.1290 -0.3941 -0.4406 B140 0.5245 1.3352 -0.3205 -0.5392 B202 0.2879 1.4270 -0.2046 -0.5103 B141 0.5696 1.2175 -0.3343 -0.4528 B203 0.3669 1.3695 -0.2850 -0.4514 B142 0.4753 1.3185 -0.2846 -0.5091 B204 1.0912 1.2119 -0.6478 -0.6554 B143 0.5632 1.2942 -0.3429 -0.5145 B205 0.3380 1.4218 -0.1933 -0.5665 B144 0.6402 1.2850 -0.3720 -0.5532 B206 0.9024 1.4280 -0.5792 -0.7512 B145 0.5508 1.2966 -0.3306 -0.5168 B207 0.9836 1.2223 -0.6171 -0.5889APPENDIX 2 41 APPENDIX 2.—Continued Sample Number End member M-l M-2 D-1 D-2 Little San Bernardino block—Continued B208 1.3615 1.1131 -0.8115 -0.6631 B209 1.3327 0.9114 -0.6194 -0.6246 B210 0.8695 0.8661 -0.3869 -0.3487 B211 1.0856 0.9661 -0.6212 -0.4305 B212 1.0926 0.9137 -0.5971 -0.4092 B2I3 0.6930 0.6728 -0.1555 -0.2103 B214 1.1242 0.4543 -0.3542 -0.2242 B215 1.1917 1.2863 -0.7610 -0.7171 B216 0.5367 0.9199 -0.1936 -0.2630 B217 0.8028 1.2984 -0.4994 -0.6017 B218 1.2019 0.7027 -0.6238 -0.2808 B219 0.8707 0.9628 -0.4292 -0.4043 B220 0.8327 1.1981 -0.5017 -0.5290 B221 0.9761 0.7734 -0.3945 -0.3550 B222 0.9861 0.7372 -0.3458 -0.3775 San Jacinto block B223 0.9859 0.6163 -0.4217 -0.1804 B224 0.5269 1.1874 -0.2494 -0.4648 B225 0.8099 0.7333 -0.3553 -0.1879 B226 0.3335 0.8727 0.0324 -0.2386 B227 0.3873 0.8940 -0.0063 -0.2749 B228 0.7662 0.9098 -0.3189 -0.3571 B229 0.3394 0.6504 -0.0075 0.0177 B230 0.7027 0.8791 -0.2533 -0.3286 B231 0.7688 0.9180 -0.3496 -0.3372 B232 0.5791 0.6307 -0.1327 -0.0771 B233 0.9301 0.7230 -0.2860 -0.3670 B234 0.8227 0.7338 -0.3230 -0.2334 B235 0.9812 0.7210 -0.4435 -0.2587 B236 0.7323 0.9621 -0.2682 -0.4262 B237 0.6757 0.6984 -0.2248 -0.1493 B238 0.7945 0.8537 -0.3348 -0.3134 B239 0.7069 0.7113 -0.2417 -0.1766 B240 0.7020 0.7919 -0.2888 -0.2051 B241 0.7869 0.7907 -0.3413 -0.2363 B242 0.7418 0.7718 -0.2798 -0.2338 B243 0.6834 0.7075 -0.2253 -0.1656 B244 0.9483 0.6390 -0.2984 -0.2888 B245 0.8296 0.7806 -0.3135 -0.2967 B246 0.8339 0.9386 -0.4089 -0.3636 B247 0.9322 0.7245 -0.4146 -0.2421 B248 1.0296 0.8927 -0.5081 -0.4142 B249 0.6344 0.9222 -0.2490 -0.3075 B250 0.5227 1.0323 -0.2031 -0.3520 B251 0.8156 0.5936 -0.2161 -0.1931 B252 0.7942 0.7917 -0.3493 -0.2366 B253 0.7450 0.7447 -0.2511 -0.2386 B254 0.8024 0.7573 -0.3275 -0.2322 B255 0.7274 1.4937 -0.4254 -0.7958 B256 0.6158 0.8424 -0.1786 -0.2796 B257 0.5904 0.7457 -0.1681 -0.1680 B258 0.7911 0.7382 -0.2928 -0.2364 B259 0.8759 1.2565 -0.4897 -0.6427 B260 1.0092 0.6565 -0.4307 -0.2350 B261 0.6684 0.7997 -0.2364 -0.2317 B262 0.6928 1.0889 -0.3289 -0.4529 B263 1.0899 0.9518 -0.5424 -0.4993 B264 0.9418 1.1636 -0.5102 -0.5951 B265 0.8129 0.6700 -0.2551 -0.2279 B266 0.1626 1.7666 -0.1621 -0.7672 B267 0.4676 1.6743 -0.2933 -0.8486 Sample Number End member M-l M-2 D-1 D-2 B268 1.0436 0.9379 -0.4854 -0.4961 B269 0.7386 0.5526 -0.2010 -0.0901 B270 1.5060 1.1208 -0.8940 -0.7327 B271 1.2724 1.1995 -0.7740 -0.6979 B272 1.7592 0.7880 -0.9433 -0.6039 B273 0.9636 0.5577 -0.3389 -0.1824 B274 0.4859 0.8500 -0.1016 -0.2343 B275 0.4204 1.1421 -0.1659 -0.3966 B276 0.6237 0.8767 -0.1874 -0.3130 B277 0.5948 0.7799 -0.1031 -0.2716 B278 0.5477 0.8156 -0.1293 -0.2340 B279 0.9333 1.0080 -0.4850 -0.4563 B280 0.9770 0.9052 -0.5050 -0.3773 B281 0.7293 1.0214 -0.3135 -0.4372 B282 0.4173 1.4874 -0.2221 -0.6825 B283 0.1947 1.1170 0.0732 -0.3850 B284 0.2175 1.1809 -0.0330 -0.3655 B285 0.3901 0.9863 -0.0627 -0.3137 B286 0.4279 0.7348 0.0064 -0.1691 B287 0.5701 1.0460 -0.2011 -0.4150 B288 0.6260 0.7331 -0.1171 -0.2420 B289 0.6420 0.9983 -0.2687 -0.3716 B290 0.6324 0.7067 -0.1502 -0.1890 B291 0.5500 1.1095 -0.2079 -0.4516 B292 0.8084 0.9822 -0.3271 -0.4634 B293 0.6962 0.9963 -0.2730 -0.4195 B294 0.7272 0.9919 -0.2983 -0.4208 B295 0.4824 0.6804 -0.0083 -0.1546 B296 0.3890 1.0707 -0.0928 -0.3669 B297 0.3528 0.9094 -0.0053 -0.2569 B298 0.4075 0.8574 -0.0096 -0.2552 B299 0.3154 0.8865 -0.0172 -0.1847 B300 0.4704 0.6234 -0.0057 -0.0882 B301 0.3563 0.9111 0.0039 -0.2713 B302 0.2546 0.9323 0.0568 -0.2437 B303 0.1444 1.0893 0.0494 -0.2832 B304 0.3335 0.6797 0.0641 -0.0773 B305 0.6179 0.5864 -0.1186 -0.0857 B306 0.0977 0.9747 0.1468 -0.2192 B307 0.3589 0.9237 -0.0018 -0.2808 B308 0.3225 0.7292 0.0577 -0.1094 B309 0.3749 1.0721 -0.0838 -0.3632 B310 0.5617 0.8145 -0.0804 -0.2958 B311 0.5814 0.7289 -0.1146 -0.1957 B312 1.2970 1.3622 -0.8076 -0.8516 B313 1.1211 1.3030 -0.6676 -0.7565 B314 1.2809 1.0455 -0.7273 -0.5991 B315 0.7331 0.6313 -0.1986 -0.1658 B317 0.7599 0.6091 -0.2398 -0.1292 B318 0.5452 0.6453 -0.0892 -0.1013 B319 0.7249 0.6300 -0.2229 -0.1320 B320 0.4567 0.6404 -0.0264 -0.0707 B321 0.2819 0.6737 0.0641 -0.0197 B322 0.7896 0.6504 -0.2489 -0.1910 B323 0.6313 0.5529 -0.1475 -0.0368 B324 1.1072 0.5289 -0.4441 -0.1920 Perris block B325 1.0270 0.6908 -0.3763 -0.3415 B326 1.2220 0.6021 -0.4297 -0.3943 B327 1.4513 0.5810 -0.6315 -0.4008 B328 1.3916 0.7255 -0.6483 -0.4688 B330 1.7953 0.9791 -0.9467 -0.8277 End member Sample Number M-l M-2 D-l D-2 B331 1.5502 0.9483 -0.8582 -0.6404 B332 1.4352 1.0775 -0.8182 -0.6945 B333 1.5518 0.9864 -0.8744 -0.6637 B334 1.4987 1.1368 -0.8786 -0.7569 B336 1.2551 0.3337 -0.3420 -0.2469 B337 1.3087 0.8591 -0.5859 -0.5819 B338 1.1805 0.3435 -0.6150 0.0910 B339 1.4054 0.3488 -0.5089 -0.2453 B340 1.5226 0.3666 -0.5774 -0.3118 B341 0.5722 0.1931 0.1996 0.0352 B342 1.3682 0.2810 -0.4121 -0.2372 B343 1.4111 0.4971 -0.5270 -0.3812 B344 1.4100 0.4402 -0.5073 -0.3428 B345 1.6812 0.4779 -0.6754 -0.4837 B346 2.0244 0.8481 -1.1517 -0.7208 B347 1.2062 0.3612 -0.3572 -0.2102 B348 2.0263 0.8663 -1.1735 -0.7192 B350 1.4198 0.6986 -0.6824 -0.4360 B351 1.6113 0.6087 -0.7936 -0.4264 B352 1.2370 0.5046 -0.3927 -0.3489 B353 1.2389 0.3859 -0.3839 -0.2410 B354 1.2425 0.7050 -0.5080 -0.4395 B355 1.2820 0.7921 -0.5759 -0.4982 B356 1.4016 0.6942 -0.6872 -0.4086 B357 1.4453 0.3245 -0.4348 -0.3350 B358 1.5523 0.4009 -0.5916 -0.3615 B359 1.1839 0.4262 -0.2977 -0.3125 B360 1.8643 0.4958 -0.9396 -0.4205 B361 1.3819 0.6714 -0.6709 -0.3825 B363 2.0647 0.3651 -0.8853 -0.5446 B364 2.0905 0.8017 -1.1718 -0.7204 B366 1.8375 0.5314 -0.9395 -0.4295 B367 2.0478 0.9741 -1.1496 -0.8722 B368 1.7444 0.9050 -0.9761 -0.6733 B369 1.8090 1.0659 -1.0442 -0.8307 B370 1.2524 0.6511 -0.4604 -0.4430 B371 1.1371 0.4786 -0.4329 -0.1828 B372 0.8761 0.6609 -0.2943 -0.2427 B373 0.8987 0.4922 -0.3191 -0.0718 B374 0.8425 0.6663 -0.2210 -0.2878 B375 0.9793 0.3972 -0.3103 -0.0662 B376 0.8043 0.6020 -0.1637 -0.2427 B377 1.4700 0.3375 -0.5895 -0.2180 B378 1.5581 0.4664 -0.6848 -0.3397 B380 1.0473 0.4904 -0.3671 -0.1706 B381 0.8412 0.4006 -0.2394 -0.0024 B382 0.9697 0.4813 -0.3410 -0.1100 B383 0.8729 0.4801 -0.2686 -0.0844 B384 0.9437 0.5656 -0.3808 -0.1286 B385 1.0219 0.4591 -0.3388 -0.1422 B386 1.7529 0.4200 -0.6622 -0.5106 B387 1.0528 1.1913 -0.6190 -0.6251 B388 1.9236 0.9767 -1.1362 -0.7641 B389 1.7926 0.9340 -1.0611 -0.6655 B390 1.4990 0.5528 -0.7762 -0.2757 B391 1.5496 0.6817 -0.8332 -0.3981 B392 0.3044 0.6424 -0.0475 0.1007 B394 1.0670 0.6235 -0.4249 -0.2656 B395 0.9725 0.5664 -0.3392 -0.1997 B396 1.0863 1.0770 -0.5842 -0.5791 B397 0.8499 0.7583 -0.3735 -0.2347 B398 0.7902 0.6071 -0.2416 -0.1558 B399 0.9724 0.7382 -0.4102 -0.3005 B400 1.6440 0.4768 -0.8424 -0.2784 B401 1.2272 0.3712 -0.3480 -0.250442 BATHOLITHIC ROCKS OF SOUTHERN CALIFORNIA APPENDIX 2.—Continued End member Sample Number M-1 M- 2 O-l 0-2 Perris block—Continued B402 0.6372 1.0920 -0.1744 -0.5547 B403 1.3768 0.3126 -0.4490 -0.2403 B404 0.9711 0.7182 -0.3703 -0.3189 B405 0.6742 0.6510 -0.1048 -0.2204 B406 0.7566 1.2203 -0.2567 -0.7202 B407 1.2854 0.9295 -0.7204 -0.4945 B408 1.1790 0.5061 -0.4077 -0.2774 B409 1.2244 0.5982 -0.5067 -0.3159 B410 1.2343 0.6591 -0.5153 -0.3781 B411 0.9239 0.5092 -0.1467 -0.2864 B412 0.4703 0.2809 0.1185 0.1304 B413 1.3006 0.6849 -0.5267 -0.4588 B415 1.1394 0.8811 -0.4749 -0.5456 B416 1.2384 1.2007 -0.7388 -0.7003 B417 1.2209 0.5101 -0.3363 -0.3948 B418 1.0601 1.2860 -0.6292 -0.7169 B420 0.6256 0.5349 0.0153 -0.1758 B421 1.3384 1.2911 -0.7400 -0.8895 B422 1.0345 1.1845 -0.5660 -0.6530 B423 1.1152 0.6262 -0.3193 -0.4221 B424 1.5740 0.9303 -0.9321 -0.5723 B425 1.1126 1.1130 -0.6041 -0.6215 B426 1.6943 0.5408 -0.7948 -0.4404 B427 1.5244 0.3121 -0.5077 -0.3287 B428 1.4103 1.3798 -0.8749 -0.9152 B429 2.6100 -0.3401 -1.3843 0.1144 B430 0.6489 0.8630 -0.1454 -0.3665 B431 1.3082 0.7015 -0.5398 -0.4698 B432 0.7463 0.6594 -0.1526 -0.2531 B433 1.2086 0.6159 -0.4664 -0.3581 B434 1.5552 0.7281 -0.8070 -0.4764 B435 1.1509 0.4707 -0.4200 -0.2016 B436 1.4615 0.3951 -0.5794 -0.2773 B437 1.0942 0.5450 -0.4529 -0.1863 B438 0.7841 0.6277 -0.2259 -0.1859 B439 1.0627 0.6097 -0.4216 -0.2508 B440 1.7744 0.5888 -0.9042 -0.4591 B441 1.0250 0.6584 -0.3863 -0.2970 B442 1.1911 0.6003 -0.4619 -0.3296 B447 1.1217 0.8412 -0.5649 -0.3980 B448 1.3424 0.2424 -0.5602 -0.0246 B449 1.2736 0.6097 -0.5074 -0.3760 B450 0.7420 1.4328 -0.4471 -0.7277 Sample Number End member M-l M- 2 0-1 0-2 B451 1.5397 0.6778 -0.7416 -0.4759 B452 0.9229 0.7731 -0.3253 -0.3706 B453 0.8337 0.7454 -0.2872 -0.2919 B454 0.7106 0.7977 -0.2244 -0.2839 B455 1.0197 0.7925 -0.4010 -0.4112 B456 1.2100 0.5883 -0.4708 -0.3274 Santa Ana block B458 1.9135 0.8606 -1.1011 -0.6730 B459 1.7670 1.0704 -1.0623 -0.7751 B460 2.0496 0.8175 -1.1759 -0.6912 B461 2.1629 0.9533 -1.2337 -0.8826 B462 1.4343 0.8453 -0.6722 -0.6074 B464 1.5710 0.9421 -0.8209 -0.6921 B465 1.6721 1.0488 -0.9802 -0.7407 B466 1.9508 0.8012 -1.1159 -0.6360 B469 1.8732 0.9285 -1.0555 -0.7461 B470 1.8369 0.9476 -1.0367 -0.7479 B471 2.0347 1.2250 -1.2098 -1.0498 B472 1.7620 0.7959 -0.9843 -0.5736 B473 1.5866 0.5076 -0.7271 -0.3671 B474 1.5042 0.5765 -0.5789 -0.5018 B475 1.7422 1.0713 -0.9861 -0.8274 B476 1.7772 0.7606 -0.8813 -0.6565 B478 1.6282 0.6953 -0.7726 -0.5509 B479 1.5948 1.0462 -0.8869 -0.7542 B480 1.8548 0.9585 -1.0452 -0.7680 B482 1.9808 1.0664 -1.2051 -0.8421 B484 1.1093 0.2486 -0.2721 -0.0859 B485 1.3515 1.0358 -0.7494 -0.6378 B486 1.6495 0.4253 -0.6999 -0.3748 B487 1.0815 0.1737 -0.2007 -0.0545 B489 1.6731 0.9822 -1.0446 -0.6107 B490 1.6631 1.1061 -0.9912 -0.7781 B491 1.8256 0.9594 -1.0607 -0.7244 B492 1.2329 0.7007 -0.4954 -0.4382 B493 2.1502 0.5189 -1.2684 -0.4007 B494 1.3882 0.5007 -0.5172 -0.3717 B495 1.5606 0.4386 -0.6090 -0.3902 B496 2.2246 0.5550 -1.1734 -0.6061 B497 1.7294 0.4977 -0.7516 -0.4754 B498 1.4506 0.3243 -0.4534 -0.3215 B499 1.7833 0.4276 -0.7683 -0.4426 End member Sample Number M-l M-2 0-1 0-2 B500 1.2032 0.4107 -0.3525 -0.2615 B502 1.2378 0.5842 -0.5065 -0.3155 B503 1.1302 0.3147 -0.2648 -0.1800 B504 1.5663 1.0441 -0.8981 -0.7124 B506 1.9503 0.6488 -1.0632 -0.5359 B507 1.9119 0.3398 -0.8939 -0.3578 B508 1.9738 0.6965 -1.1051 -0.5652 B510 1.9703 0.8275 -1.1149 -0.6828 B513 1.0880 0.0000 -0.1165 0.0285 B514 1.0534 0.1775 -0.1530 -0.0779 B515 2.1895 0.9767 -1.3064 -0.8598 B516 2.5641 0.7164 -1.4776 -0.8029 B517 1.8413 0.5840 -0.8553 -0.5699 B518 1.3600 0.2735 -0.3995 -0.2339 B519 1.5813 0.3836 -0.6068 -0.3582 B520 2.2689 0.6373 -1.2865 -0.6197 B521 1.9913 0.7746 -1.1445 -0.6214 B522 1.2333 0.2972 -0.3680 -0.1625 B523 1.1810 0.2259 -0.2744 -0.1325 B524 1.8796 0.2871 -0.7876 -0.3791 B525 1.7953 0.4217 -0.7808 -0.4362 B528 2.3401 0.5816 -1.3256 -0.5961 B529 1.0731 0.4469 -0.2912 -0.2288 B530 1.1585 0.3245 -0.2768 -0.2062 B531 1.0718 0.4949 -0.2703 -0.2964 B533 1.0072 0.0829 -0.0706 -0.0195 B536 1.3371 0.0609 -0.2591 -0.1390 B537 1.2498 0.3334 -0.3295 -0.2537 B538 1.8314 0.8159 -1.0151 -0.6322 B539 2.2172 0.5596 -1.2106 -0.5662 B540 1.7085 0.8473 -0.9539 -0.6018 B541 1.8661 0.8006 -1.0206 -0.6461 B542 2.3167 0.3555 -1.2362 -0.4360 B543 1.4461 0.0870 -0.3423 -0.1908 B544 1.4404 0.4186 -0.5373 -0.3217 8546 1.1315 0.6895 -0.4431 -0.3779 B547 0.9797 0.3356 -0.2268 -0.0885 B548 1.1964 0.4501 -0.4272 -0.2193 B549 1.7151 0.6954 -0.8955 -0.5150 B550 1.4935 1.2775 -0.9442 -0.8267 B551 1.7885 0.6883 -0.9721 -0.5048 B553 1.3364 1.3784 -0.8653 -0.8495 B555 1.2041 0.4383 -0.4052 -0.2371 B556 1.2864 0.5755 -0.5126 -0.3493 B557 0.7757 0.3869 -0.1188 -0.0438 ☆US. GOVERNMENT PRINTING OFFICE 1984-776-041/4035The Black Creek-Peedee Formational Contact (Upper Cretaceous) in the Cape Fear River Region of North Carolina By NORMAN F. SOHL and RAYMOND A. CHRISTOPHER GEOLOGICAL SURVEY PROFESSIONAL PAPER 1285 Stratigraphic and paleontologic data suggest a disconformable relationship between the Black Creek and Peedee Formations along the Cape Fear River of North Carolina UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON: 1983UNITED STATES DEPARTMENT OF THE INTERIOR JAMES G. WATT, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Sohl, Norman F. (Norman Frederick), 1924- The Black Creek-Peedee formational contact (Upper Cretaceous) in the Cape Fear River region of North Carolina. (Geological Survey professional paper; 1285) Bibliography: p. Supt. of Docs, no.: 119.16:1285 1. Geology, Stratigraphic—Cretaceous. 2. Geology—North Carolina—Cape Fear River Watershed. I. Christopher, Raymond A. II. Title. III. Series. QE688.S63 1983 551.7T09756 83-600075 For sale by the Distribution Branch, U.S. Geological Survey, 604 South Pickett Street, Alexandria, VA 22304CONTENTS Page Abstract.................................................... 1 Introduction................................................ 1 Geographic setting...................................... 2 Previous work........................................... 2 Section descriptions........................................ 6 Walkers Bluff........................................... 6 Stratigraphic column, upstream section ............ 12 Stratigraphic column, medial section.............. 12 Stratigraphic column, downstream section....... 12 Paleontology....................................... 12 Discussion ........................................ 12 Jessups Landing.................................... 17 Stratigraphic column............................... 17 Paleontology....................................... 17 Discussion ........................................ 17 Deepwater Point........................................ 17 Lithologic descriptions of the samples............. 17 Paleontology....................................... 17 Donoho Creek Landing................................... 17 Stratigraphic column............................... 17 Paleontology ...................................... 17 Discussion ........................................ 19 Section descriptions—Continued Robinsons Landing.................. Stratigraphic column........... Paleontology .................. Browns Landing..................... Stratigraphic column........... Paleontology .................. Discussion .................... Remarks........................ Interpretations........................ Age and biostratigraphic correlations. Paleoenvironmental interpretations. Physical stratigraphy.............. Conclusions............................ Measured sections...................... Walkers Bluff, upstream section.... Walkers Bluff, medial section...... Walkers Bluff, downstream section.. Jessups Landing.................... Donoho Creek Landing............... Robinsons Landing.................. Browns Landing..................... References............................. Paste 20 20 20 20 20 20 20 20 20 20 26 28 30 31 31 33 34 35 35 36 36 37 ILLUSTRATIONS Page Figure 1. Map of the Cape Fear River, N.C., between Walkers Bluff and Browns Landing, showing localities discussed...... 3 2-4. Stratigraphic columns of: 2. Upstream section at Walkers Bluff .................................................................. 13 3. Medial section at Walkers Bluff..................................................................... 14 4. Downstream section at Walkers Bluff............................................................,____ 15 5. Diagram showing physical correlation of the lithologic units exposed at the downstream, medial, and upstream sections at Walkers Bluff.............................................................................. 16 6-9. Stratigraphic columns of: 6. Jessups Landing________________________________________________________________________________________ 18 7. Donoho Creek Landing.................................................................................. 19 8. Robinsons Landing....................................................................................... 21 9. Browns Landing ________________________________________________________________________________________ 22 10. Chart showing stratigraphic ranges of 11 biostratigraphically important invertebrate fossils that are found in ex- posures along the Cape Fear River between Walkers Bluff and Browns Landing............................. 24 11. Chart showing stratigraphic ranges of 31 biostratigraphically important pollen species that are found in exposures along the Cape Fear River between Walkers Bluff and Browns Landing..................................... 25 12. Biostratigraphic correlation chart of the outcropping Cretaceous sections between Walkers Bluff and Browns Landing, Cape Fear River, N.C., and the sections of New Jersey and the Chattahoochee River, Ala. and Ga. 27 13. Diagram showing the relative frequency of occurrence (in percent) of dinoflagellates and acritarchs in palynologic samples from Walkers Bluff, Jessups Landing, Deepwater Point, Donoho Creek Landing, Robinsons Landing, and Browns Landing, Cape Fear River, N.C............................................................... 29 14. Schematic cross section along part of the Cape Fear River showing physical, paleoenvironmental, and biostrati- graphic interpretations of the relationship between the Black Creek and Peedee Formations.............. 32 TABLES Table 1. Location of the Cape Fear River sections discussed in this report............................................ 2. Distribution of megainvertebrate fossils and vertebrate remains in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Browns Landing...................................... 3. Distribution of biostratigraphically important pollen species in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Browns Landing...................................... Page 2 7 11 III THE BLACK CREEK-PEEDEE FORMATIONAL CONTACT (UPPER CRETACEOUS) IN THE CAPE FEAR RIVER REGION OF NORTH CAROLINA By Norman F. SoHLand Raymond A. Christopher1 ABSTRACT Stratigraphic and paleontologic data from six sections exposed along the Cape Fear River, N.C., suggest that the contact between the Upper Cretaceous Black Creek and overlying Peedee Formations is unconformable. The unconformity is exposed in three closely spaced sections between mileposts 49.5 and 50.25 and is indicated by the presence of a thin (2-inch-thick) zone at the base of the Peedee Formation that contains worn and abraded bone and shell, bored phosphate pebbles, teeth, and clasts reworked from the underlying Black Creek Formation in a matrix of poorly sorted coarse-grained sand. The disconformity represents a relatively short hiatus that occurred sometime between the latest Campanian and earliest Maestrichtian (that is, during the time of concurrence of Exogyra ponderosa errati-costata and E. costata). The short duration of the hiatus makes it difficult to detect by means of existing megafossil and (or) palynologic zonations. However, lithologic and paleontologic evidence suggests that a major paleoenvironmental change takes place at the disconformity. Below the disconformity, the upper part of the Black Creek Formation consists of a series of interfingering and discontinuous lithologic units that reflect rapidly changing nearshore environments. Above the disconformity, the basal part of the Peedee Formation consists of massive, glauconitic, bioturbated, muddy sands of open marine shelfal origin. Previous interpretations of the ages of the Black Creek and Peedee Formations and the nature of their contact were based, in large part, on an interpretation of the fauna contained in and the stratigraphic position of a zone of sandstone blocks of Cretaceous age that had not been recognized previously as having been reworked into the basal part of the Waccamaw Formation of early Pleistocene age. These blocks occur in an outcrop referred to as Walkers Bluff, located 10 river miles upstream from the exposed Black Creek-Peedee contact. INTRODUCTION According to Stephenson (1923), the Cretaceous System of the Carolina Coastal Plain is best exposed along the Cape Fear River of North Carolina. Here, the Upper Cretaceous Series consists of three formations; in ascending stratigraphic order, they are the Cape Fear, Black Creek, and Peedee Formations. The excellence and accessibility of the exposures along the river have prompted many workers to use the Cape Fear River section as the basis for interpreting the stratigraphic relationships among these formations in other areas (for example, Stephenson, 1912, 1923; Brett and Wheeler, 1961; Heron and Wheeler, 1964; Swift, 1964; Swift and Heron, 1967, 1969; Swift and others, 1969; Christopher and others, 1979). Stratigraphic relationships among the Upper Cretaceous units and with the underlying and overlying rocks are well established for most of the section; unconformities are recognized (1) at the base of the Cape Fear Formation, where the Cape Fear rests nonconformably on crystalline rocks of the Piedmont province, (2) between the Cape Fear and Black Creek Formations, and (3) at the top of the Peedee Formation, where post-Cretaceous rocks discon-formably overlie the Peedee. However, the nature of the contact between the Black Creek and Peedee Formations is still unresolved; it has been regarded as gradational by some workers (for example, Stephenson, 1923; Heron and Wheeler, 1964), as unconformable by others (for example, Brett and Wheeler, 1961), and as a ra-vinement by still others (for example, Swift, 1964; Swift and Heron, 1967, 1969; Swift and others, 1969). To better understand, and perhaps to help resolve, the nature of the Black Creek-Peedee contact, we visited several exposures of the upper part of the Black Creek and basal part of the Peedee Formations along the Cape Fear River. During these visits we made detailed observations of the lithostratigraphic relationships within and among the exposures and collected invertebrate and palynomorph samples for use in bio-stratigraphic and paleoenvironmental interpretations. The results of our study suggest that the Black Creek and Peedee Formations are disconformable and that the two units were dominated by distinctly different environments of deposition; the disconformity is exposed along the Cape Fear River at three closely spaced outcrops between 49.5 and 50.25 river mi upstream from Wilmington, N. C. It is our purpose to present the data upon which we base our conclusions and to suggest correlations for the upper part of the Black Creek and basal part of the Peedee Formations with Cretaceous units elsewhere in the Coastal Plain province other than the Carolinas. We make no effort in this report to relate our findings to areas outside the Cape Fear River region, nor do we imply that the ages, paleoenvironmental interpretations, and stratigraphic relationships between the Black Creek and Peedee Formations along the Cape Fear River can be extended to other areas of the Carolina Coastal Plain. We intend to treat these topics in future reports that will be prepared when we complete our ongoing investigations of the Upper Cretaceous Series as exposed along the Roanoke, Tar, Neuse, North ARCO Oil and Gas Company. Plano, TX 75221. 1BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. 2 East Cape Fear, Black, and Peedee Rivers of the Carolinas. In this report, we follow the convention of using “right” and “left” bank as one faces downstream. Geographic Setting Most studies of the contact between the Black Creek and Peedee Formations in the Carolinas are based on paleoenvironmental interpretations of and physical and biostratigraphic correlations between exposures along the Cape Fear River at Walkers Bluff and a series of three closely spaced outcrops referred to as Donoho Creek, Robinsons, and Browns Landings (fig. 1). At Walkers Bluff, located 60 river mi upstream from Wilmington, N.C., the Cape Fear River impinges against its right valley wall for a horizontal distance of approximately 0.75 mi, thereby exposing 45-55 ft of Cretaceous sediments and 10-20 ft of shell marls of Pliocene and (or) Pleistocene age. Ten river mi downstream from Walkers Bluff, the Cape Fear River again impinges against its right valley wall, exposing both Cretaceous and post-Cretaceous sediments. This bluff, located between mileposts 50.25 and 49.5, is divided into three sections by two small tributaries to the Cape Fear River; in a downstream direction, these sections are referred to as Donoho Creek Landing, Robinsons Landing, and Browns Landing, respectively. Between Walkers Bluff and the exposures at Donoho Creek, Robinsons, and Browns Landings, a horizontal distance of 7.3 mi, Stephenson (1912, 1923) noted the presence of Cretaceous sediments at three localities. Downstream (that is, upsection) from Walkers Bluff, these localities are Jessups Landing (milepost 56), an unnamed locality at milepost 53.5, and Deepwater Point (milepost 51.5). At Jessups Landing, 3-4 river mi downstream from Walkers Bluff, Cretaceous sands and clays overlain by gravelly alluvial (?) sand crop out for approximately 300 yd along the left bank of the Cape Fear River. Lock and Dam No. 2, downstream from Donoho Creek Landing, was constructed after Stephenson’s traverse of the river in 1907, and the presence of the dam has caused water level to rise sufficiently to drown the localities at milepost 53.5 and Deepwater Point; at present, Jessups Landing is the only outcrop of Cretaceous sediments between Walkers Bluff and the exposures at Donoho Creek, Robinsons, and Browns Landings. Data for our report come from stratigraphic interpretations and paleontologic analyses of samples from Walkers Bluff, Jessups Landing, Donoho Creek Landing, Robinsons Landing, and Browns Landing. In addition, we collected several samples of the fossiliferous marl and sandstone mentioned by Stephenson (1912, 1923) as occurring at Deepwater Point; these samples were collected with the use of snorkeling gear. The positions of the localities along the Cape Fear River reported on herein are illustrated in figure 1, and detailed locality information is presented in table 1. Previous Work The exposures between Walkers Bluff and Donoho Creek Landing are some of the most intensively studied Cretaceous sections in the Atlantic Coastal Plain. Several workers (for example, Stephenson, 1912, 1923; Brett and Wheeler, 1961; Heron and Wheeler, 1964) have provided measured sections and sedimentologic and (or) stratigraphic data for these exposures. Others (for example, Swift, 1964; Swift and Heron, 1967,1969; Swift and others, 1969) have provided less detailed information regarding these sections but have referred to them in interpretations of the Late Cretaceous history of the Carolinas. According to Stephenson (1912, 1923), the Black Creek Formation consists of thinly laminated, cross-bedded ferruginous sand and carbonaceous clay. Lignite, amber, marcasite, and fossil leaves are common throughout the unit; glauconite is found in the unit but is not common or abundant. Horizontal and vertical changes between sand-dominated and clay-dominated facies are common and may be abrupt. Marine invertebrate fossils have not been found in the lower part of the Black Creek Formation, but Stephenson (1912,1923) observed the laminated sands and clays toward the top of the unit to be interstratified with lenses or beds of indurated calcareous sands and marls, some of which contain an abundant marine Table 1.—Location of the Cape Fear River sections discussed in this report Sections are presented in ascending stratigraphic order Section name Milepost Bank County North latitude West longitude 15-minute quadrangle Degree Minute Second Degree Minute Second Walkers Bluff 60.0 Right.. ...Bladen . 34 33 14 78 29 20 White Lake Jessups Landing 56.0 Left... do 34 32 52 78 26 10 Do. Deepwater Point 51.5 (1) do ...34 29 31 78 24 35 Bolton Donoho Creek Landing 50.25 Right... do 34 28 26 78 24 40 Do. Robinsons Landing 49.75 do do 34 28 04 78 24 10 Do. Browns Landing 49.5 do do 34 28 00 78 23 54 Do. 'Rocks now inundated.INTRODUCTION 3 Figure 1. Map of the Cape Fear River, N.C., between Walkers Bluff (milepost 60) and Browns Landing (milepost 49.5), showing the localities discussed in this report. The numbers in parentheses refer to publications in which stratigraphic and (or) paleontologic information was provided for that section: 1, Stephenson (1912); 2, Stephenson (1923); 3, Brett and Wheeler (1961); 4, Heron and Wheeler (1964); 5, this report.4 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. fauna. Stephenson (1923) assigned these fossiliferous beds to his “Snow Hill Calcareous Member,” which is restricted to the upper part of the Black Creek Formation, and he considered them to have been deposited in a somewhat deeper water marine environment than was the lower, unfossiliferous part of the formation. Stephenson (1912, 1923) reported the presence of these fossiliferous sandstones and marls of the “Snow Hill Calcareous Member” from several localities along the Cape Fear River. At Walkers Bluff (milepost 60), he observed a 1-ft-thick bed at the top of the section where, according to Stephenson’s (1923) description, it is underlain by typical laminated, unfossiliferous sands and clays of the Black Creek Formation and is unconformably overlain by the Waccamaw Formation, which is early Pleistocene age according to Blackwelder (1979). Stephenson (1923) observed this sandstone to be discontinuous along the bluff and concluded that it changes laterally to an unfossiliferous loose sand. At the base of the exposure at Jessups Landing (milepost 56), Stephenson (1912) noted the presence of several feet of shell marl, overlain by 12 ft of laminated, unfossiliferous sands and clays that he attributed to the Black Creek Formation. At the time of publication of his description of the Jessups Landing section (1912), the term “Snow Hill Calcareous Member” had not been introduced. However, the shell marl at the base of the section falls within his later (1923) circumscription of this unit. Stephenson (1912, 1923) also reported the occurrence of fossiliferous sandstones and shell marls at Deepwater Point (milepost 51.5), where, at the time of his visit, “there is exposed at extreme low water a layer of greenish-gray, calcareous and fossiliferous rock about 1 foot in thickness, of marine origin. It is underlain by a dark, unconsolidated shell marl containing many sharks’ teeth and fragile Cretaceous fossils, about 1 foot being exposed above the water.***” (Stephenson, 1912, p. 123). At the base of the section at Donoho Creek Landing (milepost 50.25), Stephenson (1912, 1923) noted the occurrence of 10 ft of dark blue, laminated, poorly fossiliferous Black Creek clays. These clays, he noted, were overlain by 26 ft of the Peedee Formation, which he characterized throughout the Carolinas as a compact, dark-green or dark-gray, finely micaceous, glauconitic and argillaceous, fossiliferous, massive sand. Impure limestone beds and concretions and dark clays are also present in the Peedee Formation. From the massive bedding, contained fossils, and disseminated glauconite, Stephenson (1923) concluded that the Peedee Formation was deposited in open marine waters below wave base but above a depth of 50 fathoms. From his observations of the sections between Walkers Bluff and Donoho Creek Landing, Stephenson (1923) considered the contact between the Black Creek and Peedee Formations to be transitional, as he found no physical evidence of an unconformity or paleonto-logic evidence of a significant time-break between the two units. He concluded, therefore, that the Black Creek-Peedee contact represents a gradual transition from nearshore shallow marine bay and estuarine environments, represented by the lower part of the Black Creek Formation, to open marine conditions, represented by the Peedee Formation. The “Snow Hill Calcareous Member” is the transition between these two environments. In an attempt to assess the stratigraphic relationships of the “Snow Hill Calcareous Member” to the Black Creek and Peedee Formations, Brett and Wheeler (1961) examined the stratigraphy, sedimentol-ogy, and paleontology of the Walkers Bluff, Donoho Creek Landing, and Browns Landing sections, in addition to sections from other areas of North Carolina. At Donoho Creek and Browns Landings, they observed a 6-in- to 6-ft-thick, poorly indurated, fossiliferous sandstone unconformably overlying the laminated Black Creek clays that occur at the base of both sections. They considered this sandstone to be the basal bed of the Peedee Formation, and, on the basis of their paleontologic and lithologic data, they considered it to be part of the same lithostratigraphic unit and deposited at the same time as the 1-ft-thick sandstone that caps the Cretaceous section at Walkers Bluff [which Stephenson (1923) placed in the “Snow Hill Calcareous Member” of the Black Creek Formation]. As a result of their investigation, Brett and Wheeler (1961) concluded that, between Walkers Bluff and Donoho Creek Landing, the contact between the Black Creek and Peedee Formations is unconformable and that an indurated calcareous sandstone marks the base of the Peedee Formation. Despite their conclusion that an unconformity separates the Black Creek and Peedee Formations in the Cape Fear River region, Brett and Wheeler (1961) considered the upper part of the Black Creek and the basal part of the Peedee Formations to have been deposited during the same time interval (late Tayloran) and to represent a transition from a delta or near-delta distributary environment (the unfossiliferous, laminated carbonaceous clays of the Black Creek Formation) to a “beach,” “inlet,” or “upper-sound” environment (the basal sandstone of the Peedee Formation) to an “open lagoonal inlet” environment (the remainder of the Peedee Formation as exposed at Donoho Creek Landing). The sequence of transitional environments and the apparently synchronous nature of the upper part of the Black Creek and the lower part of the Peedee Formations led Brett and Wheeler (1961) to postulate that theINTRODUCTION 5 Black Creek and Peedee Formations represent part of a single marine transgression that occupied the entire Late Cretaceous in North Carolina. Heron and Wheeler (1964) agreed with Brett and Wheeler (1961) that, lithologically, the “Snow Hill Calcareous Member” as described by Stephenson (1923) should be placed in the Peedee Formation. However, in addition to the fossiliferous sandstone at the top of the Cretaceous part of the Walkers Bluff section previously noted by both Stephenson (1912, 1923) and Brett and Wheeler (1961), Heron and Wheeler (1964) also observed scattered shells in a 2- to 3-ft-thick bed of medium- to coarse-grained sand approximately 20 ft below the sandstone at the top of the section. By definition, this lower sand would be included in the “Snow Hill Calcareous Member” of Stephenson (1923), but, because Brett and Wheeler (1961) had placed the “Snow Hill” in the basal part of the Peedee Formation, Heron and Wheeler (1964) placed the Black Creek-Peedee contact at the base of this sand. As this lower sand appears to be conformable with a laminated clay below, Heron and Wheeler (1964) concluded that, at Walkers Bluff, the Black Creek-Peedee contact is conformable and that “[i]t is difficult to place a finger on an unequivocal lithologic boundary” between these units (Heron and Wheeler, 1964, p. 46). At Donoho Creek Landing, however, Heron and Wheeler (1964) recognized the unconformity that Brett and Wheeler (1961) considered to mark the Black Creek-Peedee contact. This unconformity occurs between the undisputed Black Creek clays at the base of the section and the 1-ft-thick sandstone that Brett and Wheeler (1961) considered as marking the base of the Peedee Formation. According to Heron and Wheeler (1964 p. 46), the unconformity is characterized by phosphate nodules, large specimens of Ostrea pratti, and “unusual small soft discs of sand***.” They further observed that neither the unconformity nor the basal sandstone of the Peedee Formation is as conspicuous everywhere as each is at Donoho Creek Landing. At Robinsons Landing, just 0.5 mi downstream from Donoho Creek Landing, Heron and Wheeler (1964) found no physical evidence of the unconformity, but they noted that the contact between the Black Creek and Peedee Formations is marked by the presence of bedding in the Black Creek Formation and a 3-ft-thick fossiliferous zone at the base of the Peedee Formation , that apparently occupies the same stratigraphic position as does the basal sandstone at Donoho Creek Landing. For a distance of 0.5 mi upstream from Donoho Creek Landing, however, Heron and Wheeler (1964) noted blocks of the basal part of the Peedee sandstone scattered along the water’s edge. As had Brett and Wheeler (1961), Heron and Wheeler (1964) concluded that the Black Creek and Peedee Formations are transitional, and, as such, they are evidence that the entire Upper Cretaceous Series of the Carolinas represents a single marine transgression. Their reasoning included not only the apparently transitional nature of the contact but the time of deposition of the units. According to their reasoning: 1. Stephenson (1923) placed the Black Creek Formation, including its upper “Snow Hill Calcareous Member,” in the Exogyra ponderosa Zone of Taylo-ran Age and the Peedee Formation in the E. costata Zone of Navarroan Age. 2. Brett and Wheeler (1961) removed the “Snow Hill Calcareous Member” from the Black Creek Formation and placed it in the Peedee Formation. As a result, the Peedee Formation, as defined by Brett and Wheeler (1961), includes rocks of both Tayloran Age (the “Snow Hill”) and Navarroan Age (the remainder of the Peedee Formation). 3. Therefore, because the Black Creek Formation is of Tayloran Age and the Peedee Formation, including the “Snow Hill,” is of both Tayloran and Navarroan Age, these two units must be time transgressive. Swift (1964), Swift and Heron (1969), and Swift and others (1969) accepted and expanded upon the hypotheses advanced by Brett and Wheeler (1961) and Heron and Wheeler (1964). Specifically, they agreed with the previously drawn conclusions that the Black Creek and Peedee Formations represent part of a single Late Cretaceous marine transgression and that the “Snow Hill Calcareous Member” lithologically and paleoenvironmentally belongs to the basal part of the Peedee Formation rather than to the upper part of the Black Creek Formation. In a series of sedimentologic studies aimed at interpreting the environments of deposition of the Cretaceous formations of the Carolinas, these workers recognized three environments (or lithosomes) within the Black Creek Formation (an estuarine, a lagoonal, and a littoral lithosome; Swift and Heron, 1967) and three within the Peedee Formation (a proximal-shelf, a distal-shelf, and a shelf-mud lithosome; Swift and others, 1969). With regard to these lithosomes, Swift and Heron (1967) and Swift and others (1969) interpreted the sections between Walkers Bluff and Donoho Creek Landing as follows: ; 1. The laminated sands and clays at the base of the sections at Walkers Bluff, Donoho Creek Landing, and Robinsons Landing represent an estuarine depositional environment. This lithosome is interpreted as being fluviomarine in origin; it is characteristic of tidally influenced upper-delta and estuarine environments that reflect both wave- and river-generated currents.BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. 6 2. The “Snow Hill Calcareous Member” at the top of the Cretaceous part of the Walkers Bluff section (considered as the basal beds of the Peedee Formation) represents a proximal-shelf environment, characterized by poorly sorted, medium-grained, bioturbated sands that contain horizons of coarser grained material. The coarser, gravelly horizons consist of megaclasts of tightly packed mollusk shells, shark teeth, very fine grained quartz pebbles, lignite, reptile bone fragments, and pebblesized clay clasts. 3. The Peedee Formation that forms the top of the section at Donoho Creek Landing [that is, above the unconformity recognized by Brett and Wheeler (1961) and Heron and Wheeler (1964)] represents a distal-shelf environment. This environment is characterized by very fine to fine-grained, bioturbated sands that may contain as much as 50 percent clayey silt matrix. The presence of the unconformity documented by Brett and Wheeler (1961) and Heron and Wheeler (1964) at Donoho Creek Landing suggested to Swift (1964) and Swift and Heron (1969) that the relationship among these three lithosomes, and hence the contact between the Black Creek and Peedee Formations, is a ravinement. According to this interpretation, the low-energy marsh and lagoonal sediments and the high-energy barrier-island sands, all of which were deposited at the margins of the Late Cretaceous sea, were destroyed as the sea transgressed over them. As a result, the proximal- and distal-shelf deposits of the Peedee Formation disconformably overlie the nearshore deposits of the Black Creek Formation. In support of this interpretation, Swift and Heron (1969) cited the faunal composition of the “Snow Hill Calcareous Member” as described by Stephenson (1923) from Walkers Bluff and interpreted by Brett and Wheeler (1961) as representing an open-lagoonal environment. Faunally, the “Snow Hill” contains a diverse invertebrate assemblage that reflects both lagoonal components reworked from the open lagoonal environment destroyed by the trangressing sea and components indigenous to the proximal-shelf environment of the “Snow Hill.” Hence, Swift (1964), Swift and Heron (1969), and Swift and others (1969) regarded the Black Creek and Peedee Formations as representing part of a single Late Cretaceous transgression in which the nearshore deposits of the uppermost part of the Black Creek Formation were cannibalized by the transgressing Peedee sea, resulting in a disconformity, or ravinement, between the two units. In accordance with the interpretation of the Black Creek-Peedee contact as a ravinement, the hiatus between the two units is of an insignificant duration, at least in the Cape Fear River region. SECTION DESCRIPTIONS Seven sections were measured along the Cape Fear River; three at Walkers Bluff and one each at Jessups, Donoho Creek, Robinsons, and Browns Landings. Complete descriptions of the lithologic units recognized at these exposures are presented under the heading “Measured sections” at the end of this report; diagrammatic interpretations of the sections are presented as figures 2-4 and 6-9. Because the outcrop noted by Stephenson (1923) at Deepwater Point is now under water, we could not interpret the stratigraphic relationships of the sandstones and marls present at this locality. We did, however, recover samples of both marl and sandstone from Deepwater Point; their lithologies are described in this section. A total of 23 invertebrate and 16 palynomorph samples were collected and analyzed from the seven measured sections and from Deepwater Point. The stratigraphic positions of the samples relative to the measured sections are shown in figures 2-4 and 6-9. The stratigraphic distribution of the invertebrate fossils recovered from the samples is presented in table 2, and the palynomorph data is presented in table 3. Table 2 is a complete listing of the invertebrate fauna recovered from the Walkers Bluff to Browns Landing sections. The palynomorphs listed in table 3, however, represent only those taxa we consider conspecific with those illustrated by Wolfe (1976) from the northern Atlantic Coastal Plain. Wolfe (1976) did not apply generic and specific names to his “species”; instead, he assigned each an alphanumeric code. However, his is the only pollen zonation presently available for the Campanian and Maestrichtian Stages of the Atlantic Coastal Plain. Therefore, for purposes of palynologic correlation, we find it necessary to compare the assemblages from the Cape Fear River sections with those described by Wolfe (1976) from the Raritan and Salisbury embayments of the northern Atlantic Coastal Plain. Walkers Bluff Although much of Walkers Bluff was covered with vegetation at the time of our visits (October 1979, April and August 1980, March 1981), the areas seen indicated a great deal of lateral variation in lithology and bedding character within the Cretaceous part of the section. For this reason, we made detailed measurements and observations at three localities along the face of the bluff. At the upstream end, we measured a section 100 yd from where the bluff diverges from the river. About 0.5 mi downstream from this locality was another nearly complete exposure of Cretaceous sediments,SECTION DESCRIPTIONS 7 Table 2.—Distribution of megainvertebrate fossils and vertebrate remains in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Browns Landing [+, presented but abundance not recorded; R, rare occurrence (1-3 specimens); F, few specimens recorded (4-10 specimens); C, common occurrence (11-25 specimens); A, abundant occurrence (>26 specimens)] Deepwater Walkers Bluff Point ■»T O' 3 2 © IT) 00 r*) o O' r- 3 3 IT) 90 3 3 3 O' •c 90 i & r*> r*> »*> Donoho Creek Landing Robinsons Landing Browns Landing 1 'C ir. r- vo O' »r> 90 S' r- £ ft £ 3 2 $ S s § **“««*> rn r*> Porifera: Brachiopoda: Lingula sp R Bryozoa: encrusting types branching types Coelenterata: Micrabacia sp. Mollusca (Cephalopoda): Mollusca (Bivalvia): Agerostrea falcata (Morton) Anomia argent aria Morton + R linifera Conrad + R R lintea donohuensis Stephenson olmstedi Stephenson R Aphrodina sp. R n. sp. Area hladensis (Stephenson) + F F carolinensis Conrad sp R Astarte? sp Botula sp. Caestn or hula crassiplica (Gabb) + sp. Cardium sp + R Carvoeorbula? oxvnema (Stephenson) R F R suhgibhosa Conrad? c Crassatella neusensis (Stephenson) c R + Cvmbophora cancellosa Stephenson trigonalis Stephenson sp c R R depressa Conrad gabbi Stephenson R F sp - + + + A A + A A + + R + + R + + + R R + F + + + + C + R F + F F c + c R F F F + R + R R + C + F R A C F c + + + R R R F R + F. R R F R R F ♦ F + c R + R F ♦ R R c R c + R + c F C + + R R R F c A C ' c C + + C ♦ R R R R R C + R R 8 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. TABLE 2.—Distribution of megainvertebrate fossils and vertebrate remains in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Browns Landing—Continued Deepwater Donoho Creek Robinsons Browns Walkers Bluff Point Landing Landing Landing O' 3 r*“> © tr, oc r-~ £ T- O' 00 t r~ 1 3 3 00 3 3 3 St 3 3 3 St O' F~ 3 N 00 3 3 3 r^i **> r*) t r*> Dreissena? sp.............................................. Etea? sp................................................... Exogyra cancellata Ste phe nso n .......................... costata Say ........................................... costata spinosa Stephenson ............................ ponderosa erraticostata (Stephenson) sp. Exogyra sp. ............................... Flemingosirea blackensis (Stephenson) pratti (Stephenson) subspaiulaia (Forbes) (early form) sp.............................................. sp? ...........................—....................... Glycymeris sp.............................................. Granocardium sp............................................ Idonearca sp. Inoceramus sp.............................................. Legumen carolinensis (Conrad)? concentricum Stephenson ............................... Legumen sp. ................. Leptosolen biplica Conrad ♦ Lima kerri Stephenson oxypleura (Conrad) reticulata Forbes? ................... ................ sp............................. Limposis meeki Wade? Linearia carolinensis Conrad? metastriata Conrad? sp..................................................... n. sp. A (cf. L. magnoliensis Stephenson) n. sp. B .............................................. Lineria? n. sp. Lithophaga carolinensis Conrad sp. (borings) ......................................... Lucina replevana Wade sp......................... Martesia? sp........................................ Nemodon hrevifrons (Conrad)? cf. N. hrevifrons Conrad sp................................. Nemodon? sp. + Nucula stantoni Stephenson + sp............... Nuculana kerrensis (Stephenson) cf. N. multiconcentrica (Wade) sp................................ .................... Nvmphalucina?parva (Stephenson) cf. N. parva (Stephenson) Ostrea plumosa Morton sloani Stephenson SP.................................................. + Pachycardium sp. Panopea decisa Conrad Paranomia scabra (Morton) Parmicorbula cf. P. suffalciata (Wade) Phacoides glebula (Conrad) Pholadidea? sp....... Pholadomya sp...................... Pinna sp. Postligata greenensis (Stephenson) sp..................................................... Protodonax n. sp............... Pteria? sp......................... + Radiopecten mississippiensis Conrad quinquinarius (Conrad) St abrotrigonia bartrami (Stephenson)? cf. S. bartrami (Stephenson) eufaulensis (Gabb) ............... .................... SP..................................................... + c c C F A R R F F R R F R R + R + R R R R R + R R R R k F + c C F R F R R R R R R + + + + + + R + + R F R F + ♦ c ♦ ♦ R R F R R C F R R R F F R + + + + + + F F R R + c + FSECTION DESCRIPTIONS 9 TABLE 2.—Distribution of megainvertebrate fossils and vertebrate remains in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Broums Landing—Continued Scabrotrigonia? sp................... Solyma levis Stephenson ............. sp................................. Striarca sp.......................... Striarca? sp......................... Syncyclonema Simplicius (Conrad) Tellinimeria elliptica (Conrad) stephensoni (Salisbury) cf. T. stephensoni (Salisbury) sp................................. n. sp. (cf. T. gabhi (Gardner)) Tellinimeria? sp..................... Trachycardium carolinensis (Conrad) ... donohuensis (Stephenson) .......... longstreeti (Weller) .............. vaughani (Stephenson) sp........................ Trigonarca elongata Stephenson maconensis Conrad triquetra Conrad .................. sp........ Trigonarca? sp....................... Trigonia? sp......................... Uddenia carolinensis (Conrad) ....... sp................................. Uddenia? sp.......................... “Unicardium" neusensis (Stephenson) ... Veniella conradi (Morton) ........... mullinensis Stephenson ............ Vetericardia? sp..................... Vetericardiella sp................... Mollusca (Gastropoda): Acirsa (Hemiacirsa) sp............... Amuletum? sp......................... Ampullospira? n. sp.................. Anchura sp........................... Anteglossa? sp....................... Architectonica sp.................... Ataphrus kerri Gabb ................. sp................................. Ataphrus? sp......................... Beretra sp........................... Buccinopsis globosa (Gabb) .......... sp................................. Buccinopsis? sp...................... Cerithiella cf. C. imlayi (Stephenson) cf. C. nodoliratum (Wade) ......... cf. C. semirugatum (Wade) ......... n. sp.............................. Cerithiid indet...................... Cylichna sp.......................... Euspira rectilabrum (Conrad) ........ E. sp................................ E? sp................................ Graciliala johnsoni (Stephenson)? Gyrodes sp........................... Gyrotropis kerri Gabb (plus operculum) G. squamosus Gabb G. sp............. .................. G? sp. ............. Laxispira monilifera So hi Liopeplum tarensis (Stephenson) L? sp................................ Lowenstamia liratus (Wade)? Melanatria? sp....................... Mesotoma sp.......................... Morea cancellaria (Conrad) Nacicid boreholes Napulus sp........................... Nerita sp............................ Deepwater Donoho Creek Robinsons Browns Walkers Bluff Point Landing Landing Landing O' s 1 O' ir, 00 «*■> r*) £ I ® oc 3 3 3 3 r- 3 1" 1 I rs 3 £ *r> so O' O' v© r-~ oe £ t— n in CM '© N© 00 00 oc —> CM 3 3 r+) cc fcj r*", rn rn cc> rr> C^l fC) fc> fcj + R R + F R + R F F + R R R R + R + ....J. + C A A ■f + R A + A F l F A c F R + R F F + + F R F + R R + + R R + R + F R + R R R c + R R R R F + + F + R R R + R + R R F R + + R C F + + R + R R R + R R + R R F F R R R A F C R R R R R R A R R R R F F R F 10 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. TABLE 2.—Distribution of megainvertebrate fossils and vertebrate remains in the outcropping Cretaceous sections along the Cape Fear River, N.C., between Walkers Bluff and Browns Landing—Continued Deepwater Walkers Bluff Point ■'T O' 3 <*■> rD O •d 00 r~~ o N O' t— 1 3 ID 90 3 3 3 E £ 1 £ rn fD fD £■> Donoho Creek Robinsons Browns Landing Landing Landing O' £ •D 90 fD £ r~~ id £ ID — N O' O' r~~ r~ £ so 00 00 £ S £ fD fD fD fD rD fD fD fD fD N. n. sp......................... Ornopsis sp. .................... Paladmete cf. P. gardnerae Wade Pterocerella sp. Pugnellus sp. Pyrgulifera? sp. Pyropsis sp. .................... Seila cf. 5. meeki (Wade) Trachytriion? sp. Tuba sp.......................... Turritella kerrensis (Stephenson) ... (Sohlitella) quadrilira (Johnson) (Sohlitella) trilira (Conrad) . sp............................. sp. (internal mold) ........... Vivipariidae n. gen. et sp....... Indet. neogastropods ............ Vermes: Hamulus onyx Morton walkerensis Stephenson sp............................. Serpula cretacea (Conrad) sp- ........................... Serpulid (large sized) .......... Large annulate worm tube ........ Arthropoda (Malacostraca): Crab fingers ....................... Echinodermata (Cirrepedea): Barnacle plates ................. Echinodermata (Echinoidea): Echinoid parts .................. Vertebrata: Bone fragments .................. Fish teeth ...................... Shark teeth Vertebrae .......................SECTION DESCRIPTIONS 11 Table 3.—Distribution of biostratigraphically important pollen species in the outcropping Cretaceous sections along the Cape Fear River. N.C., between Walkers Bluff and Browns Landing [The alphanumeric code assigned to each species by Wolfe (1976) is in parentheses after the binomen] Baculostephanocolpites sp. A (MPH-I) Betulaceoipollenites sp. A. (NO-3) Brevicolporites sp. B (CP3F-2) Casuarinidites sp. A (NO-2) sp. B (NO-5) ............................... Choanopollenites cf. C. conspicuus Tschudy (NA-8) sp. A (NA-3) ........................ sp. E (NA-7) ..................... Complexiopollis abdita Tschudy (NB-1) Extremipollis viva Tschudy (NJ-2) HolkopoUeniles cf. H. chemardensis Fairchild (CP3D-3) ? Holkopollenites sp. C (CP3E-1) Labrapollis sp. A (NV-1) Osculapollis aequala Tschudy (NO-1) Plicapollis usitata Tschudy (NE-3) aff. Plicapollis sp. A (NN-1) ?Plicapollis sp. C (ND-3) Proteacidites sp. A (PR-1) sp. D (PR-4) ............................ sp. G (PR-7) ............................ PseudopHcapollis endocuspa Tschudy (NC-2) serena Tschudy (NC-3) ............... sp. A(NC-l) ............................. Pseudovacuopollis involuta Tschudy (NT-1) "Retitricolpites"sp. H (C3B-3) "Retitricolpites"sp. L (C3C-3) Triatriopollenites sp. A (NP-1) sp. B (NP-2)............................. Tricolporites sp. K (CP3B-8) ............... Triporate type A (NU-1)..................... ?Trudopollis sp. A (NF-2) and Endoinfundibulapollis distincta Tschudy (NM-I) ............................. Relative frequency of abundance (in percent) of dinoflagellates and acritarchs ■ ~ Walkers Jessups £■ o Donoho Creek Bluff Landing q landing Robinsons Browns Landing Landing R2238A R2238B £ n fS r4 os o 90 Q N H os R2239A oc O' N as 1 R2411 R2240A R2240B R2240C R2240D R224IA R2241B R2241C R2242A R2242B + + + ♦ + + + + + + + + + + + + + + + + + + + + + + + + ♦ + + + ♦ + + + ♦ + + ♦ ■f + + ♦ + + + ♦ + + + + + + + + + + ♦ ♦ + + ■f + + + + + + ♦ + ♦ + + + + + + + + ♦ + + + ♦ ■f + ♦ •f + + + + + + + + + ♦ + + + + + + + + + ■f + + + ■f + + ♦ ♦ + + + ♦ + ♦ + + + + ♦ ♦ ♦ ♦ + + + + + + ♦ + + ♦ + + + •f + + ♦ + ♦ + + ♦ + ♦ + ♦ ♦ + ♦ + + + + ♦ + + + + + + ■f ♦ 38.7 10.0 28.0 23.0 20.5 20.0 15.5 13.0 35.5 73.0 58.4 30.5 72.5 72.5 37.5 69.012 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. which we refer to as the medial section. The downstream section is located 200 yd downriver from the medial section. STRATIGRAPHIC COLUMN, UPSTREAM SECTION At this exposure, we differentiated five lithologic units within the Cretaceous (units 1-5 of fig. 2) overlain by Neogene shell marls of the Waccamaw Formation of early Pleistocene age (unit 6 of fig. 2). STRATIGRAPHIC. COLUMN, MEDIAL SECTION At this exposure, which is the highest outcrop along the bluff, we recognized three Cretaceous units, over-lain by the Waccamaw Formation (units 1-4 of fig. 3). STRATIGRAPHIC COLUMN, DOWNSTREAM SECTION The Waccamaw marls and shell beds that overlie the Cretaceous deposits at the upstream and medial sections (unit 6 of fig. 2 and unit 4 of fig. 3, respectively) are absent in the main bluff face at the downstream section (fig. 4). Hence, the contact between the Cretaceous and the overlying unit, if present in the main bluff face, cannot be accurately located. At the extreme upriver end of the exposure, however, we observed one large float block of sandstone that is lithically and faun-ally like those incorporated in the basal part of the Waccamaw of the medial bluff section. At the downstream exposure, stratigraphic relationships are somewhat obscure, but there is a suggestion that the Waccamaw has been removed by channelling and that younger clastic sediment has filled the channel. This interpretation is supported by the reappearance of the Waccamaw Formation high on the heavily wooded slopes downriver, from the main downstream bluff, where the basal marl again contains transported sandstone blocks and disconformably overlies the Cretaceous sands. PALEONTOLOGY Eight of the fossiliferous horizons from the Cretaceous part of the Walkers Bluff section were sampled and analyzed for their invertebrate fauna, and four samples of the carbonaceous clay were collected for palynologi-cal analysis (see figs. 2-4 for the stratigraphic location of the paleontologic samples and tables 2 and 3 for a listing of their contained fauna and palynomorphs). In addition, analyses were made of the fossils contained in the sandstone blocks incorporated in the basal part of the Waccamaw Formation at the medial section, as preliminary field examination revealed that the blocks contain a fauna of Late Cretaceous age. Sample 31794 (U.S. Geological Survey Mesozoic Invertebrate Collection Number) was taken from a lenticular sandstone 8 to 10 ft above water level. This sandstone occurred as float on the face of the bluff, and we are uncertain as to whether it was derived from the series of discontinuous lenses of sandstone at the top of unit 1 of the downstream section or from an undetected sandstone within the laminated sands and clays of unit 1. Fossils in samples 31844, 31849, and 31850 (all from the downstream section) occur only as impressions, whereas shell material was recovered from all other samples. More than 75 species of spores and pollen were recorded from the Walkers Bluff samples. Of these, we consider only 23 to be conspecific with forms illustrated by Wolfe (1976) from the Raritan and Salisbury embayments. Dinoflagellates and acritarchs were also observed in the palynologic samples, and their abundance relative to terrestrially derived spores and pollen is presented in table 3. DISCUSSION The lithologic and paleontologic data from the Walkers Bluff sections suggest that vertical and lateral facies changes within the bluff are rapid and, in some places, abrupt, especially in the upper part of the section. The stratigraphic relationships among the three measured sections (fig. 5) were determined by walking along the face of the bluff and observing the physical relationships among the units. The only laterally persistent lithologic unit at Walkers Bluff is the laminated sand and carbonaceous clay sequence at the base of all three sections. Although the basal few feet of the bluff are covered by vegetation, scattered exposures suggest that the laminated sequence extends to, and probably below, water level. Vertical facies changes are well illustrated at the downstream section. In addition to the lithologic evidence for an abrupt change from the lower energy conditions of unit 1 to the higher energy conditions of unit 2 (and possibly unit 3), the paleontologic data provide information concerning the paleoenvironments of these units. The fauna recovered from the upper part of unit 1 (samples 31844 and 31849) is diverse; there is a dominantly infaunal assemblage of both deposit feeders (Nuculana, Tellina, Linearia) and suspension feeders (Lucina, Veniella, Cyprimeria). The presence of these elements indicates a stable substrate rich in organic material. The discontinuous sandstone lenses at the top of unit 1 contain an assemblage (sample 31843) of proportionately more epifaunal elements (Flemingostrea, Exogyra, Inoceramus) and shallow, burrowing bivalves (Scabrotrigonia, Trachycardium) than are found below,SECTION DESCRIPTIONS 13 Shell Clay Sand Medium-grained sand Medium- to coarsegrained sandstone Coarse-grained sand Wood fragments Burrows Figure 2.—Stratigraphic column of the upstream section exposed at Walkers Bluff. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left of the column. Lithologic unit numbers are to the right of the column.14 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. Sand TTJT i Burrows Figure 3.—Stratigraphic column of the medial section exposed at Walkers Bluff. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left and U.S. Geological Survey Paleobotanical Collection Numbers are to the near left of the column. Lithologic unit numbers are to the right of the column.SECTION DESCRIPTIONS 15 3. Sand, fine- to medium-grained, subangular, poorly to well-sorted, unfosslliferous, massive; a 2-foot-thick clay bed is present 10 feet from the top of the unit. 2. Sand, medium-grained, subangular, poorly sorted, fos-siliferous, crossbedded; fossils are present as impressions in clay drapes along bedding planes. 1. Sand containing laminated carbonaceous clay; discontinuous fossiliferous sandstone lenses are present at the top of the unit. METERS FEET 0—i—0 EXPLANATION Crossbedding Medium-grained sand 0 * . 0 Fine- to medium- o o • > . * . 0 • 0—■ L_ grained sand Figure 4.—Stratigraphic column of the downstream section exposed at Walkers Bluff. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left of the column. Lithologic unit numbers are to the right of the column.05 Upstream Downstream Section Figure 5.—Physical correlation of the lithologic units exposed at the downstream, medial, and upstream sections at Walkers Bluff. BLACK CREEK—PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C.SECTION DESCRIPTIONS 17 suggesting a less stable substrate and perhaps shallower water depths than those of the laminated sequence. The overlying crossbedded sands contain a less diverse, transported assemblage that, when coupled with the bedding features, confirms a vertical facies change from the low-energy, organic-rich, stable substrate environment of unit 1 to the high-energy, unstable substrate environment of unit 2. Lateral facies changes are illustrated by the fauna contained in the sands and sandstones that overlie the basal laminated sequence (samples 31848 and 31847 of the upstream section and sample 31845 of the medial section). In an upstream direction, there is an increase in the relative abundance of arcoids and cardiids, which is suggestive of increased substrate instability. The presence of Nerita (a common intertidal form) in the sandstone of the upstream section is further evidence of higher energy conditions at this locality. Jessups Landing STRATIGRAPHIC COLUMN Only two lithologic units are differentiated at Jessups Landing (units 1 and 2 of fig. 6). PALEONTOLOGY No megafossils were observed at Jessups Landing, but two samples were taken from the carbonaceous clays within the Cretaceous part of the section (that is, unit 1). The stratigraphic positions of the samples are indicated in figure 6, and the biostratigraphically important species recovered from the samples are listed in table 3. The relative abundance of dinoflagellates and acritarchs is presented in table 3. DISCUSSION We did not observe the fossiliferous marl mentioned by Stephenson (1912) as cropping out at the base of the section at Jessups Landing before construction of the lock and dam system. For this reason, we are unable to relate these marls lithologically, paleontologically, or paleoenvironmentally to any of the fossiliferous units at Walkers Bluff. As they are at Walkers Bluff, lateral and vertical facies changes from crossbedded sands to laminated clay lenses or beds are both rapid and abrupt at Jessups Landing. We did not observe any of the crossbed sets or clay beds to be continuous across the entire outcrop. Deepwater Point The fossiliferous marls and sandstones pictured by Stephenson (1912; 1923, pi. 5, fig. 13) from Deepwater Point no longer crop out; they have been covered by water that has backed up owing to the construction of Lock and Dam No. 2 on the Cape Fear River. However, we located these units on the river bottom, and, by diving, two samples of marl and one of sandstone were collected for analysis of their megafossil content. One sample of marl was also used for palynologic analysis. LITHOLOGIC DESCRIPTIONS OF THE SAMPLES Although we cannot provide a measured section for the rocks exposed along the river bottom at Deepwater Point, the lithology of the samples collected can be described as follows: Sample 31860: Silty, fine-grained, gray, fossiliferous sandstone containing pockets (burrows?) of glauconitic, carbonaceous, coarse-grained sand. Clasts of carbonaceous clay and worn phosphate nodules are present. Sample 31861: Unconsolidated, gray, sparingly glauconitic, fossiliferous, medium- to coarse-grained, angular quartz sand; some granule-sized quartz grains are present. Sample 31862: Light-gray, glauconitic, silty, fossiliferous, medium-grained sand. Small phosphate pebbles and abraded pieces of wood and bone are present. PALEONTOLOGY Samples 31860 and 31861 were analyzed only for their megafossil content; sample 31862 was analyzed for both megafossils and palynomorphs (this sample has also been assigned U.S. Geological Survey Paleobo-tanical Collection Number R2411). The megafossils present in these samples are listed in table 2, and the biostratigraphically important palynomorphs recovered from sample 31862 (R2411) are listed in table 3, where the relative abundance of dinoflagellates and acritarchs in the sample is also given. Donoho Creek Landing STRATIGRAPHIC COLUMN Approximately 30 ft of Cretaceous sediments are exposed at Donoho Creek Landing (that is, Donohue Creek Landing of Stephenson, 1912, 1923), within which we recognized seven lithologic units (units 1-7 of fig. 7). The Cretaceous units are overlain by 10 ft of alluvial(?) deposits (unit 8 of fig. 7). PALEONTOLOGY Five megafossil collections were taken from four of the fossiliferous sands that occur at Donoho Creek18 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. Sand, coarse-grained, angular to subangular, poorly sorted, massive to crossbedded; pebbles and cobbles are present at base. Sand and clay; beds and lenses of sand and carbonaceous clay interfinger and alternate throughout. METERS o- EXPLANATION Clay Sand Coarse-grained > *J sand Crossbedding 3H FEET -0 L10 Figure 6. —Stratigraphic column of the section exposed at Jessups Landing. U.S. Geological Survey Paleobotanical Collection Numbers are to the left of the column. Lithologic unit numbers are to the right of the column.SECTION DESCRIPTIONS 19 Landing, and four carbonaceous clay samples were taken for palynologic analysis. The stratigraphic positions of the samples are indicated in figure 7. The occurrences of megafossils from these collections are presented in table 2, and the biostratigraphically important pollen species recovered from the carbonaceous clays are listed in table 3, along with the relative frequency of dinoflagellates and acritarchs. DISCUSSION The megafossil assemblage from samples 31796 and 31868 of unit 2 appears to have been transported, mixed, and significantly biased by preservational (diagenetic) factors. As preserved, the molluscan ele- ments are almost entirely epifaunal suspension feeders. The most unusual aspect of these collections, however, is the abundance of bryozoa and barnacle plates. The bryozoans occur as encrusters on shells and as broken portions of branching types. Gooseneck barnacles are of occasional to common occurrence in the Gulf Coast and Western Interior chalks but have not been reported in as high an abundance from quartzose clastic sediments as we have observed them in these samples. The megafossil assemblage of sample 31795 (unit 3) has also been transported; it contains elements of bar, lower shoreface, and possibly grass-flat facies. The absence of deposit feeders reflects firm, sandy bottom conditions lacking in organic content. No marginal marine elements are present in this sample, and the Sand, medium- to coarse-grained, subangular, well-sorted, unfos-siliferous, crossbedded; very coarse grained sand and gravel is present at the base. Sand, medium-grained in a silty matrix, subangular, poorly sorted, fossiliferous, massive. Sandstone, medium- to coarse-grained, subangular, poorly sorted, fossiliferous, massive. Sand, medium to very coarse grained, subrounded, poorly sorted, fossiliferous; concentrations of quartz pebbles, bored phosphate pebbles, amber, abraded shell, and vertebrate teeth and bone common. Clay, carbonaceous; very fine grained sand partings are present throughout. Sandstone, medium- to coarse-grained, subrounded to subangular, poorly sorted, fossiliferous, planar to crossbedded; the unit is lenticular and is contained within the upper half of unit 2. Sand, medium- to coarse-grained, subrounded to subangular, poorly sorted, fossiliferous, bedding inclined in part. Laminated sand and carbonaceous clay. METERS FEET 0—|—0 3 — >—10 EXPLANATION o i Medium- to coarse- Sand grained sand ■a o\;o\°'o;c ' °-. J Medium- to coarsegrained sandstone Gravel 9 • o ' . °' Medium-grained k&%:\ Medium- to very sand P coarsegrained d sand • . **•*•' • ° • o’. . • . Clay Burrows Figure 7.—Stratigraphic column of the section exposed at Donoho Creek Landing. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left and U.S. Geological Survey Paleobotanical Collection Numbers are to the near left of the column. Lithologic unit numbers are to the right of the column.20 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. assemblage in general suggests nearshore, open marine conditions of normal salinity. The presence of only one barnacle plate in this sample, compared to the abundance of plates in the surrounding sands of unit 2, suggests a short-term increase in energy regime during deposition of unit 3; this conclusion is supported by the larger shell size in unit 3 than is found in unit 2. Robinsons Landing STRATIGRAPHIC COLUMN Here, as at Donoho Creek Landing, approximately 30 ft of Cretaceous sediments are exposed. Three lithologic units are recognized within the Cretaceous deposits (units 1-3 of fig. 8), overlain by 10 ft of alluvial(?) sediments (unit 5 of fig 8). Between the Cretaceous and alluvial(?) sediments are discontinuous lenses of shell, marl, and bored nodules (unit 5 of fig. 8) that probably belong to the Waccamaw Formation. PALEONTOLOGY At Robinsons Landing, four megafossil collections were made, and palynologic samples were taken at three horizons; the stratigraphic positions of these samples are shown in figure 8, and the megafossil and bio-stratigraphically important palynomorphs recovered from them are listed in tables 2 and 3, respectively. Dinoflagellate and acritarch abundance is also presented in table 3. Browns Landing STRATIGRAPHIC COLUMN At several places over a reach of about 150 yd along the river in the vicinity of Browns Landing, sandstone blocks litter the banks but the bluffs are overgrown with vegetation. At one point (milepost 49.2), we were able to recover a partial section by digging into a covered but near-vertical face. Three lithologic units within the Cretaceous were recognized (units 1-3 of fig. 9), but both the base and the top of the section were obscured by vegetation. PALEONTOLOGY Two collections of megafossils were made at Browns Landing. The stratigraphic positions of these samples are shown in figure 9, and the megafossils are listed in table 2. Similarly, two palynomorph samples were taken at Browns Landing; assemblages of the biostrati-graphically important species are listed in table 3, along with the relative abundance of dinoflagellates and acritarchs. The stratigraphic positions of the palynomorph samples are shown in figure 9. DISCUSSION The restricted nature of the outcrop precludes lateral tracing of the units. The intermittent distribution of the sandstone float blocks of unit 1 along the river most probably indicates a lensing unit. REMARKS Similar lithostratigraphic units are exposed at Donoho Creek, Robinsons, and Browns Landings. The lan iated sand and clay sequence at the base of the se m at Robinsons Landing (unit 1 of fig. 8) appears to be the same stratigraphic unit as the basal unit of Donoho Creek Landing (unit 1 of fig. 7). This unit, if present at Browns Landing, is covered. The sands of units 2 and 3 at Donoho Creek Landing are absent at Robinsons Landing but present (units 1 and 2) at Browns Landing, which suggests rapid lateral facies changes at this stratigraphic horizon. The Cretaceous units that overlie these variable lithologies show a greater lateral consistency throughout these sections. Present at all three sections is a thin (2-in-thick) zone of poorly sorted, relatively coarsegrained sand that contains phosphate pebbles and grains, shark teeth, vertebrate remains, rip-up clasts, and other evidence that an unconformity occurs at the base of the unit (that is, at the bases of unit 5 at Donoho Creek Landing, unit 2 at Robinsons Landing, and unit 2 at Browns Landing). The stratigraphically highest Cretaceous unit at all three sections (units 7, 3, and 3 of Donoho Creek, Robinsons, and Browns Landings, respectively) is also laterally persistent, as shown by its uniform lithologic character and contained fauna. Hence, the unconformity recognized at these localities separates lithologic units indicative of rapidly changing facies and energy conditions below from units indicative of more uniform conditions above. INTERPRETATIONS Age and Biostratigraphic Correlations Previous interpretations regarding the nature of the Black Creek-Peedee contact were based, in large part, on the ages assigned to the sections from Walkers Bluff to Browns Landing. The megafossil data provided by Stephenson (1923) and the megafossil and microfossil data provided by Brett and Wheeler (1961) were used by them and subsequent investigators as evidence that the Walkers Bluff to Browns Landing sections are synchronous or, at most, were deposited during an immeasurably short time span. The previously suggested synchroneity is based primarily on similarities in the fauna reported from Walkers Bluff and Donoho Creek Landing. However,INTERPRETATIONS 21 Sand, medium- to coarse-grained, subanguiar, well-sorted, unfossiliferous, crossbedded; very coarse grained sand and gravel are present at the base. Shell, sandy shell marl and bored nodules; the unit is found as discontinuous lenses. Sand, medium-grained in a silty matrix, subanguiar, poorly sorted, fossiliferous, massive. Sand, medium- to coarse-grained, subanguiar to subrounded, fossiliferous; phosphate grains common. Laminated sand and carbonaceous clay. R2241C 31865-------- 31864-------- R2241B 31799 -------- 31859 R2241A METERS FEET O-T-O 3 10 EXPLANATION Sand Medium-grained sand Gravel 1 Burrows ° . "o " • oJ—1—L. Shell °:;»: '!>! ii 11 ±iiii Clay Figure 8.—Stratigraphic column of the section exposed at Robinsons Landing. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left and U.S. Geological Survey Paleobotanical Collection Numbers are to the near leftof the column. Lithologic unit numbers are to the right of the column.22 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. METERS 0- FEET -0 EXPLANATION 1 • ° . .0-0 • o Fine- to mediumgrained sand ' • 0 o . ' o;o. • o o-o'; °V*?; o’ ’< Fine grained to granule-sized sandstone f'x • • ' •Ir Medium- to coarsegrained sandstone Clay 3- —10 Figure 9.—Stratigraphic column of the section exposed at Browns Landing. U.S. Geological Survey Mesozoic Invertebrate Collection Numbers are to the far left and U.S. Geological Survey Paleobotanical Collection Numbers are to the near left of the column. Lithologic unit numbers are to the right of the column. two problems that exist in the reporting of the paleonto-logic data from these localities suggest the similarities may be more apparent than real. First, the faunal lists presented by Stephenson (1912, 1923) from Donoho Creek Landing are composites; he did not distinguish the levels from which the fossils were collected. Therefore, any changes in fauna that may occur within the sections are masked. Second, the fauna reported by Stephenson (1923) and Brett and Wheeler (1961) from Walkers Bluff was derived from a sandstone that these and other workers (for example, Heron and Wheeler, 1964) considered as capping the Cretaceous section at this locality. No counterpart of the blocks is exposed in any of the sections at Walkers Bluff. With regard to the sandstone from which the Walkers Bluff fauna was collected by Stephenson (1912, 1923) and Brett and Wheeler (1961), a comparison of our measured sections with those provided by Stephenson (1912,1923), Brett and Wheeler (1961), and Heron and Wheeler (1964) indicates that the sandstone considered by these workers as capping the Cretaceous part of the section equates with the zone of sandstone blocks of Cretaceous age that we consider to have been reworked into the base of the Waccamaw Formation (see figs. 2 and 3). Evidence that these blocks of sandstone are not stratigraphically in place but are, in fact, reworked is as follows: 1. Many of the blocks are underlain by a few inches ofINTERPRETATIONS shell marl of the Waccamaw Formation. As mentioned in our description of the upstream and medial sections at Walkers Bluff, the shell marl that immediately underlies the blocks of sandstone contains phosphate pebbles, clay balls, and wood fragments, all of which indicate the presence of an unconformity below the sandstone blocks. 2. Both the upper and lower surfaces of the blocks show round-based, cylindrical (pholad?) borings that commonly are more than 1 in. long; the apertures of the borings are worn. 3. Several of the blocks we examined were encrusted on all sides by Tertiary pyncnodonts, barnacles, and colonial corals (shell material preserved) and Spon-dylus (shell material permineralized). The contrast between mineralized and original shell material suggests more than one episode of epizoan encrustation. In addition, barnacles, bryozoa, and other debris are cemented in patches on both the upper and lower surfaces. 4. The fauna contained within the blocks, although of Cretaceous age, is distinct from the fauna of any of the other collections made at Walkers Bluff (see table 2). These features indicate that the sandstone blocks have had a long and complex history that may predate their incorporation in the Waccamaw Formation. The presence of broken and worn shell, plant, and bone debris and nonarticulated bivalves together with clay casts within the sandstone suggests that the contained fauna was transported. Yet many of the bivalves are well preserved, indicating that they were derived from nearby localities. Subsequent to lithification, the blocks were exposed and bored. The blocks were then transported, during which time the apertures of the borings were eroded. At some later time, the sandstone blocks acted as a hardground attachment surface for epizoans, some of which overgrew the worn margins of the borings. Because the encrusters are present on both the upper and lower surfaces, it appears that the blocks again had been transported and (or) rolled at least once prior to their incorporation in the Waccamaw Formation. Regardless of their postdepositional history, the reworked nature of the sandstone blocks in the upper part of the Walkers Bluff section suggests that neither their stratigraphic position nor the contained fauna should be considered in any interpretation of the age of the section or in any physical correlation of the Cretaceous lithologic units exposed at Walkers Bluff with those at other localities. Therefore, the age of the in-place Cretaceous deposits at Walkers Bluff should be based only on an assessment of the in-place invertebrate fauna from samples 31844, 31849, 31843,31850, 31845, 31847, 23 and 31848, and the palynomorph samples R2238A, B, C, and D. The age of the sediments from the Cape Fear River sections as suggested by the invertebrate fauna can best be established by relating the fauna to the zonation established by Stephenson (1914, 1923) for the Upper Cretaceous Series of the Coastal Plain province. According to this zonation, two range zones are recognized in the Campanian and Maestrichtian Stages: the Exogyra ponderosa Zone below and the E. costata Zone above. A third zone, the E. cancellata Zone, is defined by the range of the nominate species and includes the lower part of the E. costata Zone. The zones and the ages assigned to them are shown in the top half of figure 10. In the bottom half of figure 10 are the ranges of several species of mollusks that Stephenson (1914, 1923), Sohl and Mello (1970), and Sohl and Smith (1980) have shown to be biostratigraphically useful and that are present in our collections from the Cape Fear River sections. A comparison of the distribution of invertebrate fossils in the Walkers Bluff to Browns Landing sections (table 2) with the ranges indicated in figure 10 suggests the following: 1. The in-place Cretaceous sediments at Walkers Bluff can be placed high in the Exogyra ponderosa Zone but not at the top of the zone. This is suggested by the concurrence of Haustator quadrilira, Pha-coides? glebula, and Flemingostrea subspatulata (early form). 2. At Deepwater Point, the concurrence of Phacoides? glebula and Flemingostrea subspatulata (early form) suggests that the sediments there, too, can be placed near, but not at, the top of the Exogyra ponderosa Zone. The first occurrence of Crenella mitch-elli and the apparent absence of Haustator quadrilira further suggests that the sediments at Deepwater Point may be younger than those exposed at Walkers Bluff. 3. The basal units that occur below the unconformity at Donoho Creek Landing (units 1-4 of fig. 7) and Browns Landing (units 1 and 2 of fig. 9) include Exogyra ponderosa erraticostata, Flemingostrea pratti, and Exogyra costata spinosa. As shown in figure 10, their concurrence suggests a biostrati-graphic equivalency with the uppermost part of the Exogyra ponderosa Zone. 4. The uppermost Cretaceous units at Donoho Creek Landing (unit 7 of fig. 7), Robinsons Landing (unit 3 of fig. 8), and Browns Landing (unit 3 of fig 9) contain Exogyra ponderosa erraticostata, E. cancellata, E. costata spinosa, and Anemia tellinoides, which suggest that these units can be assigned to the basal part of the Exogyra cancellata Zone.BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. 24 5. The sandstone blocks incorporated in the base of the Waccamaw Formation at Walkers Bluff are fau-nally similar to the units at Deepwater Point and below the unconformity at Donoho Creek and Browns Landings. This similarity indicates that the blocks at Walkers Bluff were derived from a unit situated stratigraphically above any of the in-place Cretaceous units exposed in the bluff. In support of this conclusion is the presence of Flemingostrea pratti in both the reworked sandstone blocks at Walkers Bluff and the units below the unconformity at Donoho Creek and Browns Landings. The restricted stratigraphic range of this species (see fig. 10) indicates similar ages for the units in which it is contained, an age that is younger than that suggested by the fauna of the in-place Cretaceous units at Walkers Bluff. According to the ages assigned to the invertebrate zone by Brouwers and Hazel (1978), the Walkers Bluff to Browns Landing sections are latest Campanian to earliest Maestrichtian in age and are not synchronous but chronologically sequential. Our palynologic data indicate ages that are similar to, although less precise than, those suggested by the invertebrate data for the Cape Fear River sections between Walkers Bluff and Browns Landing. The less precise nature of the palynologic data is the result of several factors, among which are the limited geographic sampling upon which the existing zonation of Wolfe (1976) was based and the lack of taxonomic descriptions of the biostratigraphically important pollen types upon which the zonation was established. Despite these drawbacks, however, broad similarities do exist between the Cape Fear River sections and the Raritan and Salisbury embayments of the northern Atlantic Coastal Plain with regard to the stratigraphic distribution of certain pollen types. Figure 11 shows the ranges of the 31 biostratigraphically important pollen species present in the Walkers Bluff to Browns Landing sections (see table 3) as they occur at the Tayloran Navarroan Campanian Maestrichtian Lower (part) Upper Lower Middle Exogyra costata Zone Exogyra ponderosa Zone Exogyra cancellata Zone ? Flemingostrea pratti £j JLUy y r w ti/owiut h^jihajou Anomia tellinoides Exogyra cancellata Exogyra costata \. Provincial Stage European Stage 2 CD CO o 3 CD c/> ■O Figure 10.—Stratigraphic ranges of 11 biostratigraphically important invertebrate fossils that are found in exposures along the Cape Fear River between Walkers Bluff and Browns Landing.INTERPRETATIONS 25 Lower Campanian Upper Campanian Maestrichtian CA-2 CA-3 CA 4 CA-5 CA-6/MA-1 A B A B A B A B Complexiopollis abdita (NB-1)_____________________ Pseudoplicapollis endocuspa (NC-2) Pseudoplicapollis sp. A (NC-1) Pseudoplicapollis serena (NC-3) Proteacidites sp. A (PR-1) IHolkopollenites sp. C (CP3E-1) IPlicapollis sp. C (ND-3) ITrudopollis sp. A and Endoinfundibulapollis distincta (NF-2 and NM-1) Osculapollis aequalu (NO-1) Choanopollenites sp. E (NA-7) "Retitricolpites" sp. L (C3C-3) aff. Plicapollis sp. A (NN-1) Extremipollis viva (NJ-2) Proteacidites sp. G (PR-7) ___________________________ Triporate type A (NU-1) Casuarinidites sp. A (NO-2) ___________________ Holkopollenites cf. H. chemardensis (CP3D-3) Baculostephanocolpites sp. A (MPH-1) _________ Triatriopollenites sp. A (NP-1) Plicapollis usitata (NE-3) Choanopollenites sp. A (NA-3) Triatriopollenites sp. B (NP-2) “Retitricolpites"sp. H (C3B-3) ____________ Proteacidites sp. D (PR-4) Tricolporites sp. K (CP3B-8) Pseudovacuopollis involuta (NT-1) _________ Choanopollenites cf. C. conspicuus (NA-8) __________ Casuarinidites sp. B (NO-5) Labrapollis sp. A (NV-1) ________ Brevicolporites sp. B (CP3F-2) Betulaceoipollenites sp. A (NO-3) Stage Pollen zone Figure 11.—Stratigraphic ranges of 31 biostratigraphically important pollen species that are found in exposures along the Cape Fear River between Walkers Bluff and Browns Landing (modified from Wolfe, 1976). The alphanumeric code assigned by Wolfe to each species is in parentheses after the binomen.26 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. stratotypes of Wolfe’s (1976) zones; the figure gives the pollen zones established by Wolfe (1976) and the ages assigned to the Campanian and Maestrichtian litho-stratigraphic units of the Raritan and Salisbury em-bayments by Brouwers and Hazel (1978). A comparison of the data presented in table 3 and figure 11 provides useful information with regard to the application of Wolfe’s (1976) zonation to the Campanian and Maestrichtian units of North Carolina. Specifically, it indicates that the ranges of some, if not all, of the guide palynomorphs must be modified. The need for extending the ranges of some species is indicated by the concurrence of species in the Cape Fear River samples that Wolfe (1976) indicated as being separated in time; most notable is the concurrence of Complexiopol-lis abdita, whose last occurrence upsection was recorded by Wolfe (1976) from zone CA-4, with Plica-pollis usitata, Pseudovacuopollis involuta, and others whose first appearance upsection was reported from subzone CA-5B. Despite the imprecise nature of the palynologic data, the Walkers Bluff to Browns Landing sections appear to be biostratigraphically situated somewhere between pollen zone CA-4 of early Campanian Age and subzone CA-5B of late Campanian to early Maestrichtian Age, a range that encompasses the ages suggested by the invertebrate data. Some palynologic events within the sections under investigation suggest possible relationships between the invertebrate and the palynologic zonations. One of these events is the last occurrence upsection of Complexiopol-lis abdita in the units below the unconformity at Donoho Creek, Robinsons, and Browns Landings. This species, along with several other species of Complexio-pollis, is present in every Cape Fear River sample examined below these beds, including that part of the section stratigraphically below Walkers Bluff. However, no representatives of the genus are present stratigraphically above their last recorded occurrence at Donoho Creek, Robinsons, and Browns Landings. Hence, the top of the range of the genus Complexiopol-lis may be a palynologic datum that corresponds to the top of the Exogyra ponderosa Zone. Similarly, Betula-ceoipollenites sp. A first appears above the unconformity at the Donoho Creek-Robinsons-Browns Landings sections. The species is present in almost all samples stratigraphically above the unconformity, and apparently it ranges to the top of the Cretaceous System of the Cape Fear River region. Its range, therefore, appears to coincide with that of Exogyra costata. In addition to suggesting possible ages for the Walkers Bluff to Browns Landing sections, both the faunal and palynomorph data can be used to correlate these sections with lithologic units in other areas of the Coastal Plain province. The biostratigraphic correlations suggested by these data for the Upper Cretaceous Series of New Jersey and the Chattahoochee River region of Alabama and Georgia are shown in figure 12. The correlation of the Cape Fear River section with the Chattahoochee River section is based on similar stratigraphic distributions of megainvertebrate fossils in both areas. Paleontologic studies in New Jersey and the Chattahoochee River region have been extensive, and we are more confident in correlating the Cape Fear River section with these areas than with others. If we are correct both in the ages assigned to the Cape Fear River sections and in the correlations indicated in figure 12, then two conclusions can be drawn regarding the unconformity that occurs at Donoho Creek, Robinsons, and Browns Landings. First, the unconformity is of limited duration, occurring sometime between late Exogyra ponderosa Zone time and early E. cancellata Zone time (that is, during the short time span of concurrence of E. ponderosa erraticostata and E. costata). Second, the unconformity is of limited geographic extent. Although we have not documented the presence of this unconformity in other areas of North and South Carolina, it appears to be geographically restricted thereto; it has not been recognized in either New Jersey or the Chattahoochee River region where sedimentation has apparently been continuous during the time represented by the unconformity in the Cape Fear River area. Paleoenvironmental Interpretations The stratigraphic, lithologic, and paleontologic data presented by us and by previous workers suggest that, in the Cape Fear River area, an abrupt paleoenvironmental change takes place across the unconformity present at the Donoho Creek-Robinsons-Browns Landings sections. Below the unconformity, the lithologic units represent rapidly changing nearshore environments, whereas a uniform, deeper water, open marine shelfal environment is suggested by the units above the unconformity. As pointed out by Swift (1964) and Swift and Heron (1967), tidal-flat deposits are represented below the unconformity by the laminated sands and carbonaceous clays at the base of Walkers Bluff (unit 1 of figs. 2-4), Donoho Creek Landing (unit 1 of fig. 7), and Robinsons Landing (unit 1 of fig. 8). This environment, which is the most frequently observed of the environments stratigraphically below the unconformity, is characterized by planar bedding and localized minor crossbedding, the presence of glauconite in the sand-dominated facies, scattered burrow structures, and relatively fine grained and well sorted sediment. In addition, theINTERPRETATIONS environment is dominated by infaunal deposit and suspension feeders (sample 31844, downstream section at Walkers Bluff), which indicate a relatively stable and organically rich substrate; such assemblages are consistent with the interpretation that this is a tidal-flat environment. In contrast to the laminated, horizontally bedded, and laterally persistent nature of the tidal-flat deposits described above, the abrupt lateral and vertical changes in the sands and clays exposed at Jessups Landing indicate an environment in which energy conditions fluctuated rapidly. There, the sands are coarser grained and entirely crossbedded; clay drapes along the bedding planes attest to rapidly changing energy conditions. These characteristics, coupled with the presence of Ophiomorpha burrows, suggest a nearshore to upper shore-face depositional environment for the sands at Jessups Landing. These sands alternate and interfinger with discontinuous, laminated, carbonaceous clay beds that contain fine-grained sand partings. The clays were deposited under relatively uniform low-energy conditions, and the entire Jessups Landing section probably represents a sequence of oscillating near- 27 shore or, as suggested by Powers (1951), fluvially dominated depositional environments. The sand units at Walkers Bluff (units 2-4 of the upstream section, fig. 2; unit 2 of the medial section, fig. 3; unit 2 and the lenticular sandstones at the top of unit 1 of the downstream section, fig. 4) and below the unconformity at Donoho Creek Landing (units 2 and 3 of fig. 7) and Browns Landing (unit 1 of fig. 9) are highly variable in their sedimentary characteristics, faunal assemblages, and trophic structure nuclei. As we have already discussed, there is a rapid lateral facies change within the sand units above the tidal-flat deposits at Walkers Bluff from lower energy, more stable substrate environments at the downstream section to higher energy, less stable substrate environments at the upstream section. Similar rapid lateral facies changes are suggested for the sand units below the unconformity at Donoho Creek, Robinsons, and Browns Landings. Channel fills are indicated by the lithology, bedding characteristics, geometry, and preservational features of the fauna of unit 3 of Donoho Creek Landing and unit 1 of Browns Landing (see figs. 7, 9). These units are discontinuous, lenticular, and crossbedded u o> to to u o> (O to _ .2 “ 8 .E a > P P E t: to a a> 0) 5 o > a E Cape Fear River section North Carolina New Jersey Chattahoochee River section Donoho Creek, Robinsons, and -Browns Landings Deepwater Point — Jessups Landing — Walkers Bluff — Peedee Formation Black Creek Formation Mount Laurel Sand Wenonah Formation Marshalltown Formation Ripley Formation Cusseta Sand Figure 12.—Biostratigraphic correlation chart of the outcropping Cretaceous sections between Walkers Bluff and Browns Landing, Cape Fear River, N.C., and the sections of New Jersey and the Chattahoochee River, Ala. and Ga.28 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. and contain phosphate pebbles and wood fragments; zones of whole shell alternate with zones of fragmented shell material, but overall clast size fines upward. These lenticular channel fills are entirely contained within other sand units (unit 2 at Donoho Creek Landing, fig. 7), which are themselves discontinuous, as evidenced by their absence at Robinsons Landing. The discontinuous nature, poor sorting, fragmented shell material, abundant barnacle plates, and domination by epifaunal suspension feeders of the contained fauna all indicate relatively high energy, nearshore environments for these confining sand bodies. Although we were unable to observe the geometry of and sedimentary structures within the sandstones that crop out in the channel of the Cape Fear River at Deepwater Point (sample 31860), their contained fauna suggest a depositional environment similar to that of the channel fill sands at Donoho Creek and Browns Landings. The assemblages are taxonomically similar, and both contain large shells and abundant barnacle plates. The sandstone blocks that have been reworked into the basal part of the Waccamaw Formation at Walkers Bluff are also indicative of nearshore, high-energy conditions. Included in the poorly sorted, coarse-grained matrix of these blocks are clay clasts, worn bone, pieces of wood, broken and worn shell debris, and disarticulated but well-preserved shells that suggest a transported assemblage derived from nearby sources. In addition, the large number of taxodont bivalves (Post-ligata, Area, Trigonarca), cardiids, ostreids (Exogyra, Flemingostrea), and intertidal neritid gastropods suggests a mixed community structure. Of particular paleoenvironmental importance is the presence of a vivi-pariid and a possible melanatriid snail in these sandstones, as they indicate that deposition occurred in proximity to fresh water, possibly a river mouth. In contrast to the rapidly changing nearshore environments discussed above, the Cretaceous units above the unconformity at Donoho Creek Landing (unit 7 of fig. 7), Robinsons Landing (unit 3 of fig. 8), and Browns Landing (unit 3 of fig. 9) suggest a deeper water, open marine shelfal depositional environment. These units are massive and highly bioturbated and contain glauconitic quartz sands in a silty clay matrix. Faunally, they are dissimilar to the fossiliferous sands below, in that venerid bivalves are common, arcoids and cardiids are rare, and oysters dominate assemblages from the upper units, as opposed to the dominance of arcoids and tellinids in assemblages from the sands below. Where the assemblage is well preserved and thus more diverse, as in unit 3 of Browns Landing (fig. 9), it is dominated by infaunal filter-feeding bivalves. The infauna includes bivalved specimens of deep-burrowing filter feeders such as Panopecu it shows no trace of sig- nificant transport. Paleoenvironmentally, the assemblages from the upper, muddy sands at Donoho Creek, Robinsons, and Browns Landings are typical of open marine shelfal deposits that were deposited below wave base. Brett and Wheeler (1961) suggested that these upper units were deposited in an inlet seaward of a barrier ridge and in water depths of 20 to 30 ft. Although it is difficult to decipher Swift’s (1964) paleoenvironmental interpretations of the lithologic units exposed at Donoho Creek, Robinsons, and Browns Landings, it appears that he agreed with the interpretations of Brett and Wheeler (1961). However, neither the sedimentary structures within these units nor their contained fauna support such conclusions. Situated between the shallow-water, nearshore, high-energy deposits below and the dark, massive, muddy, shelfal deposits above, and immediately above the unconformity at Donoho Creek, Robinsons, and Browns Landings, is a thin (2-in-thick) zone of coarse-grained sand (unit 5 of fig. 7; unit 2 of fig. 8; unit 2 of fig. 9) that represents a condensed zone of some time significance. Included in this zone are concentrations of phosphate, vertebrate remains, shark teeth, and clay clasts that had accumulated on the upper surfaces of the underlying units during the time of nondeposition represented by the unconformity. Such concentrations have been widely recognized throughout the Coastal Plain province as evidence of an unconformity (Stephenson, 1929, and many subsequent workers). The faunal assemblage of this zone (samples 31797, 31799, 31859; table 2) is similar to those of the units below and distinct from the assemblages in the units above, as the fossils were reworked into the zone from the unit below. To support our contention that a significant paleoenvironmental change takes place at the unconformity at Donoho Creek, Robinsons, and Browns Landings, we plotted the relative abundance of dinoflagellates and acritarchs in the palynologic samples examined from the sections under consideration. These data are presented in figure 13. In samples from below the unconformity, the percentage of dinoflagellates and acritarchs ranges from 10 to 40 percent and averages 25.3 percent. However, the percentage of these organisms increases dramatically in samples from above the unconformity, in which the relative frequency is more than double that in the samples below. Such a drastic change in abundance of dinoflagellates and acritarchs is further evidence that a major paleoenvironmental change takes place at the unconformity, from nearshore environments below to open shelfal environments above. Physical Stratigraphy If we are correct in asserting that the exposures between Walkers Bluff and the Donoho Creek-Robin-< oc O a) 03 -a 03 03 a_ _cp CO < o z £ o m O LLI —I < CL d d d D PERCENT DINOFLAGELLATES AND ACRITARCHS PERCENT DINOFLAGELLATES AND ACRITARCHS O! 0 10 20 30 40 50 60 70 80 PERCENT DINOFLAGELLATES AND ACRITARCHS o> c I (r2242A^ I II I II I II I II I I j^LLU-11 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80 Figure 13.—Relative frequency of occurrence (in percent) of dinoflagellates and acritarchs in palynologic samples from Walkers Bluff, Jessups Landing, Deepwater Point, Donoho Creek Landing, Robinsons Landing, and Browns Landing, Cape Fear River, N.C. Note the increase in dinoflagellates and acritarchs across the Black Creek-Peedee contact (horizontal line). to CD INTERPRETATIONS30 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. sons-Browns Landings sections are chronologically sequential and that a significant paleoenvironmental change takes place across the unconformity present at the downstream localities, then the stratigraphic relationships among the lithologic units exposed at these localities can be interpreted. With regard to the lithologic units stratigraphically below the unconformity, two lines of evidence suggest that any attempt to physically correlate any of these units from one exposure to another is futile. First, the paleontologic data indicate that the sediments at each locality were deposited at slightly different times. (In this regard, we consider the exposures at Donoho Creek, Robinsons, and Browns Landings to be one locality divided into three sections.) Secondly, because many of these units lack lateral consistency within a single exposure, it is improbable that any single unit would maintain its lithologic characteristics from one locality to the next. Even the most persistent of these units, the laminated sands and clays of tidal-flat origin, are intercalated with the less persistent sand units and are themselves chronologically repetitive. The laminated sands and clays at Walkers Bluff are directly overlain by less persistent sand units, and similar laminated sands and clays at Donoho Creek, Robinsons, and Browns Landings are apparently underlain by the sandstones and marls at Deepwater Point and those that Stephenson (1912, 1923) reported at Jessups Landing. The lithologic and faunal characteristics of the shel-fal sands above the unconformity, however, are consistent for the sections within which they occur and are distinct from those of the units below the unconformity. The lithologic units at Walkers Bluff, Jessups Landing, Deepwater Point, and the Donoho Creek Landing to Browns Landing sections below the unconformity do not show the massive bedding, high degree of bioturba-tion, and silty clay matrix that is characteristic of these shelfal sands, nor do they contain the same shelfal fauna. Hence, no evidence exists for correlating the shelfal sands of Donoho Creek, Robinsons, and Browns Landings with any lithologic unit upstream (down-section). CONCLUSIONS The ages, paleoenvironmental interpretations, and stratigraphic relationships discussed above lead us to conclude that the contact between the Black Creek and Peedee Formations is disconformable in the Cape Fear River region. The disconformity is present at Donoho Creek, Robinsons, and Browns Landings; it separates high-energy, rapidly changing, nearshore environments of the Black Creek Formation below from open marine, shelfal environments of the Peedee Formation above. Our conclusion that the Black Creek-Peedee contact is disconformable is in agreement with that of Brett and Wheeler (1961). However, their suggestion that the disconformity can be recognized for several miles along the Cape Fear River is not substantiated by our findings, as their “basal Peedee sandstone” at Walkers Bluff is apparently a series of reworked blocks of Black Creek sandstones incorporated in the base of the Wac-camaw Formation. The conclusion reached by both Stephenson (1912, 1923) and Heron and Wheeler (1964) that the contact between the Black Creek and Peedee Formations is conformable is based not only on a misinterpretation of the stratigraphic position of these reworked sandstone blocks at the top of the Walkers Bluff section but on the assumption that the fauna at Donoho Creek Landing is uniformly distributed throughout the section. Our data indicate that two distinctly different faunas occur here that are separated by the Black Creek-Peedee disconformity. Swift (1964) and Swift and others (1969) interpreted the contact between the Black Creek and Peedee Formations as a ravinement. If one accepts their definition of a ravinement as a condition where, in a “sedimentary sequence, marine sediments rest disconformably on the coastal plain sediments or on the eroded remnants of the marginal record” (Swift, 1964, p. 100), then the contact in the Donoho Creek-Robinsons-Browns Landings area is a ravinement. The basal reworked bed of the Peedee, which contains phosphate, bone, teeth, abraded fossils, and clasts of clay and sandstone reworked from the underlying beds, rests directly on a surface eroded into subjacent units of various lithologies that represent several marginal marine environments. Similar sequences have been described for Cretaceous deposits in other areas of the Coastal Plain, such as New Jersey (Owens and Sohl, 1969). The major difference between the “ravinement” unit exposed at Donoho Creek, Robinsons, and Browns Landings and similar sequences elsewhere is the thinness of the Cape Fear River unit, v which indicates an exceptionally rapid episode of transgression. Elsewhere, such basal units may be several feet thick and show graded and fining-upward character; at Donoho Creek Landing to Browns Landing, the overlying massive shelfal sand lies sharply on the cannibalistic unit. In addition, we disagree with previous workers in their placement of the “ravinement.” At Walkers Bluff, Swift (1964, fig. 62, col. 3) placed the ravinement at the base of the sandstone that is reworked into the Waccamaw Formation; thus it has no relationship to the “ravinement” at the downstream sections. The conclusions we have drawn concerning the ages, depositional environments, and stratigraphic relationships among the Walkers Bluff to Browns Landing sections are diagrammatically shown in figure 14. InCONCLUSIONS 31 this figure, the Black Creek Formation is characterized as deposits representing a series of nearshore marine environments that interfinger throughout the upper part of the formation. The Peedee Formation discon-formably overlies the Black Creek and is represented by massive, bioturbated, muddy sands of an open marine shelfal environment. Superimposed on these units in figure 14 are the approximate stratigraphic positions of the exposures at Walkers Bluff, Jessups Landing, Deepwater Point, and Donoho Creek, Robinsons, and Browns Landings. Our conclusion that the contact between the Black Creek and Peedee Formations is disconformable suggests that the Upper Cretaceous Series of North Carolina as exposed along the Cape Fear River may consist not of a single marine transgression, as suggested by Brett and Wheeler (1961), Heron and Wheeler (1964), Swift (1964), and others, but of at least two cycles of transgression and regression. This hypothesis is supported by the Santonian Age assigned to the Cape Fear Formation by Christopher and others (1979), which, when coupled with the widely recognized unconformity between the Cape Fear and Black Creek Formations (see Stephenson, 1923; Heron and Wheeler, 1964), suggests that the Upper Cretaceous Series as exposed along the Cape Fear arch consists of at least three transgressive-regressive cycles. Additional studies are needed to demonstrate the regional nature of these cycles and to determine if the Black Creek-Peedee contact described in this report represents shifting deltaic lobes or a coeval regional event. MEASURED SECTIONS Complete descriptions of the lithologic units exposed at the three measured sections at Walkers Bluff, Jessups Landing, Donoho Creek Landing, Robinsons Landing, and Browns Landing are presented below; they are intended to augment the brief descriptions that accompany the stratigraphic columns of figures 2-4 and 6-9. Walkers Bluff, upstream section Thickness Ft In Unit 1................................... 16 8 Sand and laminated clay. The sands are gray to greenish gray; iron-staining occurs along some bedding planes. The clays are carbonaceous, medium to dark gray. The sands are quartzose; mica and glauconite are each present in amounts as large as 10 percent. Comminuted plant debris is disseminated throughout, and carbonaceous matter is concentrated along some bedding planes. The sands are very fine to fine grained, well sorted, and subangular. Obvious burrows are Thickness Ft In rare, but clay laminae are disrupted along bedding planes (bioturbated), and occasional Ophiomorpha and fine Chon-drites-like burrows are present. Thallas-sinoid burrows are present in the upper 4 ft 0 in, and wood is scattered throughout the unit. The upper 4 ft 0 in is dominated by small-scale crossbedded sands; clay laminae are thin, but clay clasts occur along bedding planes. The next lower 3 ft 0 in is dominated by a “salt-and-pepper” glauconitic sand, where clay-lined burrows are present but rare. The next lower 5 ft 8 in is more highly bioturbated but has more obvious planar bedding than is found in the upper part; sand still dominates over clay. Unit 2................................... 3 0 Sand, yellow-brown to tan. The unit is dominantly quartzose, but angular to subangular feldspar clasts are common. Finely disseminated carbonaceous matter is present throughout the unit, and a few thin (1-in-thick) carbonaceous clay lenses are present. The sands are medium to mainly coarse grained, subangular, and poorly sorted; quartz pebbles as much as 0.25 in. in diameter are common. Worn and fragmented wood, bone, and shell material is present throughout, but no burrows were observed. The unit is crossbedded; the crossbedding is more apparent (that is, more highly developed?) in the upper part of the unit than it is at the base. A single 2-in-thick carbonaceous clay was observed lining the base of a channel fill near the top of the unit. The basal contact is sharp and planar. Unit 3................................... 2 0 Sandstone, yellow-brown to tan. The sandstone consists almost entirely of quartz; glauconite is present in minor amounts. The sandstone is medium to coarse grained, angular to subrounded, and poorly sorted. Shell material occurs throughout, and abraded wood, bone fragments, and clay clasts are present at the base. No burrows were observed. The unit is crossbedded to massive. The basal contact is sharp but undulating; the undulation is caused by differential cementation of the sands of units 2 and 3.▼ . ▼. .'T. Massive, glauconitic, muddy. Poorly sorted, cross-grained. • •+ -f . bioturbated sand of open-shelf. fossiliferous sand and marl rr:: VrrAv: marine origin. of near-shore, high-energy. marine origin. Laminated and bedded fine-grained sand and clay of deltaic and tidal-flat origin. Figure 14.—Schematic cross section along part of the Cape Fear River, showing relationship between the Black Creek and Peedee Formations. The numbered rectangles represent the localities discussed in this report: 1, Walkers Bluff; 2, Jessups Landing; 3, Deepwater Point; 4, Donoho Creek Landing; 5, Robinsons Landing; 6, Browns Landing. BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C.MEASURED SECTIONS 33 Thickness Ft In Unit 4.................................. 5 0 Sand, yellow-brown to tan. The unit is dominantly quartzose; glauconite is present in trace amounts. The sand is subangular, well sorted, and medium grained; no clay was observed as matrix, laminae, or lenses. The unit is unfossilif-erous. The unit is massively bedded, and the basal contact is sharp and planar. Unit 5.................................. 8 5 Laminated sand and clay. The clays are carbonaceous and dark gray to black; the sands are light gray but are commonly iron-stained yellow orange. The sands are quartzose, and mica and glauconite are present in minor amounts; comminuted plant material is disseminated throughout the sands. The sands are very fine to fine grained, subangular, and well sorted. No megafossils were observed. The upper 2 ft 3 in is dominated by sand, which occurs in beds as much as 3 in thick, and clay beds as much as 1 in thick. Clay beds thicken and clay is more abundant toward the base of the unit. The basal contact is sharp and planar. Unit 6.................................. 20 0 Shell and sandy shell marl, buff to tan to yellow-brown. Shells are present as both whole and broken specimens. The marl consists of worn and weathered pieces of shell. The sand is quartzose; finely comminuted plant material occurs throughout the unit. The quartz sands are fine to medium grained, angular to subangular, and poorly sorted. Beds of whole shell, shell marl, and marly sand alternate throughout the unit; individual beds range in thickness from 1 in to 2 ft. The basal 1-2 ft of the unit contains phosphate pebbles, clay balls, and scattered, irregularly shaped sandstone blocks as much as 1 ft in diameter. The basal 2-3 in of the unit contains carbonaceous material. The basal contact is sharp and irregular; pockets of shell debris- fill depressions cut into the underlying unit. Burrows as much as 4 in long extend down into the underlying unit and pipe down shell debris in a mushy, clayey, sand matrix. Walkers Bluff, medial section Thickness Ft In Unit 1.................................... 31 0 Sand and laminated clay. The sands are gray to greenish gray when fresh but are commonly colored orange to yellow orange by iron-staining. The clays are carbonaceous, medium to dark gray. The sands are quartzose and contain minor amounts of mica and glauconite. Carbonaceous plant material is both disseminated throughout the unit as finely comminuted particles and concentrated along some bedding planes. The sands are very fine to fine grained, angular to subangular, and well sorted. Bioturba-tion is apparent, as indicated by the presence of branching, thallassinoid-type burrows. Wood is present throughout the unit. Clay laminae are mostly planar bedded; crossbedding is present but rare. Unit 2.................................... 3 0 Sand, tan to yellow-brown. The unit consists almost entirely of quartz; finely comminuted plant material is distributed throughout. No mica or glauconite is present. The sands are fine grained, angular to subangular, and well sorted. The unit is fossiliferous and massively bedded. The basal contact is sharp and planar. Unit 3.................................... 8 0 Laminated clay and sand. The clays are carbonaceous, dark gray to black; the sands are light gray to greenish gray but are commonly iron-stained orange to yellow orange. The sands are dominantly quartzose; mica and glauconite are present in minor amounts. The sands are very fine to fine grained, angular to subangular, and well sorted. No megafossils were observed. Clay dominates the unit, but the sand content increases toward the base. The basal contact is sharp and planar. Unit 4.................................... 20 0 Shell and sandy shell marl, buff to tan to yellow-brown. Shells occur as both whole and fragmented specimens. The marl consists primarily of worn, weathered, and finely divided shell material. The sands are quartzose. Finely comminuted34 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. Thickness Ft In plant material is present throughout the unit. The sands are fine to medium grained, subangular to angular, and poorly sorted. The basal contact is sharp and irregular. The irregularities or depressions at the top of the underlying unit are shallow and are filled with as much as 2 in of a clay marl. Included in the basal clay marl are wood fragments, phosphate pebbles, and cobble-sized clay clasts. Overlying the marl is a layer of sandstone blocks that measure 6 in to 3 ft in diameter; the sandstone blocks are occasionally in point contact with the underlying unit. The long dimensions of the sandstone blocks are generally oriented horizontally, but some are oriented at low angles and are imbricate. The blocks are bored on all sides and are encrusted with mollusks, barnacles, and worm tubes. Lithologically, the sandstone is medium to coarse grained, subrounded to subangular, poorly sorted, fossiliferous, quartz sand in a silty matrix; rare granule-sized quartz grains are present, and glauconite occurs in trace amounts. Pieces of wood, bone fragments, and clay clasts as much as 0.75 in. in diameter occur sporadically in the blocks. Silty, clay-lined, circular burrows as much as 1 in. in diameter ramify the sandstone. Small (as much as 0.25 in. in diameter), branching, lined burrows are present but rare in the sandstone. Above the zone of sandstone blocks, beds dominated by large mollusks alternate with beds of small mollusks or coral beds to the top of the exposure. Walkers Bluff, downstream section Unit 1..................................... 27 0 Sand and laminated clay. The sands are gray to greenish gray but are commonly iron-stained to yellow orange; the clays are carbonaceous, medium to dark gray. The sands are quartzose and contain minor amounts of mica and glauconite; they are very fine to fine grained, subangular, and well sorted. Finely comminuted plant material is disseminated throughout the unit. Sand is dominant Thickness Ft In over clay throughout the unit and is more abundant in the uppermost 3 ft than it is in the lower part of the unit. Amber, phosphatized pebbles, and fossils (occurring only as impressions) are present in the uppermost 3 ft. Discontinuous lenses or pockets of sandstone are present at the top of the unit. The sandstones are gray and consist of fossiliferous medium- to coarse-grained quartz, sparingly glauconitic; scattered phosphate grains are also present. Unit 2.................................... 13 0 Sand, tan to yellow-brown. The sand is dominantly quartzose but contains minor amounts of mica, glauconite, and angular to subangular feldspar clasts. The sand is medium grained, subangular, and poorly sorted. The entire unit is crossbedded, and thin gray clay laminae occur both as drapes along bedding planes and as more or less continuous horizontal laminae that separate sets of crossbeds. Fossils are present as impressions within the clays. The basal contact is sharp and planar; the basal 6 in contains clay clasts, wood fragments, coarsegrained sand, and rare pebbles. Unit 3.................................... 21 0 Sand, tan to yellow-brown. The sand is quartzose and has small amounts of glauconite and clay; both glauconite and clay become abundant toward the base of the unit. Blebs of reddish sandy clay are present throughout the unit. The sand is fine to medium grained and subangular. Sorting is variable; some intervals are well sorted, and others, poorly sorted. The unit is unfossiliferous and appears to be massively bedded, although planar bedding is suggested by the presence of a few carbonaceous streaks. A 2-ft-thick, laminated, light-gray clay bed is present 10 ft from the top of the section. The basal contact of the unit cannot be accurately located due to the deep weathering of the sands of units 2 and 3; we considered the contact as occurring at the base of a 4-in-thick clay bed at which a change in slope of the outcrop face takes place.MEASURED SECTIONS 35 Jessups Landing Thickness Ft In Unit 1.................................. 12 0 Interfingering and alternating sands and clays. The sand beds are yellow brown and commonly are iron stained; the clays are carbonaceous, medium to dark gray. The sands are quartzose and contain minor amounts of glauconite and mica; the clays are silty, laminated, and often contain very fine grained micaceous sand partings. Comminuted plant material occurs both as concentrations along bedding planes and as disseminated particles throughout the unit. The sands are medium grained, subangular, and poorly sorted. No megafossils are present, but pieces of wood occur in both the sands and clays. Ophiomorpha burrows are present in the sands. The sands are crossbedded; large-scale crossbed sets as much as 2 ft thick and thin (3-inthick) clay lenses and drapes dominate the lower 6 ft, and clay beds as much as 2.5 ft thick dominate the upper 6 ft. Unit 2.................................... 4 6 Sand, yellow-orange. The sands are quartzose, angular to subangular, poorly sorted, and coarse grained. The unit is massive to crossbedded. The basal contact is sharp, and the unit has concentrations of pebbles and cobbles at the base. Donoho Creek Landing Unit 1.................................... 4 6 Laminated sands and clays; sands and clays present in approximately equal proportions. The sands are gray to greenish gray; the carbonaceous clays are dark gray. The sands are quartzose and contain minor amounts of mica. Comminuted plant material is disseminated throughout the unit. The sands are fine to medium grained, angular to subangular, and poorly sorted; clay-dominated facies contain very fine grained sand partings. A 1-ft-thick, glauconitic, massive, medium- to coarse-grained sand is present at water’s edge; the sand contains clay clasts that are piped down in burrows from the overlying clay laminae. Pieces of wood are scattered throughout the unit, but no megafossils were observed. Thickness Ft In Unit 2................................... 5 6 Sand, buff to brownish. The sand is quartzose and has minor amounts of mica; the sand is medium to coarse grained, subangular to subrounded, and poorly sorted. The unit is richly fossilif-erous; the fossils occur primarily as small, fragmented shell material that is oriented convexly up. Bedding is inclined in part. The basal contact is sharp and irregular; clay clasts at the base have apparently been reworked from the underlying unit. Unit 3___________________________________ 1 6 Sandstone. Weathered surfaces are buff to yellow brown, whereas fresh surfaces are blue gray. The sandstone is quartzose; glauconite is present but rare, and phosphate pebbles as much as 2 mm in diameter occur throughout the unit. The sand is medium to coarse grained, subrounded to subangular, and poorly sorted. Wood fragments are present throughout the sandstone but are more common toward the base. The unit is richly fossiliferous, and zones of whole larger shells alternate with zones of smaller shells and fragmented shell hash. The sandstone is lenticular, thinning both upstream and downstream, and is contained entirely within the upper half of unit 2; both its upper and its lower contacts are sharp. The unit is crossbedded at the base and grades upward to near-horizontal bedding at the top. Unit 4................................... 2 6 Clay, carbonaceous and dark gray. Very fine grained micaceous sand partings occur throughout the clay. No megafossils are present. The clay is horizontally bedded and is sandy at its base. The basal contact is sharp and planar. Unit 5................................... 0 2 Sand, gray. The sand is primarily quartzose and contains concentrations of quartz pebbles as much as 0.5 in. in diameter, phosphate pebbles, and rare pieces of amber. The phosphate pebbles are rounded and commonly bored by bivalves; one pebble is a worn coprolite.36 BLACK CREEK-PEEDEE FORMATIONAL CONTACT, CAPE FEAR RIVER REGION, N.C. Thickness Ft In Also present within the unit are tabular sandstone clasts as much as several inches in length that contain molds of mollusks and one large clast (8 in * 10 in x 2 in) that is lithically and faunally like the sandstone of unit 3. The quartz sands are medium to very coarse grained, subrounded, and poorly sorted. Shells (especially ostreids and clams), vertebrate remains, and shark teeth are common. The basal contact is sharp and planar. Unit 6.................................... 1 0 Sandstone, gray. The sand is quartzose in a silty matrix, medium to coarse grained, subangular, and poorly sorted. The sandstone is fossiliferous but contains only Anomia argentaria; it is highly bioturbated, massively bedded, and poorly cemented. The basal contact is sharp and planar. Unit 7.................................... 18 0 Sand. Weathered surfaces are gray; fresh surfaces are greenish gray. The sands are quartzose and contain minor amounts of mica and glauconite. The unit is a medium-grained sand in a silty matrix; grains are subangular, and the unit is poorly sorted. Shells are abundant, and vertebrate remains are present but rare. The unit is highly bioturbated and massively bedded. The basal contact is gradational over a vertical distance of a few inches. Unit 8.................................... 10 0 Sand, tan to yellow orange. The sand is quartzose, medium to coarse grained, subangular, and well sorted. The unit is unfossiliferous. The entire unit is cross-bedded, and its basal contact is sharp and planar. The basal 4 ft consists of very coarse grained sands and gravels. Robinsons Landing Unit 1.................................... 12 0 Laminated sands and carbonaceous clays; sands and clays present in approximately equal proportions. The sands are gray to greenish gray, and the clays are dark gray. The sands are quartzose and contain minor amounts of mica; comminuted plant material is present throughout the unit. The sands are fine Thickness Ft In to medium grained, subangular, and poorly to well sorted. Very fine grained micaceous sand partings occur within the clay laminae. No megafossils were observed, but wood fragments are scattered throughout the unit. Unit 2.................................... 0 2 Sand, dark gray. The sand is quartzose and contains minor amounts of mica, traces of glauconite, fine- to mediumgrained phosphate grains, and rare phosphate pebbles, all in a matrix of silt. The sand is medium to coarse grained, subangular to subrounded, and poorly sorted. The unit is fossiliferous; fossils occur as both well-preserved and worn shell material. The basal contact is sharp and planar. Unit 3.................................... 17 0 Sand, gray to greenish-gray. The sand is quartzose and contains minor amounts of mica and glauconite. The quartz grains are medium grained to granular, subangular, and poorly sorted and are incorporated in a silty matrix. The unit is fossiliferous, highly bioturbated, and massively bedded. The basal contact is gradational over a few inches. Unit 4.................................... 1 0 Shell and shell marl, buff to tan. The shell and marl occur in a medium- to coarse-grained quartz-sand matrix. Phosphate pebbles and nodules are present and commonly are bored. Both whole and fragmented shells are present. The unit occurs as lenticular, discontinuous pockets along the face of the bluff; the basal contact is sharp and irregular. Unit 5.................................... 10 0 Sand, yellow-orange. The sand is quartzose, medium to coarse grained, subangular, and well sorted. The unit is unfossiliferous and crossbedded. The basal contact is sharp and undulating; very coarse grained sand and gravel are present in the basal 3 ft of the unit. Browns Landing Unit 1.................................... 1 6 Sandstone, gray. The sand is quartzose and contains a minor amount of glauco-REFERENCES Thickness Ft In nite and rare mica. The quartz is medium to coarse grained and subrounded to subangular. The sandstone is fossiliferous, containing both whole and fragmented shells of invertebrates throughout and concentrations of large pieces of wood and worn bone and teeth near the base. Float blocks show the lower part of the sandstone to have inclined bedding that grades to more planar bedding above. Unit 2........................................ 0 2-4 Sand, buff. The sand is quartzose, medium to coarse grained, and contains an abundance of fine shell hash. In the upper 1 in of the unit, vertebrate remains and granule- to pebble-sized phosphate nodules are common. The base of the unit appears to be irregular, filling depressions in the underlying unit. Unit 3........................................ 0 9+ Sandstone, gray to brownish-gray. The sand is quartzose in a clayey silt matrix; it contains rare mica and scattered granule-sized, well-rounded phosphate grains. The quartz sand is poorly sorted, fine grained to granule sized, and primarily subangular. The sandstone is fossiliferous; invertebrates are abundant. The upper contact of the unit is covered. REFERENCES Blackwelder, B. W., 1979 [1980], Stratigraphic revision of lower Pleistocene marine deposits of North and South Carolina, in Sohl, N. F., and Wright, W. B., Changes in stratigraphic nomenclature by the U.S. Geological Survey, 1978: U.S. Geological Survey Bulletin 1482-A, p. A52-A61. Brett, C. E., and Wheeler, W. H„ 1961, A biostratigraphic evaluation of the Snow Hill member, Upper Cretaceous of North Carolina: Southeastern Geology, v. 3, no. 2, p. 49-132. Brouwers, E. M., and Hazel, J. E., 1978, Ostracoda and correlation of the Severn Formation (Navarroan; Maestrichtian) of Maryland: Society of Economic Paleontologists and Mineralogists, Paleontological Monograph 1, 52 p. Christopher, R. A., Owens, J. P., and Sohl, N. F., 1979, Late Cretace- 37 ous palynomorphs from the Cape Fear Formation of North Carolina: Southeastern Geology, v. 20, no. 3, p. 145-159. Heron, S. D., and Wheeler, W. H., 1964, The Cretaceous formations along the Cape Fear River, North Carolina: Guidebook, 5th Annual Field Excursion, Atlantic Coastal Plain Geological Association, October 9-10, 1964: [Durham, N.C., Duke University] 53 p. Owens, J. P., and Sohl, N. F., 1969, Shelf and deltaic paleoenviron-ments in the Cretaceous-Tertiary formations of the New Jersey Coastal Plain, in Subitzky, Seymour, ed., Geology of selected areas in New Jersey and eastern Pennsylvania and guidebooks of excursions: New Brunswick, N.J., Rutgers University Press, p. 235-278. Powers, M. C., 1951, Black Creek deposits along the Cape Fear River, North Carolina: Chapel Hill, University of North Carolina, M.S. thesis. Sohl, N. F„ and Mello, J. F., 1970, Biostratigraphic analysis, in Owens, J. P., Sohl, N. F., and Mello, J. F., Stratigraphy of the outcropping post-Magothy Upper Cretaceous formations in southern New Jersey and northern Delmarva Peninsula, Delaware and Maryland: U.S. Geological Survey Professional Paper 674, p. 28-55. Sohl, N. F., and Smith, C. C., 1980, Notes on Cretaceous biostratigraphy, in Frey, R. W„ ed., Excursions in southeastern geology, v. 2 [Guidebook]—Geological Society of America 1980 Annual Meeting, Atlanta, Georgia: Falls Church, Va„ American Geological Institute, p. 392-402. Stephenson, L. W„ 1912, The Cretaceous formations, in Clarke, W. B., and others, The Coastal Plain of North Carolina; Part I, The physiography and geology of the Coastal Plain of North Carolina: North Carolina Geological Survey [Report], v. 3, p. 73-171. ------1914, Cretaceous deposits of the eastern Gulf region and Species of Exogyra from the eastern Gulf region and the Carolinas: U.S. Geological Survey Professional Paper 81, 77 p. ------1923, The Cretaceous formations of North Carolina; part I, Invertebrate fossils of the Upper Cretaceous formations: North Carolina Geological Survey [Report], v. 5, 604 p. ------1929, Unconformities in Upper Cretaceous Series of Texas: American Association of Petroleum Geologists Bulletin, v. 13, no. 10, p. 1323-1334. Swift, D. J. P., 1964, Origin of the Cretaceous Peedee Formation of the Carolina Coastal Plain: Chapel Hill, University of North Carolina, Ph.D. dissertation, 163 p. Swift, D. J. P., and Heron, S. D., Jr., 1967, Tidal deposits in the Cretaceous of the Carolina Coastal Plain: Sedimentary Geology, v. 1, no. 3, p. 259-282. ------1969, Stratigraphy of the Carolina Cretaceous: Southeastern Geology, v. 10, no. 4, p. 201-245. Swift, D. J. P., Heron, S. D., Jr., and Dill, C. E., Jr., 1969, The Carolina Cretaceous—Petrographic reconnaissance of a graded shelf: Journal of Sedimentary Petrology, v. 39, no. 1, p. 18-33. Wolfe, J. A., 1976, Stratigraphic distribution of some pollen types from the Campanian and lower Maestrichtian rocks (Upper Cretaceous) of the Middle Atlantic States: U.S. Geological Survey Professional Paper 977, 18 p.7 DAYS rAfzT Q£ J?5> Downstream Effects of Dams on Alluvial Rivers GEOLOGICAL SURVEY PROFESSIONAL PAPER 1286Downstream Effects of Dams on Alluvial Rivers By GARNETT P. WILLIAMS and M. GORDON WOLMAN GEOLOGICAL SURVEY PROFESSIONAL PAPER 1286 UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1984UNITED STATES DEPARTMENT OF THE INTERIOR WILLIAM P. CLARK, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Williams, Garnett P. Downstream effects of dams on alluvial rivers. (Geological Survey Professional Paper; 1286) Includes bibliographical references. Supt. of Does. No.: I 19.16:1286 1. River channels. 2. Rivers—Regulation. 3. Dams. I. Wolman, M. Gordon (Markley Gordon), 1924- II. Title. III. Title: Alluvial rivers. IV. Series. TC175.W48 1983 551.48'2 82-600318 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402CONTENTS Page Abstract ....................................................... 1 Introduction ................................................... 1 Scope of study............................................. 1 Study sites and selection criteria......................... 1 Acknowledgments............................................ 2 Methods of analysis and data sources............................ 2 Water discharge ........................................... 2 Sediment load.............................................. 2 Bed and bank materials..................................... 2 Mean bed elevation......................................... 4 Measured cross sections................................ 4 Published graphs of bed-elevation changes .... 4 Gage height-discharge records (rating tables) ... 4 Channel width.............................................. 5 Time origin of channel changes............................. 5 Vegetation................................................. 5 Variability of natural channels................................. 5 Downstream effects of dams...................................... 7 Water discharges........................................... 7 Sediment loads............................................. 8 Mean bed elevation........................................ 14 General nature of changes in bed elevation ... 14 Degree of change attributable to dams................. 15 Degraded reach downstream from a dam.................. 17 General function of degradation with time at a site............................................ 17 Maximum degradation and associated time . . 22 Standardized degradation-time plot................ 23 Length of degraded reach.......................... 24 Zone of variable bed changes ......................... 26 Page Downstream effects of dams—continued Mean bed elevation—Continued Longitudinal-profile changes......................... 26 Bed material and degradation.................................. 29 Theoretical expectations.................................. 29 Variations in bed-material sizes with time at a cross section ............................................... 29 Variations in bed-material sizes with distance downstream.............................................. 31 Channel width................................................. 31 General nature of width changes........................... 31 Distance affected......................................... 35 Factors affecting changes in channel width................ 36 Alluvial-bank materials ............................. 36 Bedrock-bank controls and downstream effects . . 37 Water flow........................................... 37 Width-depth ratio ................................... 39 Time trends of channel widening at a site ............. 39 Time trends of channel narrowing at a site....... 42 Prediction of post-dam channel-width changes.............. 43 Role of a dam in effecting a change in channel width . 47 Sediment volumes removed and channel equilibrium .... 49 Vegetation.................................................... 50 Observed changes in vegetation............................ 50 Possible causes of vegetation changes..................... 52 Separating flow regulation from other factors affecting vegetation change ...................................... 54 Effects of vegetation growth ............................. 55 Conclusions................................................... 56 References.................................................... 61 Supplementary data (tables 13 and 14)......................... 65 ILLUSTRATIONS Page Figure 1. Map of conterminous United States showing location of major study sites......................................... 3 2—4. Graphs showing: 2. Variation in annual suspended-sediment loads before and after closure of Hoover Dam, Colorado River, Arizona. 14 3. Suspended-sediment loads (concentrations) transported by various discharges at successive downstream stations before and after closure of Canton Dam, North Canadian River, Oklahoma ........................... 15 4. Post-dam/pre-dam ratio of annual suspended-sediment loads versus distance downstream from Gavins Point Dam, Missouri River, South Dakota................................................................. 16 5-6. Photographs showing: 5. Time progression of bed degradation and channel armoring at the streamflow-gaging station downstream from Jemez Canyon Dam, Jemez River, New Mexico......................................................... 16 6. Degradation represented by successively lower water-intake pipes for the streamflow-gaging station 2.6 kilome- ters downstream from Fort Supply Dam, Wolf Creek, Oklahoma ............................'.......... 17 7-16. Graphs showing: 7. Changes in mean bed elevation with time at streamflow-gaging stations 48 kilometers upstream and 1.3 kilome- ters downstream from Kanopolis Dam, Smoky Hill River, Kansas...................................... 17 8. Examples of irregular rates of bed degradation with time............................................ 18 9. Representative regression curves of bed degradation with time at selected sites..................... 20 inIV CONTENTS Page Figure 10. Frequency distributions based on 111 measured cross sections on various rivers: (A) maximum expected degradation depth; (B) years needed to deepen to 95 percent of maximum depth; and (C) years needed to deepen to 50 percent of maximum depth............................................................... 23 11. Standardized degradation-time dimensionless plot of degradation curves in figure 9...................... 24 12. Longitudinal-profile changes downstream from four dams...................................................... 28 13. Variation in bed-material size with time at a site, after dam closure....................................... 30 14. Variation in bed-material diameter and bed degradation with distance downstream, at a given time after dam closure............................................................................................ 32 15. Increase in bed-material size with bed degradation downstream from three dams........................... 33 16. Variation in median bed-material diameter with distance along the Colorado River downstream from Hoover Dam, at successive times after dam closure............................................................. 33 17-22. Photographs showing: 17. Jemez River downstream from Jemez Canyon Dam, New Mexico.................................................... 34 18. Old streamflow-gaging site on Arkansas River 3 kilometers downstream from John Martin Dam, Colorado . 34 19. Wolf Creek about 2.6 kilometers downstream from Fort Supply Dam, Oklahoma....... 35 20. North Canadian River about 0.8 kilometer downstream from Canton Dam, Oklahoma........................... 35 21. Sandy bank of Missouri River about 10 kilometers downstream from Garrison Dam, North Dakota ... 36 22. Stratified sand and silt bank, Missouri River downstream from Garrison Dam, North Dakota................. 36 23-29. Graphs showing: 23. Changes in channel cross section of Sandstone Creek, Oklahoma......................................... 38 24. Examples of relative increase of channel width with time: (A) irregular rates; (B) regular rates with fitted regression curves...................................................................................... 39 25. Frequency distributions of estimated eventual increases in channel width, based on 44 measured cross sections on various rivers: (A) final values of MfJWa (B) years needed to widen to 95 percent of final Wf/W^; (C) years needed to widen to 50 percent of final W(/Wi..................................................... 43 26. Dimensionless plot of relative increase in channel width with time, for 6 representative cross sections . 43 27. Examples of relative decrease in channel width with time: (A) irregular rates; (B) regular rates with fitted regression curves.................................................................................... 44 28. Dimensionless plot of relative decrease in channel width with time, for 6 cross sections.............. 44 29. Computed versus measured values of post-dam average channel widths.................................... 46 30. Photograph showing Canadian River about 3 kilometers downstream from Ute Dam, New Mexico.......................... 47 31. Photograph showing Republican River downstream from Trenton Dam, Nebraska......................................... 48 32. Graph showing variation in cumulative net sediment volumes of channel erosion with distance downstream from Denison Dam on the Red River, Oklahoma-Texas.................................................................... 50 33-35. Photographs showing: 33. Canadian River downstream from Sanford Dam, Texas..................................................... 51 34. Washita River about 1.4 kilometers downstream from Foss Dam, Oklahoma................................. 53 35. Republican River downstream from Harlan County Dam, Nebraska, before and after dam closure ... 54 36-49. Graphs showing changes in mean streambed elevation with time at streamflow-gaging station on: 36. Colorado River 6.4 kilometers downstream from Parker Dam, Arizona ....................................... 56 37. Jemez River 1.3 kilometers downstream from Jemez Canyon Dam, New Mexico, and at the control station near Jemez 13 kilometers upstream from dam............................................................. 57 38. Missouri River 13 kilometers downstream from Fort Peck Dam, Montana...................................... 57 39. Missouri River 11 kilometers downstream from Fort Randall Dam, South Dakota............................ 57 40. Missouri River 8 kilometers downstream from Gavins Point Dam, South Dakota............................... 57 41. Smoky Hill River 1.3 kilometers downstream from Kanopolis Dam, Kansas, and at the control station at Ellsworth 48 kilometers upstream from dam ............................................................. 58 42. Republican River 2.7 kilometers downstream from Milford Dam, Kansas, and at the control station at Clay Center 49 kilometers upstream from dam................................................................. 58 43. North Canadian River 4.8 kilometers downstream from Canton Dam, Oklahoma, and at the control station near Seiling 45 kilometers upstream from dam........................................................... 58 44. Red River 4.5 kilometers downstream from Denison Dam, Oklahoma, and at the control station near Gaines- ville, Texas, 106 kilometers upstream from dam......................................................... 58 45. Neches River 0.5 kilometer downstream from Town Bluff Dam, Texas, and at the control station on Village Creek near Kountze in an adjacent drainage basin....................................................... 59 46. Chattahoochee River 4 kilometers downstream from Buford Dam, Georgia, and at the control station on the Chestatee River near Dahlonega 73 kilometers upstream from dam..................................... 59 47. Rio Grande 1.3 kilometers downstream from Caballo Dam, New Mexico........................................ 59 48. Marias River 3.2 kilometers downstream from Tiber Dam, Montana, and at the control station near Shelby 65 kilometers upstream from dam ....................................................................... 59 49. Frenchman Creek 0.3 kilometer downstream from Enders Dam, Nebraska...................................... 59CONTENTS V TABLES Page Table 1. Selected examples of rates of change of channel width in alluvial reaches.................................................... 6 2. Illustrative examples of long-term aggradation and flood deposition in alluvial reaches unaffected by manmade works . 6 3. Illustrative examples of long-term degradation and flood erosion in alluvial reaches unaffected by manmade works . 7 4. Water-discharge data for pre-dam and post-dam periods........................................................................ 10 5. Values associated with fitted degradation curves ............................................................................ 20 6. Maximum degradation downstream from various dams............................................................................. 22 7. Data on the degraded reach downstream from dams.............................................................................. 25 8. Particle-size distributions of bed and bank material, Salt Fork, Arkansas River, downstream from Great Salt Plains Dam, Oklahoma..................................................................................................... 37 9. Flow data for Sandstone Creek near Cheyenne, Oklahoma, 1951-59 .............................................................. 38 10. Values associated with hyperbolic curves fitted to changes in channel width with time, at a cross section................. 41 11. Data used to derive post-dam channel-width equation...................................................................... 47 12. Change in approximate percentages of riparian vegetation downstream from various dams ............................... 52 13. Data on channel features, as measured from resurveyed cross sections ....................................................... 67 14. Changes in streambed elevations as estimated from streamflow-gaging-station rating tables................................. 81 SYMBOLS AND DEFINITIONS LIST Ci Coefficient in empirical equations, with i = 1 to 4. d( Dominant size of bed material. r Correlation coefficient. t Time. tmax Time (years) needed to reach maximum degradation. tp Time needed for degradation to reach a designated proportion of maximum degradation. Ci Empirical coefficient in hyperbolic equation of channel change with time, with i = 1 or 2. D Bed degradation. Dmax Maximum eventual bed degradation. Qm Average daily discharge (arithmetic average of the annual mean daily flows for a f of years). Qp Arithmetic average of the annual 1-day highest averaged flows for the pre-dam period of W Channel (bankfull) width. Wt Channel width at t years after dam closure. W\ Channel width at time of dam closure: W2 Channel width as of latest resurvey after dam closure. ' DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL RIVERS By Garnett P. Williams and M. Gordon Wolman ABSTRACT This study describes changes in mean channel-bed elevation, channel width, bed-material sizes, vegetation, water discharges, and sediment loads downstream from 21 dams constructed on alluvial rivers. Most of the studied channels are in the semiarid western United States. Flood peaks generally were decreased by the dams, but in other respects the post-dam water-discharge characteristics varied from river to river. Sediment concentrations and suspended loads were decreased markedly for hundreds of kilometers downstream from dams; post-dam annual sediment loads on some rivers did not equal pre-dam loads anywhere downstream from a dam. Bed degradation varied from negligible to about 7.5 meters in the 287 cross sections studied. In general, most degradation occurred during the first decade or two after dam closure. Bed material initially coarsened as degradation proceeded, but this pattern may change during later years. Channel width can increase, decrease, or remain constant in the reach downstream from a dam. Despite major variation, changes at a cross section in streambed elevation and in channel width with time often can be described by simple hyperbolic equations. Equation coefficients need to be determined empirically. Riparian vegetation commonly increased in the reach downstream from the dams, probably because of the decrease in peak flows. INTRODUCTION Many alluvial channels are considered to be systems in equilibrium. This concept implies that the channel size, cross-sectional shape, and slope are adjusted to the quantities of sediment and water transported so that the streambed neither aggrades nor degrades. Similarly, the channel cross-sectional shape remains approximately constant. In this concept, both short-time changes (scour and fill) and long-term geologic or evolutionary changes (associated with climatic changes involving hundreds or thousands of years) are excluded. Neither the time scale nor magnitude of the changes involved in these concepts is precise. Nevertheless, the notion of adjustment and equilibrium implies that alluvial channels could be altered by significant manmade modifications, such as dams, in the regimen of water and sediment delivered. This study deals with channel changes that have taken place downstream from 21 dams on alluvial rivers. Documentation of these changes can be useful in evaluating and (or) mitigating the expected effects of dams. SCOPE OF STUDY The primary emphasis of this study is on changes in bed elevation and width of river channels after alteration of the flow regimen by closure of dams. Information availability dictated the degree of study. Evidence of changes in bed material and in vegetation is presented where the data permit. Measured water discharges and sediment loads also are discussed because of their effect on all these features. This study documents changes as they have occurred, particularly changes that have progressed for several decades. We have not been able to develop equations of sediment transport and erosion that might encompass the transient processes described, nor to produce a method of predicting the specific changes likely if a dam is built on a particular river. However, the data presented here should be useful for testing theoretical or empirical approaches. Brief discussion is devoted to the kinds of assumptions and constraints imposed on predictive models. Environmental impacts have received increasing attention during the past decade (see, for example, Turner 1971; Fraser, 1972; Gill, 1973; Sundborg, 1977; American Society of Civil Engineers, 1978; and Ward and Stanford, 1979) but will not be discussed separately here. STUDY SITES AND SELECTION CRITERIA The preferred selection criteria for a damsite and downstream reach were: 1. An alluvial bed at the time the dam was built. Gener- ally, this meant bed material in the silt-to-gravel range, as these sizes are more susceptible to erosion. 2. Monumented channel cross sections at various sites downstream from the dam, with repeat surveys (one of which was done at about the time of dam construction). 3. No significant dredging, channelization, or similar operations in the study reach. 4. No significant backwater effects from downstream dams. Data that met the above criteria were available for 21 dams (fig. 1). Most of these are in the Plains States and semiarid West. Many other dams, too numerous l2 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS to include in figure 1, also will be mentioned throughout this paper. Although resurveyed cross sections were the preferred source of data for channel changes, gage height versus discharge relations at U.S. Geological Survey streamflow-gaging stations also were used to estimate bed-level changes, if the gaging station had: (1) An erodible bed in the reach of the gage; (2) a location within about 10 km (kilometers) downstream from the dam; (3) gaging records beginning at the time of the dam closure (and preferably much earlier); and (4) a channel width that has not changed appreciably in the gaging-station reach, during the time period examined. Reaches downstream from 14 dams were found with a gaging station meeting these requirements. Eleven of these reaches have resurveyed cross sections and were among the 21 sites shown in figure 1. Most of the analysis was based on information from sites that met the criteria noted above. However, where specific information was available on bed material, special channel characteristics, sediment loads, or vegetation, this information was used to illustrate specific changes and to enlarge upon the findings. ACKNOWLEDGMENTS We are grateful to Wayne Dorough, John Turney, Kim Zahm, Ken McClung, Harry Hartwell, Cecil Cour-cier, Dave Shields, Brian Morrow, Jerry Buehre, Max Yates, Isaac Sheperdson (deceased), Joseph Caldwell (deceased), Albert Harrison, and Don Bondurant (retired) of the U.S. Army Corps of Engineers; Ernest Pemberton (retired), Willis Jones (retired), and James Blanton of the U.S. Bureau of Reclamation; Robert Wink of the U.S. Soil Conservation Service; and DeRoy Bergman, Lionel Mize, Harold Petsch, Bill Dein, Steve Blumer, Farrel Branson, Ivan Burmeister, Bob Liggett, John Borland, and John Joems (retired) of the U.S. Geological Survey, for assistance in acquiring data and photographs and in making onsite observations. William Graf of Arizona State University, Ernest Pemberton, Wayne Dorough, and H. C. Riggs reviewed the manuscript and made many helpful and constructive suggestions. We also extend our warmest thanks to many others, too numerous to mention, who helped in various ways. METHODS OF ANALYSIS AND DATA SOURCES WATER DISCHARGE Water-discharge data were available for all 21 sites from U.S. Geological Survey gaging-station records. These data were used to determine what effect the dam had on the magnitude and frequency distribution of downstream flows. Comparison of pre-dam and postdam flow records of the nearest long-term gaging station downstream from the dam indicated the overall effect of the dam on downstream flow. However, any influence of the dam needs to be separated from other factors, such as regional climatic changes and upstream operations of man. Pre-dam and post-dam flow records were examined for the nearest gaging station both upstream and downstream from the dam. The “control” station upstream from the dam reflects to a significant degree the flows that would have occurred downstream from the dam if no dam had been built. A control station is most useful located as close as possible to the dam, as long as it is not within the backwater of the dam. The flow record used for a damsite was the longest period common to both the downstream gaging station and the upstream control station. This common period sometimes was abbreviated to avoid the effects of a subsequently-built dam on the flow at one of the stations. The flow characteristics examined in this paper include average daily flow (commonly called mean annual flow), average annual flood peak, and certain flow-duration features. The average daily discharge for a given year is computed by taking the average discharge during each day, adding these for 365 consecutive days, and dividing the total by 365. We averaged these annual figures for a number of years to get a representative average daily discharge for that period. Similarly, the instantaneous annual peak discharges were averaged for the period of interest. Flow-duration values used here are the discharges equaled or exceeded 5, 50, and 95 percent of the time, where the duration curve is based on flow records for the appropriate period. These statistics represent only an approximate summary of flow characteristics and will not reveal changes in annual, seasonal, or daily mean flows. Daily variations, for example, can be large downstream from dams operated for power production. SEDIMENT LOAD Information about measured suspended-sediment loads before and after construction of dams is available for a few river reaches from U.S. Geological Survey and U.S. Army Corps of Engineers reports (some unpublished). Such data have been used in specific cases to show dam-related changes in suspended load. BED AND BANK MATERIALS Data on bed and bank materials were available for selected sites from research investigations or from pre-and post-engineering surveys for reservoir and dam112* ~1 106' 106* 104* ~r—i—‘-r- 102* ~1 96* 96' 94* 64' —r~= "7- J / / { ; L i / / NUMBER 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 EXPLANATION RIVER, DAM, AND STATE COLORADO, GLEN CANYON, ARIZONA' COLORADO, HOOVER, ARIZONA COLORADO, DAVIS, ARIZONA COLORADO, PARKER. ARIZONA JEMEZ, JEMEZ CANYON, NEW MEXICO ARKANSAS, JOHN MARTIN, COLORADO MISSOURI, FORT PECK. MONTANA MISSOURI, GARRISON, NORTH DAKOTA MISSOURI, FORT RANDALL, SOUTH DAKOTA MISSOURI, GAVINS POINT. SOUTH DAKOTA MEDICINE CREEK. MEDICINE CREEK. NEBRASKA MIDDLE LOUP, MILBURN, NEBRASKA DES MOINES. RED ROCK, IOWA SMOKY HILL, KANOPOLIS, KANSAS REPUBLICAN, MILFORD, KANSAS WOLF CREEK. FORT SUPPLY, OKLAHOMA NORTH CANADIAN, CANTON, OKLAHOMA CANADIAN, EUFAULA, OKLAHOMA RED, DENISON. OKLAHOMA-TEXAS NECHES, TOWN BLUFF, TEXAS CHATTAHOOCHEE, BUFORD, GEORGIA Figure 1.—Location of major study sites. co METHODS OF ANALYSIS AND DATA SOURCES4 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS planning and design. These data have been supplemented by samples collected by the authors. In addition, the authors made pebble counts (Wolman, 1954) of coarse particles on the beds of several rivers downstream from dams. The results of these measurements are used to illustrate some aspects of channel-and bed-material change. MEAN BED ELEVATION Mean bed elevation was determined from: (1) Measured cross sections; (2) published graphs of bed elevation at successive times after dam closure (Colorado River only); and (3) gage height-discharge relations at gaging stations. MEASURED CROSS SECTIONS The preferred method for determining mean bed elevation was based on plots of 248 resurveyed cross sections provided by the U.S. Army Corps of Engineers and U.S. Bureau of Reclamation. These 248 cross sections had been measured a total of 1,202 times. All measured cross sections were referenced to elevation above sea level (National Geodetic Vertical Datum of 1929). For each of the 1,202 cross-section surveys, we took from 15 to 30 elevation readings at equally-spaced intervals across the entire bed width, then averaged these readings for mean bed elevation. Bars and unvegetated islands, indicated in field notes, aerial photographs, and published topographic maps, were included as part of the channel-bed data. In a few cases, the edge of a bar was high enough relative to the adjacent streambed that problems arose in defining the edge of the streambed or channel. The surveyors had the same difficulty. The four chief sources of error or variability in determining mean bed elevation from plotted cross sections are: (1) Locations of placement of the stadia rod; (2) natural changes in the bed configuration with time; (3) recognition of the bed as opposed to the bank on the plotted cross section; and (4) operator error in choosing and averaging many bed elevations to get a mean value. Error due to location of the stadia rod can be assumed to be minor. Bed configurations do change with time, quite apart from scour and fill, because of passage of bedforms and redistribution of sediment. River surveys normally are conducted during low flow (wading conditions). Resurveys associated with the passage of a flood on the Colorado River near Lees Ferry, Arizona, showed about 2 m (meters) of change in mean bed level (Leopold and others, 1964, p. 228); low-flow resurveys of the present study undoubtedly involve changes considerably less than this. Exact error from changes in bed configuration with time is unknown. Recognition of the streambed and banks on plotted cross sections was facilitated by the original notes of surveyors. Operator error was considered by comparing two operator’s determination of the average of many elevations across the bed; differences of 0. to 0.4 m appeared, which is not a geomorphically significant error. Because mean bed elevations naturally fluctuate with time at any alluvial cross section, fluctuations of less than about 0.1 or 0.2 m were considered insignificant in this study. Significance of a measured absolute change in bed elevation depends not only on measuring precision but also on the scatter in elevations, the rate of change of elevation with time, and the period of record. For example, for the magnitudes of changes occurring at one cross section downstream from Fort Peck Dam on the Missouri River, Wolman (1967, p. 90) estimated that about 10 years of record would be needed to reliably show a degradation rate of 0.08 m/yr (meter per year), and 30 years would be needed to show a degradation rate of about 0.01 m/yr. These values will vary from site to site. PUBLISHED GRAPHS OF BED-ELEVATION CHANGES For an additional 39 resurveyed cross sections, downstream from Hoover, Davis, and Parker Dams on the Colorado River, U.S. Bureau of Reclamation reports for various years provide graphs of mean bed elevation versus time. We read bed elevations for selected times directly from the plotted curves, for all sections downstream from these dams. The authors of those reports derived the curves from measured cross sections by: (1) Planimetering the cross-sectional area below an arbitrarily chosen low bank-to-bank horizontal baseline, the elevation of which is constant with time for each site; (2) dividing this area by the baseline width (which stayed virtually constant with time), and (3) subtracting the mean depth thus obtained from the elevation of the baseline. In almost every case, no islands were present at the cross sections. The 39 cross sections in this category had been measured a total of 615 times. The total number of resurveyed cross sections for the study thus was 287, and these had been measured a total of 1,817 times. On the average, then, each cross section in the study was measured about 6 times, at intervals ranging from about 1 to many years. GAGE HEIGHT-DISCHARGE RECORDS (RATING TABLES) Within the criteria listed earlier, the gage height corresponding to an arbitrarily chosen discharge is approximately proportional to the bed level. Lowering of such a gage height with time would indicate lowering of the streambed. For the reference discharge, a low flow is better than a high one, because the low-flow part of the gage height-discharge relation is more sensitive to changes in bed level and is better defined than the high-METHODS OF ANALYSIS AND DATA SOURCES 5 flow part of the relation. (The elevation corresponding to zero discharge probably would be best, but it is not defined for many gaging stations.) Where possible, we used the discharge exceeded 95 percent of the time as the reference discharge. Where this discharge was not defined on a significant number of rating tables, the lowest discharge common to most of the tables was used. Although this method can show general trends in bed elevation, it is not as accurate as measured cross sections. Water-surface elevations can be affected by changes in channel shape, channel roughness, and downstream features, even where width has remained approximately constant. Where the rating-table method was used, a control station upstream from the dam, if available, also was examined. Control stations usually were located more than 10 km upstream from the dam in an attempt to avoid any effects of the reservoir. CHANNEL WIDTH Channel (banktop) width was measured directly from plotted cross sections. The survey notes in some cases were used to help define the banks. Defining the banks usually was not difficult. Regulation of discharge by several dams reduced the channel-forming flows to such an extent that the postdam channel became narrower. The new banks, as well as the original banks, then appeared on a plotted cross section. In such instances, we measured the width between the newer banks, even though occasional flow releases could overtop those banks. TIME ORIGIN OF CHANNEL CHANGES As a preliminary step to constructing a dam across a channel, major or minor rearrangement of the stream and its channel usually is made. Thus, the normal movement of sediment and water are interfered with from the early stages of construction. Such interference can cause channel changes downstream. The extent of these changes will vary from one dam to another, according to the nature and rate of progress on the project. Several years usually are needed to complete construction and officially close a dam. Furthermore, storage in the reservoir generally begins before the dam is closed officially. The date of dam closure, therefore, may represent a rather belated time from which to date channel changes. A more logical date might be the date construction began. However, channel cross sections generally were not established at such an early stage. The available original cross-sectional measurements were made at times ranging from several years prior to the beginning of construction to a year or two after the dam was closed. The year of dam closure is used as the reference date in this study because it is the only date commonly available to all sites. A cross-sectional measurement made no later than about 1 year after dam closure usually was accepted as representative of the channel at the time of dam closure. VEGETATION Analysis of vegetation changes in this study is limited to a gross quantitative approach, with little attention to individual plant types. Differences in vegetation cover for a number of study sites were determined in one or more of three ways: (1) Onsite mapping; (2) successive aerial photographs; and (3) successive ground photographs. Onsite, the simple method used consisted of comparing exposed areas of channel bars and islands clearly discernible on earlier aerial photographs with existing stands of vegetation in the same reaches. Mapping was confined to the channel itself and did not include the entire valley bottom. VARIABILITY OF NATURAL CHANNELS To evaluate the effect of manmade alterations on natural environment, the natural variability of an environment needs to be considered. A few observations of the characteristics and changes in alluvial rivers virtually unaffected by manmade structures are reviewed briefly here to provide a reference for subsequent analyses of apparent changes associated with dams. Two kinds of variability are involved in any analysis of channel changes. First, at any time a channel’s width, depth, and slope vary in space. For example, although the mean width of the Missouri River downstream from Garrison Dam in North Dakota in 1957 was 415 m for a reach 87 km long, the standard deviation of 24 measurements was approximately 122 m or 29 percent. The actual width ranged from a minimum of 255 m to a maximum of 845 m. This variability also shows that, in comparing present and past widths of the channels, a change needs to be demonstrable statistically and, thus, outside of the range of natural variability in any one set of measurements. The second, more complex type of variability occurs with time at a given river cross section. Some selected, representative data from the literature on naturally-occurring changes in channel width and bed elevation are summarized in tables 1-3. These changes can be large. For example, within several weeks the Yellow River of China at any one spot may widen by as much as hundreds of meters (Chien, 1961). Another of the world’s largest and most sediment-laden rivers, the Brahmaputra in India, also has extreme changes in width with time (Coleman, 1969). The rates of change range from a few meters to hundreds of meters per6 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 1.—Selected examples of rates of change of channel width in alluvial reaches [ +, increase; decrease; USA, United States of America] River, location Approximate width of initial channel (meters) Flood changes Long-term changes Time Change in width (meters) (percent) Rate of change Years of ----------------------- observation (meters (percent per per year) year) Reference Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Brahmaputra River, Bangladesh Katjuri River, India Gila River, USA Gila River, USA Gila River, USA Gila River, USA Gila River, USA Rio Salado, USA Rio Salado, USA Cimarron River, USA Cimarron River, USA Cimarron River, USA Cimarron River, USA Cimarron River, USA Red River, USA (average of 20 sites) Red River, USA (average of 20 sites) Patuxent River, USA Patuxent River, USA Trinity River, USA (many sites) 6.700 11,800 12,200 12,600 11,000 6.700 3,100 7,400 — — 588 — — — 98 225 — — — 88 — — — 174 4.0 14.9 18 — — 427 — — 198 20 884 1,200 1,070 -- 26 hours +6.2 — — About 2 0 0 days 105 — 0 to +45 0 to 8 +70 +1.0 8 -65 -.55 11 0 0 8 +118 +.93 8 +98 +.89 8 +15 +.22 8 -42 -1.4 133 +36 +.49 100 +16 — 21 -21 -3.5 9 +12.5 +12.8 1 -64 -28 1 +85 +97 1 -73 -42 36 +4.3 +108 36 +4.2 +28 25 +16.4 +91 15 -15.3 -3.6 6 0 0 25 +34.6 +173 15 -42.7 -4.8 16 -25 -2 4 +33 +3 Coleman, 1969, p. 161. Coleman, 1969, p. 161. Coleman, 1969, p. 161. Coleman, 1969, p. 161. Coleman, 1969, p. 161. Coleman, 1969, p. 161. Coleman, 1969, p. 161. Latif, 1969, p. 1689. Inglis, 1949, p. 67. Burkharo, 1972, p. 5. Burkham, 1972, p. 5. Burkham, 1972, p. 5. Burkham, 1972, p. 5. Burkham, 1972, p. 5. Bryan, 1927, p. 18. Bryan, 1927, p. 18. Schumm and Lichty, 1963, P- 73-74 Schumm and Lichty, 1963, P- 73-74 Schumm and Lichty, 1963, P* 73-74 Schumm and Lichty, 1963, P- 73-74 Schumm and Lichty, 1963, P* 73-74 Schumm and Lichty, 1963, P* 86. Schumm and Lichty, 1963, P- 86. Gupta and Fox, 1974, p. 503. Gupta and Fox, 1974, p. 503. Ritter, 1968, p. 17-52. Table 2.—Illustrative examples of long-term aggradation and flood deposition in alluvial reaches unaffected by manmade works [USA, United States of America; USSR, Union of Soviet Socialist Republics] Flood deposition Long-term aggradation River, location Time Depth (meters) Years of observation Rate (meters per year) Reference Colorado River, USA — — 31 0.03 Cory, 1913, p. 1212. Yellow River, China — — — .03 Todd and Eliassen, 1940, p. 446. Alexandra-North Saskatchewan River, Canada — — 358-2,400 .0007-.003 Smith, 1972, p. 182. Kodori River, USSR — — 32 .03 Mandych and Chalov, 1970, p. 35. Last Day Gully, USA — — 11 .006 Emmett, 1974, p. 58. Arroyo de Los Frijoles, USA — — 6 .01 Leopold and others, 1966, p. 219. Nile River, Egypt — — 1,900-2,800 .00096-.0016 Lyons, 1906, p. 315. Mu Kwa River, Formosa — — 3 4 Lane, 1955, p. 745-747. Brahmaputra River, Bangladesh 6 months 0.6-8 — — Coleman, 1969, p. 178. Rio Guacalate, Guatemala 2 hours 1 — — Foley and others, 1978, p. 114. James River tributaries, USA Several hours 0-1.5 — — Williams and Guy, 1973, p. 42. Van Duzen River, USA About 3 days .3-3 — — Kelsey, 1977, p. 284-301. Little Larrabee Creek, USA About 3 days 2.4 “ " Kelsey, 1977, p. 284-301. Trinity River tributaries, USA — 0-3.4 — — Ritter, 1968, p. 53-54. Waiho River, New Zealand — 3-24 — — Gage, 1970, p. 621. Centre Creek, New Zealand 8 months .44-.55 — — O'Loughlin, 1969, p. 697. year; however, most of the changes are less than 1 per- width more significantly in arid climates than in humid cent of the channel width per year. The channel width of the Cimarron River in Kansas fluctuated significantly from 1874 to 1954 (Schumm and Lichty, 1963). Wolman and Gerson (1978) suggested that floods affect river climates, but the magnitudes are not well-defined. A few instances are noted in table 1 for rivers comparable to those included in the present study. Natural bed aggradation measured during manyDOWNSTREAM EFFECTS OF DAMS 7 Table 3.—Illustrative examples of long-term degradation and flood erosion in alluvial reaches unaffected by manmade works [USA, United States of America; USSR, Union of Soviet Socialist Republics] Flood erosion Long-term degradation River, location Time Depth (meters) Years of observation Rate (meters per year) Reference Castaic Creek, USA i/ioo 0.01 Lustig, 1965, p. 8. Red Creek tributary, USA — — about 815 .004 LaMarche, 1966, p. 83. Beatton River, Canada— — — 250 .011 Hickin and Nanson, 1975, p. 490. Lena River, USSR^ 20 .0005 Borsuk and Chalov, 1973, p. 461. Klaralven River, Sweden — — about 7,000 .007 de Geer, 1910, p. 161. Trinity River and tributaries, USA — 0-0.5 — — Ritter, 1968, p. 54. Wills Cove, USA Several hours 1/0-3 — ” Williams and Guy, 1973, P. 35. Yellow River, China About 12 hours 5-9 ~ Todd and Eliassen, 1940, p. 376. Pickens Creek, USA TGbout 1 hour 0-6 — — Troxell and Peterson, 1937, p. 93. Centre Creek, New Zealand 8 months .2-.5 — — O'Loughlin, 1969, p. 697 Klaralven River, Sweden About 1 4.7 — — de Geer, 1910, p. 174. month —^Estimated. 2/ — Classification uncertain. years can be very small (table 2). Examples are 0.0007 to 0.0034 m/yr, or 1 m every 290 to 1,430 years (Alexandra-North Saskatchewan River, Canada) and about 0.001 m/yr, or 1 m every 1,000 years, for the Nile River near Aswan in Egypt. Values of about 0.03 m/yr (1 m about every 30 years) have been given for the Colorado River in the United States, the Yellow River in China, and the Kodori River in the Soviet Union. The most rapid reported rate is about 4 m/yr for the Mu Kwa River in Formosa, where sediment from landslides during 3 years raised the streambed about 12 m. In contrast to long-term average rates, bed aggradation during floods can be enormous. Some observed maximum depths of fill for a single flood are about 8 m on the Brahmaputra River and 24 m in the Waiho River in New Zealand (table 2). Depths of 1 to 3 m are common for the cases reported in the literature. Reported measurements of long-term natural degradation (table 3) range from about 0.0005 to 0.011 m/yr. For example, Borsuk and Chalov (1973) gave an averaged bed lowering of 0.0005 m/yr during 20 years for the Lena River, Soviet Union. LaMarche (1966) used vegetation to estimate an average of 0.004 m/yr during 815 years for a small channel in Utah. The longest period examined seems to be 7,000 years by de Geer (1910), who counted varved clays and estimated an average bed degradation of 0.007 m/yr for the River Klaralven, Sweden. Hickin and Nanson (1975) reported an average degradation rate of 0.011 m/yr for the Beat-ton River, Canada, for 250 years. During floods, streambeds in southern California in a-matter of hours have eroded as much as 6 m (Troxell and Peterson, 1937), and the Yellow River in China has degraded by as much as 9 m (Todd and Eliassen, 1940) (table 3). In some cases, the bed refills during the waning stages of the flood; in others, the bed refills during a number of years, and along some reaches the channel seems to be changed permanently. Some data used in this study of channels downstream from dams may not demonstrate a cause and effect relation; instead, they may show a sequential or natural change. Cause and effect in certain cases needs to be inferred from the timing of the changes and from their nature and persistence; such proof can be demonstrated only occasionally. Commonly, the precise magnitude of the changes and the separation of manmade causes from those changes associated with climate and other natural phenomena may be difficult, as discussed below. DOWNSTREAM EFFECTS OF DAMS WATER DISCHARGES A number of papers in recent years (for example, Lauterbach and Leder, 1969; Moore, 1969; Huggins and Griek, 1974; DeCoursey, 1975; Petts and Lewin, 1979; and Schoof and others, 1980) have discussed the effects of dams on downstream flows. Because of the various purposes for which dams are built, there are large variations from one dam to another in the magnitude and duration of flow releases. At some dams (for example, Sanford Dam on the Canadian River, Texas, and Conchas Dam on the Canadian River in New Mexico), all8 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS or almost all the water is withheld from the downstream reach. Only drainage through the dam, tributary inflow, springs, ground water, and other downstream sources provide water downstream from the dam. At other dams, water is released only a few times per year. Discharge at hydropower dams may be stopped or curtailed for part of a day, and then a relatively large flow released during another part of the day (Fort Peck Dam on the Missouri River, Montana). At diversion dams, such as Milburn Dam on the Middle Loup River in Nebraska, large quantities of water may be diverted during the irrigation season, but all flows (and some sediment) may be passed directly through during the rest of the year. Even at dams built solely for irrigation, water may be released in a variety of patterns. At one extreme, virtually no water is ever released, and all irrigation diversions are made directly from the reservoir (Sanford Dam, Canadian River, Texas). Near the other extreme, practically no water is released during the winter storage period, but relatively large flows are released steadily during the irrigation season, with irrigation diversions made from various points downstream (Caballo Dam on the Rio Grande, New Mexico). Each dam, because of its purposes and the arrival of floods from upstream, has a unique history of daily, seasonal, and annual flow releases. Whatever the pattern of controlled releases, they are almost certain to be distributed differently from the natural flows. The uniqueness of release policy at each dam precludes simple generalizations about the discharge distributions, except that flood peaks will be decreased (table 4). For the 29 dams of table 4, average annual peak discharges were decreased to 3 to 91 percent of their pre-dam values (averaging 39 percent). The flow exceeded only 5 percent of the time was reduced in many (but not all) cases. High flows may be important, especially in controlling channel size and vegetation. Average daily discharge in a reach may increase, remain the same, or decrease after a dam has been built (table 4). Low flows (equaled or exceeded 95 percent of the time) also were diminished in some instances and increased in others. Judging from the records at the control stations (table 4), some, and possibly all, of the changes in average daily flow (but not necessarily in other flow statistics) at a number stations in our sample would have occurred in the absence of regulation. Changes in climate, ground-water withdrawals, flow diversions, vegetation, or combinations of these factors could have been the causes. SEDIMENT LOADS In addition to changing the flow regimen, dams are effective sediment traps. The curtailment of sediment supply, as with the change in water discharge, could have an important effect on the downstream channel. With some dams, such as those built mainly for hydro-power generation, the sediment may be trapped as an incidental consequence of the dam’s overall structure and operation. On other dams, sediment control may be a specific intent or purpose in building the dam. For example, Cochiti, Abiquiu, Jemez Canyon, and Galisteo Dams have been built on the Rio Grande and its major tributaries in an effort to reduce or eliminate aggradation on the Rio Grande. A dam’s role in trapping sediment can be shown by periodic reservoir surveys, by sediment-transport measurements, or by both. Sediment-transport measurements generally are given either as sediment concentrations (weight of sediment per unit volume of water-sediment mixture) or as annual sediment loads, in tons per year. Hoover Dam on the Colorado River is a good example. Suspended loads in the Colorado River have been measured upstream and downstream from Hoover Dam. The upstream station is near Grand Canyon, Arizona, 430 km from the dam; the downstream station is near Topock, 180 km downstream from the dam. Two characteristics of the suspended load under natural conditions— the large quantities and the very large annual variations—are shown in figure 2. Before closure of Hoover Dam in 1936, annual loads at the two stations were similar. After closure, sediment inflow, represented by the data for the Grand Canyon station, continued to be large and variable. Downstream from the dam, at Topock, however, both the load and the annual variations were markedly decreased. Data for several other dams also indicate a significant decrease in sediment load. For Glen Canyon Dam on the Colorado River (U.S. Bureau of Reclamation, 1976) the average annual pre- and post-dam suspended-sediment loads, as measured 150 km downstream at Grand Canyon, are as follows: Pre-dam (1926-62), 126 million megagrams; post-dam (1963-72), 17 million megagrams. This is a reduction of about 87 percent. On the Missouri River at Bismarck, North Dakota, 121 km downstream from Garrison Dam, sediment loads during 1949-52 averaged 48.6 million megagrams per year. The dam closed in 1953. During 1955, the sediment was 9.8 million megagrams, and during 1959, it was only 5.3 million megagrams. At Yankton, South Dakota, 7 km downstream from Gavins Point Dam, which began storing water in 1955, the Missouri River’s pre-dam annual sediment load was about 121 million megagrams. The load then diminished to 8.1 million megagrams during 1955 and was only 1.5 million megagrams during 1960. Data for the above examples may not reflect accurately the actual trap efficiency, because the measuringDOWNSTREAM EFFECTS OF DAMS 9 stations are a considerable distance downstream from the dam. The entrance of major tributaries, the erosion of sediment from the bed and banks immediately downstream from the dam, and various other factors (Howard and Dolan, 1981) can affect the apparent trends. Measurements made at or just downstream from the dam are much more suitable for an indication of trap efficiency. Such measurements show that the trap efficiency of large reservoirs commonly is greater than 99 percent. For example, during the first 19 years after closure of Canton Dam on the North Canadian River in Oklahoma, a total of 20.5 million megagrams of sediment arrived in the reservoir, and only 0.11 million megagrams went past the outlet works of the dam (U.S. Army Corps of Engineers, 1972, p. 6-8). The dam, therefore, trapped about 99.5 percent of the total sediment load. The trap efficiency of Denison Dam on the Red River, Oklahoma-Texas, during the first 12 years after closure was 99.2 percent (U.S. Army Corps of Engineers, 1960, p. 11). These examples illustrate the efficiency of dams that do not sluice appreciable volumes of sediment through the dam. Many diversion dams and some sediment-storage dams, however, are built and operated to permit sediment to be flushed out of the reservoir. For example, Milburn Dam on the Middle Loup River and other irrigation-type diversion dams such are those on the Rio Grande and Imperial Dam on the Colorado River are designed for flushing sediment either continuously or periodically through the dam to the downstream channel. Less commonly, a reservoir is emptied approximately once per year, such as at John Martin Dam on the Arkansas River in Colorado. The entire reservoir water storage at John Martin Dam typically has been released each spring during the irrigation season. The escaping water carves a channel in the stored sediment and transports sediment out with it. From 1943 to 1972, the annual trap efficiencies at this dam varied randomly between 0 and 99 percent (U.S. Army Corps of Engineers, 1973). At John Martin Dam and at similar dams, annual trap efficiency (sediment storage) can vary with: (1) Volume of water stored during the winter and released (mainly a function of rainfall); (2) volume of sediment entering the reservoir since the previous year’s release; (3) rate at which the reservoir release in made; (4) bottom topography of the pool (deep versus relatively shallow); (5) type and location of outlet gates; and (6) sizes of sediment particles (coarse versus very fine) entering the reservoir. After dam closure, the downstream sediment loads at a particular site do not appear to recover from their greatly decreased values. Data on pre- and post-dam annual suspended loads were available for five stations downstream from Gavins Point Dam on the Missouri River (U.S Army Corps of Engineers, unpublished data, various years). This dam does not sluice appreciable quantities of sediment. Further, several major reservoirs were built on the Missouri River upstream from Gavins Point Dam at about the same time that Gavins Point Dam was constructed. The post-dam annual loads for a given station downstream from Gavins Point Dam were relatively small and showed no significant change with time, for the 1 to 3 decades after dam closure for which data are available. Instead, the loads only fluctuate within the same relatively narrow range from year to year (as for the Colorado River downstream from Hoover Dam, mentioned above). Similarly, data from various sources show that sediment concentrations for a given discharge at four sites downstream from Canton Dam also have not changed significantly with time for as long as 3 decades (the period of record) after dam closure. What river distance downstream from a dam is required for a river to recover to its normal pre-dam or upstream-from-the-dam sediment loads or concentrations? Sediment in the channel bed and banks and in tributary inflows are major factors in determining the length of channel needed. This distance for the North Canadian River downstream from Canton Dam is illustrated in figure 3. Upstream from the dam, at Seiling, Oklahoma, a given discharge transported about the same volume of sediment before and after the 1948 dam closure. Reduction in concentration 5 km downstream from the dam is dramatic. A significant post-dam decrease still is quite noticeable 140 km downstream from the dam. Even at Oklahoma City, 182 km downstream from the dam, sediment concentration for a given discharge is not as much as it was prior to dam construction. Finally at Wetumka, 499 km downstream from the dam, with a drainage area some 4,640 km2 (square kilometers) larger than that at the dam, sediment concentrations have recovered and may even be greater at high flows. Thus, the river required more than 182 km, and possibly as much as about 500 km, of channel distance for bed and bank erosion, coupled with tributary inflows, to provide sediment concentrations equivalent to those transported in the same reach at a given water discharge prior to closure of Canton Dam. Curves similar to those in figure 3 for the Red River downstream from Denison Dam (U.S. Army Corps of Engineers, 1960, plates 64 and 65) indicate the same order of magnitude of channel distance (or possibly even a longer required reach) for recovery of pre-dam sediment concentrations. At Arthur City, Texas, 150 km downstream from the dam, post-dam sediment concentrations for the 17 years after dam closure were only about 20 to 55 percent of the pre-dam concentrations for the same water discharge. At Index, Arkansas, 38710 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 4.—Water-discharge data [km, kilometers; m3/s, cubic Dam Year number River, dam, State of dam (fig. 1) closure Downstream gaging station and control station^ River distance of station 2/ from dam— (km) Period used (water years) Pre-dam Post-dam 1. Colorado, Glen Canyon, 1963 Colorado River at Lees Ferry, Arizona 26 1922-62 1963-78 Arizona C: No suitable station 2. Colorado, Hoover, Arizona 1965 Colorado River near Topock, Arizona 180 1923-34 1935-49 C: Colorado River near Grand Canyon, Arizona 430 1923-34 1935-49 3. Colorado, Davis, Arizona 1950 Colorado River near Topock, Arizona 72 1935-49^ 1950-78 C: Colorado River below Hoover Dam, Arizona 108 1935-M-^ 1950-7&2/ 4. Colorado, Parker, Arizona 1938 Colorado River below Parker Dam, Arizona- 0-6.4 1936-37-/ 1938-78 California C: Colorado River near Topock, Arizona 63 1936-37-/ 1938-782/ 5. Jemez, Jemez Canyon, 1953 Jemez River below Jemez Canyon Dam, 1.3 1937;1944-52 1953-78 New Mexico New Mexico C: Jemez River near Jemez. New Mexico 43 1937-41; 1950^ 1953-78 6. Arkansas, John Martin, 1942 Arkansas River at Lamar, Colorado 34 1914-41 1942-55, 1960-73 Colorado C: Arkansas River at LaJunta, Colorado 70 1914-41 1942-55, 1960-73 7. Missouri, Fort Peck, Montana 1937 Missouri River near Wolf Point, Montana 100 1929-36 1937-78 C: No suitable station 8. Missouri, Garrison, 1953 Missouri River at Bismarck, North Dakota 120 1929-52-2/ 1953-78 North Dakota C: No suitable station 9. Missouri, Fort Randall, 1952 Missouri River at Fort Randall, South Dakota 0-11 1948-51-2/ 1952-78 South Dakota C: No suitable station 10. Missouri, Gavins Point, 1955 South Dakota 11. Medicine Creek, Medicine 1949 Creek, Nebraska 12. Middle Loup, Milburn, Nebraska 1955 13. Des Moines, Red Rock, Iowa 1969 14. Smoky Hill, Kanopolis, Kansas 1948 15. Republican, Milford, Kansas 1967 16. Wolf Creek, Fort Supply, 1942 Oklahoma 17. North Canadian, Canton, 1948 Oklahoma Missouri River at Yankton, South Dakota 8 1948-54-/ 1955-78 C: Missouri River at Fort Randall, 110 1948-54^ 1955-7^ South Dakota Medicine Creek at Cambridge, Nebraska 10-15 1938-48 - Medicine Creek below H. Strunk Lake, Nebraska 0.8 - 1951-78 C: No suitable station Middle Loup River at Walworth, Nebraska 19 1946-54 1955-60 C: Middle Loup River at Dunning, Nebraska 31 1946-54 1955-60 Des Moines River near Tracy, Iowa 19 1941-68 1969-76 C: Des Moines River below Raccoon River at 94 1941-68 1969-76 Des Moines, Iowa Smoky Hill River near Langley, Kansas 1.3 1941-47 1948-77 C: Smoky Hill River at Ellsworth, Kansas 48 1941-47 1948-7 7-/ Republican River below Milford Dam, Kansas 2.7 1964-66 1967-77 C: Republican River at Clay Center, Kansas 49 1964-66 1967-77 Wolf Creek near Fort Supply, Oklahoma 2.6 1938-41 1942-78 C: No suitable station North Canadian River at Canton, Oklahoma .8 1939-47-/ 1948-78 C: North Canadian River at Woodward, Oklahoma 106 1939-47—^ 1948-78DOWNSTREAM EFFECTS OF DAMS 11 for pre-dam aitd post-dam periods meters per second; C, control station] Flow equaled or exceeded "x" percent of the time, Average daily 3 discharge (m /s) Average annual peak 3 discharge (m /s) (m3/s) x — 5 percent x = 50 percent x = 95 percent Pre-dam Post-dam Pre-dam Post-dam Pre-dam Post-dam Pre-dam Post-dam Pre-dam Post-dam 480 320 2,200 800 1,800 560 230 310 100 31 520 400 2,200 640 1,800 700 270 400 120 145 520 500 2,300 2,300 1,800 1,800 270 250 105 120 400 340 640 550 700 560 400 340 145 140 410 360 650^ 615^ 700 590 400 370 155 140 230 340 850 640 300 630 250 360 125 140 230 380 400 600 300 620 250 370 135 155 1.5 1.5 160 39 8.5 6.2 0.4 0.5 0.006 0.0 2.0 1.9 50 52 7.0 6.7 .9 .8 .5 .4 7.3 4.8 560 190 29 16.0 .2 .6 .05 .07 7.4 6.6 500 340 20 20 2.3 1.6 .3 .4 200 280 770 690 500 630 140 240 70 40 600 660 3,900^' 1,100 1,600 1,100 450 650 140 250 880 680 6,300^ 1,500 2,000 1,400 820 680 195 155 930 740 5,200^ 1,200 2,100 1,400 820 760 250 220 860 670 s.ioo^ 1,400 1,900 1,300 760 680 220 145 2.7 - 530 - 4.6 - 1.6 - 0.8 _ - 1.9 - 13.5 - 8.2 - 1.0 - .02 23 22 58 53 31 30 22 22 16.5 16.5 11.0 11.5 18 20 13.5 14.5 11.0 11.5 9.1 9.3 140 200 1,200 800 530 560 62 115 7.6 13.0 105 155 950 900 390 560 46 82 4.5 10.0 8.7 9.9 320 135 35 54 2.4 2.3 .5 .5 7.6 8.9 330 320 27 37 1.8 2.1 .4 .4 23 24 290 150 69 90 13.5 10.5 4.5 1.2 19.5 23 300 450 60 79 11.5 10.5 4.2 3.4 2.5 1.7 240 35 8.5 6.0 .5 .1 .006 .009 7.7 4.7 280 44 29 26 1.8 .2 .0006 .03 7.2 5.0 400 155 26 19.5 1.2 1.1 .0006 .012 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 4.—Water-discharge data for pre- Dam number River, dam, State (fig. D Year of dam closure River distance Downstream gaging station and control station^ from dam— (km) Period used (water years) Pre-dam Post-dam 18. Canadian, Eufaula, Oklahoma 1963 Canadian River near Whitefield, Oklahoma 13 1939-62-/ 1963-78 C: Canadian River at Calvin, Oklahoma 108 1939-62-/ 1963-78^ 19. Red, Denison, Texas-Oklahoma 1943 Red River at Denison Dam near Denison, Texas 0.5-4.0 1937-42 1943-78 C: Red River near Gainesville, Texas 106 1937-42 1943-78^ 20. Neches, Town Bluff, Texas 1951 Neches River at Evadale, Texas 93 1922-50 1951-64 C; Neches River near Rockland, Texas 63 1922-50 1951-64 21. Chattahoochee, Buford, 1956 Chattahoochee River near Buford, Georgia 4.0 1942-55 1956-71 Georgia C: Chestatee River near Dahlonega, Georgia 73 1942-55 1956-71 -- Rio Grande, Caballo, 1938 Rio Grande below Caballo Dam, New Mexico 1.3 1939-77 New Mexico C: Not needed -- Marias, Tiber, Montana 1955 Marias River near Chester, Montana 3-8 1946-47 1956-78 C: Marias River near Shelby, Montana 65 1946-47 1956-78 — Canadian, Sanford, Texas 1964 Canadian River near Canadian, Texas 120 1939-69-/ 1964-78^ C: Canadian River near Amarillo, Texas 47 1939-63-/ 1964-78^ — Canadian, Conchas, New Mexico 1938 Canadian River below Conchas Dam, 4.5-5.6 1937-38 1943-72 New Mexico C: Canadian River near Sanchez, New Mexico 50 1937-38 1943-72 - Canadian, Ute, New Mexico 1963 Canadian River at Logan, New Mexico 3.2 1943-62 1963-72 C: Canadian River below Conchas Dam, 112 1943-62 1963-72 New Mexico — Republican, Trenton, Nebraska 1953 Republican River at Trenton, Nebraska 1.5 1948-52-/ 1953-78 C: Republican River at Benkelman, Nebraska 50 1948-52 1953-78 - Republican, Harlan County, 1952 Republican River near Hardy, Nebraska 115 1948-51—^ 1952-77 Nebraska C: Republican River near Orleans, Nebraska 37 19 48-51-/ 1952-77-/ -- Washita, Foss, Oklahoma 1961 Washita River near Clinton, Oklahoma 43 1938-60 1961-78 C: Washita River near Cheyenne, Oklahoma 112 1938-60 1961-78 —long-term gaging station upstream from the dam. 2/ — Main station is downstream from dam and control station is upstream from dam on same river. 3/ — Flows affected by one or more upstream dams. 4/„ Highest mean daily flow used, rather than instantaneous peak. Only years available. — Only data for water years 1961-78 available. —^Only data for water years 1921 and 1946 used (only data available). 8 / —All post-dam flow is from seepage at dam and from springs and tributaries downstream from dam; no releases are made from the dam. 9/ — Water years 1936-39 (before dam closure) used. km downstream from the dam, a given water discharge after dam construction transported about 50 percent of the volume of sediment it did before the dam. On the Red River, too, then, the deficit persists for hundreds of kilometers. The actual length of reach required for complete recovery on the Red River cannot be determined from the above data. Five stations downstream from Gavins Point DamDOWNSTREAM EFFECTS OF DAMS 13 dam and post-dam periods—Continued Average daily 3 discharge (m /s) Average annual peak 3 discharge (m /s) Flow equaled or exceeded (nJ/s) rx" percent of the time, x = 5 percent X - 50 percent x = 95 percent Pre-dam Post-dara Pre-dam Post-dam Pre-dam Post-dam Pre-dam Post-dam Pre-dam Post-dam 175 120 3,600 740 760 480 48 54 3.1 1.6 56 29 2,300 1,300 270 120 9.6 6.8 .3 .1 185 120 3,000 950 720 400 56 74 7.1 3.2 120 70 2,400 1,400 540 280 26 19.0 3.1 4.2 200 130 1,100 800 700 500 93 54 8.8 5.9 77 48 550 360 290 195 31 17.5 1.2 .7 60 54 660 270 140 150 46 40 19.0 12.0 10.0 10.5 220 260 23 24 7.6 8.2 2.8 3.4 - 24 - 77 - 63^ - 16.0*' - .03^' 25 26 n & 90 79 74 11.6 18.0 4.2 2.8 25 27 1151' 560 76 105 11.5 11.5 4.2 4.0 16.0 2.6 1,100 370 66 7.6 .9 .7 .01 .003 12.0 5.4 1,100 640 48 24 .7 .6 .08 .05 12.0 .8 1,000^' 100 35 1.1 .5 .1 .02 .009 6.5 4.3 640^' 420 22 16.0 .6 1.0 .05 .02 3.5^' 1.2 550 66 14.0 8.6 .2 .07 .006 .02 1.1*' .3 145 32 1.7 .2 .2 .1 .01 .05 6.0 1.8 290 24 16.0 7.1 4.0 0.08 .001 .02 3.4 2.4 105 60 7.7 4.9 2.8 2.3 .3 .03 32 11.0 530 185 135 42 14.5 4.8 4.0 1.6 16.5 8.0 380 125 52 23 9.3 5.4 2.0 .5 4.1 1.6 290 69 15.5 5.4 .7 .7 .01 .1 1.2 .4 290 66 4.0 1.4 .2 .1 .0003 .0 provide an example of the degree of downstream recovery of suspended-sediment loads. Three dams—Fort Randall (1952), Garrison (1953), and Gavins Point (1955)—were closed on the Missouri River within 3 years during the 1950’s. Inspection of the yearly sediment data downstream from Gavins Point Dam shows that annual loads consistently decreased during this period (water years 1953-56), as expected. These years were excluded here in computing pre- and post-dam average loads. For the five downstream stations, the available water years of pre- and post-dam data, respectively, were: Yankton—1940-52, 1957-69; Omaha— 1940-52, 1957-73; St. Joseph, Kansas City, and Hermann—1949-52, 1957-76. These stations are 8 (Yankton), 314 (Omaha), 584 (St. Joseph), 716 (Kansas City), and 1,147 (Hermann) km downstream from Gavins Point Dam. From the annual suspended-sediment loads, an average annual load was computed for the pre-dam period and again for the post-dam period. The ratio of these average loads as a function of distance downstream from the dam is shown in figure 4. At Yankton, just 8 km downstream from the dam, the average post-dam annual load was less than 1 percent of that for the pre-dam period. Even 1,147 km14 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS WATER YEAR Figure 2.—Variation in annual suspended-sediment loads before and after closure of Hoover Dam, Colorado River, Arizona, at a station upstream from the dam (Grand Canyon) and downstream from the dam (Topock). downstream from the dam, the post-dam average annual load was only 30 percent of the pre-dam load. Data from the Mississippi River at St. Louis, downstream from the confluence of the Missouri and Mississippi Rivers, and about 1,300 km downstream from Gavins Point Dam, show that the mean annual suspended load decreased from about 320 million megagrams during 1949-52 to 109 million megagrams during 1957-80, after closure of the dams on the Missouri River. Changes elsewhere in the Mississippi River basin also may have contributed to the decrease in sediment load in the Mississippi. However, along the nearly 1,300 km of the Missouri River downstream from Gavins Point Dam, post-dam average suspended loads have not approached the much larger pre-dam average values. Hammad’s (1972, p. 601) data for the Nile River downstream from Aswan High Dam show that, even 965 km downstream from the dam, annual sediment loads 2 years after dam closure were only about 20 percent of pre-dam values. The above examples indicate that, in some major rivers, sediment concentrations and annual sediment loads may not achieve pre-dam values for hundreds or thousands of kilometers downstream, if at all. The cases documented here are large dams and reservoirs from specific geographic regions. As noted earlier, a variety of conditions will control the response on different rivers. MEAN BED ELEVATION The results of the bed-elevation analyses based on resurveyed cross sections are listed in table 13 at the end of this report. The data for changes in mean bed elevation determined from gaging-station rating tables are listed in table 14 also at the end of this report. From the latter rating tables, changes in mean bed elevation with time were plotted for each gage site (figs. 36-49 at the end of this report). (Such graphic relations proceed in “stairsteps” because a constant bed elevation is assumed for the period during which a given rating table is in effect. The change to a new rating table brings what appears in figures 36-49 as a sudden switch to a new constant bed level. The actual change in bed level with time probably follows a smoother curve.) Similar plots of change in bed level with time were made for all resurveyed cross sections from the voluminous data in table 13, and representative examples are shown below. GENERAL NATURE OF CHANGES IN BED ELEVATION For all 21 channels (fig. 1) having resurveyed cross sections, a lowering of the mean bed level—here called degradation—occurred immediately downstream from the dam (figs. 5 and 6), unless constrained by very coarse material or bedrock. Such bed degradation downstream from dams is a well-known phenomenon on alluvial streams (Lane, 1934; Gottschalk, 1964, p. 17-5). Analytical studies of open-channel bed degradation include those by Lane (1948), Mostafa (1957), Tin-ney (1962), Breusers (1967), Komura and Simons (1967), Aksoy (1970, 1971), Hales and others (1970), Komura (1971), Rzhanitzin and others (1971), de Vries (1973), Hwang (1975), and Strand (1977). Special flume studies of bed degradation have been conducted by Schoklitsch (1950), Harrison (1950), Newton (1951), Ahmad (1953), Willis (1965), Garde and Hasan (1967), Ashida and Michiue (1971), and others. In some reaches, degradation can occur simply by the removal of bars in the absence of replenishment of sediment from upstream. This was observed on the Red River in the region about 10 to 15 km downstream from Denison Dam. Koch and others (1977) reported a similar removal of bars in the reach downstream from Yellowtail Dam on the Bighorn River, Montana.DOWNSTREAM EFFECTS OF DAMS 15 X e> LU >- CO Q LU (/> O oc LU CL DISCHARGE, ll\l CUBIC METERS PER SECOND Figure 3.—Suspended-sediment loads (concentrations) transported by various discharges at successive downstream stations before and after closure of Canton Dam, North Canadian River, Oklahoma. Control-station curve based on unpublished U.S. Geological Survey data; other curves redrawn from U.S. Army Corps of Engineers, 1958. On four rivers—Jemez River, Arkansas River, Wolf Creek, and the North Canadian River—post-dam flow releases were so much less than pre-dam discharges that the channel became considerably narrower. In such instances, lowering of the mean bed elevation can result not only from bed erosion but also because the post-dam narrowed river occupies only the lowest part of the original channel. Should a dam release little or no water, the bed downstream might not degrade. In fact, local aggradation sometimes occurs, because the controlled flows are not strong enough to remove deposits left by tributary flash-floods; by main-channel, sediment-removal works associated with canals; or by wind. Examples are found on the Rio Grande in New Mexico (Lawson, 1925; Lagasse, 1980) and on the Peace River in Canada (Bray and Kellerhals, 1979). Downstream from Elephant Butte Dam on the Rio Grande the controlled releases are depleted systematically by irrigation intakes. In a reach beginning about 265 km downstream from the dam this decrease in flow strength, together with the deposits delivered by tributaries, brought the river bed in many places to an elevation higher than the adjoining farm area (Lawson, 1925). DEGREE OF CHANGE ATTRIBUTABLE TO DAMS The magnitude of the measured changes (as much as 7 m, as described below) greatly exceeds the expectable errors in measurement and analysis. Furthermore, occurring as they do during periods ranging from a few16 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS 0C Figure 4.—Post-dam/pre-dam ratio of annual suspended-sediment loads versus distance downstream from Gavins Point Dam, Missouri River, South Dakota. years to 2 or 3 decades, the changes greatly exceed those that would be expected as part of a temporal fluctuation around a mean bed level and those generally observed to occur naturally (table 3). Several considerations support the view that the measured changes in alluvial channels downstream from the dams studied here are due to the dam and reservoir upstream: 1. As the longitudinal profiles discussed below show, degradation generally was greatest at or near the dams and usually decreased somewhat progressively downstream, though with local exceptions. 2. From the rating tables for the 14 streamflow-gaging stations downstream from dams (table 14 and figs. 36-49), the relation of water-surface elevation to discharge for a reference low flow indicates that the channels generally were relatively stable prior to dam construction and began degrading just after the dams were built. This timing is illustrated by the bed changes, as assumed from the stage-discharge relation, for the Smoky Hill River near Langley, Kansas, about 1.3 km downstream from Kanopolis Dam (fig. 7). 3. Whereas the river bed downstream from the dam tended to erode, the elevation of the bed at eight Figure 5.—Time progression of bed degradation and channel armoring at the streamflow-gaging station downstream from Jemez Canyon Dam, Jemez River, New Mexico. A, 1952; B, 1957; C and D, 1980. Dam was closed in 1953. Station is 1.3 kilometers downstream from the dam. White dashed line is at a constant elevation for reference.DOWNSTREAM EFFECTS OF DAMS 17 Figure 6.—Degradation represented by successively lower water-intake pipes for streamflow-gaging station 2.6 kilometers downstream from Fort Supply Dam, Wolf Creek, Oklahoma. Dam was closed in 1942; photograph was taken in 1951. control stations upstream from dams for which data were available did not change significantly during the years after dam closure (fig. 7; table 14). 4. For the channels having resurveyed cross sections, and for those with gaging-station records, extrapolation of the post-dam degradation rates back into the pre-dam years would place the pre-dam streambeds at unrealistically high elevations. Thus, the timing, magnitude, and spatial distribution of the measured changes in the alluvial channels studied here indicate that the dams and upstream reservoirs are responsible for the measured degradation. DEGRADED REACH DOWNSTREAM FROM A DAM Given the capacity of flow releases to entrain sediment from channel bed and banks, erosion of the bed and banks should continue downstream from a dam until some factor or a combination of factors results in establishment of a new stable channel. These factors may include: (1) Local controls of bed elevation (emergence of bedrock; development of armor by winnowing of fines); (2) downstream base-level controls (ocean, lake, or larger river; manmade structure such as a dam; barrier of deposited sediment); (3) decrease in flow competence (flattening of slope by progressive degradation; expansion of channel width, resulting in decreased depth, redistributed flow velocities, or both); (4) infusion of enough sediment to restore the balance between arriving and departing sediment (upstream erosion; sluicing from the upstream dam; inflow from tributaries); and (5) growth of vegetation. Several of these changes or processes are considered or illustrated WATER YEAR Figure 7.—Changes in mean bed elevation with time at streamflow-gaging stations 48 kilometers upstream and 1.3 kilometers downstream from Kanopolis Dam, Smoky Hill River, Kansas. Bed elevations assumed proportional to the gage height that corresponds to a constant low discharge of 0.42 cubic meter per second at the upstream gage and 0.51 cubic meter per second at the downstream gage. separately here; any number of them may occur together along any given reach of river. GENERAL FUNC TION OF DEGRADA TION WI TH TIME A T A SI TE Except for cross sections underlain only by sand beds of unlimited depth, the degradation rate at a section would be expected to decrease with time as the bed becomes armored, or until the channel slope in that reach becomes too flat for the bed material to be moved. Eventually, an equilibrium bed elevation should be reached, as postulated in many analytical studies cited earlier. A number of the sections for which data are given in table 13 show this trend. At many other cross sections, however, the rate of degradation with time varies considerably (table 13). For instance, one or more temporary periods of aggradation may be included within a long-term trend of degradation. Or, after some initial degradation, the bed level may become constant rather abruptly with time at a certain depth, probably an indication that bedrock was reached or that armor had developed. Other sites have an S-shaped curve, where initial degradation rates were slow, then increased with time for some years, and then reversed this trend to decrease in later years. (A possible cause of such a curve might be minimal releases the first few years after dam closure to fill the reservoir, and greater releases thereafter.) Some of these irregular degradation-time trends are shown in figure 8. Besides variations in flow releases with time, departures from a regular degradation curve could be due to differences in bed material with depth, to changes in cross-sectional shape, and to development and death or eradication of vegetation.BED DEGRADATION, IN METERS 18 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Missouri River, South Dakota 2 — 1.6 kilometers downstream “ from Fort Randall Dam i_—• i' Red River, Oklahoma-Texas 27 kilometers downstream _ from Denison Dam I_______I________I_______ 0 10 20 30 40 YEARS AFTER DAM CLOSURE Figure 8.—Examples of irregular rates of bed degradation with time. Data from table 13. To search for a possible general function of bed degradation with time at a site, all 287 resurveyed cross sections were used except for those: 1. That did not show a general trend of bed lowering (84 sections, mostly in the zone of varied bed changes downstream from the degraded zone); 2. Of the remaining 203 sites that lacked enough data points to justify fitting a curve, our arbitrary requirement being at least three resurveys after the onset of degradation, not counting the original survey (49 sites); and 3. Survivors of the above two requirements that showed marked aberrations in general degradation, such as abrupt cessation of bed erosion or even substantial aggradation after inital erosion (40 sites, exemplified in figure 8). One hundred fourteen cross sections were left, after the above exclusions, for use in the degradation-time analysis. Plots of bed lowering with time (representative examples given below) for the 114 cross sections generally show that the rate of degradation is fastest immediately after erosion begins and gradually slows with time, becoming asymptotic toward some new stable bed elevation. Some of the plots have large scatter, scarcity of points, or an irregular trend; any of several types of functions could be fitted to such data with a large standard error. Empirically analyzing the trends for the more regular, better-defined curves, either ofDOWNSTREAM EFFECTS OF DAMS 19 two functions appeared to fit most cases: (1) Logarithmic, with degradation D (arithmetic scale) as a function of the logarithm of time, t\ or (2) hyperbolic, with 1 ID as a function of 1 It, both on arithmetic scales. Least-squares regressions were calculated applying each of these functions to each of the 114 cross sections. For the prediction of D (not 1 ID) at the observed times, the square of the correlation coefficient (r2) is as follows: for the logarithmic function, a range of 0.16 to 1.00 with an average of 0.82; for the hyperbolic function, a range of 0.10 to 1.00 with an average of 0.81. The average of 0.82 corresponds to an r-value of about 0. 91. and the average of 0.81 corresponds to an r-value of about 0.90; they indicate a reasonably good fit. The logarithmic relation had a greater r2 for 55 of the 114 cross sections; the hyperbolic relation had a greater r2 for 50 sections; and, at the 9 remaining cross sections, r2 was the same for both equations. In most cases, there was little difference in the two correlation coefficients for a given cross section. However, at a few cross sections the hyperbolic equation predicted values of D that diverged greatly from the measured values; whereas, no such grossly disparate predictions resulted from the logarithmic relation. Analytical considerations indicate that the bed erosion should decrease with time. The degradation-time plots indicate that this decrease or cessation of degradation tends to occur within decades or a few centuries. For the data of this study, the hyperbolic equation generally predicts this approximate time much more closely than the logarithmic equation, the latter in some instances predicting billions of years for degradation to cease. The relative advantages of using each type of equation seem to be: Logarithmic Equation: 1. Slightly—but probably not too significantly—greater correlation coefficient. 2. Reasonable predictions for a few cross sections at which the hyperbolic equation gives a very poor fit. Hyperbolic Equation: 1. Better calculated-versus-measured agreement of the time within which approximate eventual maximum degradation occurs, and of the magnitude of this maximum limiting degradation. 2. The practical benefit of providing a reasonable value of maximum degradation for planning purposes. 3. Better consistency in the sign of the first coefficient (intercept) of the regression equation. (With degradation considered positive, only 3 of the 114 cross sections have a negative coefficient using the hyperbolic equation, as opposed to 24 such cross sections for the logarithmic equation. Reasons for such negative coefficients are mentioned below.) On the basis of the previous discussion, the hyperbolic equation seems more suitable as a model for bed degradation with time at the many sites analyzed here. This equation has the form: where D is degradation, in meters, at t years after the start of bed erosion; and Ci and c2 are constants for a given cross section or graph. Such a hyperbola is asymptotic to a line parallel to the x axis (time). Its equation (eq. 1) plots as a straight, line on arithmetic scales when written in the form: 0-ID) = c2+Ci (1/0 (la) where c2 is the intercept; and Cj is the slope of the best-fit straight line. For convenience, degradation is considered to be in a positive direction. Degradation can be considered negative simply by making the signs of Ci and c2 negative. Regression coefficents for each cross section are given in table 5. The reciprocal of c2 is the asymptote on a plot of D versus f; that is, l/c2 is the eventual limit of degradation. The reciprocal of Ci on the same plot is the slope or tangent just after degradation has first begun; that is, 1 lei is the initial degradation rate, in meters per year. Therefore, both cx and c2 have an important practical significance. To fit a curve of this type, the time origin (/=0 years) needs to be taken as the year at which degradation began. Degradation at sites close to the dam can be taken as beginning at the time of dam closure. However, for some downstream sites, there is a response time or lag time between the date of dam closure and the start of degradation. In some instances the year in which degradation began was not determined accurately and had to be estimated from the plots of degradation versus time. Twelve typical curves, using the hyperbolic function as the general model, are shown in figure 9. The data to which the least-squares regressions were applied are listed in table 13. These particular examples were chosen to reflect the range of r2 values of the 114 applicable cross sections.20 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 5.—Values associated with fitted degradation curves1 Table 5.—Values associated with fitted degradation curves1—Continued Distance Slope Max- of cross Correl- of imum Estimated Estimated section Response ation Inter- best- expected time to time to downstream time=? coeff- cept fit degra- achieve achieve from (years) icient— c9 straight dation 0.95 D max 0.5 D max dam r line D (years) (years) (kilometers) ci (meters) Colorado River, Arizona, Glen Canyon Dam 2.6 0 0.88 0.30 2.03 3.3 130 7 A.3 0 .77 .24 .75 4.2 60 3 6.4 0 .93 .52 1.78 1.9 65 3 10.5 2.0 .99 .33 2.54 3.0 150 8 19.5 2.0 1.00 .29 1.93 3.5 130 7 Colorado River, Arizona, Hoover Dam 2.3 .5 .69 .52 .57 1.9 20 1.1 3.2 0 .87 .19 .17 5.3 17 .9 4.5 0 .81 .37 .13 2.7 7 .4 6.1 0 .56 .41 .18 2.4 8 .4 7.1 0 .84 .27 .26 3.7 18 1.0 8.0 0 .77 .23 .14 4.4 12 .6 9.7 0 .94 .18 .56 5.6 60 3 11.0 0 .94 .48 .59 2.1 24 1.2 12.5 0 .73 .18 .23 5.6 24 1.3 13.5 0 .54 .32 .16 3.1 10 .5 15.5 0 .93 .18 .26 5.6 28 1.4 16.5 .5 .79 .31 .20 3.2 12 .6 18.0 .5 .92 .37 .25 2.7 12 .7 19.5 .5 .74 .25 .23 4.0 18 .9 21 .5 .84 .36 .34 2.8 18 .9 28 1.5 .98 .20 1.38 5.0 130 7 36 1.0 .91 .22 1.97 4.6 170 9 42 .5 .84 .34 1.62 2.9 90 5 51 2.0 .88 -.05 3.40 — — — 57 1.3 .91 .10 2.30 10.0 440 22 63 2.0 .80 .18 1.31 5.6 140 7 70 1.0 .94 .18 1.05 5.6 110 6 77 2.6 .93 .28 1.93 3.6 130 7 87 3.5 .96 .35 1.90 2.9 100 5 94 3.0 .84 .20 1.71 5.0 160 9 104 3.0 .94 .23 1.65 4.4 140 7 110 3.2 .96 .08 2.29 12.5 540 28 117 4 .98 .031 2.00 32 1,200 65 Colorado River, Arizona, Davis Dam 1.1 0 0.97 0.15 0.69 6.7 85 5 8.8 .5 .95 .32 2.46 3.1 150 8 Colorado River, Arizona, Parker Dam 27 0 .97 .26 1.29 3.9 95 5 39 2.0 .95 .19 1.05 5.3 100 6 46 1.0 .93 .17 1.29 5.9 140 8 66 1.0 .77 .20 1.31 5.0 120 7 80 1.75 .66 .45 1.11 2.2 45 2.5 95 1.0 .63 .43 2.30 2.3 100 5 Arkansas River, Colorado, John Martin Dam 12.0 6 1.00 .57 13.29 1.8 440 24 15.5 6 1.00 .66 10.02 1.5 290 15 22 3.0 .77 .73 5.55 1.4 140 8 26 7 .91 .94 6.10 1.1 120 6 Missouri River, Montana, Fort Peck Dam 9.2 0 .10 1.28 1.97 .78 30 1.5 13.0 0 .64 .74 11.45 1.4 290 15 16.5 0 .48 .58 6.33 1.7 210 11 23 9 .83 .49 6.18 2.0 240 13 Missouri River, North Dakota, Garrison Dam 2.7 0 .99 .13 4.86 7.7 710 38 6.4 0 1.00 .17 2.06 5.9 230 12 8.0 0 .90 .33 1.22 3.0 70 4 10.5 0 .71 .46 .89 2.2 36 1.9 12.0 2.0 .87 .39 3.09 2.6 150 8 15.0 2.0 .92 .60 4.99 1.7 160 8 24 0 .96 .51 .74 2.0 28 1.5 32 0 .28 1.82 .77 .55 8 .4 36 0 .82 .66 2.21 1.5 65 3 38 0 .37 1.65 3.49 .61 40 2 51 1.0 .37 .73 2.09 1.4 55 3 A small r2 (presumably indicative of a minimum correlation between variables for the type of function being used) can result not only from large scatter about the Distance of cross section downstream from dam (kilometers) Correl-Response ation 2/ time—' coeff-37 (years) icient— 2 r Inter- cept c2 Slope of best- fit straight line ci Max- imum expected degra- dation D max (meters) Estimated time to achieve 0.95 D max (years) Estimated time to achieve 0.5 D max (years) Missouri River, South Dakota, Fort Randall Dam 3.1 0 0.77 0.74 4.25 1.4 110 6 4.2 1.0 .90 .70 3.23 1.4 90 5 5.1 2.5 .77 1.26 3.61 .8 55 2.9 6.6 1.0 .84 .40 4.59 2.5 220 11 7.7 1.0 .81 .52 3.46 1.9 130 7 11.0 0 .33 .52 .48 1.9 18 .9 12.5 2.0 .50 .96 1.52 1.0 30 1.6 29 0 .33 1.67 1.25 .60 14 .8 Missouri River, South Dakota, Gavins Point Dam 2.3 0 .86 .33 2.37 3.0 140 7 3.4 0 .99 .25 3.78 4.0 290 15 4.3 4 .98 .28 3.71 3.6 250 13 5.3 0 .99 .026 8.86 38 6,500 340 6.8 0 .87 .44 10.59 2.3 460 24 7.9 0 .96 .30 7.84 3.3 500 26 8.4 0 .13 .88 1.04 1.1 22 1.1 8.5 3.5 .98 .18 5.76 5.6 610 32 9.5 0 .94 .23 10.35 4.4 860 4.5 11.0 0 .75 .82 6.68 1.2 160 8 12.5 0 .45 .92 6.72 1.1 140 7 36 0 .48 .60 5.03 1.7 160 8 44 0 1.00 1.51 12.53 .66 160 8 Middle Loup River, Nebraska, Milburn Dam .2 0 .59 .48 .24 2.1 10 .5 1.6 6.3 .97 .64 2.76 1.6 80 4 5.6 6.3 .98 .78 2.00 1.3 48 2.6 Smoky Hill River, Kansas, Kanopolis Dam .8 0 .92 .61 1.68 1.6 50 2.8 2.9 2 .97 .66 4.35 1.5 130 7 Wolf Creek, Oklahoma, Fort Supply Dam 0.3 2? 1.00 0.25 1.21 4.0 90 5 1.0 2? .48 .46 .29 2.2 12 .6 1.3 2? .94 .33 1.92 3.0 110 6 1.6 6? .95 .31 3.70 3.2 230 12 2.9 2? .96 .42 4.10 2.4 190 10 3.9 2? .70 .51 1.68 2.0 65 3 4.7 2? 1.00 .29 9.67 3.5 630 34 6.6 2? .96 .16 13.44 6.3 1,600 85 North Canadian River , Oklahoma , Canton Dam 1.8 0 .93 .35 .83 2.9 46 2.4 3.1 0 .79 .63 1.00 1.6 30 1.6 5.0 0 .40 .93 .80 1.1 16 .9 5.6 0 .54 .51 2.15 2.0 80 4 7.4 0 .90 .33 2.43 3.0 140 7 12.0 0 .63 .25 6.01 4.0 460 24 14.5 1.0 .85 1.05 3.08 .95 55 3 35 0 .59 2.26 2.77 .44 24 r, 1.2 125 0 .99 1.34 1.99 .75 28 1.5 Red River, Oklahoma-Texas, Denison Dam .6 0 .87 .64 .52 1.6 15 .8 2.1 0 .97 .36 1.15 2.8 60 3 7.2 3 .97 .82 .92 1.2 22 1.1 8.4 0 .84 .59 .87 1.7 28 1.5 11.5 0 .98 .45 1.15 2.2 48 2.6 15.0 0 .90 .31 1.20 3.2 75 4 109 4 .79 1.52 2.63 .66 32 1.7 Chattahoochee River , Georgia, Buford Dam .5 0 .31 .88 1.12 1.14 24 1.3 1.9 0 .99 .10 4.32 10.0 820 44 2.9 0 .95 -.02 7.96 — — — 4.0 0 .88 -.11 11.49 — — — -'(1/D) ' c2 + C1 d/t). where D = measured degradation, in meters, at t years after start of degradation. 2/ — Years between dam closure and start of degradation. 3/ 2 —Listed r is for estimation of D rather than for 1/D.DOWNSTREAM EFFECTS OF DAMS 21 C/5 cc < O < cc O LU Q Q LU 00 < 1 1 ; 1 1 1 1 1 1 1 1 V. • w "V \ " -M- -r -M _ Missouri River, North Dakota 32 kilometers downstream Chattahoochee River, Georgia — Missouri River, South Dakota — from Garrison Dam — — 0.5 kilometer downstream 8.4 kilometers downstream from Buford Dam from Gavins Point Dam r2 = 0.28 r2 = 0.13 I 1 1 1 I 1 r2 =0.31 I I I ^ 1 1 | I 1 1 1 Colorado River. Arizona 1 6.1 kilometers downstream lb from Hoover Dam * 1 1 ‘ Vt„ • _ Missouri River, Montana • __ - • ' '• 16.5 kilometers downstream Colorado River. Arizona • from Fort Peck Dam •v 95 kilometers downstream r2 = 0 48 I I I r2 =0.56 I I I from Parker Dam ! r2 = 0.63 | 1 1 1 I * 1 1 1 V \ \ "I «- • r —•- ‘--.J North Canadian River, Oklahoma 14.5 kilometers downstream _ North Canadian River. Oklahoma Red River, Oklahoma-Texas from Canton Dam 3.1 kilometers downstream 8.4 kilometers downstream from Canton Dam from Denison Dam r2 = 0.85 _| | | i ^z9 i i r210 84 i \I I I 1 1 1 1 \ 1 1 1 X \ Colorado River. Arizona T 1.1 kilometers downstream \ \ \ x from Davis Dam \ Smoky Hill River, Kansas i V 0.8 kilometer downstream X o from Kanopolis Dam \m r^ - 0.97 \# ~~ — -• - - \* Wolf Creek, Oklahoma ” r2 =0.92 ~ 0,3 kilometer downstream from Fort Supply Dam I I I i r2 = 1.00 _J | | 10 20 30 40 0 10 20 30 YEARS AFTER DAM CLOSURE 40 0 10 20 30 40 Figure 9.—Representative regression curves (dashed lines) of bed degradation with time at selected sites. Data from table 13. line, but from other factors, such as: (1) Small depths of degradation (flat slope of best-fit straight line); (2) small number of data points (especially during the first few years after bed erosion begins); (3) errors in estimating any response time; and (4) irregularities in the trend of the curve, such as a rather abrupt cessation of degradation or an S-shaped curve. These features also contribute to a negative value of the coefficient c2 (intercept). A negative c2 precludes estimation of the maximum limit of degradation and associated values. The type equation used here could predict degradation with time at a cross section if the coefficients cx and c2 could be predicted. These certainly are functions at least of bed material and water discharge. A third factor might be distance downstream from the dam. At present (1982), the coefficients cannot be determined in advance. Therefore, prior to dam closure any estimates of bed erosion need to be based on some type of degradation analysis using measured bed-material sizes and expected water discharges (Strand, 1977; Priest and Shindala, 1969a). This approach requires: (1) Adequate Sampling of the bed material with depth, distance across the section, and distance along the channel; and (2) accurate predictions of future flow releases. On the other hand, with a few years of measurements after the start of degradation, the model described above might be used with due caution. (References cited earlier include models of degradation based on transport equations, particle-size measurements, and the assumption of winnowing.)22 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS MAXIMUM DEGRADATION AND ASSOCIATED TIME At a given cross section within the degraded reach, the maximum observed depth of bed erosion (table 6) was negligible on some channels but was as much as 7.5 m on others (Colorado River 12 km downstream from Hoover Dam, Arizona). (Maximum values for a cross section do not necessarily apply to an entire reach, as local features can affect the extent of degradation at any one site.) Had Davis Dam not been built downstream, more degradation than 7.5 m probably would have occurred at the Colorado river section just mentioned. The data (table 13) show that the bed at this site was still degrading at a rapid rate 13 years after dam closure, when measurements were discontinued due to backwater from Davis Dam. The maximum possible degradation in some cases can be limited or restricted by a fixed base level. For example, bedrock is encountered in places on the Smoky Hill River downstream from Kanopolis Dam (Kansas) and on the Republican River downstream from Harlan County Dam (Nebraska). On the Rio Grande and along reaches of the Colorado River, fans of very coarse debris are controlling. In contrast, certain wide, shallow cross sections in some reaches on the Missouri River downstream from Fort Randall Dam (South Dakota) may no longer degrade because flow velocities are too slow. If one assumes that the bed does not contain a subsurface layer of erosion-resistant material and that the same discharges will continue, it is interesting to extrapolate the empirical hyperbolic equation for each of the 114 applicable cross sections discussed above into the future. With due regard both for the uncertainties Table 6.—Maximum degradation downstream from various dams [Data from table 13, except last two entries which are from unpublished sources] River, dam, State Years since closure Maximum lowering of bed elevation (meters) Colorado, Glen Canyon, Arizona 9 7.3 Colorado, Hoover, Arizona 13 7.5 Colorado, Davis, Arizona 26 5.8 Colorado, Parker, Arizona 27 4.6 Jemez, Jemez Canyon, New Mexico 12 2.8 Arkansas, John Martin, Colorado 30 .9 Missouri, Fort Peck, Montana 36 1.8 Missouri, Garrison, North Dakota 23 1.7 Missouri, Fort Randall, South Dakota 23 2.6 Missouri, Gavins Point, South Dakota 19 2.5 Medicine Creek, Medicine Creek, Nebraska 3 .6 Middle Loup, Milburn, Nebraska 16 2.4 Des Moines, Red Rock, Iowa 9 1.9 Smoky Hill, Kanopolis, Kansas 23 1.5 Republican, Milford, Kansas 7 .9 Wolf Creek, Fort Supply, Oklahoma 27 3.4 North Canadian, Canton, Oklahoma 28 3.0 Canadian, Eufaula, Oklahoma 6 5.1 Red, Denison, Oklahoma-Texas 16 3.0 Neches, Town Bluff, Texas 14 .9 Chattahoochee, Buford, Georgia 15 2.6 South Canadian, Conchas, New Mexico 7 3.0 Salt Fork, Arkansas, Great Salt Plains, Oklahoma 9 .6 of the assumptions and for the risks of extrapolation, we have nevertheless done this (table 5) to estimate: (1) Maximum eventual degradation Z)max; (2) years needed to achieve 95 percent of the eventual maximum degradation (the function goes to infinity at maximum degradation); and (3) years needed for the bed to erode to 50 percent of its eventual maximum degradation. All estimates were computed from the regression coefficients Cj and c2 and rounded off appropriately. Z)max values (l/c2) were estimated for all 114 cross sections. Three of these Dmax values obviously were unreasonable and were not considered further. A frequency distribution of Dmax for the remaining 111 sections is shown in figure 10. The distribution is virtually the same if the 21 cross sections that narrowed considerably are excluded. Ordinarily degradation needs to be viewed in relation to the size of the channel rather than in absolute values. Thus, Dmax needs to be adjusted by a scaling factor, such as the channel width. Because widths were not available for 35 cross sections on the Colorado River (about 33 percent of the total), the frequency distribution was drawn without applying any scaling factor. The 111 cross sections used to compile figure 10A are downstream from the following dams (number of cross sections in parentheses): Glen Canyon (5), Hoover (27), Davis (2), Parker (6), John Martin (4), Fort Peck (4), Garrison (11), Fort Randall (8), Gavins Point (13), Milbum (3), Kanopolis (2), Fort Supply (8), Canton (9), Denison (7), and Buford (2). A variety of rivers and channel conditions is reflected in figure 10A. According to the data in figure 10A, the modal or average maximum expectable degradation for the cross sections represented on the graph is about 2 m. The range is from about 0.4 to 38 m; about 98 percent of the values are less than 10 m. Accuracy of these predictions is related to the fit from the data themselves, the number and duration of measurements, the assumed nature of the subsurface sediment, and the validity of the many other assumptions. If the coefficients in equation 1 are known, then the time needed for the bed to degrade to any proportion of the maximum eventual degradation depth can be estimated quickly by the following method. The actual depth value need not be known. Let p = the decimal proportion of the maximum degradation depth, for example, 0.95 if the depth of interest is 0.95 Dmax. The time tp needed to reach any designated proportion of ^max I® tp Cp C j C o where cp=(^~)-l. For example, the Colorado River 2.6 km downstream from Glen Canyon Dam, at whichDOWNSTREAM EFFECTS OF DAMS 23 30 25 UJ in < 20 U- o 15 t— 2 10 O cc LLJ Q- Figure 10.—Frequency distributions based on 111 measured cross sections on various rivers: A, maximum expected degradation depth; B, years needed to deepen to 95 percent of maximum depth; and C, years needed to deepen to 50 percent of maximum depth. c, = 2.03 and c2 = 0.30 (table 5), would be predicted to reach 0.95 Z)max in 19x2.03/0.30 = 129 years (or about 130 years) after the start of degradation. This shortcut method is preferable to equation 1 for estimating times needed to reach a given depth because rounding of D in equation 1 causes variations in the computed times, compared to the times calculated from the coefficients alone. Variations are insignificant for the steeper part of the degradation-time curve (early years, shallow depths) but can be as much as 60 percent where the curve flattens in later years. The number of years predicted for the bed to achieve 95 percent of its eventual total degradation (0.95 fmax) for the 111 cross sections ranges from 7 to 6,500 years (fig. 10Z?). About 91 percent of the values are between 7 and 500 years. The modal time is about 140 years. Data for many of the cross sections (table 13) indicate that these individual estimates of 0.95 fmax are of the right order of magnitude. Each such computation, however, is based on the assumption that the remaining subsurface material does not differ substantially from the original channel bed sediment and that the same flow pattern will continue. The adjustment period at a site, or predicted time required to reach the new stable depth (0.95 Dmax), does not seem to have any consistent relation to distance downstream from the dam, for the few reaches where this aspect could be assessed. Any relation probably is obscured by the irregular differences in degradation from one cross section to another along a river. Most of the 111 cross sections eroded one-half of their predicted eventual maximum depths within the first few years after the start of degradation. The range of these predicted times (0.5 fmax) is from 0.4 to 340 years (fig. 10C); modal value is about 7 years. All distributions in figure 10 are skewed, with a preponderance of smaller values within the respective range. Initial degradation rates (the reciprocal of the coefficient Cj) range from virtually negligible to as much as 7.7 m/yr. Even downstream from the same dam, different sites show different initial degradation rates. No direct relation between initial degradation rate and predicted eventual maximum depth of degradation could be established. ST ANDARDIZED DEGRADAT ION T IME PLOT The degradation-versus-time plots (fig. 9) can be standardized and made dimensionless by converting the D axis to D/0.95 Z/max and the t axis to t/0.95 >120 — 3 — 7 -»120 — 15 — 12 —L>120 -- 13 - Jemez River, New Mexico, Jemez Canyon Dam 6 y>2.i .6 .6 12 y> 2.1 1.1 22 y> 2.1 -- 1.0 — Arkansas River, Colorado, John Martin Dam 9 26 3 22 4 24 26 0 4 30 26 0 22 — Missouri River, Montana, Fort Peck Dam 13 — >75 — 17 9 18 ^>75 — 17 — 23 —/>75 — 23 — 36 ->75 “ 17 — Missouri River, North Dakota, Garrison Dam 1 12 12 28 3 7 19 1.2 6 11 18 -.25 6 17 21 .5 6 23 21 0 6 — Missouri River, South Dakota. , Fort Randall Dam 2 5 2.5 11 1.6 5 13 2.3 1.6 — 8 14 .3 11 15 14 0 11 23 15 .1 11 — Missouri River , South Dakota, Gavins Point Dam 5 15 3 2 1.5 10 14 -.2 4 15 23 1.8 2 19 23 0 2 — Des Moines River, Iowa, Red Rock Dam 9 20 2.2 12 2.3 Smoky Hill River, Kansas, Kanopolis Dam 3 5 1.7 .8 .8 4 4 -1.0 .8 13 14 1.1 .8 23 — .8 — Wolf Creek, Oklahoma, Fort Supply Dam 7 y> 7 — .3 .3 19 y> 7 .3 27 y> 7 — .3 (Mostafa, 1957; Albertson and Liu, 1957; Hales and others, 1970). Such lengthening occurred downstream from 9 of the 11 dams for which this feature could be determined (table 7). Lengths as of the latest resurvey ranged from 4 km on the Neches River downstream from Town Bluff Dam, Texas, to more than 120 km on the Colorado River downstream from Hoover Dam, Table 7.—Data on the degraded reach downstream from dams—Continued Years after dam closure Location of front of degraded reach downstream from dam (kilometers) Rate of advance (kilometers per year) Distance from dam to site of maximum degradation (kilometers) Location of first measured section downstream from dam North Canadian River, Oklahoma, Canton Dam 1 7 <1.8 1.8 2.8 7 — <1.8 3.4 7 — <1.8 11 7 — <1.8 18 7 — <1.8 — Canadian River, Oklahoma, Eufaula Dam 6 16 2.7 .8 .8 14 29 1.6 .8 Red River, , Oklahoma, Denison Dam 3 7 2.3 15 .6 6 —^>27 15 16 y> 27 1 27 y>n - 15 — Neches River, Texas, Town Bluff Dam 9 — 14 4 — .2 Chattahoochee River, Georgia , Buford Dam 7 7 1 2 .5 9 9 1 2 12 — 2 15 10 .2 2 - Distance of farthest cross section that was established at time of dam closure. Califomia-Arizona. In most of these cases, there is no indication that the reach had stopped lengthening by the time of the most recent survey. This means that the zone of degradation can continue increasing in length for at least 30 years or more after dam closure, although it could stop sooner. The migration rate and the final length of the degradation zone should vary with flow releases, bed-material sizes, and topography. Consequently, growth rate and eventual length are likely to vary from one dam to another. Migration of the front of the degraded zone means that at a downstream site a response time or lag time occurs before the bed reacts to the dam, if it is going to react. For some dams, this response time (and hence the migration rate of the edge of the degraded zone) could not be determined, because: (1) Cross sections were not established far enough downstream; (2) downstream measurements were not started until too many years after dam closure; or (3) a downstream base level interrupted or controlled the normal degradation process. A probable example of the latter is Wolf Creek downstream from Fort Supply Dam, Oklahoma. This stream joins the North Canadian River 6 km downstream from the dam. Successive profiles showed a hinge or base-level control at or near the confluence with the larger river. Similarly, the zero degradation point downstream from Town Bluff Dam on the Neches River, Texas, is sea level (Gulf of Mexico). Bedrock outcrops appear along the Republican River26 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS downstream from Harlan County Dam, and there are cobble riffles that act as controls on the Red River downstream from Denison Dam. Within a year or two after dam closure, the length of the degraded reach can range from little or nothing to as much as 50 km. After 2 or 3 decades, the length downstream from some dams remained as short as a few kilometers (and theoretically could be much less), but downstream from Hoover Dam it was more than 120 km (table 7). Hales and others (1970) proposed a method for predicting the temporary length of the degraded zone, based on 15 years or less of data for the Missouri River downstream from Garrison, Fort Randall, and Gavins Point Dams. This channel length in their treatment is a function of an average peak discharge, the time during which the degradation has been occurring, the dominant size of the bed material, and the area of the channel cross section. We have not tested their relation, mainly because of uncertainties in their definition of the peak discharge, uncertainties in the location of the cross section at which the area and particle sizes are to be measured, and their definition or way of determining the dominant grain size. Similarly, because of the few instances in which the degraded reach had stopped lengthening, a test of the method that Priest and Shin-dala (1969b) proposed for predicting that ultimate distance was not possible. Migration rates of the leading edge ranged from very little to as much as 42 km/yr (kilometers per year) immediately after dam closure (table 7). Slower rates for subsequent periods ranged from virtually negligible to about 29 km/yr. Most of the travel rates range from about 0 to 2 km/yr. According to Makkaveev (1970, p. 109), his countryman Fedorov determined migration rates of several kilometers per year on large lowland rivers, and several tens of kilometers per year on mountain rivers in the Soviet Union. The rate of advance of the downstream edge of the degraded zone depends on the flow releases and bed materials; these vary widely from one stream to another. According to the data in table 7, the rate on any river is not constant; the front occasionally may appear to retreat for isolated periods, even though the long-term trend is downstream. Migration rates appear to be fastest during the years immediately after dam closure. The relatively slow rates of subsequent years might be expected in some cases due to a flattening of gradient (discussed below); however, variable flow releases also will affect the rate with time. Whether the rates eventually become constant or continue to get slower with time cannot be determined from available data. ZONE OF VARIABLE BED CHANGES Cross sections downstream from the degraded zone may aggrade, degrade, or stay at the same level (table 13). There is some uncertainty as to whether bed-elevation changes in this downstream zone are due to the dam. Cross sections were not established prior to dam construction; therefore, the investigator does not have the benefit of this control. Marked trends, such as sudden and deep degradation typical of many cross sections near the dam and of the time when changes began, are not readily apparent on many measured sections. Most bed changes shown by the gaging-station data for control stations (table 14 and figs. 36-49) do not show trends. For these reasons, there is little basis for believing that the dam caused any observed changes in bed elevation beyond the degraded zone. Availability of ground and aerial photographs eliminates much of this uncertainty in regard to channel width and density of vegetation, but does not help to define bed elevations. We, therefore, have not evaluated observed fluctuations in bed level in the reach beyond the degraded zone. It is possible that degradation results in aggradation at some point downstream. Borland and Miller (1960, p. 70) noted that after closure of Hoover Dam in 1935 and Davis Dam about 1950 on the Colorado River, degradation downstream from the dams increased the aggradation in a reach farther downstream at Needles, California. Similarly, while only small changes in the overall longitudinal profile of the Rio Grande occurred after closure of Elephant Butte Dam and reservoir, J. F. Friedkin (International Boundary Commission, written commun., 1959) noted degradation of 1 to 2 m just downstream from the dam and deposition of about 1.5 m at El Paso, Texas, about 225 km downstream. These data are suggestive but are too limited to support a generalization regarding downstream aggradation associated with upstream degradation. Our study provides no additional data. A related intriguing possibility is that enlargements in channel width (discussed in detail below) could result in downstream aggradation. On the Missouri River downstream from Garrison, Fort Randall, and Gavins Point Dams, and on the Red River downstream from Denison Dam, significant increases in channel width at some cross sections are associated with bed aggradation near the approximate downstream edge of the degraded reach. LONGITUDINAL-PROFILE CHANGES To analyze changes in bed elevation with distance downstream (longitudinal profiles), we required at least four cross sections downstream from the dam andDOWNSTREAM EFFECTS OF DAMS 27 enough post-dam resurveys, bed degradation, inclusive time, and total downstream distance to reveal trends and features. Of the 21 dams (fig. 1), these requirements eliminated Davis, Jemez Canyon, John Martin, Medicine Creek, Milburn, Milford, and Red Rock Dams, leaving 14 dams for this particular analysis. Degradation models based on flume studies generally show maximum bed erosion at or near the dam, relative to the total reach undergoing bed changes (Ahmad, 1953; Mostafa, 1957; Aksoy, 1970; Hwang, 1975). In a general way, our data support that finding. The cross section of greatest degradation at a given time was the closest section to the dam in five cases (Gavins Point, Kanopolis, Fort Supply, Canton, and Eufaula Dams). Downstream from seven other dams, the greatest degradation was some distance, generally 2 to 16 km, downstream from the dam, but still generally nearer the upstream than downstream end of the degrading reach. (Variations in the downstream location of maximum bed erosion were mentioned by Wolman, 1967). For the two remaining dams, the location of maximum degradation was indeterminate. Due to the spacing of the cross sections and the natural variations of bed and bank erodibility with distance downstream, the data do not reveal how close to the dam the maximum degradation will occur when bed material is homogeneous with depth and distance. Results of Ahmad’s (1953) flume study indicate that the greatest degradation takes place closer to the upstream than to the downstream end of the degraded zone, but not right at the dam. Data in table 13 at least show more degradation closer to the upstream than downstream end of the degraded zone. Whether maximum degradation occurs immediately downstream from the dam needs to be determined by new measurements. Flume studies also indicate progressively less degradation with distance downstream, at a given time. For our data, this occurs in some reaches, but others do not seem to have a well-defined trend of degradation with distance. Instead, downstream from some dams, varying depths of bed erosion seem to be distributed almost randomly. For example, the data for the Colorado River downstream from Hoover Dam (table 13) show considerable variability in degradation depths with distance downstream. Within the general degraded zone, some cross sections had only minor bed erosion, while others degraded many meters by the same year. Because flows were the same for all sections and channel width did not vary significantly, such degradation differences probably are due to differences in bed erodibility (Stanley, 1951, p. 945). Variations with time also occur. If degradation is a maximum at or near the dam, then the channel’s downstream longitudinal profile should flatten with time as degradation proceeds. This process has been observed in the laboratory, along with the expected decline in the rate of degradation. At a given cross section, the sediment-transport rate decreases progressively with time as the bed slope (and hence stream competence) decreases. Transport eventually should cease if the slope becomes sufficiently flat (Tinney, 1962). Where no bed controls exist, Ahmad’s (1953) flume studies show that the point of maximum degradation migrates downstream with time. For most of our cross sections, maximum degradation either stayed at the same cross section with time (six dams) or varied from one cross section to another while showing a general preference for one site (seven dams), with one dam indeterminate. In several of the seven instances where the location varied with time, the first resurvey after dam closure showed maximum erosion at the cross section nearest the dam, but for later resurveys, the greatest bed degradation occurred at some fixed downstream cross section. In general, then, the site of greatest bed erosion tends to remain constant with time for the dams of this study, in which there is probably great variability of bed materials at or close to the surface. In nature, the bed profile downstream from a dam is affected by differences in bed material with both depth and distance, the presence of local controls, the history of flow releases, tributary contributions of water and sediment, and other factors. The profile downstream from a dam varies irregularly with time, and a uniform flattening of slope is not common. In most cases, the rate and depth of degradation are greater closer to the dam, but, in other respects, each dam is unique in regard to profile adjustment. Four examples are shown in figure 12. The Smoky Hill River downstream from Kanopolis Dam perhaps most closely approaches laboratory results and theoretical expectation, at least for the first 10 km or so downstream from the dam. Beginning at or immediately downstream from the dam, degradation decreases progressively downstream (fig. 12). The profile of the Colorado River downstream from Parker Dam is remarkably different in that only to a very slight extent is the expected flattening of the slope evident (fig. 12). Instead, degradation seems to be almost uniform throughout a reach at least 60 to 70 km long. In contrast, the channel profile downstream from Fort Randall Dam on the Missouri River, though generally tending to flatten with time, has widely varying degrees of bed-level change with time from one cross section to another (fig. 12). Degradation, no change, and aggradation all have happened at different downstream locations.28 < > DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS 30 T > o GQ < 00 CE 27 - 24 - 21 - 18 15 - 12 - 6 - i--------1-------T COLORADO RIVER, ARIZONA PARKER DAM Year of dam closure -------4 years after dam closure 13 years after dam closure \-----37 years after dam closure 25 50 75 100 125 i-------r COLORADO RIVER, ARIZONA GLEN CANYON DAM --Year of dam closure ■ vH--- years after dam closure KILOMETERS DOWNSTREAM FROM DAM Figure 12.—Longitudinal-profile changes downstream from four dams. Finally, the Colorado River channel throughout about a 25 km-long reach downstream from Glen Canyon Dam has undergone both a decrease and increase in slope with time (fig. 12). By 3 years after dam closure, the expected trend in degradation and flattening of slope had developed. However, degradation then ceased near the dam and, despite local irregularities such as cobble riffles, increased in the downstream direction. By 9 years after dam closure, this process had produced an average slope steeper than the slope at the time of dam closure. No change in the profile occurred during the following 10 years. Coldwell (1948) shows other examples of variability in the development of post-dam longitudinal profiles. The profile along the Green River downstream from Flaming Gorge Dam, Utah, is changing due to the development of rapids (Graf, 1980). The reduced (postdam) high flows are no longer able to move the coarse material. Some bed degradation near the rapids might accentuate the profile changes. Detailed records from the Rio Grande (J. F. Fried-kin, written commun., 1959) provide one of the bestBED MATERIAL AND DEGRADATION 29 illustrations of the variability of degradation and aggradation and their effect on the longitudinal profile. The reach for this example extends from Elephant Butte Dam, New Mexico, to and including a cross section downstream from El Paso, Texas; however, the upper reaches of the Rio Grande have similar problems (Lagasse, 1980). This complex case demonstrates both the effects of man (diversion dams) and the effects of sediment contributions from tributaries. From 1917 to 1932, immediately downstream from Elephant Butte Dam, the streambed degraded to a depth of about 1.8 m. Similarly, downstream from each of a number of diversion dams that control the channel elevation but provide little storage, degradation during 1917-32 ranged from 0.3 to 2.0 m. Due to this degradation and the downstream controls, the slope flattened downstream from each diversion structure. For example, in a reach 17 km long downstream from Percha Dam (46 km downstream from Elephant Butte Dam), from 1917 to 1932, the slope decreased from 0.00080 to 0.00065. Maximum depth of scour downstream from Percha Dam was about 2.0 m. In addition to the effects of these diversion dams, a number of steep arroyos, with intermittent large flows and large quantities of coarse material, periodically deliver that sediment to the Rio Grande. Because Elephant Butte Dam virtually eliminated downstream floods along the Rio Grande, the main channel can no longer transport the coarse material brought in by the arroyos. These sediment accumulations along some reaches block the channel and divert it completely. Along other reaches, such as those controlled by bank-protection works and jetties, such sediment accumulations provide a control by raising the elevation of the main stream at the confluence. This, in turn, induces deposition in the main channel for short distances upstream. The gradient of the Rio Grande is about 0.00028 to 0.00076, so deposition of coarse material can significantly flatten the local gradient. A river’s longitudinal profile and slope also can be affected by changes in river length or sinuosity. An increase in sinuosity (or in river length) has been noted in connection with local aggradation and vice versa (Hathaway, 1948; Ahmad, 1951; Frederiksen, Kamine and Associates, Inc., 1979). BED MATERIAL AND DEGRADATION THEORETICAL EXPECTATIONS Few if any natural channels are underlain by perfectly uniform sediments. Because magnitude and frequency of high flows are significantly decreased by dams, and because released flows may not be able to transport sizes previously moved by higher flows, suc- cessive flows can winnow finer materials from the bed. Progressive winnowing concentrates the coarser fraction. As degradation proceeds, the average particle size on the bed increases, possibly eventually resulting in a surface covering or armor of coarse particles alone. This idealized theory has long been accepted in engineering planning. Onsite and laboratory studies (Pemberton, 1976; Harrison, 1950; Little and Mayer, 1972) have demonstrated the importance of armoring in limiting degradation. In a general way, the number or extent of coarser particles should govern partly the depth of degradation in the cross section. Livesey (1965) has shown that as little as 10 percent coarse material in a standard sieved sample may be sufficient to provide the bed armor. (This underlines the importance of adequately sampling the surface and subsurface material for predictive purposes, before the dam is built. Representative sampling is difficult.) Livesey’s observations show further that a postdam armored bed need not be covered entirely by coarse material, and that the percentage covered is about 50 percent. The estimated gravel cover for the bed of the Red River downstream from Denison Dam, as obtained by pebble counts throughout long reaches of the river, indicates that 30 to 50 percent cover limits or controls degradation. Armor is a veneer underlain by normal or unwinnowed material. To date, onsite studies have not provided any proven examples of unravelling or unrolling of the veneer and reexposure of the subsurface sands. Assuming releases of large discharges from a dam, one would expect some unravelling of the surface. The extent should depend on the magnitude and duration of such excessive flows. Presumably restabilization and rearmoring of the bed should follow. The progressive changes in particle size in the vertical should have their counterparts along the longitudinal profile, as degradation moves progressively downstream with time. Thus, armoring of the bed should appear first close to the dam, then disappear somewhere downstream. VARIATIONS IN BED-MATERIAL SIZES WITH TIME AT A CROSS SECTION An unpublished U.S. Army Corps of Engineers report gives median grain size (d50) at different years for two sites downstream from Gavins Point Dam on the Missouri River. Various U.S. Bureau of Reclamation reports, for example U.S. Bureau of Reclamation (1948), show size-frequency curves for the bed material at different locations downstream from Hoover, Davis, and Parker Dams on the Colorado River. The variation of d50 with time for these Missouri River and Colorado River sites is shown in figure 13.MEDIAN BED-MATERIAL DIAMETER IN MILLIMETERS 30 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS YEARS AFTER DAM CLOSURE Figure 13.—Variation in bed-material size with time at a site, after dam closure. Initially, drM increased with time following dam closure, at least at cross sections near the dam. The magnitude of this increase can be more than a factor of 100 (and theoretically much more) in the value of d50, depending on the sizes and number of coarse particles in the reach. Within about 1 to 10 years after the start of the coarsening, the particle sizes seemed to stabilize. From the graphs in figure 13, stabilization occurred relatively abruptly rather than gradually, but such an impression may be due to the sampling intervals. In a few instances, the data suggest a subsequent reversal of the trend, that is, a decrease of d50 with time following the initial increase. Possible explanations, all speculative, are the arrival of finer material from upstream or from tributary inflow, the uncovering of finer material at some depth below the original surface, lateral movement of the channel, and sampling inaccuracies. Such a decrease in d50 may or may not reach a new stable value, judging from figure 13. Thus the changes in median size of bed material at these sites, while initially tending to coarsen as expected, did not follow an ideal or common pattern thereafter, but varied in several ways during later periods. Near the downstream end of the degraded zone, an increase of d50 with time may or may not occur. Where it does occur, it may lag behind the time of dam closure. Coarsening at these downstream sites seems to be less than that occurring near the dam. Whether this relatively limited coarsening is partly a function of distance from the dam, in addition to the distribution of particleCHANNEL WIDTH 31 sizes in the subsurface material and other factors, cannot be determined from the available data. Particle-size distributions given in U.S. Bureau of Reclamation publications for Colorado River reaches show that sorting as well as median size of streambed material downstream from dams varies with time. Some finer material is present in all samples, but later samples tend to have larger sizes (hence a wider range of sizes) than the earlier ones, as well as a greater percentage of coarse particles. Sorting, therefore, decreases with time. The presence of fine material in all samples may mean that such small grains really are on the bed surface, or it may result from the sampling technique, that is, the sample could include both surface and subsurface material. VARIATIONS IN BED-MATERIAL SIZES WITH DISTANCE DOWNSTREAM Some streams, such as the North Canadian River, have nearly constant sediment sizes for long distances downstream. Others, such as the Republican River downstream from Harlan County Dam in Nebraska and the Salt Fork of the Arkansas River downstream from Great Salt Plains Dam in Oklahoma, show great downstream variability in bed-material sizes. Such variability in these last two examples results in part from the sediment contributed by cliffs that abut the channel in places. Thus local geology can mask changes that might occur from dam construction. Where bedrock controls are absent and the bed of the river has a mixture of grain sizes, the postulated succession of particle size with distance occurs. Kira (1972, fig. 11) showed a gradual decrease in the mean diameter of bed-surface particles with distance downstream from Huchu Dam on the Aya River, Japan, as of 5 years after dam closure. Downstream from Kanopolis Dam on the Smoky Hill River and Denison Dam on the Red River, pebble counts of the sediment on gravel bars exposed at low water were obtained in 1960 throughout long reaches. Sieve analyses also were available for the bed material of the Colorado River downstream from Hoover Dam from U.S. Bureau of Reclamation sources. For these three rivers, the upper part of each of the three plots in figure 14 shows relatively coarse particles nearest the dam and a gradual grain-size decrease in the downstream direction. Bed-sediment analyses (discussed below) made when Hoover Dam was closed show that the bed material at that time was much finer than it was 6V2 years later (the year of the data plotted in fig. 14). The post-dam decrease of particle size with distance downstream, therefore, is reasonably attributable to the dam. Downstream from Kanopolis and Denison Dams, bed-material sizes were not measured at the time of dam closure. Therefore, one cannot say with certainty whether the post-dam trend resulted from the dam or whether it occurred naturally. However, the similarity of the two grain-size versus distance curves to one another and to that for the Hoover Dam data, along with qualitative agreement with theoretical expectations, indicate that the decrease in grain size probably is due to the dams. The lower plot for each dam in figure 14 shows variation in bed elevation with distance downstream, using the data of table 13 for the same year as the sediment-size data. The relative changes in grain size, degradation, and distance downstream then can be compared. If one assumes that the sizes of pre-dam channel sediment downstream from these three dams did not vary significantly with distance within the reach examined, then the relation between bed-material changes, degradation, and distance downstream agrees with the theoretical model described above. Reading the associated values of grain size and degradation at successive distances from the smoothed curves in figure 14, the curves in figure 15 were drawn to show the increase in bed-material grain size writh degradation for each study reach. This shows more graphically the increase in bed-material sizes relative to the depth of bed degradation. The curves in figures 14 and 15 might have been different in position on the graph if the data had been measured at some other time after dam closure; however, the trend would not be affected. U.S. Bureau of Reclamation data permit an evaluation of how the grain size-distance relation varies with time for the Colorado River downstream from Hoover Dam (fig. 16). During the first year or so after dam closure, the reach that underwent changes (coarsening) in bed-sediment sizes was somewhat less than 10 km long. After 3 years, coarsening was quite noticeable at 20 km but not at 70 km downstream from the dam; and by about 6 or 7 years after closure, coarsening was apparent 70 km, but not at 135 km, downstream from the dam. Coarsening did not seem to progress to the site 135 km downstream from the dam until about 13 years after closure. CHANNEL WIDTH GENERAL NATURE OF WIDTH CHANGES Channel widths downstream from the dams of this study narrowed, widened, or remained constant, depending on the site, in the years following dam closure (table 13). In general, the post-dam changes in channel width at a cross section as documented by measured cross sections, photographs, and maps, can be divided into five categories.DEGRADATION, IN METERS GRAIN SIZE, IN MILLIMETERS 32 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS 100 p- l------r SMOKY HILL RIVER, KANSAS KANOPOLIS DAM ABOUT 13 YEARS AFTER DAM CLOSURE' 10 - J_______L I RED RIVER, OKLAHOMA-TEXAS . DENISON DAM ABOUT 16 YEARS AFTER DAM CLOSURE DISTANCE DOWNSTREAM FROM DAM, IN KILOMETERS Figure 14.—Variation in bed-material diameter and bed degradation with distance downstream, at a given time after dam closure. Smoky Hill and Red River data are median diameter from pebble counts on gravel bars; Colorado River size data are dM for entire sample of sieved bed material. Degradation data are from table 13.CHANNEL WIDTH 33 DEGRADATION, IN METERS Figure 15.—Increase in bed-material size with bed degradation downstream from three dams (as estimated from the smoothed curves in figure 14). DISTANCE DOWNSTREAM FROM DAM. IN KILOMETERS Figure 16.—Variation in median bed-material diameter with distance along the Colorado River downstream from Hoover Dam, at successive times after dam closure. The first category is a statistically constant width, in which the width at successive times is within about ±4 percent (an arbitrary figure) of the width at the time of dam closure. In table 13, 231 cross sections downstream from 17 dams have meaningful width data. The width has remained virtually constant at 51 of these sections (about 22 percent of the total). (Such percentages are affected by the number of measuring sections downstream from the various dams and do not necessarily reflect relative frequency of the five categories of width change.) Channel width in canyons, such as occur along some reaches downstream from Colorado River dams, obviously is constrained. Such sections were excluded from the total of 231 considered here. However, Howard and Dolan (1981) report that fine-grained terrace materials in depositional reaches on the Colorado River downstream from Glen Canyon Dam are being reworked by flow releases, resulting in slight channel widening. A second category of channels widened, where widening arbitrarily is defined here as the most recently measured width being at least 5 percent greater than the width at the time of dam closure. About 46 percent (105 cross sections) are in this category. Although sometimes the channel has become about twice as wide during the post-dam period, most increases as of the latest resurvey were less than about 50 percent. Pronounced widening occurred at some cross sections downstream from Fort Peck, Gavins Point, Medicine Creek, Town Bluff, and Fort Randall Dams, but widths at other sites downstream from these dams did not change significantly. Also, changes in width were not consistent with distance downstream. Minor increases in width (less than about 15 percent) happened at a number of cross sections downstream from Milburn, Milford, Kanopolis, Red Rock, and Buford Dams. However, the magnitude varied considerably with distance along the river. Category three consists of channels that have become narrower. Using the arbitrary 5 percent criterion, 59 cross sections (about 26 percent) are in this group. About one-half of these are located downstream from Jemez Canyon, John Martin, Fort Supply, and Canton Dams (figs. 17-20). These channels are now only about 17 to 50 percent of their pre-dam widths. The fourth category includes channels that widened initially after dam closure, but later reversed this trend, and were most recently narrower than at the time of dam closure. Twelve cross sections (about 5 percent) are in this group. The North Canadian River downstream from Canton Dam, Oklahoma, has several such sections. The fifth category, including only 4 of the 231 cross34 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Figure 17.—Jemez River downstream from Jemez Dam, New Mexico, A, April 1936; B, spring 1951; C, June 1980. Dam was closed in 1953. sections, shows an initial channel narrowing followed by widening. The channel as of the latest resurvey was wider than at dam closure. Changes in width seem to have occurred at least from the time of dam closure; such changes tend to accompany changes in bed elevation. However, as with bed Figure 18.—Old streamflow-gaging site on Arkansas River 3 kilometers downstream from John Martin Dam, Colorado. A, March 1946; B, September 1959; C, July 1980. Dam was closed in 1943. degradation, there can be a considerable lag time before effects become noticeable at some of the downstream sections. Examples occur along the Red River downstream from Denison Dam.CHANNEL WIDTH 35 Figure 19.—Wolf Creek about 2.6 kilometers downstream from Fort Supply Dam, Oklahoma. A, April 1940; B, September 1958; C, August 1972. Dam was closed in 1942. Figure 20.—North Canadian River about 0.8 kilometer downstream from Canton Dam, Oklahoma. A, about 1938; B, July 1980. Dam was closed in 1948. Both scenes are looking downstream at the highway bridge. DISTANCE AFFECTED For those rivers having significant increases or decreases in width, the changes extend at least to the farthest measured cross section. (In most instances this was well beyond the zone of bed degradation.) Thus, the extent of a reach over which width has changed cannot be determined due to lack of data; however, it can be many tens of kilometers. No downstreamward trend in the magnitude of change in width is discernible for most reaches. This is true whether one considers the degraded zone alone or the entire reach for which data are available. For example, width changes do not seem to be greater near36 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS the dam. Rather, changes in most cases appear to vary randomly with distance or to remain about constant. FACTORS AFFECTING CHANGES IN CHANNEL WIDTH ALLUVIAL-BANK MATERIALS Data on bank materials were not available for most of the cross sections and reaches described here. However, some analyses and onsite observations illustrate the variety of bank-material factors that affect channel widening. At different locations along the Missouri River downstream from Garrison Dam, two distinctive types of channel bank occur. In one location (fig. 21) about 10 km downstream from the dam, the entire bank is composed of almost uniform sand (median diameter 0.17 mm (millimeter), standard deviation 0.027 mm). Erosion of these sugar sands, as they are called, appears to be a function of the shear of the flow against the surface of the bank. The rate of erosion is likely to be proportionate to the discharge and time the bank is subjected to the flow. Moderate fluctuations in flow, without major high flows, result in erosion of the sand deposit at the base of the bank, forming a narrow beach. With minor changes in water stage, this beach provides some protection against further erosion of the bank. The river bank in figure 21 eroded at a rate of about 3.6 m/yr between 1946 and 1957. (This approximate rate is mentioned only in a general sense and is not meant to show any effect of Garrison Dam, which was closed in 1953.) In contrast to the uniform sands in the bank in figure 21, the banks at other cross sections consist of layers of sand interbedded with finer-grained strata (fig. 22). The sand, about 1 m thick, is overlain by stratified silts (median diameter 0.009 mm, 99 percent finer than 0.074 mm) about 5 m thick. Low water on the outside of the bend impinges directly upon the sand, which is eroded readily by the continuing flow, even at low stages. The bank collapses by undercutting, with large blocks dropping vertically into the flow. Such silt-clay blocks retard bank erosion for a time, but eventually disintegrate and then are transported by the continuing flow. The bank in figure 22 eroded at a rate of 73.2 m/yr during 1946— 57. Similar banks composed entirely of sand erode even more rapidly. This is a very rapid rate of erosion, but even at other cross sections downstream from Garrison and other Missouri River dams, the erosion rates generally exceed 20 or 30 m/yr (table 13; Rahn, 1977). All manner of permutations and combinations of bank materials and stratigraphy occurs on the Plains rivers, which dominate the sample of rivers studied here. On the Missouri River in the 100 km reach downstream Figure 21.—Sandy bank of Missouri River about 10 kilometers downstream from Garrison Dam, North Dakota. Figure 22.—Stratified sand and silt bank, Missouri River downstream from Garrison Dam, North Dakota. from Garrison Dam, the percentage of silt (particles less than 0.074 mm) in the banks ranges from 3 to 100 percent. The bank commonly has thin strata containing large percentages of silt and some clay; however, on the average, silt and smaller sizes constitute no more than 33 percent. In this reach, the average of samples of bed material contained less than 2 to 10 percent siltCHANNEL WIDTH 37 size or finer. In contrast, a representative sample of bank material from 1 m above low water on the Smoky Hill River downstream from Kanopolis Dam had 75 percent of the particles finer than 0.074 mm. Samples from the bed and banks of the Salt Fork of the Arkansas River illustrate both the stratigraphy of the flood plain or channel banks and the contrasting character of bed and bank materials. As the data in table 8 show, 54 percent of the bed material is larger than 0.5 mm (coarse sand). With increasing distance above the bed, the proportion of silt-clay in the channel perimeter increases; that is, the percentage of coarse and medium sand decreases. Only in the upper 0.5 m of the flood plain is the percentage of silt and clay appreciable, a fact clearly evident in the stratigraphy of the bank as seen at the site. For the 2 m-high bank as a whole, 75 percent of the vertical section is composed of sand coarser than 0.125 mm. The remaining 25 percent is very fine sand or smaller. Considering the entire bank as a whole, the percentage of silt and clay (weighted according to the proportion of the vertical section described by the sample) is about 12 percent. For the few rivers where bank materials were examined in detail, no general and simple correlation could be made between erosion rates and the percentage of sands or silt and clay in the banks, except for isolated examples along the Missouri River and for straight reaches several kilometers downstream from Kanopolis Dam on the Smoky Hill River, where erosion of the silty banks appeared minimal. (Bank erosion on the Smoky Hill River, however, was significant at bends or where the thalweg of the channel meandered.) Although cohesive banks retard erosion, tests of stability criteria based on a weighted silt-clay content in bed and banks, using the method proposed by Schumm (1960, fig. 10, p. 23), indicate that measured channel sections known to be either aggrading, widening, stable, or unstable are not distinguishable on the basis of the width-depth ratio and weighted mean percentage of silt and clay. The difficulty appears to be that weighting of the particle size of the sediments by the channel width significantly distorts a controlling relationship between actual differences in bed and bank sediments. Generally, a cohesive bank will limit both channel width and the tendency to bank erosion or lateral migration; however, many other factors occurring simultaneously appear to dominate in the control of bank erosion. BEDROCK-BANK CONTROLS AND DOWNSTREAM EFFECTS Several cross sections on the Missouri River downstream from Garrison Dam indicate that channel shifting and bank erosion may increase downstream from cross sections at which bank erosion is controlled or retarded. Bedrock on one or both sides of the valley constricts the valley and channel in places. Lateral erosion at such constrictions usually is minimal, but in the expanding valley width downstream from such controls, erosion of one or both banks is relatively much greater. Further work is needed to determine whether the lateral erosion downstream from the constricted sections is greater than it would be without the constrictions. WATER FLOW In a detailed analysis, Chien (1961, p. 751) showed that the shifting of a river’s course varies directly with the rate of rise and fall of flood flows, bed shear stress, relative width of water surface at peak floods and at bankfull stage, width-depth ratio at bankfull stage, and varies inversely with particle size. Chien also noted (1961, p. 744) that channel shifting is related to the Table 8.—Particle-size distributions of bed and bank material. Salt Fork, Arkansas River, domistream from Great Salt Plains Dam, Oklahoma (Total height of flood plain above water surface is 1.83 meters] River bank and bed features (Distance below flood plain, in meters) Percent finer than indicated size (millimeters) 0.002 0.004 0.008 0.016 0.031 0.062 0.125 0.25 0.50 1.0 2.0 4.0 Top surface of clay band in 42 50 59 69 79 87 93 99 100 silt 0.30-0.43 9 9 11 16 24 44 68 90 99 100 — — 0.49-0.67 4 36 96 100 1.47-1.22 1 6 77 100 1.49-1.77 0 4 25 84 98 99 100 Channel bed^ 0 10 46 88 99 100 Bar at water level 0 7 81 98 100 — —^Sand sample beneath shallow water in braided channel.38 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS downstream spacing of control points. These relationships probably are not precise correlatives of bank erosion, but channel shifting is related closely to bank erosion. It has not been possible to obtain sufficiently complete data with which to verify equations provided by Chien. However, many observed phenomena in the alluvial channels described here qualitatively support his conclusions. Observations on the Missouri River downstream from Fort Peck Dam in Montana indicate that bank erosion increases markedly with discharges equal to or greater than about 500 m3/s (cubic meters per second). This is equivalent to flows that occurred equal to or less than about 12 percent of the time prior to closure of the dam (U.S. Army Corps of Engineers, 1952, p. 37). The increase in erosion accompanying increases in flow, particularly at or near the bankfull stage, also is documented elsewhere (Chien, 1961, p. 741; Leopold and others, 1964, p. 88). Net erosion in terms of enlargements in cross-sectional area (width and depth increases) can occur even with decreases in certain flow statistics. A decrease in mean daily discharges and in peak discharges during the years immediately after dam closure on the Red River downstream from Denison Dam nevertheless was accompanied by about a 25 percent increase in the average bankfull cross-sectional area of the downstream channel. Reductions in those same flows in the North Platte River downstream from Guernsey Dam in Wyoming were accompanied by a doubling of the average cross-sectional area throughout a 5 km-long reach downstream from the dam. Thus in these cases the mean daily flow and the annual peaks do not reflect adequately the erosive flows. Along other channels, decreases in flows have been accompanied by decreases in cross-sectional area and in width. Flow reductions due largely to various dams probably have caused the observed decrease in width of the Platte River in much of Nebraska to as little as 10 to 20 percent of its 1865 width (Williams, 1978). Where the decrease in flow has been significant, as in the lower Rio Grande, J. F. Friedkin (International Boundary Commission, written commun., 1959) has shown that the channel almost may disappear as vegetation, windblown sand, and sediment deposited by low flows clog the channel. Comparable changes on the Canadian River downstream from Sanford Dam, Texas, are described later in this report. Sandstone Creek near Cheyenne, Oklahoma, provides one of the better-documented examples of cross-section decreases and channel narrowing due to dams, in this instance, a combination of dams (Bergman and Sullivan, 1963). Sandstone Creek has a drainage area of 277 km2. Land-treatment measures were begun during the 1940’s; by 1952, 24 floodwater-retarding structures and 17 gully plugs had been built in the watershed. Further construction continued during the 1950’s and 1960’s. During the 1950’s the hydrologic regimen of the stream was altered significantly (table 9). From 1951 to 1959, mean daily flow tended to increase as the number of days of zero flow decreased from almost two-thirds of the year to zero. In addition, a significant increase in the number of peak flows occurred during 1953-56, suggesting a brief period of increased rainfall. In 1954, the channel cross section still retained the box-like form characteristic of an arroyo (fig. 23). By 1961, however, a much narrower channel (about one-third the former width), stabilized by vegetation (grass, shrubs, and some trees) had formed within the original cross section. A new flood plain had been created, virtually as an inset fill. The effect of the new channel cross section and vegetation is illustrated by the decreases in cross-sectional area and flow at successive stages (fig. 23). The metamorphosis of Sandstone Creek seems to follow a pattern typical of a number of other dammed streams (see Frickel, 1972, p. 29; Gregory and Park, 1974; Petts, 1977). Once the larger flows are eliminated, the flows occupy a somewhat narrower channel. Vegetation commonly tends to become established on the lesser-used part of the old streambed. This plant growth probably traps sediment during any inunda- Table 9.—Flow data for Sandstone Creek near Cheyenne, Ok-lahoma, 1951-59 [m3/s = cubic meters per second] Water year Mean daily flow (m3/s) Days of zero flow Number of peaks greater 3 than 14 m /s 1952 0.028 222 0 1953 .020 184 2 1954 .36 114 5 1955 .14 39 7 1956 .075 87 3 1957 .25 62 2? 1958 .11 7 1 1959 .45 0 1 PERCENT DECREASE I £ CD cc ~ LU LU h- I g LU CD 2 < ~ CD AREA FLOW I 1 1 1 — 4 32 l \ .1961 \ \\ / 1 1 / “ 15 40 ~ 1954-->' * .'~V s ✓ — 12 23 0 3 6 9 12 15 METERS 1 _I___I---1---1--1 Figure 23.—Changes in channel cross section of Sandstone Creek, Oklahoma, at the streamflow-gaging station, 1954-61, caused by many upstream flood-detention dams (modified from Bergman and Sullivan, 1963).CHANNEL WIDTH 39 tions. The vegetated zone thereby aggrades (fig. 23) and becomes a new flood plain. The old flood plain becomes inactive (a terrace), rarely or never flooded. In this manner, the stream channelizes itself, commonly in more stable banks from the binding properties of the vegetation. Although it is intuitively obvious that the magnitude and frequency of flows must affect bank erosion, a precise characterization of such flows for purposes of a general equation has not yet been obtained. Some preliminary efforts to develop a general equation are described later in this discussion. WIDTH-DEPTH RATIO Analysis of a number of cross sections indicated that wide, shallow channels tend to increase in width at a somewhat greater rate than relatively narrow, deep sections. A large initial width-to-depth ratio indicates that bank material in such sections may be more erodi-ble, and that these sections are likely to predominate in braided reaches. Because such a process cannot continue forever, channels may narrow by taking a new course or by developing several distributary sections. TIME TRENDS OF CHANNEL WIDENING AT A SITE A dimensionless relative change in width can be defined as WtfWi, where W, is the bankfull channel width at the time of dam closure at the cross section of interest and Wt is the bankfull channel width t years later at the same section. A plot of this ratio with time was made for each cross section downstream from the 17 dams for which data (table 13) were available. On these plots, nearly 50 percent of the 105 cross sections that became wider have either too many aberrations, no noticeable pattern, or insufficient data to warrant an attempt to fit a line to the points (see fig. 24A for some typical examples of such cross sections). The trend of relative increase in width with time for the remaining 54 cross sections can be described by a simple hyperbolic equation of the same type used for bed degradation. As applied to relative channel-width changes, this equation has the straight-line form (WifW,) = c3+ c4 (1 It) (3) where c3 is the intercept; and c4 is the slope of the fitted straight line on a plot of WxIWt versus 1 It. ii x i o LU > I- < LU CC (A) Irregular rates of increase 1.5 0.5 ... T~ T Missouri River. North Dakota 61 kilometers downstream from Garrison Dam ____i________i____ i--------r Missouri River, South I th Dakot^* * 1.6 kilometers downstream from Fort Randall Dam _i________i_____ 15 0.5 10 20 30 0 10 20 1 1 Red River, Oklahoma-Texas 80 ki lometers — downstream from Denison Dam — i i Red 1 River, 1 Oklahoma Texas Missouri River, South Dakota ^ 11 kilometers downstream from downstream from Denison Dam i i Fort Randall Dam i i (B) "Regular" rates of increase 2.0 1,5 I I Medicine Creek. Nebraska 0.8 kilometer downstream from Medicine Creek Dam r2 =0,33 2.0 - 1.5 Missouri River, South Dakota 26 kilometers downstream / from Gavins Point Dam >-/ • f r2 =0.81 I J______L 2.0 - 1.5 30 1.0 YEARS AFTER -------1--------1--------1------- Missouri River, North Dakota j 47 kilometers downstream r from Garrison Dam — ,2 = _l = 0.91 J______ I -------1--------1------1------- Missouri River, South Dakota 14.5 kilometers downstream from Gavins Point Dam __--------- = 0 79 ________i______i_______ —i—i—i— Missouri River. South Dakota 43 kilometers downstream from Fort Randall Dam r2 =0.90 :_______i_______i________l -------1--------1--------1------ Missouri River South Dakota 48 kilometers downstream from Gavins Point Dam r2 =0.99 0 10 20 30 40 0 DAM CLOSURE 10 20 30 40 Figure 24.—Examples of relative increase of channel width with time: A, irregular rates; B, regular rates with fitted regression curves (dashed lines). Data from table 13.40 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS The coefficients are positive where width has increased, as in the present discussion, and negative where width has decreased. The reciprocal of c4 is the initial rate of change, in relative-change (WyWi) units per year. The reciprocal of c3 ordinarily would be the asymptote or eventual new value of WtIW\. However, with the present application, the value of WtfWi at t = 0 is 1.0 rather than 0. To adjust the data to an origin of 0, 1.0 first needs to be subtracted from each WtIWi before performing the regression. Consequently, the asymptote or final predicted WtfWx is (l/c3) + 1.0. Similarly, the value of WtfWi at any time, t, is (WW= —+1.0 (c3+T) Values associated with these channel-width regression curves, such as coefficients c3 and c4, estimated final equilibrium values of W(/W1( time needed to complete the estimated total change (here given as 95 percent of the estimated change), time needed to attain 50 percent of the estimated total change, and the square of the correlation coefficient (r2) are given in table 10. As discussed in connection with bed degradation, several features that do not directly affect the goodness of fit influence the value of r2. For example, the value of r2 is somewhat sensitive to the location of the origin; that is, to the specification of the response time where such a lag period occurs. Hence, measurements in the first few years after (and before) dam closure are very important in defining the curve. The asymptote of the curve, or extrapolated eventual value of Wf/Wi, needs to be treated with caution. Where the basic data fit the curve, the extrapolated final value is valid. However, in several instances, the data show enough departures from a smooth curve that illogical values of the asymptote obtain. These few cases are noted in table 10. Based on the coefficients and the observed fit of the curve to the data points, the equations for 10 of the 54 sections are questionable. The remaining 44 cross sections, listed by dam and number of sections, were downstream from Fort Peck (3), Garrison (11), Fort Randall (3), Gavins Point (20), Medicine Creek (1), Kanopolis (3), and Denison Dams (3). The regression features for these widening cross sections show a wide range in initial rate of increase of channel width, predicted (or observed) final relative increase, and time required for the new width to develop. Some representative trends are shown in figure 24B. The initial rate of increase (reciprocal of the coefficient c4) for the 44 cross sections ranges from 0.0032 to 4.0 relative-change units per year. The predicted final values of W^/Wj (called (W,/Wi)max in table 10) as extrapolated from the regression curves (again keeping in mind the risks of extrapolation) range from very slight (1.05) to about 2.8. (The latter number would indicate that the final width would be 2.8 times the width at the time of dam closure.) The frequency distribution of these 44 values (fig. 25A) shows most of them closer to the smaller end of the range, with the mode at about 1.12. The estimated time needed for completion of 95 percent of the eventual change in width (see bed-degradation section for computation details) ranges from about 2 to nearly 1,900 years (table 10). The modal value of the 44 estimates is about 35 years (fig. 25B). Assuming no radical changes in the flow regime, most sections are predicted to need from about 1 decade to 600 years to complete their widening. Within this range, the longer durations (as much as hundreds of years) of course are mathematical results. We have no evidence that channel widening continues for such durations, and there is considerable evidence of discontinuity and change. As with bed degradation, much of the estimated widening occurs relatively quickly. One-half the total estimated overall increase in channel width can occur in as little as 1 or 2 months (table 10). For the 44 cross sections, the maximum estimate of the time needed for a section to complete 50 percent of its widening was 100 years. The distribution within this range (fig. 25C) has its mode at about lVz to 2 years. At most cross sections, 50 percent of the total eventual increase in width probably occurs within 2 or 3 decades after dam closure, according to these data. The above estimates of magnitudes of eventual widening and of adjustment time also apply to many cross sections for which the data were not fitted by a regression curve, judging from plots of WtrW\ versus time (fig. 24). Curves of relative increase in width with time (fig. 2AB) can all be combined onto one general, dimensionless curve (fig. 26) similar to the one for bed degradation. The ordinate in this case is the ratio of observed relative change in width (W/WO at a given time to the extrapolated maximum expectable relative change, the latter being approximated by 0.95 (W^/W^ma*. The abscissa on the plot is the proportion of total adjustment time that has elapsed, f/0.95 £max. Here the denominator (0.95 tmax) is 19 c4/c3 (as explained earlier). The dimensionless equation, referred to as the derived equation in figure 26, is identical to equation 2, for degradation, with the new dependent variable inserted. To the extent that the data fit the standardized curve, the same tendencies that described bed degradation with time also apply to the rate of channel widening. One-half the total change occurs during the firstCHANNEL WIDTH 41 Table 10.—Values associated with hyperbolic curves fitted to changes in channel width with time, at a cross section' [km, kilometer; yr. year; r2, square of correlation coefficient; c;t, coefficient (intercept) of fitted straight line on plot of W,/Wt versus 1/t; e4, coefficient (slope) of fitted straight line on plot ofW,/W, versus 1/t; W„ channel width at t years after dam closure; W,, channel width at time of dam closure] Distance of cross section downstream from dam (km) Response time (yr) & r c4 v 4/ 5/ Time to reach (wt) °.95 I-< > m I- cr ? < X o < oc Q. < DC 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 T7 1 1 - 9 CURVE OF - /DERIVED EQUATION^ A - /■ o 0,000 A - o £>/ A * 0.8 kilometer downstream from Medicine Creek Dam Nebraska ? A 14.5 kilometers downstream from Gavins Point Dam, South Dakota I □ 26 kilometers downstream from — Gavins Point Dam. South Dakota 1 o 43 kilometers downstream from 3 Fort Randall Dam. South Dakota ~ A 47 kilometers downstream from — Garrison Dam. North Dakota ■ 48 kilometers downstream from Gavins Point Dam, South Dakota 1 i j i i 0 0 0.2 0.4 0.6 t 0 95tmax 0.8 1.0 Figure 26.—Dimensionless plot of relative increase in channel width with time, for the 6 representative cross sections of figure 24B. Data from table 13. (At sections not describable by a hyperbolic curve downstream from these same four dams, the new width generally was only about 25 to 50 percent of the initial width, except for a few of the cross sections downstream from Canton Dam.) Theoretically, the relative decrease in width for a channel can range from almost 1.00 to 0, depending on flow regulation. The estimated time needed for the channel to reach its new, narrow width varies from 4 to 230 years for the 11 cross sections (table 10). Most estimates from the fitted curves are about a few decades or less. One-half the total adjustment can occur virtually immediately or within as much as about a decade. Less than 1 or 2 years is typical for the available data. The dimensionless standardized curve of the type applied above to bed degradation and channel widening is shown in figure 28. The derived equation is that of equation 2 with the appropriate dependent variable (proportional relative decrease in width). PREDICTION OF POST-DAM CHANNEL-WIDTH CHANGES Channel width depends primarily on water discharges and the boundary sediment. A multitude of regime- and hydraulic-geometry equations relate width to discharge. Unfortunately, most of those that are not site-specific require a resistance coefficient, a characteristic or dominant discharge, or both. Bed-material sizes change with time during the armoring process downstream from many dams (fig. 13), so even in the rare case where the size distribution had been measured adequately, it would be hard to build this changing particle-size variable into a resistance coefficient to predict eventual channel width. Similarly, identification of the most diagnostic or dominant discharge to use in an equation for44 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS (A) Irregular rates of decrease (B) "Regular" rates of decrease ~i i r~ North Canadian River 1 1 1 Wolf Creek, Oklahoma Oklahoma 134 kilometers 3.9 kilometers downstream ► f ♦ 1 from Fort Supply Dam downstream from \ r2 = 0.34 Canton Dam \ r2 -0.30 l I I 1 1 !_l ~\ i r~ Arkansas River, Colorado I I I Jemez River. New Mexico 36 kilometers downstream 1.0 kilometer downstream \ from Jemez Canyon Dam — '***-• 4. — from John Martin Dam ■\ r2 =0.37 r2 =0.91 L I I I* ~T~ | Wolf Creek, Oklahoma 2.9 kilometers downstrean from Fort Supply Dam r2 = 0.98 J_ I I I North Canadian River. Oklahoma 1.8 kilometers downstream from Canton Dan r2 = = 0.81 _L _L Figure 27.- 0 10 20 30 40 0 10 20 30 40 YEARS AFTER DAM CLOSURE -Examples of relative decrease in channel width with time: A, irregular rates; B, regular rates with fitted regression curves (dashed lines). Data from table 13. -------1-------1-------1-------1-------1-------1 I T • 134 KILOMETERS DOWNSTREAM FROM CANTON DAM. OKLAHOMA O 36 KILOMETERS DOWNSTREAM FROM JOHN MARTIN DAM. COLORADO ■ 2.9 KILOMETERS DOWNSTREAM FROM FORT SUPPLY DAM. OKLAHOMA 3.9 KILOMETERS DOWNSTREAM FROM FORT SUPPLY DAM. OKLAHOMA 1.0 KILOMETER DOWNSTREAM FROM JEMEZ CANYON DAM. NEW MEXICO 1.8 KILOMETERS DOWNSTREAM FROM CANTON DAM. OKLAHOMA 1.6 1 8 2.0 2.2 2 4 2.6 2.8 Figure 28.—Dimensionless plot of relative decrease in channel width with time, for the 6 cross sections of figure 21B. One point (for the section 134 kilometers downstream from Canton Dam) plots off the graph. Data from table 13.CHANNEL WIDTH 45 channel form remains an unsolved problem. An empirical effort was made (no acceptable theory being available) to determine those measures of discharge best related to width and to changes in width. Due to lack of data on bank cohesiveness, the search involved only water discharge. We used stepforward multiple regressions to test possible correlations between channel width and magnitudes, frequencies, and characteristics of discharge, namely: (1) Mean daily discharge; (2) average annual instantaneous peak flow; (3) single highest and lowest instantaneous annual peak flow; (4) highest average daily flow for consecutive periods of 1, 3, 7, 15, 30, 60, 90, 120, and 183 days for each year; (5) flow equaled or exceeded 1, 5, 10, 25, and 50 percent of the time; and (6) variability of flows within periods ranging from 1 to many years. Many different possible expressions for flow variability (item 6) involving the ratio of a high flow to a standardized flow were tested. Variations within a day, however, could not be considered, and such variations could be important on channels where flows are regulated by dams for hydroelectric power. Seasonal sequences were not explored. All of the above discharges, including ratios thereof, were examined for the pre- and post-dam periods, separately. The many discharge statistics, plus the log of each, amounted to 115 independent variables. For each river, the average width for all cross sections as a group was taken as the representative channel width for the particular year. These reaches in general have little significant tributary inflow throughout their lengths. Average width was calculated for the year of dam closure (first surveys of cross sections), yielding W\, and for the year of the latest resurvey, W2. The relative change in width is then WJW{. Nine cross sections had special local topographic features and were not included in the calculation of the average change in width for the entire reach downstream from a dam. These nine sections are downstream from a'total of 6 dams and probably do not affect significantly the regression results described here. Along some reaches, sparseness of cross sections is a drawback of this sampling approach to generalizing the change in width of a reach. Locations of cross sections is another possible disadvantage, in regard to: (1) Position around or near meander bends versus straight reaches, and (2) spacing with river distance downstream. Usually, the sections are close together immediately downstream from the dam and become farther apart with distance downstream. Ideally, the length of river reach within which the change applies needs to be standardized for the entire group of rivers. The first standardized length of reach we considered was 47 channel widths (from the most recent resurvey), this being the longest distance common to the 15 reaches for which enough data were available. Second, we tried defining the standard reach as the zone of bed degradation, again from the most recent resurvey. A third reach used—the entire distance covered by the measured cross sections—was not standardized for the group. Best correlations came from this last approach, probably because the greater number of cross sections provided a better representation. A general estimate of W2 downstream from the 15 dams is given by W2 = 13 + 0.5 Q,„ + 0.1 Q, (4) where W2 is the average bankfull width at the time of the latest resurvey, in meters; Qni is the arithmetic average of the annual mean daily flows during the post-dam period from dam closure to the latest resurvey, in cubic meters per second; and Qp is the arithmetic average of annual 1-day highest average flows for the pre-dam period of record, in cubic meters per second. Thus, both pre- and post-dam flows are represented, though by different flow statistics. As with many empirical expressions, the relation is not correct dimensionally. The r for the regression equation is 0.99, and the average absolute error in the predicted W2 is ± 19 percent. Computed versus observed values of W> are compared in figure 29. The ranges of values used in determining equation 4 are 30 < W2 < 939 m, 22 < Qp < 5,000 m'Vs; and 1.6 < Qm < 830 m'Vs (table 11). Average daily discharges differ slightly from those of table 4 because only flow data up through the latest channel resurvey were used for equation 4. Also, filling of the reservoirs for Fort Peck, Garrison, Fort Randall and Gavins Point Dams was not completed until about 1964, so mean daily discharges were computed beginning with 1965 for these dams. The period of reservoir filling for the other dams was assumed to be negligible. This empirical equation applies only to the ranges of data included in the analysis. For example, the equation may not be valid for dams which release little or no flow. We have no explanation for why the post-dam mean daily flow and pre-dam average annual 1-day high flow turned out to be the significant variables. Two sites with the required flow data were found to test equation 4. The tests are only approximate because the measurement or estimate of post-dam width is not made for a long reach of the river. The Canadian River at 3 km downstream from Ute Dam (closed in 1963) is shown in figure 30. Three measurements of46 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS MEASURED POST DAM AVERAGE CHANNEL WIDTH, IN METERS EXPLANATION NUMBER DAM AND STATE 1 Jemez Canyon, New Mexico 2 John Martin. Colorado 3 Fort Peck, Montana 4 Garrison, North Dakota 5 Fort Randall, South Dakota 6 Gavins Point. South Dakota 7 Medicine Creek, Nebraska 8 Red River, Iowa 9 Kanopolis, Kansas 10 Fort Supply, Oklahoma 11 Canton. Oklahoma 12 Eufaula, Oklahoma 13 Denison, Oklahoma-Texas 14 Town Bluff, Texas 15 Buford, Georgia Figure 29.—Computed (by the equation WL= 13+0.5Q,„ +0.1 Qp) versus measured values of post-dam average channel widths. channel width were made in 1981 at different sections along a nearly uniform Q.3-km reach (fig. 30). This reach is typical in regard to present channel width, according to the local U.S. Geological Survey engineers. The measured widths ranged from 43 to 46 m, averaging 44 m. According to equation 4, the width should be 55 m (Qm = 1.2 m3/s, Qp = 410 m3/s). This estimate is therefore slightly (25 percent) too large. The second test area is the Republican River downstream from Trenton Dam, Nebraska (fig. 31). The flow records provide Qm = 1.8 m3/s and Qp = 112 m3/s, for which equation 4 yields a width of 25 m. Judging from figure 31, the present width is an estimated 30 percent less than 25 m. None of the 115 variables correlated particularly well with the relative change in channel width (W2/Wj), although some approximate correlations will be mentioned below. With respect to W2/Wx, the 15 damsites can be divided into two distinct groups. The first includes the 11 dams downstream from which the channel either has undergone a slight widening, on the average, or has not changed appreciably. All hydropower dams, and some others, are in this group. The channels in the second group (downstream from Jemez Canyon, John Martin, Fort Supply, and Canton Dams) have narrowed considerably (as described earlier). The distinctive feature of the post-dam flow regime for the latter group seems to be that these channels convey little or no flow during a large part of the year. In contrast, the channels in which the width has remained constant or has enlarged are rarely dry and generally convey substantial (though not overbank) flows. Osterkamp and Hedman (1981) studied regulated Kansas streams on which the flow releases, though sustained, were not large and erosive. The channels tended to be narrower downstream from the dams than upstream. A general knowledge of a proposed dam’s release policy, therefore, might indicate whether significant channel widening or narrowing is likely to occur. Further study needs to be given to this possibility. Some of the narrowed channels may have conveyed little water during much of the year even during the pre-dam era; however, periodic floods then probably kept the channels wider. With the virtual elimination, or marked curtailment, of such high flows (table 4), low-flow periods appear to have assumed much greater importance. Such prevailing low flows form their own new (narrower) channel. Those high post-dam flows that are released may not be sufficient to maintain the former channel, especially since such flows generally are lower than pre-dam high flows (table 4). Vegetation has a better chance to become established on the lesser-used part of the streambed, and the course of events described above in connection with Sandstone Creek (table 9; fig. 23) can occur. Northrop (1965) reported similar processes on the Republican River in Nebraska, although flows there have been greater. Ws/Wi did show an approximate correlation with flow durations of low flows and also of certain high flows, namely: (1) The percent of time which a low flow equal to about 0.06 Qm was equaled or exceeded; (2) the percent of time a high flow equal to 8 Qm was equaled or exceeded; and (3) the percent of time a high flow equal to 0.1 times an estimated bankfull discharge was equaled or exceeded. In all three cases, correlation was improved by adding the average bankfull width-depth ratio as of the year of dam closure as a second independent variable. From these tests, it seems quite possible that flow durations help determine relative channel changes (W2/Wi); however, the general cause-and-effect relation remains unsolved. Part of the difficulty lies in the fact that the mechanisms are erosional in someCHANNEL WIDTH 47 Table 11.—Data used to derive post-dam channel-width equation [F, flood control; I, irrigation and water conservation; L, low-flow augmentation; M, municipal and industrial supply; N, navigation; P, hydropower; R, re-regulation of flow; S, sediment control; m, meters; m'Vs, cubic meters per second) Dam 1/ Dam no Year of Main purpose of Water years included in analysis Latest average width Relative change in width Post-dam average daily discharge % (m3/s) Pre-dam average 1-day high flows QP (m3/s) closure dam Pre-dam Post-dam W2 (m) Vwi 5. Jemez Canyon 1953 S,F 1937; 1944-52 1954-75 46 0.22 1.64 22.2 6. John Martin 1943 I, F 1914-41 1943-72 50 0.31 3.37 283 7. Fort Peck 1937 N,P,F 1929-36 1965-73 299 1.16 332 746 8. Garrison 1953 N,P,F 1929-52 1965-76 703 1.08 795 3,420 9. Fort Randall 1952 N,P,F 1948-51 1965-75 820 1.12 779 4,460 10. Gavins Point 1955 N,P,F 1948-54 1965-74 939 1.18 830 4,990 11. Medicine Creek 1949 I 1938-48 1951-78 47 1.18 1.88 170 13. Red Rock 1969 F,L, I 1941-68 1970-78 167 1.03 170 1,160 14. Kanopolis 1948 L,F 1941-47 1949-71 40 1.03 10.1 228 16. Fort Supply 1942 F,M 1938-41 1943-69 31 0.15 1.93 119 17. Canton 1948 F,M 1939-47 1949-71 30 0.47 5.32 219 18. Eufaula 1963 P,S,F 1939-62 1965-77 357 0.97 135 2,920 19. Denison 1943 P, F 1937-42 1944-69 373 1.10 120 2,760 20. Town Bluff 1951 R,I,M 1922-50 1952-65 126 1.19 119 1,110 21. Buford 1956 P,F 1942-55 1957-71 73 1.04 56.4 566 1/ In figure 1 and table 4. Figure 30.—Canadian River about 3 kilometers downstream from Ute Dam, New Mexico. A, August 1954; B, April 1980. Dam was closed in 1963. channels (those that have widened) but not in others (those that have narrowed). ROLE OF A DAM IN EFFECTING A CHANGE IN CHANNEL WIDTH Through control of water and sediment flow, the change in hydrologic regime associated with reservoir releases could result in an increase, decrease, or no change in downstream channel width. Channel widening conceivably might result from : (1) A decreased sedi- ment load in the flow, enhancing the capacity of the flow to entrain sediment from the bed and banks; (2) a decrease in the volume of sediment brought to, and deposited on or near, the banks, due to the reduced sediment transport and decreased high flows (net removal of material); (3) diurnal flow fluctuations (power or other controlled releases) causing consistent bank wetting and promoting greater bank erodibility; (4) bed degradation, where it occurs, resulting in flows impinging at a lower level on the banks, undermining vegetation and the higher section of the banks; and (5) rapid changes in flow releases (common with power dams) causing the river position to wander indiscriminately from one side of the channel to the other, encouraging48 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Figure 31.—Republican River downstream from Trenton Dam, Nebraska. Dam was closed in 1953. Looking downstream from bridge at Trenton (4 kilometers downstream from the dam), about 1949 (A) and July 1980 (B). Looking downstream from bridge at Culbertson (19 kilometers downstream from dam), July 1932 (C) and July 1980 (D). periodic erosion of first one bank and then the other without compensatory deposition. Whether the specific increases in width reported in this study are due to the dams cannot be determined because of lack of conclusive data, especially pre-dam cross-section measurements. The dam’s role on the four channels that have become narrower is clearer. Photographs of the Jemez River (fig. 17) show that little if any significant narrowing of the channel occurred from 1936 to 1951, while a very marked reduction in width took place sometime between 1951 and 1980 (dam closure was 1953). Measured cross sections (table 13) indicate a relatively wide channel around the time of dam closure and a striking decrease in width in the years immediately thereafter. No major reduction in water discharge occurred at the control station upstream from the dam during the postdam period (table 4), and 1978 aerial photographs show that the channel upstream from the reservoir is still relatively wide (about as wide as the pre-dam channel downstream from the dam). This upstream-downstream aerial-photograph comparison of reaches which are geographically near one another rules out any climatic effects or other factors that might be noticeable on other mountain streams in the western United States. Finally, the data of table 13 indicate that the overall bed degradation and channel narrowing on the Jemez River downstream from Jemez Canyon Dam have not been affected significantly by any changes farther downstream, such as on the Rio Grande. The channel narrowing since 1951 downstream from Jemez Canyon Dam, therefore, must be due to the dam. Wolf Creek downstream from Fort Supply Dam (fig. 19) in 1969 was only about 15 percent as wide as it was when the dam was closed in 1942. Aerial photographs taken in 1973 show a relatively wide channel upstream from the reservoir, compared to the narrow channel downstream from the dam. The similarity of the present upstream reach to the pre-dam channel upstream and downstream from the dam, coupled with the decreases in width shown by onsite measurements (table 13) and photographs (fig. 19) indicate that the radical post-dam decrease in width downstream fromSEDIMENT VOLUMES REMOVED AND CHANNEL EQUILIBRIUM 49 the dam very probably is due to the altered flow regime controlled by the dam. Measured cross sections (table 13) and photographs (fig. 18) of reaches downstream from John Martin Dam on the Arkansas River show a pronounced decrease in width (an average of nearly 70 percent for the WV^i values) after the 1943 dam closure. Such a radical change has not occurred upstream. Cableway discharge measurements at Las Animas, about 25 km upstream from the dam, show no change in the channel width during 1946-57, the period for which usable data are available. At Nepesta, about 100 km upstream from the dam, a similar analysis of cableway measurements for 1943-65 shows only about a 5 percent decrease in channel width. Two ground photographs of the latter site, taken in 1938 and 1963, also indicate no significant change in width. Aerial photographs taken in 1950 and 1970 seem to show a slight channel narrowing and an increase in vegetation for many tens of kilometers upstream from the reservoir during that period. The vegetation change had been occurring since at least 1936 (Bittinger and Stringham, 1963). Due to man’s extensive effect on the hydrology of the Arkansas River, some channel narrowing and vegetation growth probably would have occurred even without the dam. The differences upstream and downstream from the dam are large enough, however, that most of the channel narrowing downstream from the dam probably has resulted from the dam. Bankfull width at the streamflow-gaging station 4.8 km downstream from the site of Canton Dam was about 60 m in 1938, according to the station description of that year. In 1947, the first cross-section surveys downstream from Canton Dam (closed in 1948) showed channel widths of 65 m 3.1 km downstream from the dam and 47 m 5.0 km downstream from the dam. These figures indicate some, but not a major, decrease in channel width in the reach 5 km downstream from the dam during the 9 years before construction of the dam. According to the 1976 resurvey, the channel by then was 74 percent and 37 percent of its 1947 width at the same two cross sections. No control station at which water discharges and channel changes are unaffected by flow regulation is available for the North Canadian River at Canton Dam. However, the fairly stable predam width compared to the decrease in post-dam width indicates that much of the decrease in channel width is due to Canton Dam. SEDIMENT VOLUMES REMOVED AND CHANNEL EQUILIBRIUM Year-to-year estimates of volumes of sediment removed from the entire channel boundary within a finite reach can be determined from end-area measurements of cross sections. Such estimates, made by the Corps of Engineeers and Bureau of Reclamation, show many of the same features as bed degradation. For example, U.S. Bureau of Reclamation (1976) computations of this type for separate reaches on the Colorado River downstream from Davis Dam show that the largest volumes of sediment removal per year take place soon after dam closure. As years go by, the estimated volumes removed tend to approach zero net change. These tendencies agree with observed degradation and channel-width changes with time, described by the hyperbolic curve discussed above. Large differences, however, can be found from one year to the next, and in some years net deposition takes place. Net erosion in one reach can occur during the same year as net deposition in an adjoining reach. Similar data obtained by the U.S. Army Corps of Engineers for the Red River show how the volume of sediment removed varies with distance downstream. Successive times after dam closure also can be compared. A plot of cumulative volumes of sediment removed from the channel boundary as a function of distance downstream is shown in figure 32. A steep line on the plot indicates a large increase in the volume removed from one cross section to the next, or, in other words, a large volume of erosion has occurred throughout a unit downstream distance during the inclusive period represented by the plotted line. The steepness of the curve is proportional to the erosion rate for the unit reach. A horizontal line indicates that the cumulative volume removed, as of the survey year, no longer changes with distance downstream. In the latter case neither net erosion nor deposition occurs with distance, presumably an indication of a stable channel unaffected by the dam. Both curves in figure 32 show maximum channel erosion in the reaches closest to the dam, with the volume of erosion (slope of line) decreasing with distance downstream. In 1948, 6 years after closure, the reach of appreciable sediment removal extended downstream about 55 km. By 1958, the steep curve extended to about 90 km; even 160 km downstream, it had not become horizontal. From 1942 to 1948, the first 6 years after dam closure, the average rate of sediment removal from the first 25 km downstream from the dam was about 863,000 m'Vyr. By 1958, this rate had decreased to about 620,000 m'Vyr. The downstream patterns of degradation and channel widening discussed earlier show that the relative volumes of erosion of bed and banks along a given river are variable. The contribution from the bed appears to be greater closer to the dam; therefore, the longer the eroded reach (or the farther the subreach of interest50 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS O LU > O Figure 32.—Variation in cumulative net sediment volumes of channel erosion with distance downstream from Denison Dam on the Red River, Oklahoma-Texas. from the dam), the greater the relative contribution from the banks. Variations from one river to another also are considerable, as has been shown by Petts (1979) for British rivers. The U.S. Army Corps of Engineers (1952, p. 37) examined measured cross sections and estimated that, of the channel erosion downstream from Fort Peck Dam, about 60 to 70 percent of the sediment removed came from the banks and 30 to 40 percent from the bed. Estimates for the Red and North Canadian Rivers would make this percentage about 80 to 95 percent from the banks. In comparison, in certain reaches downstream from several dams on the Colorado River, the width is constrained, directing most of the erosion to the bed alone. The persistence of disequilibrium or the reestablishment of equilibrium in the relation of sediment inflow and outflow in reaches downstream from dams probably varies considerably from river to river. Clear water released from the dam may receive a new supply of sediment mainly from the channel bed, the channel banks, or from tributary inflows. Unless tributary inflows supply a relatively large proportion of the sediment, the regulated river has difficulty in regaining its former sediment load from the bed and banks alone. VEGETATION OBSERVED CHANGES IN VEGETATION Vegetation cover in and along channels downstream from the dams of this study either remained about the same or (most commonly) increased, following dam clo- sure. A decrease in vegetation after a dam was built was reported by other investigators in only one case, cited below. Noticeable, and in some cases very extensive, encroachment of vegetation onto former streambeds is apparent downstream from dams on the Jemez River (fig. 17), Arkansas River (fig. 18), Wolf Creek (fig. 19), North Canadian River (fig. 20), Canadian River (fig. 30), Republican River (fig. 31), and others shown below. Considerable vegetation has grown on the Platte River downstream from Kingsley Dam in Nebraska (Williams, 1978). A striking increase in vegetation has occurred on the Canadian River downstream from Sanford Dam, Texas (fig. 33), where virtually no releases of any magnitude have been made since dam closure in 1964. Due to the scarcity of major tributaries, the effect still is very pronounced 120 km downstream from the dam and probably much farther. Vegetation cover increased in direct proportion to the reduction in channel width. Studying the flood plain rather than the channel, Johnson and others (1976) reported a post-dam decrease in overall extent of forest cover and in certain kinds of trees downstream from Garrison Dam on the Missouri River. Green ash (Fraxinus pennsylvanica), however, increased. Vegetation changes in selected reaches downstream from 10 dams were mapped in the present study. Vegetated zones were marked on aerial photographs taken about the time of dam closure. About 7 to 13 years after the date of the aerial photograph, the same areas were visited, and vegetated areas again were mapped on the same aerial photographs. Of the 10 reaches examined, vegetation had covered as much as 90 percent of the channel bottomland in some cases (table 12). Seven of the 10 areas showed an increased growth of more than 50 percent. The alternative presence of willow (Salix sp.) or saltcedar (tamarisk sp.) for the sites in table 12 appears to be dictated at least in part by water quality. For example, saltcedar seems to thrive in the saline water of the Salt Fork of the Arkansas River in Oklahoma, while willow covers large areas on the Republican River in Nebraska. Differences are less apparent between the Arkansas, Canadian, and Republican Rivers. Distribution of vegetation in and along channel areas appears in at least three common patterns. In the first pattern, the increase in vegetation occurs in a strip along each bank. Turner and Karpiscak (1980) beautifully document such increases in riparian vegetation on the Colorado River between Glen Canyon Dam and Lake Mead, Arizona. Of their many sets of photographs, even those that were taken 0 to 13 years priorVEGETATION 51 Figure 33.—Canadian River downstream from Sanford Dam, Texas. Dam was closed in 1964. Looking downstream from about 400 meters downstream from damsite, October 1960 (A, prior to dam) and April 1980 (B). White arrow points to 1980 channel, about 5 meters wide. Looking upstream from railroad bridge near Canadian, about 120 kilometers downstream from dam, October 1960 (C, before Sanford Dam) and March 1980 (D). (Photograph credits: A and C, U.S. Bureau of Reclamation; B, U.S. Soil Conservation Service; D, U.S. Geological Survey.) to closure of Glen Canyon Dam (1963), compared to recent (1972-76) photographs almost all show a definite increase in vegetation. The authors note (p. 19) that “in the short period of 13 years the zone of post-dam fluvial deposits has been transformed from a barren skirt on both sides of the river to a dynamic double strip of vegetation. ” The Des Moines River downstream from Red Rock Dam, Iowa, also exemplifies this kind of distribution. Overbank areas that formerly had relatively frequent flooding now have significantly more trees. Most of the trees sprouted naturally, the remaining few having been planted by residents who found the land along the riverside much more habitable after dam closure upstream. In the second pattern, vegetation encroachment occurs within and adjacent to the former channel, leaving a much narrower single channel to carry the decreased post-dam flows. The succession of changes that a reach undergoes in this transformation is illustrated by the Washita River 1.4 km downstream from Foss Dam, Ok- lahoma (fig. 34). (Other smaller dams, farther upstream, also have affected this and other streams in this part of Oklahoma, as discussed below.) The Republican River downstream from Harlan County Dam in Nebraska exemplifies a third characteristic pattern. This is shown on aerial photographs taken in 1949 and in 1956 (fig. 35). The dam was closed in 1952. The discharge on the day of the 1949 photograph was 29 m3/s, whereas the regulated flow on the day of the 1956 photograph was only 1.8 m3/s. The vegetation changes are quite evident, nevertheless. In 1949 the channel shown in the photograph had the typical island and bar topography of a braided channel, with exposed expanses of clean white sand. In contrast, in 1956 the channel consisted of thin threads of open water in channels converging and diverging around “dark” islands fixed by vegetation. The vegetation consists of a dense growth of willows that form a virtually impenetrable jungle (see also fig. 31, Republican River downstream from Trenton Dam, Nebraska).52 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 12.—ChaTige in approximate percentages of riparian vegetation downstream from various dams River, dam, location Year of dam closure Post-dam time period analyzed Length of reach (kilometers) Average percent change in area covered by vegetation Type of vegetation Arkansas, John Martin, above 1942 1947-60 40 90 Saltcedar Lamar, Colorado Republican, Trenton, below 1953 1952-60 6 50-60 Willow Trenton, Nebraska 1960-80 6 85-95^ Willow Republican, Harlan County, 1952 1949-56 32 60-80 Willow near Franklin, Nebraska 1956-80 32 85-95^ Willow Republican, Harlan County, 1952 1950-56 26 65 Willow Superior, Nebraska Red, Denison, Denison, Texas 1943 1948-55 35 6 — Salt Fork, Arkansas, Great Salt 1941 1941-54 31 33 Saltcedar Plains, Jet, Oklahoma 1960 16 60 Saltcedar North Canadian, Canton, 1948 1960 16 30-50 Willow Oklahoma (local) Wolf Creek, Fort Supply, Fort 1942 1951-59 5 0 — Supply, Oklahoma 1959-72 5 80-90^ Grass, shrub, willow —^Estimated for short reach from ground photographs. POSSIBLE CAUSES OF VEGETATION CHANGES The roots of a plant are vital to its survival; therefore, the scouring effect of high flows can be devastating to vegetation. (The root depth and strength, the age and size of the plant and its trunk flexibility all affect a plant’s ability to withstand the scouring action of floods.) Even when a plant is not uprooted completely by a flood, germination and seedling survival generally depend on species flood tolerance. This in turn is a function of flood magnitude, frequency, and duration (Turner, 1974; Teskey and Hinckley, 1977). A reduction in such flood characteristics, therefore, often enhances vegetation survival and growth. If one deals only with the flood plain as opposed to the channel and banks, the effect of floods is less clear. Some trees, for example, may grow better under periodic flooding, especially where vigorous scouring is less active or less effective than gentler inundation. Johnson and others (1976) attributed a post-dam decrease in cottonwood (Populus deltoides Marsh), box elder (Acer negundo L.), and American elm (Ulmus americana L.) on the flood plain of the Missouri River downstream from Garrison Dam in part to the reduction of floods that formerly brought more nutrients and produced a higher water table. An increase in low flows has been thought to increase riparian plant growth. Such augmentation would raise the water table and increase the soil moisture, thus effecting an increase in vegetation. Some of the dams listed in table 12, such as Wolf Creek downstream from Fort Supply Dam, seem to support this thesis, insofar as an increase of both vegetation and low flows has occurred. However, a number of other dammed rivers, such as the Jemez and part of the Republican, have considerably reduced low flows, and yet vegetation also increased downstream from the dams on these rivers (figs. 17, 31, and 35). Thus, while increased low flow can encourage the spread of riparian vegetation along rivers, it does not appear to be a requirement, provided moisture is available. Ground-water withdrawals downstream from some dams have increased in recent years. Such withdrawals theoretically should lower the water table and decrease soil moisture, tending to inhibit many plant species. The importance of ground-water withdrawals in regard to post-dam vegetation changes could not be determined for the rivers studied here. Climatic changes could bring new conditions of temperature, humidity, and rainfall. The reaction of vegetation type and density to such changes may not be readily apparent. A period of less annual rainfall, forVEGETATION 53 Figure 34.—Washita River about 1.4 kilometers downstream from Foss Dam, Oklahoma. A, February 1958; B, May 1962; C, March 1967; D, February 1970. Dam was closed in 1961. example, could mean fewer flood peaks and an attendant establishment of vegetation, or it could mean less moisture in the ground and less vegetative growth. (Flood intensity and spatial distribution, which in turn depend on the intensity and distribution of precipitation, may be as important for plant survival as flood frequency. Total annual rainfall might not show changes in any of these factors.) In any event, changes in plant species might accompany climatic changes. Channel shape also could be a factor in vegetation changes. Little change can be expected on a narrow, deep channel. In comparison, a wide, shallow channel offers a better opportunity for vegetation to become established. Rate of channel meandering has not been treated separately in this paper. However, if sinuosity is affected by a dam (as mentioned briefly above), then rate of channel meandering also would change. Gill (1973) explains that the nature of the flood-plain plant community is very closely related to lateral migration of the channel. Johnson and others (1976) attributed a lack of young stands of cottonwood (Populus sp.) along the Missouri River downstream from Garrison Dam to a lesser rate of meandering after dam construction. The U.S. Fish and Wildlife Service studied the seed germination and seedling establishment of willows and cottonwoods along the Platte River in Nebraska,54 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Figure 35.—Republican River downstream from Harlan County Dam, Nebraska, before and after dam closure. Reach shown is near Bloomington, Nebraska. Flow was 29 cubic meters per second at time of 1949 photograph and only 1.8 cubic meters per second at time of 1956 photograph, but the increase in vegetation (dark areas) and the change in channel pattern are nonetheless apparent. downstream from Kingsley Dam (U.S. Fish and Wildlife Service, 1981). Favorable features were a significant soil-moisture content, bed material in the fine sand-to-sand size class, at least 1 to 2 weeks of ground exposure during the time when viable seed is available (mid-May to August), and moderate water depth during flooding. New seedlings survived only on bare sandbars, and not in upland shrub and woodland areas. Many factors, most of which could be affected or controlled by flow regulation, affect plant growth. SEPARATING FLOW REGULATION FROM OTHER FACTORS AFFECTING VEGETATION CHANGE A downstream increase in vegetation following dam construction does not necessarily mean the dam caused the change. As noted above, a number of factors, not all of which are dam-related, can affect vegetation. In addition, a particular plant may spread rapidly and even achieve dominance in a given region. Saltcedar growth, for example, is highly suspect as an indicator of the effects of flow regulation. This plant has been spreading at a rapid rate along innumerable valleys in the southwestern United States since its introduction late in the 18th century (Everitt, 1980). Although many of the valleys into which it has spread have been subjected to flow regulation, reaches or sections of many others have not. Larner and others (1974) concluded that, in view of the regional spread of saltcedar in west-central Texas since about the 1920’s, the observed accelerated increase in saltcedar downstream from various dams in that area meant that flow regulation by dams contributed to, but was not solely responsible for, the increase in riparian vegetation. On the Arkansas River, infestation by saltcedar is, if anything, more extensive on flood plain and channel bottom upstream from John Martin Dam, including several hundred kilometers beyond the backwater reach, than it is downstream from the dam (Bittinger and Stringham, 1963). FlowVEGETATION 55 regulation at John Martin Dam does not appear to be a sufficient explanation of the spread of saltcedar along this reach of the Arkansas. Similar growth has been observed on the Pecos River both upstream and downstream from Red Bluff Reservoir near the New Mexico-Texas State line (memorandum and photographs of Trigg Twichell, U.S. Geological Survey, December 21, 1961). In low areas along a 25-km reach of the Gila River valley in Arizona, saltcedar has become a dominant species of vegetation from 1944 to 1964. Turner (1974, p. 10) notes that changes in vegetation since 1914 have not coincided with channel changes. Moreover, while natural changes in flow regime have reduced winter flood frequency, and although increased summer low flow would enhance saltcedar growth, a decrease in cottonwoods does not correspond with changes in hydrologic regime (Turner, 1974 p. 13). Turner (1974, p. 18) concludes that, for the reach studied on the Gila River, neither disruption of the channel nor changes in flow regime account for the ascension of saltcedar to dominance over the indigenous vegetation. Rather, saltcedar competed succesfully with native plants and appears to be able to sustain its position indefinitely (Turner, 1974, p. 19). Climatic variability can complicate any attempt to determine the extent of channel changes and of increased vegetation growth attributable to dams. For example, a number of major dams were built in Oklahoma in the 1950’s and 1960’s. At the same time, hundreds of smaller flood-detention reservoirs were installed on tributaries. In the Washita River basin, 476 such reservoirs were completed from 1952 to 1972 (Carr and Bergman, 1976). Average annual rainfall in west-central Oklahoma during 1961-71 was about 12 percent less than during 1938-60. This reduced rainfall alone could have resulted in decreased streamflows and in observed changes in channels. In fact, streamflows during 1961-70 were decreased by as much as 60 percent, compared to the earlier period. Both dam construction and less rainfall probably were responsible for this reduction; although, given the very large changes in flow regime associated with the dams, their effect may well have been more significant than the change in rainfall. Even discounting possible effects of rainfall variability, the extensive simultaneous construction of small flood-detention dams and of dams on major rivers in parts of Oklahoma means that observed channel changes on the bigger rivers in those areas cannot be assumed to be entirely due to just the one dam immediately upstream. Thus, for example, some of the drastic changes on the Washita River downstream from Foss Dam (fig. 34) might have occurred even without Foss Dam because of the many upstream detention dams. Because vegetation can increase regardless of changes in post-dam low flows, an increase in vegetation cannot be attributed necessarily to low-flow augmentation from reservoir regulation. Regulation of high flows (magnitude, frequency, and duration) seems to be the only dam-related factor that is reasonably certain to encourage an increase in vegetation. Even with this feature, evidence on the extent to which the dam is accountable commonly may be absent. Little information exists on the response of riparian vegetation to changes in climate and hydrology unaffected by man. Vegetation changes comparable to those observed downstream from dams have occurred in the past in the absence of dams, though not as abundantly. Examples are on the Gila River (Turner, 1974; Burkham, 1972) and on the Cimarron River (Schumm and Lichty, 1963). Thus, although an increase in riparian vegetation due to flow regulation might logically be expected, the degree of the change ascribable to the dam cannot always be fixed from available data. Regulation of high flows in some cases could be virtually the sole cause of the change, while in other cases, it could be only a contributory part of the cause. In general, however, information from this and other studies indicates that the reduction of high flows by dams, if not controlling, often contributes significantly to the downstream growth of riparian vegetation, especially in cases where the channel has become narrower (figs. 17-20, 30-31, 33-35). EFFECTS OF VEGETATION GROWTH Channel vegetation blocks part of the channel, resulting in reduced channel conveyance, faster flow velocities in the channel thalweg or both. Conveyance is decreased both by physical reduction of flow area by the vegetation and by impeding the sediment transport process and inducing bed aggradation. On the Republican River in Nebraska, vegetation decreased the channel capacity by 50 to 60 percent in some reaches (Northrop, 1965). Such reduced conveyance leads to more frequent and longer-lasting overbank flooding. Faster velocities in the channel thalweg have been observed in some California streams, resulting in chutes where riffles used to be (John Hayes, California Fish and Game Commission, oral commun., 1980). Vegetation also enhances greater bed stability. Not only does vegetation impede the flow, but the roots help bind the sediment. Sediment within vegetated areas can be extremely difficult to erode. A potential effect of vegetation, though not documented specifically in this study, is greater bank stability, due to the binding and protective effects of the vegetation. Such bank stability would be enhanced by decreases in damaging flood flows.56 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Another potential major effect of significant new vegetation growth is an increase in water losses by evapo-transpiration. No comparative studies of water losses from sand channels before and after vegetation growth have been made. The only work done seems to have dealt with flood plain rather than channel vegetation. Similarly, it is still unclear whether more water is lost from a plain water surface than from one with a plant cover. On flood plains, comparisons of evapotranspira-tion before and after phreatophyte removal, as well as studies using evapotranspirometers, indicate possibly significant increase in consumptive use of water by phreatophytes compared to volumes for sand and bare soil (Van Hylckama, 1970; Culler and others, 1982; Lep-panen, 1981). Several studies (Meyboom, 1964; Bowie and Kam, 1968; Ingebo, 1971) suggest that increased vegetation depletes streamflow, but the variety of conditions under which this occurs is not yet established. The elevation of the water table also has an effect. Evaporation is decreased significantly if the water table is lowered 0.6 m (Hellwig, 1973, p. 106). CONCLUSIONS The large data set compiled and examined in this report includes 287 measured cross sections downstream from 21 dams. Each cross section was resurveyed periodically, under the auspices of the U.S. Army Corps of Engineers or the U.S. Bureau of Reclamation, since about the time of dam closure. We have analyzed 1,817 such cross-section surveys (table 13). For each resurvey, we determined the mean bed elevation and measured the bankfull channel width. In addition, gage height-water discharge relations at 14 streamflow-gag-ing stations (table 14, figs. 36-^49) were inspected. Thirdly, numerous supplementary observations and measurements (such as time-sequential photographs, grain-size measurements, and vegetation mapping) downstream from other damsites have been included in the study. Data published here and in many other reports show that the construction of dams on alluvial channels, by altering the flow and sediment regimen, is likely to result in a number of hydrologic and morphologic changes downstream. For example, average annual peak discharges for the rivers of this study were reduced by from 3 to 91 percent of their pre-dam values by the dams. Mean daily flows and average annual low flows were decreased in some instances and increased in others. On most of the alluvial rivers surveyed, the channel bed degraded in the reach immediately downstream from the dam. Channel width in some cases showed no appreciable change, but in others, increases of as Figure 36.—Changes in mean streambed elevation with time at streamflow-gaging station on Colorado River 6.4 kilometers downstream from Parker Dam, Arizona. Plotted points represent elevation corresponding to a discharge of 90.6 cubic meters per second, as determined from rating tables. No upstream control station available. much as 100 percent or decreases of as much as 90 percent were observed. At many cross sections, the changes in bed elevation and in channel width proceeded irregularly with time. At other cross sections, however, the average rates of degradation and also of changes in channel width can be described by a simple hyperbolic equation of the form: (l/y)=C,+C2(l/0CONCLUSIONS 57 Figure 37.—Changes in mean streambed elevation with time at streamflow-gaging station on Jemez River 1.3 kilometers downstream from Jemez Canyon Dam, New Mexico, and at the control station near Jemez 13 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 0.034 cubic meter per second downstream from dam and 0.37 cubic meter per second upstream from dam, as determined from rating tables. WATER YEAR Figure 38.—Changes in mean streambed elevation with time at streamflow-gaging station on Missouri River 13 kilometers downstream from Fort Peck Dam, Montana. Plotted points represent elevation corresponding to a discharge of 85 cubic meters per second as determined from rating tables. No upstream control station available. where Y is either bed degradation in meters or relative change in channel width; C ] and C2 are empirical coefficients; and t is time in years after the onset of the particular channel change. This model equation at present only describes observed channel changes. However, it perhaps could be- WATER YEAR Figure 39.—Changes in mean streambed elevation with time at streamflow-gaging station on Missouri River 11 kilometers downstream from Fort Randall Dam, South Dakota. Plotted points represent elevation corresponding to a discharge of 464 cubic meters per second, as determined from rating tables. No upstream control station available. D LU DO < LU cc I— CO LU C3 2 < X o WATER YEAR Figure 40.—Changes in mean streambed elevation with time at streamflow-gaging station on Missouri River 8 kilometers downstream from Gavins Point Dam, South Dakota. Plotted points represent elevation corresponding to a discharge of 312 cubic meters per second, as determined from rating tables. No upstream control station available. come usable for pre-construction estimates if a way could be found to predict the two coefficients, at least where subsurface and bank controls are absent. These coefficients probably are functions, at least, of flow releases and boundary materials. Research is needed to find a way of determining the coefficients prior to dam closure. Without a predictive equation, estimates of expected degradation need to be based on sediment-transport equations. The applicability of sediment-transport equations will depend on the channel-bed material, hydraulic characteristics, and depth to bedrock. The subsurface conditions are assessed best by detailed engineering and geologic surveys, such as excavations and core borings. It is difficult, however, to conduct such surveysdownstream effects of dams on alluvial fans 58 WATER YEAR Figure 41.—Changes in mean streambed elevation with time at streamflow-gaging station on Smoky Hill River 1.3 kilometers downstream from Kanopolis Dam, Kansas, and at the control station at Ellsworth 48 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 0.51 cubic meter per second downstream from dam and 0.43 cubic meter per second upstream from dam, as determined from rating tables. Figure 42.—Changes in mean streambed elevation with time at streamflow-gaging station on Republican River 2.7 kilometers downstream from Milford Dam, Kansas, and at control station at Clay Center 49 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 1.2 cubic meters per second downstream from dam and 3.4 cubic meters per second upstream from dam, as determined from rating tables. accurately. Core borings might fail to disclose coarse sediments at depth, and excavations or more detailed examinations may be required to find any controls. Even excavations may be insufficient if not suitably located. Extrapolation of a fitted hyperbolic curve to estimate future bed degradation or changes in width at a site probably will give reliable estimates in a number of cases, assuming no major hydraulic changes are introduced. However, bed degradation at some (possibly many) cross sections will not be as deep as the predicted bed degradation because of unassessed subsurface controls (coarse sediment or bedrock). Similarly, o LU CO S < UJ cc I— Cfi a i o WATER YEAR Figure 43.—Changes in mean streambed elevation with time at streamflow-gaging station on North Canadian River 4.8 kilometers downstream from Canton Dam, Oklahoma, and at the control station near Seiling 45 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 0.031 cubic meter per second downstream from dam and 0.00057 cubic meter per second upstream from dam, as determined from rating tables. Figure 44.—Changes in mean streambed elevation with time at streamflow-gaging station on Red River 4.5 kilometers downstream from Denison Dam, Oklahoma, and at the control station near Gainesville, Texas, 106 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 3.7 cubic meters per second downstream from dam and 4.2 cubic meters per second upstream from dam, as determined from rating tables. unassessed variations in bank erodibility can affect the predicted width changes. In the sites studied here, rates of degradation during the initial period following dam closure are about 0.1 to 1.0 m/yr, but ranged from negligible to as much as 7.7 m/yr. (Such rapid rates generally did not last for more than a few months). Rates at many sites became very slow after 5 to 10 years. The maximum depth of degradation varied considerably from one cross section to another and ranged from less than 1 m to as much as 7.5 m. On rivers havingCONCLUSIONS 59 D LU g WATER YEAR Figure 45.—Changes in mean streambed elevation with time at streamflow-gaging station on Neches River 0.5 kilometer downstream from Town Bluff Dam, Texas, and at the control station on Village Creek near Kountze in an adjacent drainage basin. Plotted points represent elevation corresponding to a discharge of 4.2 cubic meters per second downstream from dam and 1.5 cubic meters per second at the control station, as determined from rating tables. X o Figure 48.—Changes in mean streambed elevation with time at streamflow-gaging station on Marias River 3.2 kilometers downstream from Tiber Dam, Montana, and at the control station near Shelby 65 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 2.8 cubic meters per second downstream from dam and 4.0 cubic meters per second upstream from dam, as determined from rating tables. O WATER YEAR O LU m O Figure 46.—Changes in mean streambed elevation with time at streamflow-gaging station on Chattahoochee River 4 kilometers downstream from Buford Dam, Georgia, and at the control station on the Chestatee River near Dahlonega 73 kilometers upstream from dam. Plotted points represent elevation corresponding to a discharge of 12.2 cubic meters per second downstream from dam and 3.4 cubic meters per second upstream from dam, as determined from rating tables. O Figure 47.—Changes in mean streambed elevation with time at streamflow-gaging station on Rio Grande 1.3 kilometers downstream from Caballo Dam, New Mexico. Plotted points represent elevation corresponding to a discharge of 28.3 cubic meters per second, as determined from rating tables. No control station. slopes of about 1 to 3 m/km, degradation of as little as 1 m significantly decreases the gradient. Figure 49.—Changes in mean streambed elevation with time at streamflow gaging station on Frenchman Creek 0.3 kilometer downstream from Enders Dam, Nebraska. Plotted points represent elevation corresponding to a discharge of 1.3 cubic meters per second, as determined from rating tables. No control station available. Some of the rates and volumes of degradation in this study may appear small in the abstract. However, on a channel only 90 m wide and 15 km long, about 2 billion megagrams of sediment would be removed within 10 years from the bed of the channel alone, at the rates described. The consequences of such degradation can include undermining of structures, abandonment of water intakes, reduced channel conveyance due to flatter gradients, and a decreased capacity for the transport of sediment contributed by tributaries. Commonly, the section of maximum degradation in most cases was close to the dam, and degradation then decreased progressively downstream. However, large and small depths of degradation commonly were distributed somewhat irregularly with distance downstream from the dam. Also, the downstream location of zero degradation ranged from several to about 2,000 channel widths (4 to 125 km). For these reasons a smooth longitudinal profile is rare. In some cases not even the60 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS anticipated downstream decline in degradation was observed within the distance covered by the cross sections. Further, although the longitudinal profile downstream from many dams tended to flatten with time as expected, this did not occur in all cases. Changes in channel elevation limited even to 1 or 2 m can significantly affect the longitudinal profile on many rivers. Many analyses were performed in seeking correlations of variables that would characterize conditions before and after dam closure. No simple correlations could be established between channel size, channel gradient, particle size, or quantities of flow, with the exception of a tentative relation for channel width. This reflects the number of variables and great variability of conditions in the sample. In several of the rivers studied, bank erosion appears to account for more than 50 percent of the sediment eroded from a given reach. Bank erosion is related to bank composition. Erosion may be particularly severe where the river impinges on a bank of readily erodible sand. Fine-grained cohesive sediments may slow the rate of erosion at specific points. In large rivers flowing on sand beds, such as those found in many areas of the western plains of the United States, the location of controls, discharge, and fluctuations of discharge appear to be principally responsible for varying rates of bank erosion. Many large dams trap virtually all (about 99 percent) of the incoming sediment. The erosion of sediment immediately downstream from the dam, therefore, is not accompanied by replacement. Thus, although the rate of removal by the post-dam regulated flows may be less than that prevailing prior to regulation within a reach, the process does not result quickly in a new equilibrium. Both lateral erosion and degradation cease when the flow no longer transports the available sediments. Such cessation of net erosion may occur through local controls on boundary erosion, downstream base-level controls, decrease in flow competence (generally associated with armoring), infusion of additional transportable sediment, and through the development of channel vegetation. Armoring (increase in d50) appeared to be approximately proportional to the depth of bed degradation downstream from three dams for which data were available (fig. 15). Hundreds of kilometers of river distance downstream from a dam may be required before a river regains, by boundary erosion and tributary sediment contributions, the same annual suspended load or sediment concentration that it transported at any given site prior to dam construction. On the North Canadian River downstream from Canton Dam, this distance is about 200 to 500 km. On the Red River downstream from Denison Dam, the distance is about the same or possibly longer. On the Missouri River, 1,300 km downstream from Gavins Point Dam, the post-dam average annual suspended loads are only about 30 percent of the pre-dam loads. The Missouri and some other rivers probably are not long enough for complete recovery. Evaluation of the effects of dams on downstream channels is made difficult by the absence of adequate observations on the changes of natural channels in different climatic and physiographic regions under unregulated conditions. Natural variability that characterizes such changes (tables 1-3) may mask the response of the channel to flow regulation. To the extent that it is known, the geologic record indicates that small changes in climatic factors can produce significant alterations in channel morphology. This potential effect also complicates the identification of those changes in channel morphology and vegetation that can be ascribed solely to the effect of manmade structures. Some of the channel changes documented here might well have occurred during the period of observation even in the absence of human interference. However, several common trends should be noted, namely: (1) Frequent occurrence of major changes right after dam closure; (2) appearance in many cases of the greatest change just downstream from the dam with progressive decrease or recovery downstream; (3) progressive change toward an apparent new stability at a site, in the years after dam closure; (4) continuous or non-reversible character of the change at many locations; and (5) diversity of climatic and physiographic regions in which the process has been observed. These trends point to the installation of water-regulating dams and reservoirs and to the consequent elimination or significant decrease of sediment into downstream reaches as primary causes of the progressive channel change in a number of instances. Vegetation generally increased in the reaches downstream from the dams studied here, covering as much as 90 percent of the channel bars and banks along some rivers. In some cases, part of this increased growth might have occurred even without the dam. That is, vegetation in the region may have proliferated as a result of climate changes or for other reasons not fully understood. Decreases of high flows by the dam seem to contribute to an increased downstream growth of riparian vegetation in many cases. Most of the rivers investigated here are in a semiarid environment where the effective annual precipitation is between 20 and 40 cm. This is precisely the precipitation zone that Langbein and Schumm (1958, p. 1080) suggested is the critical point at which sediment yield may either decrease or increase, depending upon whether vegetation increases or decreases in response to a change in precipitation. The changes of the alluvialREFERENCES 61 channels downstream from dams, and, in particular, the changes in vegetation and channel morphology observed at a number of locations, indicate the sensitivity of these relationships to small changes within short times, or to the effects of unusually large changes at a given moment in time. These effects may be mitigated or reversed in several decades. However, it is still difficult to predict what the effect of a persistent but small change in runoff, for example, would be on a given reach of channel. Interestingly, environmental-impact analyses require predictions of just such changes. Where downstream channels are surveyed following dam construction, the usual method consists of topographic resurveys of fixed cross sections. These are measured either at predetermined, approximately equal time intervals (usually every 5 or 10 years), or on rare and sporadic occasions as funds permit. Such surveys instead need to be scheduled at frequent intervals (at least every 1 or 2 years) during the first 5 or 10 years after dam closure, because most of the channel changes occur during this period. Later surveys can be done much more infrequently, because much less change takes place during a unit time in these later years. The scatter on a plot of channel change versus time reflects, to some extent, the desirable frequency of resurveys. Where the scatter is large, shorter time intervals (more data) are needed to define a trend, and vice versa. Although successive surveys of cross sections provide essential data for analyzing sediment and channel changes, repetitive aerial photography keyed to specific water stages might well provide more satisfactory data for some purposes at less cost. Consideration needs to be given to monitoring some major streams by means of aerial photography, perhaps using infrared techniques, where photographs can be taken at specific water stages and seasons to make successive sets of photography comparable. Such comparability is virtually non-existent at the present time. REFERENCES Ahmad, Mushtaq, 1951, Causes of retrogression below Islam barrage and remedial measures to cause accretion and push the hydraulic jump to the toe of the Glacis: International Association for Hydraulic Research Meeting, 4th, Bombay, India, 1951, Appendix 17, p. 365-393. Ahmad, Nazir, 1953, Mechanism of erosion below hydraulic works: Minnesota International Hydraulics Convention, University of Minnesota, Minneapolis, Minnesota, Sept. 14, 1953, Proceedings, p. 133-143. Aksoy, Sahap, 1970, River bed degradation of downstream of large dams: International Congress on Large Dams, 10th, Montreal, Canada, June 1-5, 1970, Transactions, v. 3, p. 835-844. -----1971, Riverbed degradation downstream of dams: Congress of the International Association for Hydraulic Research, 14th, Paris, France, Aug. 29-Sept. 5, 1971, Proceedings, v. 3, p. 275-282. Albertson, M. L., and Liu, Hsin-Kuan, 1957, Discussion of Mostafa, M. G., River-bed degradation below large-capacity reservoirs: American Society of Civil Engineers Transactions, v. 122, 699-701. American Society of Civil Engineers, 1978, Environmental effects of large dams: New York, American Society of Civil Engineers, 225 p. Ashida, Kazuo, and Michiue, Masanori, 1971, An investigation of riverbed degradation downstream of a dam: Congress of the International Association for Hydraulic Research, 14th, Paris, France, Aug. 29-Sept. 5, 1971, Proceedings, v. 3, p. 247-255. Bergman, D. L., and Sullivan, C. W., 1963, Channel changes on Sandstone Creek near Cheyenne, Oklahoma: U.S. Geological Survey Professional Paper 475-C, p. C145-C148. Bittinger, M. W., and Stringham, G. E., 1963, A study of phreatophyte growth in the Lower Arkansas River valley of Colorado: Fort Collins, Colorado, Colorado State University, Civil Engineering Section Report CER63MWB-GES6, 31 p. Borland, W. M., and Miller, C. R. 1960, Sediment problems of the Lower Colorado River: Proceedings of the American Society of Civil Engineers, Journal of the Hydraulics Division, v. 86, no. HY4, p. 61-87. Borsuk, O. A., and Chalov, R. S., 1973, Downcutting on the Lena River channel: Soviet Hydrology: Selected Papers, no. 5, p. 461— 465. Bowie, J. E., and Kam, William, 1968, Use of water by riparian vegetation: U.S. Geological Survey Water-Supply Paper 1858, 62 p. Bray, D. I., and Kellerhals, Rolf, 1979, Some Canadian examples of the response of rivers to man-made changes, in Rhodes, D. D., and Williams, G. P., eds., Adjustments of the fluvial system: Dubuque, Iowa, Kendall/Hunt, p. 351-372. Breusers, H. N. C., 1967, Time scale of two dimensional local scour: Congress of the International Association for Hydraulic Research, 12th, Fort Collins, Colorado, Sept. 11-14, 1967, Proceedings, v. 3, p. 275-282. Bryan, Kirk, 1927, Channel erosion of the Rio Salado, Socorro County, New Mexico: U.S. Geological Survey Bulletin 790, p. 17-19. Burkham, D. E., 1972, Channel changes of the Gila River in Safford Valley, Arizona, 1846-1970: U.S. Geological Survey Professional Paper 655-G, 24 p. Carr, J. E., and Bergman, D. L. 1976, Reconnaissance of the water resources of the Clinton Quadrangle, west-central Oklahoma: Oklahoma Geological Survey Hydrologic Atlas 5, scale 1:250,000, 4 sheets. Chien, Ning, 1961, The braided stream of the Yellow River: Scientia Sinica, v. 10, p. 734-754. Coldwell, A. E., 1948, Effects of sediment on design and operation of dams and reservoirs: Federal Inter-Agency Sedimentation Conference, Denver, Colorado, May 6-8, 1947, Proceedings, p. 142-161. Coleman, J. M., 1969, Brahmaputra River: Channel processes and sedimentation: Sedimentary Geology, v. 3, p. 129-239. Cory, H. T., 1913, Irrigation and river control in the Colorado River delta: American Society of Civil Engineers Transactions, v. 76, p. 1204-1571. Culler, R. C., and others, 1982, Evapotranspiration before and after clearing phreatophytes, Gila River flood plain, Graham County, Arizona: U.S. Geological Survey Professional Paper 655-P, 67 p. DeCoursey, D. G., 1975, Implications of floodwater-retarding structures: American Society of Agricultural Engineers Transactions, v. 18, no. 5, p. 897-904.62 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS De Geer, Sten, 1910, Flaralfvens serpentinlopp och flodplan: Sveriges Geologiska Undersokning, ser. C, no. 236, (Arsbok 4, no. 8), 198 p. de Vries, M., 1973, River-bed variations - aggradation and degradation: International Seminar on Hydraulics of Alluvial Streams, New Delhi, India, Jan. 15-17, 1973, Publication no. 107, 20 p. Emmett, W. W., 1974, Channel aggradation in western United States as indicated by observations at vigil network sites: Zeitschrift fur Geomorphologie, suppl. bd. [supplementary volume] 21, p. 52-62. Everitt, B. L., 1980, Ecology of saltcedar—a plea for research: Environmental Geology, v. 3, p. 77-84. Foley, M. G., and others, 1978, Bed-load transport mechanisms during flash floods, in Conference on Hydrometeorological Aspects of Flash Floods, Los Angeles, California, May 2-3, 1978, preprint volume: Boston, Massachusetts, American Meteorological Society, p. 109-116. Fraser, J. C., 1972, Regulated discharge and the stream environment, in Oglesby, R. T., and others, eds., River ecology and man: New York, Academic Press, p. 263-285. Frederiksen, Famine and Associates, Inc., 1979, Proposed Trinity River basin and wildlife management program—Appendix B, Sediment and related analysis: Sacramento, Frederiksen, Famine and Associates, 107 p. Frickel, D. G., 1972, Hydrology and effects of conservation structures, Willow Creek Basin, Valley County, Montana, 1954-68: U.S. Geological Survey Water-Supply Paper 1532-G, 34 p. Gage, Maxwell, 1970, The tempo of geomorphic change: Journal of Geology, v. 78, p. 619-625. Garde, R. J., and Hasan, S. M., 1967, An experimental investigation of degradation in alluvial channels: Congress of the International Association for Hydraulic Research, 12th, Fort Collins, Colorado, Sept. 11-14, 1967, Proceedings, v. 3, p. 38-45. Gill, Don, 1973, Modification of northern alluvial habitats by river development: The Canadian Geographer, v. 17, no. 2, p. 138-153. Gottschalk, L. C., 1964, Reservoir sedimentation, in Chow, V. T., ed., Handbook of applied hydrology: New York, McGraw-Hill, section 17-1. Graf, W. L., 1980, The effect of dam closure on downstream rapids: Water Resources Research, v. 16, no. 1, p. 129-136. Gregory, F. J., and Park, C. C., 1974, Adjustment of river channel capacity downstream from a reservoir: Water Resources Research, v. 10, no. 4, p. 870-873. Gupta, Avijit, and Fox, H., 1974, Effects of high-magnitude floods on channel form: a case study in Maryland Piedmont: Water Resources Research, v. 10, no. 3, p. 499-509. Hales, Z. L., Shindala, Adnan, and Denson, F. H., 1970, Riverbed degradation prediction: Water Resources Research, v. 6, no. 2, p. 549-556. Hammad, H. Y., 1972, River bed degradation after closure of dams: Proceedings of the American Society of Civil Engineers, Journal of the Hydraulics Division, v. 98, no. HY4, p. 591-607. Harrison, A. S., 1950, Report on special investigation of bed sediment segregation in a degrading bed: University of California at Berkeley, Institute of Engineering Research, ser. 33, issue 1, 205 p. Hathaway, G. A., 1948, Observations on channel changes, degradation, and scour below dams: International Association for Hydraulic Structures Research Meeting, 2nd, Stockholm, Sweden, June 7-9, 1948, Appendix 16, p. 287-307. Hellwig, D. H. R., 1973, Evaporation of water from sand, 1: experimental set-up and climatic influences: Journal of Hydrology, v. 18, p. 93-108. Hickin, E. J., and Nanson, G. C., 1975, The character of channel migration on the Beatton River, northeast British Columbia, Canada: Geological Society of America Bulletin, v. 86, p. 487-494. Howard, Alan, and Dolan, Robert, 1981, Geomorphology of the Colorado River in the Grand Canyon: Journal of Geology, v. 89, no. 3, p. 269-298. Huggins, A. F., and Griek, M. R., 1974, River regulations as influence on peak discharge: Proceedings of the American Society of Civil Engineers, Journal of the Hydraulics Division, v. 100, no. HY7, p. 901-918. Hwang, J. C., 1975, Degradation of sand-bed channel: Congress of the International Association for Hydraulic Research, 16th, Sao Paulo, Brasil, July 27-Aug. 1, 1975, Proceedings, v. 2, p. 181— 188. Ingebo, P. A., 1971, Suppression of channel-side chaparral cover increases streamflow: Journal of Soil and Water Conservation, v. 26, no. 2, p. 79-81. Inglis, C. C., 1949, The behaviour and control of rivers and canals: Poona, India, Central Waterpower Irrigation and Navigation Research Station, Research Publication no. 13, pt. 1, 279 p. Johnson, W. C., Burgess, R. L., and Feammerer, W. R., 1976, Forest overstory vegetation and environment on the Missouri River floodplain in North Dakota: Ecological Monographs, v. 46, no. 1, p. 59-84. Felsey, H. M., 1977, Landsliding, channel changes, sediment yield, and land use in the Van Duzen River basin, north coastal California, 1941-1975: Ph. D. dissertation, University of California at Santa Cruz, 370 p. Fira, Hachiro, 1972, Factors influencing the river behaviour— riverbed variation due to dam construction and gravel gathering: Congress on International Commission on Irrigation and Drainage Transport, 8th, Varna, Bulgaria, 1972, Transactions, v. 5, Question 29.1, p. 405-432. Foch, Roy, Curry, Robert, and Weber, Mark, 1977, The effect of altered streamflow on the hydrology and geomorphology of the Yellowstone River Basin, Montana: Montana Department of Natural Resources and Conservation, Yellowstone Impact Study Technical Report no. 2, 163 p. Fomura, Saburo, 1971, Prediction of river-bed degradation below dams: Congress of the International Association for Hydraulic Research, 14th, Paris, France, Aug. 29-Sept. 5, 1971, Proceedings, v. 3, p. 257-264. Fomura, Saburo, and Simons, D. B., 1967, River-bed degradation below dams: Proceedings of the American Society of Civil Engineers, Journal of the Hydraulics Division, v. 93, no. HY4, p. 1-14. Lagasse, P. F., 1980, An assessment of the response of the Rio Grande to dam construction—Cochiti to Isleta reach: Albuquerque, New Mexico, U.S. Army Corps of Engineers, Technical Report, 133 p. LaMarche, V. C., Jr., 1966, An 800-year history of stream erosion as indicated by botanical evidence: U.S. Geological Survey Professional Paper 500-D, p. 83-86. Lane, E. W., 1934, Retrogression of levels in riverbeds below dams: Engineering News-Record, v. 112, June 28, 1934, p. 836-838. ------1948, An estimated of the magnitude of the degradation which will result in the middle Rio Grande channel from the construction of the proposed sediment storage basins and contraction works: U.S. Bureau of Reclamation, Hydraulic Laboratory Report no. Hyd-290, 22 p. ------1955, The importance of fluvial morphology in hydraulic engineering: Proceedings of the American Society of Civil Engineers, v. 81, Separate Paper no. 745, 17 p.REFERENCES 63 Langbein, W. B., and Schumm, S. A., 1958, Yield of sediment in relation to mean annual precipitation: Transactions of the American Geophysical Union, v. 39, p. 1076-1084. Larner, D. C., and others, 1974, Woody phreatophytes along the Colorado River from southeast Runnels County to the headwaters in Borden County, Texas: Austin, Texas, Texas Water Development Board Report 182, 19 p. Latif, Abdul, 1969, Investigation of Brahmaputra River: Proceedings of the American Society of Civil Engineers, Journal of the Hydraulics Division, v. 95, no HY5, p. 1687-1698. Lauterbach, D., and Leder, A., 1969, The influence of reservoir storage on statistical peak flows: Publication of the International Association of Scientific Hydrology, v. 2, no. 85, p. 821-826. Lawson, L. M., 1925, Effect of Rio Grande storage on river erosion and deposition: Engineering News-Record, v. 95, no. 10, p. 372-374. Leopold, L. B., Emmett, W. W., and Myrick, R. M., 1966, Channel and hillslope processes in a semiarid area - New Mexico: U.S. Geological Survey Professional Paper 352-G, p. 193-253. Leopold, L. B., Wolman, M. G., and Miller, J. P., 1964, Fluvial processes in geomorphology: San Francisco, California, W. H. Freeman and Company, 522 p. Leppanen, 0. E., 1981, Evapotranspiration from rapidly growing young saltcedar in the Gila River valley of Arizona: U.S. Geological Survey Open-File Report 81-485, 26 p. Little, W. C., and Mayer, P. G., 1972, The role of sediment gradation on channel armoring: Atlanta, Georgia, Georgia Institute of Technology, School of Civil Engineering, 104 p. Livesey, R. H., 1965, Channel armoring below Fort Randall Dam, in Federal Inter-Agency Sedimentation Conference, Jackson, Mississippi, Jan. 28-Feb. 1, 1963, Proceedings: U.S. Department of Agriculture Miscellaneous Publication 970, p. 461-469. Lustig, L. K., 1965, Sediment yield of the Castaic watershed, western Los Angeles County, California—a quantitative geomorphic approach: U.S. Geological Survey Professional Paper 422-F, 23 p. Lyons, H. G., 1906, The physiography of the River Nile and its basin: Cairo, Egypt, Cairo Survey Department, 411 p. Makkaveev, N. I., 1970, Effect of major dam and reservoir construction on geomorphological processes in river valleys: Geomorphology, no. 2, p. 106-110. Mandych, A. F., and Chalov, R. S., 1970, Trend of present-day erosion-accumulation processes in the lower reaches of rivers in the western Transcaucasus: Soviet Hydrology: Selected Papers, no. 1, p. 33-37. Meyboom, P., 1964, Three observations on streamflow depletion by phreatophytes: Journal of Hydrology, v. 2, p. 248-261. Moore, C. M., 1969, Effects of small structures on peak flow, in Moore, W. L., and Morgan, C. W., eds., Effects of watershed changes on streamflow: Austin, Texas, University of Texas Press, p. 101-117. Mostafa, M. G., 1957, River-bed degradation below large-capacity reservoirs: American Society of Civil Engineers Transactions, v. 122, p. 688-695. Newton, C. T., 1951, An experimental investigation of bed degradation in an open channel: Journal of the Boston Society of Civil Engineers, v. 38, no. 1, p. 28-60. Northrop, W. L., 1965, Republican River channel deterioration, in Federal Inter-Agency Sedimentation Conference, Jackson, Mississippi, Jan. 28-Feb. 1, 1963, Proceedings: U.S. Department of Agriculture Miscellaneous Publication 970, p. 409-424. O’Loughlin, C. L., 1969, Streambed investigations in a small mountain catchment: New Zealand Journal of Geology and Geophysics, v. 12, no. 4, p. 648-706. Osterkamp, W. R., and Hedman, E. K., 1981, Channel geometry of regulated streams in Kansas as related to mean discharges, 1970-80: Kansas Water Office Technical Report no. 15, 32 p. Pemberton, E. L., 1976, Channel changes in the Colorado River below Glen Canyon Dam: Federal Inter-Agency Sedimentation Conference, Denver, Colorado, Mar. 22-25, 1976, Proceedings, p. 5-61-5-73. Petts, G. E., 1977, Channel response to flow-regulation: The case of the River Derwent, Derbyshire, in Gregory, K. J., ed., River channel changes: Chichester, England, John Wiley and Sons, p. 145-164. ------1979, Complex response of river channel morphology subsequent to reservoir construction: Progress in Physical Geography, v. 3, no. 3, p. 329-362. Petts, G. E., and Lewin, John, 1979, Physical effects of reservoirs on river systems, in Hollis, G. E., ed., Man’s impact on the hydrological cycle in the United Kingdom: Norwich, England, Geo Abstracts Ltd., p. 79-91. Priest, M. S., and Shindala, Adnan, 1969a, Channel degradation downstream from large dams: Water Power, v. 21, p. 436-437. ------1969b, Distance downstream from a large dam to the limit of ultimate channel degradation: Congress of the International Association for Hydraulic Research, 13th, Kyoto, Japan, Aug. 31-Sept. 5, 1969, Proceedings, v. 5-1, p. 241-243. Rahn, P. H., 1977, Erosion below main stem dams on the Missouri River: Bulletin of the Association of Engineering Geologists, v. 14, no. 3, p. 157-181. Ritter, J. R., 1968, Changes in the channel morphology of Trinity River and eight tributaries, California, 1961-65: U.S. Geological Survey open-file report, 60 p. Rzhanitzin, N. A., Rabkova, E. K., and Artemiev, P. A., 1971, Deformation of alluvial channel downstream from large hydro-projects: Congress of the International Association for Hydraulic Research, 14th, Paris, France, Aug. 29-Sept. 5, 1971, Proceedings, v. 3, p. 265-274. Schoklitsch, Armin, 1950, Handbuch des Wasserbaues, bd. [v.] 1, (2d ed., revised): Vienna, Austria, Springer-Verlag, 475 p. Schoof, R. R., Thomas, W. O., Jr., and Boxley, W. M., 1980, Hydro-logic effects of the flood abatement program in southwestern Oklahoma: Water Resources Bulletin, v. 16, no. 2, p. 348-352. Schumm, S. A., 1960, The shape of alluvial channels in relation to sediment type: U.S. Geological Survey Professional Paper 352-B, p. 17-30. Schumm, S. A., and Lichty, R. W., 1963, Channel widening and flood-plain construction along Cimarron River in southwestern Kansas: U.S. Geological Survey Professional Paper 352-D, p. 71-88. Smith, D. G., 1972, Aggradation and channel patterns of the Alexandra-North Saskatchewan River, Banff National Park, Alberta, Canada, in Slaymaker, Olaf, and McPherson, H. J., eds., Mountain geomorphology: Vancouver, Canada, Tantalus, p. 117— 185. Stanley, J. W., 1951, Retrogression on the Lower Colorado River after 1935: American Society of Civil Engineers Transactions, v. 116, p. 943-957. Strand, R. I., 1977, Sedimentation, in U.S. Bureau of Reclamation, Design of small dams: Washington, D. C., U.S. Government Printing Office, p. 767-796. Sundborg, Ake, 1977, Alv-Kraft-Miljo [River-Hydropower-Environment]: Stockholm, Uppsala Universitet Naturgeografiska Institut and Statens Naturv&rdsverk, 150 p. Teskey, R. O., and Hinckley, T. M., 1977, Impact of water level changes on woody riparian and wetland communities, volume 1— plant and soil responses: Columbia, Missouri, U.S. Fish and Wildlife Service, 30 p.64 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Tinney, E. R., 1962, The process of channel degradation: Journal of Geophysical Research, v. 67, no. 4, p. 1475-1483. Todd, 0. J., and Eliassen, S., 1940, The Yellow River problem: American Society of Civil Engineers Transactions, v. 105, p. 346-453. Troxell, H. C., and Peterson, J. Q., 1937, Flood in La Canada valley, California, January 1, 1934: U.S. Geological Survey Water-Supply Paper 796-C, p. 53-98. Turner, D. S., 1971, Dams and ecology: Civil Engineering, v. 41, no. 9, p. 76-80. Turner, R. M., 1974, Quantitative and historical evidence of vegetation changes along the upper Gila River, Arizona: U.S. Geological Survey Professional Paper 655-H, 20 p. Turner, R. M., and Karpiscak, M. M., 1980, Recent vegetation changes along the Colorado River between Glen Canyon Dam and Lake Mead, Arizona: U.S. Geological Survey Professional Paper 1132, 125 p. U.S. Army Corps of Engineers, 1952, Report on degradation observations, Missouri River downstream from Fort Peck Dam: Fort Peck, Montana, U.S. Army Engineer District, 71 p. ------1958, Report of sedimentation survey, Canton Reservoir 1953: Tulsa, Oklahoma, U.S. Army Engineer District, 24 p. ------1960, Report on second resurvey of sedimentation, Denison Dam and Reservoir (Lake Texoma), Red River, Texas and Oklahoma: Tulsa, Oklahoma, U.S. Army Engineer District, 16 p. ------1972, Report of sedimentation and degradation from September 1966 resurvey, Canton Lake, North Canadian River, Oklahoma: Tulsa, Oklahoma, U.S. Army Corps of Engineers, 9 p. ------1973, Report on sedimentation John Martin Reservoir, Arkansas River Basin, Colorado, resurvey of March 1972: Albuquerque, New Mexico, U.S. Army Corps of Engineers, 24 p. U.S. Bureau of Reclamation, 1948, Report of river control work and investigations, lower Colorado River basin, calendar years 1946 and 1947: Boulder City, Nevada, U.S. Bureau of Reclamation, 50 p. ------1976, Report on river control work and investigations, lower Colorado River basin, 1974-1975: Boulder City, Nevada, U.S. Bureau of Reclamation, 34 p. U.S. Fish and Wildlife Service, 1981, The Platte River ecology study—special research report: Jamestown, North Dakota, U.S. Fish and Wildlife Service, 187 p. Van Hylckama, T. E. A., 1970, Water use by salt cedar: Water Resources Research, v. 6, no. 3, p. 728-735. Ward, J. V., and Stanford, J. A., eds, 1979, The ecology of regulated streams: New York, Plenum Press, 398 p. Williams, G. P., 1978, The case of the shrinking channels—the North Platte and Platte Rivers in Nebraska: U.S. Geological Survey Circular 781, 48 p. Williams, G. P., and Guy H. P., 1973, Erosional and depositional aspects of Hurricane Camille in Virginia, 1969: U.S. Geological Survey Professional Paper 804, 80 p. Willis, J. C., 1965, A laboratory flume study of sandbed degradation: U.S. Department of Agriculture, Research Report no. 379, 36 p. Wolman, M. G., 1954, A method of sampling coarse river-bed material: Transactions of the American Geophysical Union, v. 35, no. 6, p. 951-956. ------1967, Two problems involving river channel changes and background observations: Evanston, Illinois, Northwestern University Studies in Geography, no. 14, pt. 2, p. 67-107. Wolman, M. G., and Gerson, Ran, 1978, Relative scales of time and effectiveness of climate in watershed geomorphology: Earth Surface Processes, v. 3, p. 189-208.TABLES 13, 14TABLES 13, 14 67 Table 13.—Data on channel features, as measured from resurveyed cross sections [Footnotes on last page of tableJ Year of data collection Distance of cross section downstream Total change in mean bed Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Glen Canyon Dam Year of dam closure 1956^ 1956 0 1.1 0 —7 (104) 1959 3 1.1 -1.75 (137) 1963 7 1.1 -1.30 (141) 1965 9 1.1 -1.70 (141)(?) 1975 19 1.1 -1.90 (141) 1956 0 2.6 0 (183) 1959 3 2.6 -1.00 (183) 1963 7 2.6 -1.85 (183) 1965 9' 2.6 -2.00 (183) 1975 19 2.6 -2.15 (183) 1956 0 4.3 0 (169) 1959 3 4.3 -2.10 (167) 1963 7 4.3 -2.45 (167) 1965 9 4.3 -3.65 (167) 1975 19 4.3 -3.70 (167) 1956 0 6.4 0 (272) 1959 3 6.4 -.90 (272) 1963 7 6.4 -1.20 (272) 1965 9 6.4 -1.50 (272) 1975 19 6.4 -1.60 (272) 1956 0 8.0 0 (140) 1959 3 8.0 -1.05 (143) 1963 7 8.0 -1.70 (146) 1965 9 8.0 -4.35 (146) 1975 19 8.0 -4.10 (146) 1956 0 10.5 0 (252) 1959 3 10.5 -.35 (252) 1963 7 10.5 -1.00 (280) 1965 9 10.5 -1.90 (285) 1975 19 10.5 -2.05 (285) 1956 0 13.0 0 (97.5) 1959 3 13.0 -.75 (99.0) 1963 7 13.0 -.25 (95.5) 1965 9 13.0 -4.05 (99.0) 1975 19 13.0 -4.50 (99.0) Year of data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Glen Canyon Dam—Continued Year of dam closure 1956±' 1956 0 16.0 0 (95.5) 1959 3 16.0 -.85 (95.5) 1965 9 16.0 -7.25 (95.5) 1975 19 16.0 -7.00 (95.5) 1956 0 19.5 0 (189) 1959 3 19.5 -.45 (189) 1965 9 19.5 -2.00 (189) 1975 19 19.5 -2.20 (189) 1956 0 25 0 (109) 1959 3 25 0 (107) 1965 9 25 -5.20 (108) 1975 19 25 -3.80 (107) Colorado River, Arizona, Hoover Dam Year of dam closure 1935 1935 0 1.9 0 2/ 1935 .5 1.9 -1.20 — 1936 1 1.9 -1.30 — 1937 2 1.9 -1.35 — 1938 3 1.9 -.95 — 1939 4 1.9 -1.15 — 1940 5 1.9 -1.65 — 1941 6 1.9 -1.55 — 1942 7 1.9 -1.50 — 1943 8 1.9 -1.45 — 1944 9 1.9 -1.35 1945 10 1.9 -1.35 — 1946 11 1.9 -1.20 — 1947 12 1.9 -1.30 — 1948 13 1.9 -1.50 — 1935 0 2.3 0 — 1935 .5 2.3 -.10 — 1936 1 2.3 -.60 — 1937 2 2.3 -1.25 — 1938 3 2.3 -1.25 — Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Colorado River, Arizona, Hoover Year of dam closure Dam—Continued 1935 1939 4 2.3 -1.20 1940 5 2.3 -1.70 — 1941 6 2.3 -1.90 — 1942 7 2.3 -2.05 — 1943 8 2.3 -2.05 — 1944 9 2.3 -1.65 1945 10 2.3 -1.50 — 1946 11 2.3 -1.70 — 1947 12 2.3 -1.65 — 1948 13 2.3 -1.60 — 1935 0 3.2 0 1935 .5 3.2 -1.95 — 1936 1 3.2 -2.70 — 1937 2 3.2 -3.60 — 1938 3 3.2 -3.60 — 1939 4 3.2 -3.65 1940 5 3.2 -5.25 — 1941 6 3.2 -5.10 — 1942 7 3.2 -5.10 — 1943 8 3.2 -4.90 — 1944 9 3.2 -5.00 1945 10 3.2 -4.70 — 1946 11 3.2 -4.70 — 1947 12 3.2 -4.90 — 1948 13 3.2 -5.20 — 1935 0 4.5 0 1935 .5 4.5 -1.60 — 1936 1 4.5 -2.15 — 1937 2 4.5 -2.15 — 1938 3 4.5 -2.25 — 1939 4 4.5 -2.25 1940 5 4.5 -2.45 — 1941 6 4.5 -2.45 — 1942 7 4.5 -2.55 — 1943 8 4.5 -2.50 — data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Colorado River, Arizona, Hoover Dam—Con t inued Year of dam closure 1935 1944 9 4.5 -2.75 1945 10 4.5 -2.80 — 1946 11 4.5 -2.80 — 1947 12 4.5 -2.85 — 1948 13 4.5 -2.75 — 1935 0 5.5 0 — 1935 .5 5.5 -1.10 — 1936 1 5.5 -1.20 — 1937 2 5.5 -1.15 — 1938 3 5.5 -1.20 — 1939 4 5.5 -1.15 — 1940 5 5.5 -1.25 — 1941 6 5.5 -1.25 — 1942 7 5.5 -1.30 — 1943 8 5.5 -1.45 — 1944 9 5.5 -1.25 — 1945 10 5.5 -1.00 — 1946 11 5.5 -.80 — 1947 12 5.5 -.95 — 1948 13 5.5 -1.15 — 1935 0 6.1 0 — 1935 .5 6.1 -1.35 — 1936 1 6.1 -1.45 — 1937 2 6.1 -2.25 — 1938 3 6.1 -2.15 — 1939 4 6.1 -1.75 — 1940 5 6.1 -2.60 — 1941 6 6.1 -2.60 — 1942 7 6.1 -2.80 — 1943 8 6.1 -2.65 — 1944 9 6.1 -2.45 — 1945 10 6.1 -2.20 -r- 1946 11 6.1 -2.00 — 1947 12 6.1 -2.15 — 1948 13 6.1 -2.35 —68 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Table 13.—Data on channel features, as measured from cross sections—Continued resurveyed Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) data Year of collection Distance of crd€s section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Year Years after dam closure Colorado River, Arizona, Hoover Dam—Cont inued Colorado River, Arizona, Hoover Dam—Con t inued Year of dam closure 1935 Year of dam closure 1935 1935 0 7.1 0 1944 9 11.0 -1.90 1935 .5 7.1 -1.20 — 1945 10 11.0 -1.85 1936 1 7.1 -2.05 — 1946 11 11.0 -1.85 1937 2 7.1 -2.70 — 1947 12 11.0 -1.90 1938 3 7.1 -3.15 — 1948 13 11.0 -1.90 — 1939 4 7.1 -3.05 1935 0 12.5 0 1940 5 7.1 -3.00 — 1935 .5 12.5 -1.75 1941 6 7.1 -3.05 — 1936 1 12.5 -2.15 1942 7 7.1 -3.85 — 1937 2 12.5 -2.55 1943 8 7.1 -3.20 — 1938 3 12.5 -2.95 — 1944 9 7.1 -3.20 1939 4 12.5 -2.90 1945 10 7.1 -3.25 — 1940 5 12.5 -4.35 1946 11 7.1 -3.15 — 1941 6 12.5 -3.95 1947 12 7.1 -3.15 — 1942 7 12.5 -5.25 1948 13 7.1 -3.10 — 1943 8 12.5 -5.70 — 1935 0 8.0 0 1944 9 12.5 -5.90 1935 .5 8.0 -2.20 — 1945 10 12.5 -6.20 1936 1 8.0 -2.30 — 1946 11 12.5 -6.35 1937 2 8.0 -2.70 — 1947 12 12.5 -6.90 1938 3 8.0 -2.95 — 1948 13 12.5 -7.45 — 1939 4 8.0 -3.35 1935 0 13.5 0 1940 5 8.0 -4.45 — 1935 .5 13.5 -1.90 1941 6 8.0 -4.65 — 1936 1 13.5 -1.70 1942 7 8.0 -4.70 — 1937 2 13.5 -1.90 1943 8 8.0 -4.55 — 1938 3 13.5 -1.85 — 1944 9 8.0 -4.35 — 1939 4 13.5 -1.70 1945 10 8.0 -4.35 — 1940 5 13.5 -3.30 1946 11 8.0 -4.35 — 1941 6 13.5 -3.55 1947 12 8.0 -4.35 — 1942 7 13.5 -3.45 1948 13 8.0 -4.35 — 1943 8 13.5 -3.65 — 1935 0 9.7 0 1944 9 13.5 -3.65 1935 .5 9.7 -.75 — 1945 10 13.5 -3.70 1936 1 9.7 -1.60 — 1946 11 13.5 -3.65 1937 2 9.7 -1.85 — 1947 12 13.5 -3.70 1938 3 9.7 -2.35 1948 13 13.5 -3.70 ” Year of Year of data collection Distance of cross Total change data collection Distance of cross Total change — section downstream in mean bed Channel width section downstream in mean bed Channel width after i from dam elevation (meters) from dam elevation (meters) closure (kilometers) (meters) closure (kilometers) (meters) Colorado River, Arizona, Hoover Dam—Continued Colo'rado River, Arizona, Hoover Dam—Continued Year of dam closure 1935 Year of dam closure 1935 1939 4 9.7 -2.75 — 1935 0 15.5 0 __ 1940 5 9.7 -3.75 — 1935 .5 15.5 -1.50 1941 6 9.7 -3.70 — 1936 1 15.5 -2.15 1942 7 9.7 -4.40 — 1937 2 15.5 -2.50 1943 8 9.7 -4.50 — 1938 3 15.5 -3.30 — 1944 9 9.7 -4.55 — 1939 4 15.5 -4.40 1945 10 9.7 -4.50 — 1940 5 15.5 -4.35 1946 11 9.7 -4.35 — 1941 6 15.5 -4.45 — 1947 12 9.7 -4.35 — 1942 7 15.5 -5.50 — 1948 13 9.7 -4.35 — 1943 8 15.5 -5.10 * 1935 0 10.5 0 — 1944 9 15.5 -5.20 1935 .5 10.5 -.50 — 1945 10 15.5 -5.10 1936 1 10.5 -1.20 — 1946 11 15.5 -5.15 1937 2 10.5 -1.10 — 1947 12 15.5 -5.25 1938 3 10.5 -1.05 1948 13 15.5 -5.25 — 1939 4 10.5 -.95 — 1935 0 16.5 0 1940 5 10.5 -1.15 — 1935 .5 16.5 -.25 1941 6 10.5 -1.15 — 1936 1 16.5 -1.50 1942 7 10.5 -1.15 — 1937 2 16.5 -1.90 1943 8 10.5 -1.15 — 1938 3 16.5 -2.00 — 1944 9 10.5 -1.05 — 1939 4 16.5 -2.25 1945 10 10.5 -1.05 — 1940 5 16.5 -3.05 1946 11 10.5 -1.15 — 1941 6 16.5 -3.05 1947 12 10.5 -1.15 — 1942 7 16.5 -3.10 1948 13 10.5 -1.15 — 1943 8 16.5 -3.15 — 1935 0 11.0 0 — 1944 9 16.5 -3.10 1935 .5 11.0 -.60 — 1945 10 16.5 -3.10 1936 1 11.0 -1.00 — 1946 11 16.5 -3.10 1937 2 11.0 -1.20 — 1947 12 16.5 -3.35 / 1938 3 11.0 -1.40 — 1948 13 16.5 -3.65 — 1939 4 11.0 -1.35 — 1935 0 18.0 0 1940 5 11.0 -1.70 — 1935 .5 18.0 -.25 — 1941 6 11.0 -1.80 — 1936 1 18.0 -1.15 — 1942 7 11.0 -1.85 — j 1937 2 18.0 -1.85 — 1943 8 11.0 -2.00 1938 3 18.0 -2.00 —TABLES 13, 14 69 Table 13.—Data on channel features, as measured from resurveyed Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued cross sections—Continued Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Hoover Dam—Continued Colorado River, Arizona, Hoover Dam—Cont inued Year of dam closure 1935 Year of dam closure 1935 1939 4 18.0 -2.15 1935 0 42 Vo" 1940 5 18.0 -2.60 — 1935 .5 42 — 1941 6 18.0 -2.55 — 1936 1 42 -.30 — 1942 7 18.0 -2.45 — 1937 2 42 -.50 — 1943 8 18.0 -2.60 — 1938 3 42 -.75 — 1944 9 18.0 -2.60 — 1939 4 42 -1.00 — 1945 10 18.0 -2.50 — 1940 5 42 -1.20 — 1946 11 18.0 -2.40 — 1941 6 42 -1.05 — 1947 12 18.0 -2.45 — 1942 7 42 -2.70 — 1948 13 18.0 -2.55 — 1943 8 42 -2.60 — 1935 0 19.5 0 — 1944 9 42 -2.60 — 1935 .5 19.5 -.20 — 1945 10 42 -2.90 — 1936 1 19.5 -1.65 — 1946 11 42 -3.30 — 1937 2 19.5 -1.65 — 1947 12 42 -3.40 — 1938 3 19.5 -1.85 — 1948 13 42 -3.40 — 1939 4 19.5 -2.55 — 1935 0 51 — — 1940 5 19.5 -3.55 — 1935 .5 51 l'o~ — 1941 6 19.5 -3.65 — 1936 1.1 51 — 1942 7 19.5 -4.15 — 1937 2 51 -.15 — 1943 8 19.5 -4.30 — 1938 3 51 -.30 — 1944 9 19.5 -4.40 — 1939 4 51 -.60 — 1945 10 19.5 -4.50 — 1940 5 51 -.85 — 1946 11 19.5 -4.50 — 1941 6 51 -1.05 — 1947 12 19.5 -4.65 — 1942 7 51 -2.30 — 1948 13 19.5 -4.80 — 1943 8 51 -2.20 1935 0 21 0 — 1944 9 51 -2.75 — 1935 .5 21 0 — 1945 10 51 -2.80 — 1936 1 21 -1.00 — 1946 11 51 -2.80 — 1937 2 21 -1.50 — 1947 12 51 -3.00 — 1938 3 21 -1.65 — 1948 13 51 -3.10 — 1939 4 21 -2.05 — 1935 0 57 — — 1940 5 21 -2.40 — 1935 .5 57 l'o~ — 1941 6 21 -2.55 — 1936 1.1 57 — 1942 7 21 -2.60 — 1937 2 57 -.30 — 1943 8 21 -2.60 1938 3 S7 -.65 Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Hoover Dam—Continued Colorado River, Arizona, Hoover Dam—Continued Year of dam closure 1935 Year of dam closure 1935 1944 9 21 -2.60 1939 4 57 -1.00 1945 10 21 -2.60 — 1940 5 57 -1.50 — 1946 11 21 -2.65 — 1941 6 57 -1.35 — 1947 12 21 -2.75 — 1942 7 57 -2.05 — 1948 13 21 -2.75 — 1943 8 57 -2.95 — 1935 0 28 l'o" 1944 9 57 -2.95 — 1935 .5 28 — 1945 10 57 -3.10 — 1936 1 28 -.10 — 1946 11 57 -3.55 — 1937 2 28 -.35 — 1947 12 57 -3.00 — 1938 3 28 -.75 — 1948 13 57 -3.00 — 1939 4 28 -1.00 1935 0 63 — — 1940 5 28 -1.75 — 1935 .5 63 l'o" — 1941 6 28 -1.95 — 1936 1.1 63 — 1942 7 28 -3.10 — 1937 2 63 -.10 — 1943 8 28 -3.10 — 1938 3 63 -.75 — 1944 9 28 -2.95 __ 1939 4 63 -.90 — 1945 10 28 -2.95 — 1940 5 63 -1.35 — 1946 11 28 -3.25 — 1941 6 63 -1.50 — 1947 12 28 -3.25 — 1942 7 63 -3.55 — 1948 13 28 -3.35 — 1943 8 63 -3.30 — 1935 0 36 -- 1944 9 63 -3.15 — 1935 .5 36 a/S — 1945 10 63 -3.60 — 1936 1 36 — 1946 11 63 -3.50 — 1937 2 36 -.45 — 1947 12 63 -2.80 — 1938 3 36 -.90 — 1948 13 63 -4.50 — 1939 4 36 -1.20 — 1935 0 70 0 — 1940 5 36 -1.20 — 1935 .5 70 +.05 — 1941 6 36 -1.40 — 1936 1 70 -.15 — 1942 7 36 -2.15 — 1937 2 70 -.90 — 1943 8 36 -2.30 — 1938 3 70 -1.20 — 1944 9 36 -2.30 — 1939 4 70 -1.45 — 1945 10 36 -2.40 — 1940 5 70 -1.95 — 1946 11 36 -2.30 — 1941 6 70 -2.20 — 1947 12 36 -2.30 — 1942 7 70 -3.15 — 1948 13 36 -2.30 — 1943 if 70 -3.55 —70 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed Table 13 —Data on channel features, as measured from resurveyed cross sections—Continued cross sections—Continued Year of Year Of data collection Distance of cross Total change data collection Distance of cross Total change section downstream in mean bed Channel width section downstream in mean bed Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Hoover River, Arizona, Hoover Year of dam closure 1935 Year of dam closure 1935 1944 9 70 -3.95 — 1939 4 110 -0.35 1945 10 70 -4.15 — 1940 5 110 -.60 — 1946 11 70 -4.35 — 1941 6 110 -1.15 — 1947 12 70 -4.35 — 1942 7 110 -1.80 — 1948 13 70 -4.40 — 1943 8 110 -1.85 — 1935 0 77 — — 1944 9 110 -2.40 — 1935 .5 77 — — 1945 10 110 -2.55 — 1936 1 77 -0 — 1946 11 110 -2.65 — 1937 2.6 77 — 1947 12 110 -2.75 — 1938 3 77 -.20 — 1948 13 110 -2.95 — 1939 4 77 -.55 — 1935 0 117 — — 1940 5 77 -.70 — 1935 .5 117 — — 1941 6 77 -.90 — 1936 1 117 — — 1942 7 77 -1.65 — 1937 2 117 -0 — 1943 8 77 -1.70 — 1938 3.2 117 — 1944 9 77 -2.45 — 1939 4 117 0 — 1945 10 77 -2.45 — 1940 5 117 -.50 — 1946 11 77 -2.30 — 1941 6 117 -.85 — 1947 12 77 -2.45 — 1942 7 117 -1.85 — 1948 13 77 -2.50 — 1943 8 117 -1.85 — 1935 0 87 — — 1944 9 117 -2.45 — 1935 .5 87 — — 1945 10 117 -2.65 — 1936 1 87 -0 — 1946 11 117 -3.05 — 1937 2.6 87 — 1947 12 117 -3.55 — 1938 3 87 -.10 — 1948 13 117 -3.70 1939 4 87 -.25 — 1940 5 87 — 194&V 1941 6 87 -.75 — Year of dam closure 1942 7 87 -.85 — 1948 0 1.1 0 2/ 1943 8 87 1948 .5 1.1 -.65 — 1944 9 87 -2.05 — 1949 1 1.1 -1.20 — 1945 10 87 -2.10 — 1950 2 1.1 -2.20 — 1946 11 87 -2.15 — 1951 3 1.1 -2.45 — 1947 12 87 -2.20 — 1948 13 87 -2.40 Year of Year of data collection Distance of cross Total change data collection Distance of cross Total change in mean bed Channel width section downstream in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Hoover Dam—Con t inued Colorado River, Arizona, Davis Dam—Continued Year of dam closure 1935 Year of dam closure 1948^ 1935 0 94 — — 1952 4 1.1 -2.60 1935 .5 94 — — 1953 5 1.1 -3.25 — 1936 1 94 V0~ — 1954 6 1.1 -3.40 — 1937 2.6 94 — 1955 7 1.1 -3.80 — 1938 3 94 -.25 — 1956 8 1.1 -4.85 — 1939 4 94 -.60 — 1957 9 1.1 -4.85 — 1940 5 94 -.65 — 1958 10 1.1 -4.40 — 1941 6 94 -1.00 — 1959 11 1.1 -4.95 — 1942 7 94 -1.70 — 1960 12 1.1 -5.05 — 1943 8 94 -2.10 — 1961 13 1.1 -5.10 — 1944 9 94 -2.55 — 1962 14 1.1 -5.05 __ 1945 10 94 -2.75 — 1963 15 1.1 -4.90 — 1946 11 94 -2.80 — 1964 16 1.1 -4.90 — 1947 12 94 -3.00 — 1965 17 1.1 -5.05 — 1948 13 94 -3.10 — 1966 18 1.1 -5.10 — 1935 0 104 — — 1967 19 1.1 -5.05 — 1935 .5 104 — — 1968 20 1.1 -5.10 — 1936 1 104 — — 1969 21 1.1 -5.35 — 1937 2 104 3/tT — 1970 22 1.1 -5.65 — 1938 3 104 — 1971 23 1.1 -5.65 -- 1939 4 104 -.65 — 1972 24 1.1 -5.50 1940 5 104 -.55 — 1973 25 1.1 -5.60 — 1941 6 104 -1.05 — 1974 26 1.1 -5.75 — 1942 7 104 -1.70 — 1975 27 1.1 -5.65 — 1943 8 104 -2.05 — 1948 0 8.8 0 1944 9 104 -2.15 — 1949 1 8.8 -.20 — 1945 10 104 -2.60 — 1950 2 8.8 -.45 — 1946 11 104 -3.10 — 1951 3 8.8 -.50 — 1947 12 104 -3.30 — 1952 4 8.8 -.65 — 1948 13 104 -3.10 — 1953 5 8.8 -1.40 1935 0 110 — — 1954 6 8.8 -1.45 — 1935 .5 110 — — 1955 7 8.8 -1.45 — 1936 1 110 — — 1956 8 8.8 -1.55 — 1937 2 110 l'o“ — 1957 9 8.8 -1.70 — 1938 3.2 110 TABLES 13, 14 71 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Colorado River, Arizona, Davis Dam—Continued Year of dam closure 1948=-/ 1958 10 8 8 -2.00 1959 11 8 8 -2.00 — 1960 12 8 8 -2.05 — 1961 13 8 8 -2.05 — 1962 14 8 8 -2.15 — 1963 15 8 8 -2.15 1964 16 8 8 -2.40 — 1965 17 8 8 -2.60 — 1966 18 8 8 -2.70 — 1967 19 8 8 -2.75 — 1968 20 8 8 -2.80 1969 21 8 8 -2.65 — 1970 22 8 8 -2.40 — 1971 23 8 8 -2.30 — 1972 24 8 8 -2.45 — 1973 25 8 8 -2.60 — 1974 26 8 8 -2.70 — 1975 27 8 8 -2.75 — Colorado River, Arizona, Parker Dam Year of dam closure 1938 1938 0 27 Sfo 2/ 1939 1 27 -.65 — 1940 2 27 -1.00 — 1941 3 27 -1.50 — 1942 4 27 -1.80 — 1943 5 27 -2.05 — 1944 6 27 -2.30 — 1945 7 27 -2.45 — 1947 9 27 -2.60 — 1949 11 27 -2.55 — 1951 13 27 -2.60 — 1955 17 27 -2.85 — 1960 22 27 -2.95 — 1965 27 27 -3.00 — 1970 32 27 -3.05 — 1975 37 27 -3.15 — Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Colorado River, Arizona, Parker Dam—Con t inued Year of dam closure 1938 1938 0 39 0 1939 1 39 -.05 — 1940 2 39 -.10 — 1941 3 39 -.75 . — 1942 4 39 -1.80 — 1943 5 39 -2.05 1944 6 39 -2.60 — 1945 7 39 -2.45 — 1947 9 39 -2.75 — 1949 11 39 -2.80 — 1951 13 39 -3.50 1955 17 39 -3.80 — 1960 22 39 -3.85 — 1965 27 39 -3.65 — 1970 32 39 -4.35 — 1975 37 39 -4.35 — 1938 0 46 0 1939 1 46 -.15 — 1940 2 46 -.65 — 1941 3 46 -1.30 — 1942 4 46 -2.25 — 1943 5 46 -2.45 1944 6 46 -2.60 — 1945 7 46 -2.70 — 1947 9 46 -2.85 — 1949 11 46 -2.85 — 1951 13 46 -2.85 — 1955 17 46 -3.65 — 1960 22 46 -4.20 — 1965 27 46 -4.60 — 1970 32 46 -4.15 — 1975 37 46 -4.25 — 1938 0 66 0 — 1939 1 66 -.30 — 1940 2 66 -.60 — 1941 3 66 -1.35 — 1942 4 66 -2.25 — Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Colorado River, Arizona, Parker Dam—Con t inued Year of dam closure 1938 1943 5 66 -2.30 1944 6 66 -2.55 — 1945 7 66 -2.90 — 1947 9 66 -2.85 — 1949 11 66 -2.70 — 1951 13 66 -3.00 — 1955 17 66 -2.75 — 1960 22 66 -3.30 — 1965 27 66 -2.70 — 1970 32 66 -3.30 — 1975 37 66 -3.45 — 1938 0 80 0 — 1939 1 80 -.10 — 1940 2 80 -.20 — 1941 3 80 -1.05 — 1942 4 80 -1.40 — 1943 5 80 -1.75 — 1944 6 80 -1.50 — 1945 7 80 -1.50 — 1947 9 80 -1.40 — 1949 11 80 -1.05 — 1951 13 80 -1.60 — 1955 17 80 -1.50 — 1960 22 80 -1.70 — 1965 27 80 -2.05 — 1970 32 80 -2.05 — 1975 37 80 -2.40 — • 1938 0 95 0 — 1939 1 95 +.05 — 1940 2 95 -.35 — 1941 3 95 -.65 — 1942 4 95 -.80 — 1943 5 95 -1.50 — 1944 6 95 -1.90 — 1945 7 95 -1.35 — 1947 9 95 -1.20 — 1949 11 95 -1.25 — data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Colorado River, Arizona, Parker Dam—Continued Year of dam closure 1938 1951 13 95 -1.35 1955 17 95 -1.00 1960 22 95 -1.95 — 1965 27 95 -2.05 — 1970 32 95 -2.20 — 1975 37 95 -2.05 — Jemez River, New Mexico, Jemez Canyon Dam Year of dam closure 1953 1952 0 1.0 0 142 1959 6 1.0 -2.7 49.0 1965 12 1.0 -2.7 11.5 1975 22 1.0 -1.7 31.0 1952 0 1.3 0 272 1959 6 1.3 -2.4 70.0 1965 12 1.3 -2.7 17.0 1975 22 1.3 -1.5 24.0 1952 0 1.6 0 270 1959 6 1.6 -1.8 138 1965 12 1.6 -2.8 21.5 1975 22 1.6 -2.1 20.0 1952 0 1.8 0 216 1959 6 1.8 -2.2 105 1965 12 1.8 -3.0 48.5 1975 22 1.8 -1.9 49.5 1952 0 2.4 0 190 1959 6 2.4 -1.4 133 1965 12 2.4 -2.7 18.5 1975 22 2.4 -1.6 29.5 1952 0 2.7 0 220 1959 6 2.7 -1.8 39.5 1965 12 2.7 -2.1 42.0 1975 22 2.7 -1.3 59.072 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure from dam (kilometers) Jemez River, , New Mexico, Jemez Canyon Dam—Continued Year of dam closure 1953 1952 0 3.1 0 214 1959 6 3.1 -1.5 75.5 1965 12 3.1 -1.8 74.5 1975 22 3.1 -1.0 47.0 1952 0 3.4 0 178 1959 6 3.4 -1.1 74.5 1965 12 3.4 -1.3 100 1975 22 3.4 -.6 Arkansas River, Colorado, John Martin Dam Year of dam closure 1942 110 12/43- 1 3.5 *0 146 2/44 1951 9 3.5 -.10 142 1966 24 3.5 -1.95 30.5 1972 30 3.5 -.40 27.0 12/43- 1 5.0 — 0 128 2/44 1951 9 5.0 -.10 131 1966 24 5.0 -1.05 46.5 1972 30 5.0 -.35 44.0 12/43- 1 8.5 -0 76.0 2/44 1951 9 8.5 -.30 69.5 1966 24 8.5 -.80 39.5 1972 30 8.5 -.35 34.0 12/43- 1 12.0 — 0 100 2/44 1951 9 12.0 -.20 95.5 1966 24 12.0 -.80 30.0 1972 30 12.0 -.85 35.0 12/43- 1 15.5 a/0 157 2/44 1951 9 15.5 -.25 88.0 1966 24 15.5 -.85 40.5 1972 30 15.5 -.90 38.5 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Arkansas River, Colorado, Year of dam John Martin Dam—Continued closure 1942 12/43- 1 19.0 — 0 144 2/44 1951 9 19.0 -.05 144 1966 24 19.0 -.15 96.5 1972 30 19.0 -.50 43.5 12/43- 1 22 I'o 288 2/44 1951 9 22 -.60 165 1966 24 22 -1.15 74.0 1972 30 22 -.95 72.5 12/43- 1 26 -0 230 2/44 1951 9 26 -.25 241 1966 24 26 -.85 127 1972 30 26 -.75 86.5 12/43- 1 29 -0 168 2/44 1951 9 29 +.20 165 1966 24 29 +.25 46.0 1972 30 29 +1.30 50.0 12/43- 1 33 -0 201 2/44 1951 9 33 -.65 130 1966 24 33 -.40 99.5 1972 30 33 -.45 59 0 12/43- 1 36 0 110 2/44 1951 9 36 -.15 75.5 1966 24 36 -.20 56.5 1972 30 36 -.20 59.0 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued data Year of collection Distance of cross Total change Channel width (meters) Year Years after dam closure from dam (kilometers) elevation (meters) 1936 Missouri River, Montana, Fort Peck Dam Year of dam closure 1937—^ 0 9.2 0 348 1950 13 9.2 -.80 398 1955 18 9.2 -.65 402 1956 19 9.2 -.70 402 1958 21 9.2 -.70 408 1960 23 9.2 -.65 408 1966 29 9.2 -.75 408 1973 36 9.2 -.90 408 1936 0 13.0 0 234 1950 13 13.0 -.65 238 1955 18 13.0 -.80 238 1956 19 13.0 -.75 238 1958 21 13.0 -.60 236 1960 23 13.0 -.75 236 1966 29 13.0 -1.00 238 1973 36 13.0 -1.05 238 1936 0 16.5 0 248 1950 13 16.5 -1.00 304 1955 18 16.5 -1.26 336 1956 19 16.5 -1.00 336 1958 21 16.5 -1.00 336 1960 23 16.5 -1.05 336 1966 29 16.5 -1.15 340 1973 36 16.5 -1.75 340 1936 0 23 0 256 1950 13 23 -.50 262 1955 18 23 -.70 268 1956 19 23 -1.00 268 1958 21 23 -1.05 272 1960 23 23 -1.15 272 1966 29 23 -1.15 272 1973 36 23 -1.50 274 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River, Montana, Fort Peck Dam—Continued Year of dam closure 1937-' 1936 0 45 0 190 1950 13 45 -.15 202 1955 18 45 -.45 212 1956 19 45 -.60 212 1958 21 45 -.10 212 1960 23 45 -.20 216 1966 29 45 -.40 238 1973 36 45 -.75 238 1936 0 75 0 274 1950 13 75 -.20 286 1955 18 75 -.40 286 1956 19 75 -.25 288 1958 21 75 +.05 290 1960 23 75 -.10 292 1966 29 75 -.20 292 1973 36 75 -.25 298 Missouri River, North Dakota, , Garrison Dam Year of dam closure 1953 1946 (0) 2.7 0 530 1954 1 2.7 -.20 550 1960 7 2.7 --1.35 505 1964 11 2.7 -t-1.60 505 1970 17 2.7 --2.30 505 1976 23 2.7 --2.S0 500 1946 (0) 6.4 0 450 1954 1 6.4 -.45 525 1960 7 6.4 D-2.10 388 1964 11 6.4 — -2.13 390 1970 17 6.4 -/-3.65 392 1976 23 6.4 U-3.95 402TABLES 13, 14 73 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River, North Dakota, Garrison Dam—Continued Year of dam closure 1953 1946 (0) 8.0 0 458 1954 1 8.0 -.65 424 1960 7 8.0 -1.70 428 1964 11 8.0 -2.20 428 1970 17 8.0 -2.70 428 1976 23 8.0 -3.25 428 1946 (0) 10.5 0 585 1954 1 10.5 -.75 520 1960 7 10.5 -1.35 525 1964 11 10.5 -1.60 540 1970 17 10.5 -2.35 555 1976 23 10.5 -2.75 565 1946 (0) 12.0 0 492 1954 1 12.0 +.05 520 1960 7 12.0 -1.00 520 1964 11 12.0 -1.35 530 1970 17 12.0 -1.50 545 1976 23 12.0 -2.10 540 1946 (0) 15.0 0 505 1948 (0) 15.0 -.35 535 1954 1 15.0 + .05 580 1960 7 15.0 -.65 630 1964 11 15.0 -.75 715 1970 17 15.0 -1.15 790 1976 23 15.0 -1.25 835 1946 (0) 17.5 0 310 1954 1 17.5 0 570 1960 7 17.5 -.65 600 1964 11 17.5 -.15 610 1970 17 17.5 -1.45 680 1976 23 17.5 -.90 705 1946 (0) 21 0 895 1954 1 21 -.35 915 1960 7 21 +.65 925 1964 11 21 +.45 930 1970 17 21 +.10 945 1976 23 21 +.05 960 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River, North Dakota, Garrison Dam—Continued Year of dam closure 1953 1946 (0) 24 0 1954 1 24 -.80 1,295 1960 7 24 -1.60 1,300 1964 11 24 -1.60 1,305 1970 17 24 -1.90 1,310 1976 23 24 -1.95 1,315 1949 (0) 28 0 300 1954 1 28 -1.00 300 1960 7 28 -.70 296 1964 11 28 -1.05 298 1970 17 28 -2.15 300 1976 23 28 -3.05 306 1949 (0) 32 0 1,290 1954 1 32 -.40 1,395 1960 7 32 -.35 1,425 1964 11 32 -.50 1,430 1970 17 32 -.65 1,435 1976 23 32 -.80 1,430 1949 (0) 36 0 1,325 1954 1 36 -.35 865 1960 7 36 -1.00 855 1964 11 36 -.90 885 1970 17 36 -1.60 955 1976 23 36 -1.50 1,005 1949 (0) 38 0 448 1954 1 38 -.20 520 1960 7 38 -.30 525 1964 11 38 -.40 540 1970 17 38 -.70 555 1976 23 38 -1.50 595 1949 (0) 44 0 462 1954 1 44 +.10 505 1960 7 44 -.20 685 1964 11 44 -.05 740 1970 17 44 -.20 790 1976 23 44 -.85 805 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River, North Dakota, Garrison Dam—Continued Year of dam closure 1953 1949 (0) 47 0 525 1954 1 47 +.45 710 1960 7 47 +.45 930 1964 11 47 +.25 1,140 1970 17 47 -.35 1,145 1976 23 47 -1.05 1,150 1949 (0) 51 0 840 1954 1 51 +.05 845 1960 7 51 -1.05 488 1964 11 51 -.85 550 1970 17 51 -1.05 600 1976 23 51 -1.65 625 1949 (0) 54 0 376 1954 1 54 +.90 645 1960 7 54 +.65 690 1964 11 54 +.95 700 1970 17 54 +.35 725 1976 23 54 -.23 790 1949 (0) 58 0 635 1954 1 58 +.25 680 1960 7 58 -.45 715 1964 11 58 -.15 725 1970 17 58 -.35 740 1976 23 58 -.50 765 1949 (0) 61 0 565 1954 1 61 0 510 1960 7 61 + .30 595 1964 11 61 + .30 620 1970 17 61 -.45 635 1976 23 61 -.30 670 1949 (0) 70 0 416 1954 1 70 +.05 420 1960 7 70 +.25 424 1964 11 70 -.15 422 1970 17 70 -.45 428 1976 23 70 -.50 430 Year of data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Missouri River, North Dakota, Garrison Dam—Continued Year of dam closure 1953 1949 (0) 78 0 448 1954 1 78 + .45 490 1960 7 78 +.45 505 1964 11 78 +.25 525 1970 17 78 -.20 545 1976 23 78 -.35 560 1946 (0) 87 0 434 (?) 1954 1 87 +.75 595 1960 7 87 +.65 605 1964 11 87 +.60 605 1970 17 87 +.10 610 1976 23 87 +.20 615 Missouri River, South Dakota, Fort Randall Dam Year of dam closure 1952 1952 0 1.6 0 484 1954 2 1.6 -1.00 484 1957 5 1.6 -1.30 472 1960 8 1.6 -.85 448 1962 10 1.6 -.80 458 1967 15 1.6 -.90 458 1970 18 1.6 -.25 585 1975 23 1.6 -.45 590 1952 0 3.1 0 645 1954 2 3.1 -.35 645 1957 5 3.1 -.70 650 1960 8 3.1 -.60 650 1962 10 3.1 -.95 650 1967 15 3.1 -.90 655 1970 18 3.1 -1.00 655 1975 23 3.1 -1.45 655 1952 0 4.2 0 675 1954 2 4.2 -.25 675 1956 4 4.2 -.65 67574 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed Table 13.—Data on channel features, as measured from resurveyed Year of Year of data collection Distance of cross section downstream Total change in mean bed data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Missouri River, South Dakota, Fort landall Dam—Continued Missouri River , South Dakota, Fort Randall Dam—Continued Year of dam closure 1952 Year of dam closure 1952 1962 10 4.2 -0.90 685 1962 10 29 -0.50 1,065 1967 15 4.2 -.90 690 1967 15 29 -.50 1,075 1970 18 4.2 -1.05 695 1970 18 29 -.75 1,075 1975 23 4.2 -1.35 690 1975 23 29 -.65 1,075 1952 0 5.1 0 720 1952 0 35 0 695 1954 2 5.1 +.20 745 1954 2 35 + .05 695 1957 5 5.1 -.40 730 1957 5 35 +.10 695 1960 8 5.1 -.40 735 1960 8 35 +.05 695 1962 10 5.1 -.60 740 1962 10 35 -.10 695 1967 15 5.1 -.60 750 1967 15 35 -.25 695 1970 18 5.1 -.80 755 1970 18 35 0 695 1975 23 5.1 -.80 755 1975 23 35 -.45 695 1952 0 6.6 0 1,060 1952 0 43 0 760 1954 2 6.6 -.20 1,075 1954 2 43 + .35 895 1956 4 6.6 -.50 1,095 1957 5 43 -.55 1,035 1960 8 6.6 -1.15 1,115 1960 8 43 -.45 1,050 1962 10 6.6 -1.30 1,130 1962 10 43 -.05 1,055 1967 15 6.6 -1.15 1,130 1965 13 43 + .20 1,060 1970 18 6.6 -1.15 1,135 1967 15 43 +.10 1,060 1975 23 6.6 -1.85 1,165 1970 18 43 + .70 1,070 1952 0 7.7 0 1,070 1975 23 43 +.10 1,115 1954 2 7.7 -.25 1,115 1952 0 53 0 685 1957 5 7.7 -.75 1,130 1954 2 53 +.35 690 1960 8 7.7 -1.10 1,145 1957 5 53 +.60 690 1962 10 7.7 -1.35 1,160 1960 8 53 +.60 700 1967 15 7.7 -1.15 1,245 1962 10 53 +.70 705 1970 18 7.7 -1.05 1,260 1965 13 53 +.75 705 1975 23 7.7 -1.60 1,280 1967 15 53 +.70 705 11.0 11.0 404 406 18 53 +1.00 710 1952 1954 0 2 -1.30 1975 23 53 +.60 710 1956 4 11.0 -1.75 406 1952 0 58 0 810 1960 8 11.0 -1.50 408 1954 2 58 +.50 835 410 420 5 58 +.25 835 1962 1967 10 15 11.0 11.0 -1.50 -1.65 1961 9 58 +.50 845 1975 23 11.0 -2.60 462 Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Year after dam closure from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Missouri River, South Dakota, Fort Randall Dam—Continued Missouri River , South Dakota, Fort Randall Dam—Continued Year of dam closure 1952 Year of dam closure 1952 1952 0 12.5 0 565 1962 10 58 +0.70 845 1954 2 12.5 0 565 1967 15 58 +.75 860 1956 4 12.5 -.60 570 1970 18 58 +.75 870 1960 8 12.5 -.75 570 1975 23 58 +.85 885 10 12.5 -.80 1967 15 12.5 -.80 585 Missouri River, South Dakota Gavins Point Dam 1975 23 12.5 -1.40 605 Year of dam closure 1955^ 1952 0 14.5 0 1,080 1955 0 2.3 0 374 1954 2 14.5 +.50 1,070 1960 5 2.3 -1.30 374 1957 5 14.5 +.20 1,060 1965 10 2.3 -1.50 380 1960 8 14.5 +.25 1,065 1970 15 2.3 -2.15 380 1962 10 14.5 +.35 1,065 1974 19 2.3 -2.50 374 1967 15 14.5 +.20 1,065 1955 0 3.4 0 525 1970 18 14.5 +.10 1,065 1960 5 3.4 -1.00 525 1975 23 14.5 0 1,035 1965 10 3.4 -1.50 525 0 19.0 0 366 1970 15 3.4 -2.00 525 1954 2 19.0 -.10 366 1974 19 3.4 -2.30 525 1957 5 19.0 -.05 406 1955 0 4.3 0 344 1960 8 19.0 -.45 645 1960 5 4.3 -.25 416 1962 10 19.0 -.20 700 1965 10 4.3 -1.20 420 1967 15 19.0 +.35 735 4.3 -1.45 420 1975 23 19.0 -.15 750 1974 19 4.3 -1.90 426 0 24 0 640 1955 0 5.3 0 630 1954 2 24 -.30 640 1960 5 5.3 -.55 645 1957 5 24 +.10 650 1965 10 5.3 -1.20 650 1960 8 24 +.30 645 1970 15 5.3 -1.50 655 1974 19 5.3 -2.00 660 1962 10 24 +.25 645 1955 1967 15 24 -.35 645 0 6.8 0 980 1975 23 24 -.75 645 5 6.8 -.40 1,155 1965 10 6.8 -.60 1,160 1952 0 29 0 1,040 1970 15 6.8 -.80 1,170 1954 2 29 -.45 1,070 1974 19 6.8 -1.25 1,175 1956 4 29 -.50 1,075 1960 8 29 -.50 1,070 TABLES 13, 14 75 Table 13.—Data on channel features, as measured from resurveyed Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued cross sections—Continued Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) Missouri River , South Dakota, Gavins Point Dam—Continued Missouri River , South Dakota, Gavins Point Dam—Continued Year of dam closure 1955^ Year of dam closure 1955^ 1955 0 7.9 0 885 1955 0 30 0 460 1960 5 7.9 -.55 885 1960 5 30 +.15 570 1965 10 7.9 -.80 885 1965 10 30 -.15 575 1970 15 7.9 -1.35 885 1970 15 30 +.35 575 1974 19 7.9 -1.50 885 1974 19 30 +.35 580 1955 0 8.4 0 478 1955 0 32 0 505 1960 5 8.4 -1.10 478 1960 5 32 +.25 585 1965 10 8.4 -.65 478 1965 10 32 +.50 600 1970 15 8.4 -1.05 478 1970 15 32 +.60 620 1974 19 8.4 -1.80 478 1974 19 32 +.05 625 1955 0 8.5 0 366 1955 0 34 0 790 1960 5 8.5 -.25 368 1960 5 34 -.25 845 1965 10 8.5 -.80 366 1965 10 34 + .05 880 1970 15 8.5 -1.70 368 1970 15 34 +.05 910 1974 19 8.5 -2.05 362 1974 19 34 -.20 980 1955 0 9.5 0 • 464 1955 0 36 0 1,780 1960 5 9.5 -.45 456 1960 5 36 -.60 1,785 1965 10 9.5 -.65 464 1965 10 36 -1.20 1,815 1970 15 9.5 -1.15 466 1970 15 36 -.80 1,835 1974 19 9.5 -1.55 466 1974 19 36 -1.30 1,840 1955 0 11.0 0 880 1955 0 38 0 655 1960 5 11.0 -.50 1,020 1960 5 38 + .05 660 1965 10 11.0 -.50 1,035 1965 10 38 -.10 665 1970 15 11.0 -.90 1,060 1970 15 38 +.30 670 1974 19 11.0 -1.05 1,065 1974 19 38 + .10 680 1955 0 12.5 0 348 1955 0 39 0 368 1960 5 12.5 -.45 412 1960 5 39 +.25 378 1965 10 12.5 -.65 438 1965 10 39 +.40 380 1970 15 12.5 -.50 446 1970 15 39 +.35 380 1974 19 12.5 -1.35 470 1974 19 39 -.10 380 1955 0 14.5 0 790 1955 0 41 0 890 1960 5 14.5 0 880 1960 5 41 -.10 890 1965 10 14.5 + .45 880 1965 10 41 -.15 905 1970 15 14.5 -.45 1,045 1970 15 41 -.20 925 1974 19 14.5 -.50 1,050 1974 19 41 -.60 935 data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year of data collection Distance of cross Total change Year Years after dam closure Year Years after dam closure from dam (kilometers) elevation (meters) (meters) Missouri River, South Dakota , Gavins Point Dam- -Continued Missouri River, South Dakota, Gavins Point Dam- -Continued Year of dam closure 1955^ 1955 0 16.5 0 845 1955 0 44 0 1,600 1960 5 16.5 -1.00 850 1960 5 44 -.25 1,600 1965 10 16.5 -.50 1,125 1965 10 44 -.35 1,600 1970 15 16.5 -.45 1,160 1970 15 44 -.45 1,605 1974 19 16.5 -.90 1,190 1974 19 44 -.45 1,605 1955 0 18.0 0 905 1957 2(0?) 46 -0 945 1960 5 18.0 +.05 905 1960 5 46 -.45 960 1965 10 18.0 -.20 920 1965 10 46 -.45 970 1970 15 18.0 -.20 920 1970 15 46 -.35 975 1974 19 18.0 -.25 930 1974 19 46 -1.00 975 1955 0 22 0 615 1957 2 48 i'o 895 1960 5 22 + .10 625 1960 5 48 +.15 1965 10 22 +.15 635 1965 10 48 + .35 1,145 1970 15 22 -.25 645 1970 15 48 -.10 1,180 1974 19 22 -.60 645 1974 19 48 +.15 1,190 1955 0 23 0 520 1958 3 52 -^0 1,040 1960 5 23 +.05 605 1960 5 52 -.10 1965 10 23 +.50 780 1965 10 52 + .15 1,290 1970 15 23 +.05 950 1970 15 52 -.05 1,415 1974 19 23 +.10 975 1974 19 52 -.60 1955 0 26 0 326 1957 2 55 -^0 675 1960 5 26 +.65 466 1960 5 55 -.05 755 10 26 +.50 480 1965 10 55 + .20 1,030 1970 15 26 +1.30 675 1970 15 55 + .25 1,105 1974 19 26 +.80 690 1974 19 55 -.30 1,130 1955 0 27 0 960 1957 2 57 — 1960 5 27 +.40 1,165 1960 5 57 + .20 — 1965 10 27 +.10 1,215 1965 10 57 + .10 — 1970 15 27 -.20 1,215 1970 15 57 -.35 — 1974 19 27 -.05 1,220 1974 19 57 -.50 — 1955 0 28 0 805 1958 3 61 y<> 865 5 28 -1.05 975 1960 5 61 -.35 865 1965 10 28 -.05 1,050 1965 10 61 -.20 15 28 -.10 1,080 1970 15 61 -.20 935 1974 19 28 -.25 1,095 1974 19 61 -.80 94076 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilomaters) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River , South Dakota, Gavins Year of dam closure Point Dam—Continued 1955—^ 1959 4 64 3fo 960 1960 5 64 +.05 975 1965 10 64 +.05 1,110 1970 15 64 + .10 1,140 1974 19 64 +.75 1,140 1958 3 69 -0 755 1960 5 69 -.45 785 1965 10 69 -1.00 785 1970 15 69 -1.85 795 1974 19 69 -1.45 840 1959 4 72 -0 1,415 1960 5 72 + .10 1,445 1965 10 72 -.20 1,555 1970 15 72 -.35 1,640 1974 19 72 -.65 1,645 1959 4 78 -0 535 1960 5 78 -.25 535 1965 10 78 -.30 535 1970 15 78 -.65 545 1974 19 78 -1.15 545 1959 4 82 3/o 1,300 1960 5 82 -.05 1,300 1965 10 82 + .05 1,335 1970 15 82 +.25 1,365 1974 19 82 + .20 1,390 1959 4 85 -0 1,355 1960 5 85 + .20 1,535 1965 10 85 + .65 1,905 1970 15 85 + .40 1,925 1974 19 85 -.10 1,935 1959 4 89 -^0 680 1960 5 89 + .30 685 1965 10 89 -.35 765 1970 15 89 +.25 865 1974 19 89 + .35 895 data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Missouri River , South Dakota, Gavins Point Dam—Continued Year of dam closure 19555' 1959 4 93 -0 330 1960 5 93 -.15 392 1965 10 93 +1.70 710 1970 15 93 +1.15 740 1974 19 93 +.85 755 Medicine Creek, Nebraska, Medicine Creek Dam Year of dam closure 1949 1950 (0) 0.8 0 83.5 1952 3 .8 -.10 93.0 1962 13 .8 -.20 107 (?) 1963 14 .8 -.20 91.5 1971 22 .8 -.20 99.0 1977 28 .8 -.20 107 1950 (0) 13.0 0 30.5 1952 3 13.0 -.05 32.0 1962 13 13.0 +.50 38.5 1963 14 13.0 +.30 36.5 1971 22 13.0 +.45 36.5 1977 28 13.0 +.40 35.5 1950 (0) 16.0 0 — 1952 3 16.0 -.55 20.5 1962 13 16.0 -.30 21.0 1964 15 16.0 -.25 20.5 1971 22 16.0 +.20 21.5 1978 29 16.0 +.25 21.0 1950 (0) 16.5 0 25.5 1952 3 16.5 -.25 25.5 1962 13 16.5 + .30 26.0 1964 15 16.5 -.10 25.5 1971 22 16.5 + .45 25.5 1978 29 16.5 +.15 25.5 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Middle Loup River, Nebraska, , Milburn Dam Year of dam closure 1955 9/„ 9/ 1950 (0) 0.2 1955 .2 .2 -.60 7.6 1956 1.3 .2 -1.20 29.0 1957 2.3 .2 -2.30 44.5 1957 2.7 .2 -2.30 44.5 1957 2.8 .2 -1.35 57.5 1961 6.3 .2 -1.45 96.5 1961 6.6 .2 -1.80 96.5 1962 7.5 .2 -1.90 109 1964 9.3 .2 -2.15 110 1964 9.7 .2 -2.15 111 1967 12.3 .2 -2.65 114 1969 14.3 .2 -2.15 117 1971 16.3 .2 -2.40 124 1961 6.3 1.6 — 0 230 1962 7.5 1.6 -.35 234 1964 9.3 1.6 -.60 232 1964 9.7 1.6 -.60 238 1967 12.3 1.6 -1.00 234 1969 14.3 1.6 -1.10 234 1971 16.3 1.6 -1.20 234 1964 9.3 3.1 ^(0) 118 1964 9.7 3.1 0 118 1967 12.3 3.1 -.20 120 1969 14.3 3.1 -.15 118 1971 16.3 3.1 -.25 123 1961 6.3 5.6 -'m 90.5 1962 7.5 5.6 -.40 91.0 1964 9.3 5.6 -.65 91.0 1964 9.7 5.6 -.70 91.0 1967 12.3 5.6 -.90 92.0 1969 14.3 5.6 -.90 91.5 1971 16.3 5.6 -1.05 92.0 1967 12.3 7.4 ^(0) 163 1969 14.3 7.4 -.05 166 1971 16.3 7.4 -.25 174 data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Smoky Hill River, Kansas, Kanopolis Dam Year of dam closure 1948 1946 (0) 0.8 0 46.5 1951 3 .8 -.80 45.0 1952 4 .8 -1.10 44.5 1961 13 .8 -1.30 45.0 1971 23 .8 -1.45 48.0 1946 (0) 2.9 0 41.0 1951 3 2.9 -.20 42.0 1952 4 2.9 -.35 42.0 1961 13 2.9 -1.05 41.0 1971 23 2.9 -1.05 45.0 1946 (0) 4.8 0 40.0 1951 3 4.8 +.05 40.5 1952 4 4.8 +.15 39.0 1961 13 4.8 -.50 39.5 1971 23 4.8 -.50 42.0 1946 (0) 6.8 0 50.0 1951 3 6.8 -.05 49.5 1952 4 6.8 0 49.5 1961 13 6.8 -.45 46.5 1971 23 6.8 -.20 47.0 1946 (0) 8.7 0 39.5 1951 3 8.7 +.25 41.5 1952 4 8.7 0 46.0 1961 13 8.7 -.25 47.0 1971 23 8.7 -.20 50.0 1946 (0) 13.0 0 34.5 1951 3 13.0 -.05 38.5 1952 4 13.0 -.10 39.0 1961 13 13.0 -.25 39.5 1946 (0) 16.5 0 39.5 1951 3 16.5 +.35 41.0 1952 4 16.5 +.25 41.0 1961 13 16.5 +.35 40.5TABLES 13, 14 77 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Smoky Hill River, Kansas, Kanopolis Dam—Continued Year of dam closure 1948 1946 (0) 18.5 0 39.5 1951 3 18.5 +.10 39.5 1952 4 18.5 + .15 35.5 1961 13 18.5 +.10 38.0 1946 (0) 23 0 35.5 1951 3 23 -.05 38.0 1952 4 23 +.15 37.0 1961 13 23 -.10 36.5 1946 (0) 25 0 38.5 1951 3 25 -.05 39.5 1952 4 25 -.15 40.0 1961 13 25 -.25 41.5 1946 (0) 35 0 36.5 1951 3 35 +.05 36.0 1952 4 35 +.20 34.5 1961 13 35 +.15 37.0 1946 (0) 42 0 30.0 1951 3 42 +.05 30.0 1952 4 42 0 33.5 1961 13 42 -.10 32.5 1946 (0) 50 0 36.5 1951 3 50 -.05 36.5 1952 4 50 -.25 35.0 1961 13 50 — — 1946 (0) 56 0 54.5 1951 3 56 +.05 34.0 1952 4 56 — 63.0 1961 13 56 +.05 33.0 1946 (0) 73 0 34.0 1951 3 73 +.05 38.0 1952 4 73 -.15 39.0 1961 13 73 -.20 39.5 Year of data collection Distance of cross Total change section downstream in mean bed Channel width from dam elevation (meters) closure (kilometers) (meters) Smoky Hill River, Kansas, , Kanopolis Dam—Continued Year of dam closure 1948 1946 (0) 92 0 35.0 1951 3 92 0 35.0 1952 4 92 -.20 40.0 1961 13 92 — — 1946 (0) 108 0 30.0 1951 3 108 + .25 30.0 1952 4 108 +.20 29.5 1961 13 108 + .25 31.0 Republican River, Kansas, Milford Dam Year of dam closure 1967 1967 0 2.7 0 156 1974 7 2.7 -.85 165 1967 0 4.0 0 98.0 1975 8 4.0 -.15 116 Wolf Creek, Oklahoma, Fort Supply Dam Year of dam closure 1942 1944 2 .3 0 242 1949 7 .3 -2.05 26.5 1961 19 .3 -3.15 32.5 1969 27 .3 -3.40 23.0 1944 2 1.0 0 137 1949 7 1.0 -1.90 30.5 1961 19 1.0 -2.20 56.0 1969 27 1.0 -2.00 57.5 1944 2 1.3 0 158 1949 7 1.3 -1.40 46.0 1961 19 1.3 -2.10 63.5 1969 27 1.3 -2.60 28.5 1944 2 1.6 0 172 1949 7 1.6 -.25 163 1961 19 1.6 -1.50 90.0 1969 27 1.6 -2.45 15.0 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Wolf Creek, Oklahoma, Fort Supply Dam—Continued Year of dam closure 1942 1944 2 1.8 0 296 1949 7 1.8 — — 1961 19 1.8 -1.35 191 1969 27 1.8 -2.30 20.0 1944 2 2.6 0 107 1949 7 2.6 -.45 88.5 1961 19 2.6 -1.05 120 1969 27 2.6 — — 1944 2 2.9 0 242 1949 7 2.9 -.80 81.5 1961 19 2.9 -1.60 52.5 1969 27 2.9 -1.60 55.0 1944 2 3.9 0 246 1949 7 3.9 -1.20 79.0 1961 19 3.9 -1.45 84.5 1969 27 3.9 -2.00 24.5 1944 2 4.7 0 272 1949 7 4.7 -.45 166 1961 19 4.7 -1.15 97.0 1969 27 4.7 -1.50 26.0 1944 2 6.6 0 240 1949 7 6.6 -.35 121 1961 19 6.6 -1.15 38.0 1969 27 6.6 -1.30 30.0 North Canadian River, Oklahoma, Canton Dam Year of dam closure 1948 1947 0 1.8 0 64.5 1949 1 1.8 -.90 61.0 1951 2.8 1.8 -1.20 62.0 1951 3.4 1.8 -1.55 67.0 1959 11 1.8 -2.90 29.5 1966 18 1.8 -2.75 17.5 1976 28 1.8 -3.00 18.5 Year of data collection Distance of cross Total change section downstream in mean bed Channel width from dam elevation (meters) closure (kilometers) (meters) North Canadian River, Oklahoma, Canton Dam—Continued Year of dam closure 1948 1947 0 3.1 0 65.0 1949 1 3.1 -.60 53.0 1951 2.8 3.1 -1.00 47.0 1951 3.4 3.1 -1.30 56.5 1959 11 3.1 -1.50 55.5 1966 18 3.1 -1.20 56.5 1976 28 3.1 -1.50 48.0 1947 0 5.0 0 47.0 1949 1 5.0 -.60 48.5 1951 2.8 5.0 -.65 46.5 1951 3.4 5.0 -1.05 45.5 1959 11 5.0 -.80 49.0 1966 18 5.0 -.95 27.5 1976 28 5.0 -1.65 17.5 1947 0 5.6 0 45.5 1949 1 5.6 -.35 44.0 1951 2.8 5.6 -1.05 44.0 1951 3.4 5.6 -1.35 35.0 1959 11 5.6 -.85 29.5 1966 18 5.6 -1.60 16.5 1976 28 5.6 -1.50 20.0 1947 0 10.5 Wo -^35.5 1949 1 10.5 + .75 53.5 1951 2.8 10.5 +1.05 91.5 1951 3.4 10.5 +1.05 91.5 1959 11 10.5 +.85 97.0 1966 18 10.5 +.90 76.0 1947 0 12.0 0 75.0 1949 1 12.0 -.15 77.0 1951 2.8 12.0 -.65 63.5 1951 3.4 12.0 -.75 61.0 1959 11 12.0 -.55 85.0 1966 18 12.0 -1.45 12.578 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued cross sections—Continued Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) North Canadian River, Oklahoma, Canton Dam—Continued North Canadian River, Oklahoma, Canton Dam—Continued Year of dam closure 1948 Year of dam closure 1948 1947 0 14.5 0 93.0 1947 0 154 0 39.0 1949 1 14.5 -.10 93.5 1949 1 154 -.05 39.5 1951 2.8 14.5 -.35 74.0 1951 2.8 154 + .05 39.5 1951 3.4 14.5 -.45 72.0 1951 3.4 154 +.05 38.5 1959 11 14.5 -.85 42.0 1959 11 154 -.35 28.5 1966 18 14.5 -.70 23.0 1966 18 154 -.50 31.0 1947 1949 0 1 27 27 0 +.20 105 116 Canadian River, Oklahoma, Eufaula Dam 1951 2.8 27 + .05 105 Year of dam closure 1963 1951 3.4 27 + .05 106 1964 1 .8 0 252 1959 11 27 -.65 39.5 1969 6 .8 -5.05 177 1966 18 27 -.55 40.0 1977 14 .8 -4.95 234 1947 0 35 0 52.5 1964 1 2.1 0 460 1949 1 35 -.20 57.5 1969 6 2.1 -2.30 208 1951 2.8 35 -.25 57.5 1977 14 2.1 -3.20 234 1951 3.4 35 -.45 56.5 1964 1 3.4 0 284 1959 11 35 -.35 49.5 1969 6 3.4 -2.10 183 1966 18 35 -.45 30.0 1977 14 3.4 -2.80 220 1947 0 50 0 76.0 1964 1 4.7 0 560 1949 1 50 -.10 70.5 1969 6 4.7 -1.15 560 1951 2.8 50 -.10 69.0 1977 14 4.7 -2.15 400 1951 3.4 50 + .05 99.5 1964 1 6.6 0 218 1959 11 50 + .45 100 1969 6 6.6 -.65 280 1966 18 50 -.05 59.5 1977 14 6.6 -1.20 284 1947 0 58 0 96.5 1964 1 8.0 0 362 1949 1 58 -.30 95.0 1969 6 8.0 -.35 515 1951 2.8 58 -.10 113 1977 14 8.0 -1.00 402 1951 3.4 58 -.15 106 1964 1 11.5 0 446 1959 11 58 -1.05 40.5 1969 6 11.5 -.35 462 1966 18 58 -.85 39.5 1977 14 11.5 -1.40 420 1947 0 68 0 — 1964 1 14.0 0 505 1949 1 68 +.45 — 1969 6 14.0 -.35 620 1951 2.8 68 + .15 1977 14 14.0 -1.15 494 Year of Year of data collection Distance of cross section downstream Total change in mean bed Channel width data collection Distance of cross section downstream Total change in mean bed Channel width Years after dam closure Years after dam closure Year from dam (kilometers) elevation (meters) (meters) Year from dam (kilometers) elevation (meters) (meters) North Canadian River, Oklahoma, Canton Dam—Continued Canadian River, Oklahoma, Eufaula Dam—Continued Year of dam closure 1948 Year of dam closure 1963 1951 3.4 68 -0.65 _ 1964 1 16.0 0 422 1959 11 68 + .05 — 1969 6 16.0 0 438 1966 18 68 +.35 — 1977 14 16.0 -.60 374 1947 0 92 0 — 1964 1 18.5 0 149(?) 1949 1 92 +.10 — 1969 6 18.5 -.90 198 1951 2.8 92 +.10 -- 1977 14 18.5 -1.10 280 1951 3.4 92 +.30 — 1964 1 20 0 446 1959 11 92 -.30 — 1969 6 20 -.70 454 1966 18 92 +.25 — 1977 14 20 -1.35 428 1947 0 104 0 — 1964 1 23 0 400 1949 1 104 -.25 — 1969 6 23 -.45 460 1951 2.8 104 -.20 — 1977 14 23 -.60 580 1951 3.4 104 -.10 — 1964 1 34 0 260 1959 11 104 -.50 — 1969 6 34 -.30 260 1966 18 104 -.55 — 1977 14 34 +.40 187 1947 0 114 0 44.5 1964 1 37 0 360 1949 1 114 -.10 45.0 1969 6 37 -.70 350 1951 2.8 114 -.20 44.0 1977 14 37 0 346 1951 3.4 114 -.05 35.5 1964 1 40 0 406 1959 11 114 -.05 31.5 1969 6 40 +.05 470 1966 18 114 -.20 27.5 1977 14 40 +1.20 478 0 38.5 1949 1 125 -.30 38.0 Red River, Oklahoma-Texas, Denison Dam 1951 2.8 125 -.50 38.5 Year of dam closure 1942 1951 3.4 125 -.50 34.0 1942 0 .6 0 228 1959 11 125 -.65 26.5 1945 3 .6 -1.25 244 1966 18 125 -.70 30.0 1948 6 .6 -1.35 278 1947 1949 0 1 134 134 0 0 40.0 33.0 1958 1969 16 27 .6 .6 -1.45 -1.60 280 282 1951 2.8 134 +.10 36.5 1942 0 1.1 0 236 1951 1959 3.4 11 134 134 -.05 -.35 33.5 33.0 1945 1948 3 6 1.1 1.1 -1.40 230 11/ 1966 18 134 -.35 27.5 1958 16 1.1 -3.00 1969 27 1.1 -2.40 11/TABLES 13, 14 79 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Red River. , Oklahoma-Texas, Denison Dam—Continued Year of dam closure 1942 1942 0 2.1 0 228 1945 3 2.1 -1.35 230 1948 6 2.1 — — 1958 16 2.1 -2.45 240 1969 27 2.1 -2.40 238 1942 0 3.2 0 ^284 1945 3 3.2 -.80 284 1948 6 3.2 -1.60 284 1958 16 3.2 -2.40 284 1969 27 3.2 -2.00 284 1942 0 5.1 0 210 1945 3 5.1 -.45 218 1948 6 5.1 — — 1958 16 5.1 -2.20 222 1969 27 5.1 -1.65 214 1942 0 7.2 0 274 1945 3 7.2 -.10 280 1948 6 7.2 -.75 280 1958 16 7.2 -1.30 280 1969 27 7.2 -1.30 296 1942 0 8.4 0 396 1945 3 8.4 -1.15 398 1948 6 8.4 — — 1958 16 8.4 -1.45 400 1969 27 8.4 -1.75 400 1942 0 11.5 0 151 1945 3 11.5 -1.20 135 1948 6 11.5 — — 1958 16 11.5 -1.85 — 1969 27 11.5 -2.10 140 1942 0 15.0 0 224 1945 3 15.0 -1.45 149 1948 6 15.0 -1.85 149 1958 16 15.0 -2.45 151 1969 27 15.0 -3.25 152 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Red River, , Oklahoma-Texas, Denison Dam—Continued Year of dam closure 1942 1942 0 18.5 0 318 1945 3 18.5 + .15 324 1948 6 18.5 — — 1958 16 18.5 -1.40 336 1969 27 18.5 -.60 342 1942 0 22 0 244 1945 3 22 -.20 244 1948 6 22 — — 1958 16 22 -1.20 256 1969 27 22 -.70 246 1942 0 27 0 282 1945 3 27 -.30 328 1948 6 27 -.45 360 1958 16 27 -.40 382 1969 27 27 -.20 372 1946 4 34 218 1948 6 34 + .05 228 1958 16 34 -.25 292 1969 27 34 -.20 296 1946 4 41 — 0 308 1948 6 41 + .35 308 1958 16 41 -.45 292 1969 27 41 -.30 300 1946 4 48 -0 376 1948 6 48 + .10 530 1958 16 48 +.75 775 1969 27 48 + .75 780 1946 4 65 480 1948 6 65 -.10 488 1958 16 65 -.10 630 1969 27 65 + .10 640 1946 4 80 i'o 705 1948 6 80 +1.05 1,050 1958 16 80 -.10 990 1969 27 80 + .35 1,025 Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued data Year of collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Red River, Oklahoma-Texas, Denison Dam—Continued Year of dam closure 1942 1946 4 90 -0 191 1948 6 90 -.80 198 1958 16 90 -.70 226 1970 28 90 -.55 214 1946 4 101 -0 368 1948 6 101 -.20 382 1958 16 101 +.45 408 1970 28 101 -.10 360 1946 4 109 -0 234 1948 6 109 -.35 268 1958 16 109 -.65 262 1970 28 109 -.55 266 1946 4 122 -0 1,085 1948 6 122 -1.00 1,195 1958 16 122 +.20 910 1970 28 122 -1.35 1,025 1946 4 132 -70 312 1948 6 132 -.05 322 1958 16 132 + .30 366 1970 28 132 -.50 374 1946 4 142 ±'0 464 1948 6 142 -.35 336 1958 16 142 -.10 452 1970 28 142 -.65 384 1946 4 150 1/q 270 1948 6 150 +.40 294 1958 16 150 +.10 324 1970 28 150 -.25 258 Neches River, Texas, Town i Bluff Dam Year of dam closure 1951 1951 0 .2 0 94.5 1965 14 .2 -2.25 127 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam closure Neches River, Texas, Town Bluff Dam—Continued Year of dam closure 1951 1951 0 1.4 0 111 1960 9 1.4 -.20 127 1965 14 1.4 -.90 127 1951 0 2.9 0 101 1960 9 2.9 -.10 100 1965 14 2.9 -.60 100 1951 0 4.7 0 90. ,5 1960 9 4.7 -.90 97. ,5 1965 14 4.7 + .50 101 1951 0 6.3 0 117 1960 9 6.3 -.60 133 1965 14 6.3 -.65 145 1951 0 8.0 0 121 1960 9 8.0 -.25 151 1965 14 8.0 -.05 157 Des Moines River, Iowa, Red Rock Dam Year of dam closure 1969 1962 (0) 2.3 0 185 1978 9 2.3 -1.00 214 1962 (0) 4.7 0 155 1978 9 4.7 -1.15 162 1962 (0) 6.1 0 180 1978 9 6.1 -1.05 171 1962 (0) 12.0 0 184 1978 9 12.0 -1.75 — 1962 (0) 14.0 0 131 1978 9 14.0 -.60 145 1962 (0) 22 0 199 1978 9 22 + .20 212 1962 (0) 25 0 158 1978 9 25 + .05 168 80 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued Table 13.—Data on channel features, as measured from resurveyed cross sections—Continued data Year of collection Distance of cross Total change in mean bed elevation (meters) Channel width (meters) Year Years after dam *5°“ dam , closure (kilometers) Des Moines River, Iowa, Red Rock Dam—Continued Year of dam closure 1969 1962 (0) 29 0 153 1978 9 29 +.65 185 1962 (0) 33 0 160 1978 9 33 +.10 181 1962 (0) 36 0 154 1978 9 36 -.45 165 1962 (0) 38 0 146 1978 9 38 -.65 154 1962 (0) 40 0 200 1978 9 40 -1.85 146 1962 (0) 42 0 171 1978 9 42 -.35 172 1962 (0) 48 0 148 1978 9 48 -.45 155 1962 (0) 50 0 108 1978 9 50 -.60 110 1962 (0) 52 0 141 1978 9 52 -1.30 119 1962 (0) 55 0 129 1978 9 55 -.05 152 1962 (0) 62 0 206 1978 9 62 +.25 208 1962 (0) 68 0 179 1978 9 68 +1.00 185 1962 (0) 72 0 162 1978 9 72 +1.05 164 Chattahoochee River, Georgia , Buford Dam Year of dam closure 1956 1956 0 .5 0 95.0 1963 7 .5 -.95 98.0 1964 8 .5 -1.00 98.5 Year of data collection Distance of cross section downstream from dam (kilometers) Total change in mean bed elevation (meters) Year Years after dam closure (meters) Chattahoochee River, Georgia, Buford Dam—Continued Year of dam closure 1956 1965 9 0.5 -0.95 99.0 1968 12 ' .5 -1.10 100 1971 15 .5 -1.00 103 1956 0 1.9 0 76.0 1963 7 1.9 -1.40 77.0 1964 8 1.9 -1.50 74.0 1965 9 1.9 -1.80 76.0 1968 12 1.9 -2.15 77.5 1971 15 1.9 -2.55 75.5 1956 0 2.9 0 71.5 1963 7 2.9 -.90 74.0 1964 8 2.9 -.95 79.5 1965 9 2.9 -1.30 74.0 1968 12 2.9 -1.60 74.5 1971 15 2.9 -1.85 76.0 1956 0 4.0 0 63.0 1963 7 4.0 -.75 67.0 1964 8 4.0 -.60 68.0 1965 9 4.0 -.90 68.5 1968 12 4.0 -1.35 67.0 1971 15 4.0 -1.45 — 1956 0 5.8 0 68.5 1963 7 5.8 -.30 67.0 1964 8 5.8 -.20 68.0 1965 9 5.8 -.45 67.0 1968 12 5.8 -.75 69.5 1971 15 5.8 -.70 69.0 1956 0 7.6 0 91.0 1963 7 7.6 + .05 98.0 1964 8 7.6 +.10 97.0 1965 9 7.6 -.05 95.5 1968 12 7.6 -.20 98.0 1971 15 7.6 -.30 — data Year of collection Distance of cross Total change Channel width (meters) Year Years after dam closure from dam (kilometers) elevation (meters) Chattahoochee River, Georgia, Buford Dam—Continued Year of dam closure 1956 1956 0 11.0 0 70.5 1963 7 11.0 -.90 70.5 1964 8 11.0 +.05 70.5 1965 9 11.0 + .05 71.5 1971 15 11.0 +.10 — 1957 1 13.5 3/ 0 69.5 1963 7 13.5 -.20 64.0 1964 8 13.5 -.10 67.5 1965 9 13.5 -.10 68.0 1968 12 13.5 -.35 72.0 1971 15 13.5 + .05 74.0 ’957 1 16.5 3/ 0 59.0 1964 8 16.5 -.40 69.0 1965 9 16.5 -.60 67.0 1968 12 16.5 -.55 68.0 1957 1 18.0 3/ 0 57.5 1963 7 18.0 + .20 60.0 1964 8 18.0 + .20 61.0 1965 9 18.0 +.25 63.5 1968 12 18.0 + .15 62.0 1971 15 18.0 + .30 — 1957 1 21 3/ 0 57.0 1963 7 21 +.25 54.0 1964 8 21 +.15 55.0 1965 9 21 +.05 54.0 1968 12 21 +.15 55.0 1971 15 21 + .25 — 1957 1 24 3/ 0 60.5 1963 7 24 +.15 58.5 1964 8 24 + .25 59.0 1965 9 24 +.15 61.0 1968 12 24- +.10 60.0 1971 15 24 +.15 58.5 FOOTNOTES TO TABLE 13 —^Cofferdam closure 1956; official closure of Glen Canyon Dam was 1963. —^Channel confined in rock-walled canyon. Widths, if listed, meaningful only for general order of magnitude. —^First measurement of this cross section was later than year of dam closure. Total changes in bed elevation were measured from this later year. —^Year of initial diversion. Official closure was in 1951. —diversion dam (Headgate Rock Dam), located about 24 kilometers below Parker Dam and closed in 1942, may have some unknown influence on the cross sections listed here from about 1942 on. —^Year storage began. Dam completed in 1939. -Vit all is bed degradation; arrival of bar or spit near left bank severed part of previous channel. —^Storage began 1952. —^All data for this cross section apply only to the thalweg rather than to the entire channel. During dam construction most of the flow was diverted into the thalweg, and it gradually grew to become the new main channel. —^Pronounced lateral migration of channel at this section at least from 1947 through 1966. — Right bank washed out by tributary changing course in 1957. 12/ —Bridge section; width constrained.TABLES 13, 14 81 Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables [Footnotes on last paf?e of table] Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables—Continued River Name of distance downstream gaging station of station and control station from dam (kilometers) Reference discharge (cubic meters per second) Change in streambed elevation a Period change from initial gage height _____________(meters) River Name of distance downstream gaging station of station and control station from dam (kilometers) Reference discharge (cubic meters per second) Change in streambed elevation * Period change from initial gage height _____________(meters) Colorado River, Arizona, Parker Dam Year of dam closure 1938___________ Colorado River below 6.4 i/90.6 10/34-11/35 0 Parker Dam 12/35-2/37 +.18 2/37-12/37 +.061 12/37-3/38 +.18 10/38-1/39 -.91 1/39-12/39 -1.92 1/40-12/40 -1.74 1/41-9/41 -2.07 10/41-5/42 -2.59 No suitable control station Jemez River, New Mexico, Jemez Canyon Dam Year of dam closure 1953 Jemez River below 1.3 -^.034 4/51-3/52 0 Jemez Canyon Dam 3/55-7/55 -.55 •8/55-9/55 -.95 10/55-2/56 -.98 2/56-8/56 -.73 8/56-5/57 -.98 5/57-10/57 -1.49 10/57-3/58 -2.07 3/58-6/58 -2.32 Jemez River near 43 .37 6/36-9/36 0 Jemez (control station) 10/36-5/37 -.18 6/37-9/38 -.12 10/38-9/39 -.061 10/39-5/41 -.12 8/49-5/52 -.40 4/58-4/60 -.30 10/60-5/61 -.43 10/61-3/63 -.37 Missouri River , South Dakota, Fort Randall Year of dam closure 1952 Dam—Continued 10/60-12/63 -0.40 12/63-9/64 -.43 10/64-11/65 -.40 11/65-10/67 -.43 10/67-6/69 -.61 No suitable control station Missouri River, South Dakota, Gavins Point Dam Year of dam closure 1955 Missouri River at 8 —^312 3/32-9/33 0 Yankton 10/33-7/34 -.15 8/34-3/37 -.061 3/37-9/38 -.18 10/38-5/39 -.15 5/39-3/40 -.27 3/40-3/41 -.15 3/41-6/41 -.091 6/41-9/41 -.061 10/41-5/42 -.12 5/42-3/43 -.34 6/43-3/44 -.49 10/45-3/47 -.61 3/47-9/48 -.61 10/48-3/51 -.37 3/51-3/52 -.49 5/53-11/54 -.58 11/54-4/55 -.61 5/55-9/55 -.46 10/55-9/56 -.55 10/56-9/57 -.67 10/57-1/59 -.70 1/59-12/60 -.76 12/60-9/61 -.82 10/61-9/62 -.88 River Name of distance downstream gaging station of station and control station from dam (kilometers) Reference discharge (cubic meters per second) Period Change in streambed elevation s change from initial gage height (meters) Name of downstream gaging station and control station River distance of station from dam (kilometers) Reference discharge (cubic meters per second) Period Change in streambed elevation -change from initial gage height (meters) Missouri River, Montana, Fort Peck Dam Missouri River, South Dakota , Gavins Point Dam—Continued Year of dam closure 1937 Year of dam closure 1955 Missouri River below 13 -^ss.o 4/38-10/38 0 10/62-12/63 -0.98 Fort Peck Dam 10/38-4/39 -.030 12/63-9/65 -1.01 10/39-9/40 -.21 10/65-9/68 -1.19 10/40-9/41 -.24 10/68-9/71 -1.25 10/41-10/44 -.30 10/73-9/76 -1.92 10/44-9/45 -.52 10/76-9/79 -2.16 10/47-9/48 -.67 No suitable control 10/48-9/51 -.85 station 10/51-9/52 -1.01 2/54-9/55 -1.19 Smoky Hill River, Kansas, Kanopolis Dam 10/55-2/56 -1.25 Year of dam closure 1948 3/56-9/56 -1.37 Smoky Hill River near 1.3 0.51 10/40-9/41 0 10/57-9/58 -1.31 Langley 10/41-9/42 +.12 10/59-2/61 -1.34 10/42-10/46 + .30 10/61-9/65 -1.37 10/46-5/47 +.12 10/65-11/66 -1.46 5/47-9/47 -.030 11/66-9/79 -1.49 10/47-3/49 0 No suitable control 10/49-9/50 -.24 station 9/50-3/51 -.40 4/52-12/52 -.91 Missouri River, South Dakota, Fort Randall Dam 10/53-6/54 -.91 Year of dam 6/54-9/54 -.88 Missouri River below 11 -'464 5/47-9/51 0 10/54-10/55 -.91 Fort Randall Dam 10/52-11/52 + .030 10/55-7/57 -.88 3/53-5/53 +.030 10/59-4/60 -1.01 7/53-11/53 -.30 4/60-6/61 -1.04 5/54-9/54 -.30 10/61-9/64 -1.04 10/54-3/55 -.37 10/64-9/68 -1.07 3/55-9/55 -.24 10/68-5/70 -1.10 10/55-9/56 -.30 10/70-10/71 -1.10 10/56-9/59 -.24 10/71-3/73 -1.13 10/59-9/60 -.34 3/73-10/73 -1.19 10/73-10/74 -1.34 10/74-9/76 -1.40 10/76-12/77 -1.3782 DOWNSTREAM EFFECTS OF DAMS ON ALLUVIAL FANS Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables- —Continued Change in . . Reference Name of distance elevation - downstream gaging station of station . , . , e f 6 c , (cubic meters and control station from dam ,. (kilometers) per sec°nd) Period change from initial gage height (meters) Smoky Hill River, Kansas, Kanopolis Dam— -Continued Year of dam closure 1948 Smoky Hill River at 48 0.43 7/40-9/45 0 Ellsworth (control 10/45-7/46 + .030 station) 7/46-9/49 0 10/49-8/50 +.030 8/50-4/51 -.061 4/51-9/51 + .030 10/51-9/53 0 10/53-6/55 +.030 10/56-7/57 -.030 10/57-9/61 -.061 10/61-9/62 -.030 10/62-9/63 + .030 10/63-6/64 -.030 6/64-5/65 0 7/65-11/65 +.15 2/66-8/66 0 7/67-11/68 + .030 11/68-4/69 0 6/69-12/69 + .12 1/70-6/70 +.061 6/70-10/70 +.21 10/70-3/71 + .18 7/71-1/72 +.061 1/72-10/73 0 2/75-? + .091 Republican River, Kansas, Milford Dam Year of dam closure 1967 Republican River below 2.7 1.2 10/63-9/64 0 Milford Dam 10/64-7/65 +.27 7/65-2/66 + .15 2/66-7/67 +.12 7/67-11/67 +.091 River _ , Seme of distance ^f"ence downstream gaging station of station . , .c arg , _ , 7 ... c (cubic meters and control station from dam (kilometers) per second) Change in streambed elevation * Period change from initial gage height (meters) Republican River, Kansas, Milford Dam— Continued Year of dam closure 1967 11/67-2/69 0 2/69-5/70 -.27 10/70-6/72 -.49 6/72-4/73 -.58 4/73-11/73 -.95 11/73-4/74 -1.25 4/74-6/75 -1.34 6/75-1/76 -1.37 1/76-9/77 -1.40 10/77-6/78 -1.43 6/78-3/79 -1.49 3/79-1/80 -1.59 Republican River at 49 3.4 10/53-2/55 0 Clay Center 2/55-6/55 -.030 (control station) 6/55-9/55 -.061 10/55-9/56 -.030 10/56-9/58 -.061 10/58-2/59 -.21 2/59-9/59 -.18 10/59-3/60 -.15 3/60-9/62 -.18 10/62-9/63 -.061 10/63-1/68 -.091 2/68-7/69 -.12 10/69-5/71 -.15 5/71-5/72 -.12 5/72-9/73 -.15 10/73-9/77 -.27 10/77-1/80 -.34 North Canadian River, Oklahoma, Canton Dam Year of dam closure 1948 irth Canadian River 4.8 .031 10/37-9/41 0 at Canton 10/42-9/43 + .21 11/46-5/47 +.21 2/48-5/49 +.18 6/49-9/50 -.43 Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables—Continued Change in River streambed Name of distance discharge (cubic meters per second) elevation = downstream gaging station of station Period change from and control station from dam (kilometers) initial gage height (meters) North Canadian River, Oklahoma, Canton Dam—Continued Year of dam closure 1948 10/50-9/51 -0.70 10/51-9/53 -1.01 10/53-5/54 -.88 5/54-9/54 -1.01 North Canadian River 45 0.00057 7/46-2/48 0 near Selling 2/48-9/50 -.030 (control station) 5/51-3/53 -.030 4/53-9/53 0 10/53-9/54 -.27 10/54-5/55 -.27 5/55-9/65 -.40 Red River, Texas-Oklahoma, Denison Dam Year of dam closure 1943 Red River near Colbert, 4.5 3.7 7/42-10/42 0 Oklahoma 11/42-4/44 -.37 4/44-3/45 -.40 10/45-6/46 -.88 6/46-7/47 -1.01 10/47-1/48 -.91 1/48-7/48 -.98 7/48-9/48 -1.04 10/48-6/49 -1.16 10/49-9/51 -1.16 10/52-8/54 -1.13 8/54-3/55 -1.16 3/55-9/55 -1.19 10/55-7/57 -1.13 10/57-8/58 -1.31 8/58-11/58 -1.34 11/58-4/59 -1.28 9/59-2/60 -1.28 2/60-4/60 -1.34 4/60-7/60 -1.37 Change in Name of River distance Reference discharge (cubic meters streambed elevation a downstream gaging station of station Period change f rom and control station from dam (kilometers) gage height (meters) Red River, Texas-Oklahoma, Denison Dam- -Continued Year of dam closure 1943 Red River near 106 4.2 10/36-5/38 0 Gainesville, Texas 5/38-5/40 +.21 (control station) 5/40-9/40 +.30 10/40-5/41 + .37 4/43-5/43 -.061 10/43-4/44 -.061 6/44-1/46 0 6/46-7/47 +.24 7/47-6/48 + .030 7/48-10/49 zs +.27 10/49-? + .15 10/50-5/51 +.30 6/52-6/57 + .091 6/57-11/57 + .37 5/58-? + .73 10/58-5/59 +.70 6/59-11/59 + .95 12/59-5/62 +.70 Neches River, Texas, Town Bluff Dam Year of dam closure 1951 Neches River at Town .5 4.2 3/51-5/52 0 Bluff 5/52-11/54 -.061 11/54-9/55 -.15 10/55-9/58 -.37 10/58-12/59 -.46 12/59-9/62 -.55 10/62-9/63 -.67 10/63-9/70 -.73 10/70-10/71 -.85 10/71- -.95 Village Creek near Kountze — 1.5 4/39-12/47 0 (control station) 12/47-9/49 -.061 1/51-9/51 -.030 10/51-2/56 0 2/56-11/61 -.061 11/61-9/66 -.15TABLES 13, 14 83 Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables—Continued Name of downstream gaging station and control station River distance of station from dam (kilometers) Reference discharge (cubic meters per second) Period Change in streambed elevation * change from initial gage height (meters) Chattahoochee River, Georgia, Buford Dam Year of dam closure 1956 Chattahoochee River near 4.0 12.2 10/50-9/53 0 Buford 10/53-9/55 -.061 10/55-5/57 0 5/57-9/57 -.091 10/57-10/58 -.061 10/58-9/59 -.12 10/59-12/60 -.15 1/61-5/61 -.24 5/61-9/62 -.27 10/62-11/63 -.40 11/63-1/64 -.49 1/64-2/64 -.40 2/64-7/64 -.49 7/64-3/65 -.55 3/65-4/65 -.58 4/65-5/65 -.52 6/65-8/65 -.58 8/65-10/66 -.67 10/66-9/68 -.76 10/68-1/70 -.88 1/70-4/71 -.98 Chestatee River near 73 3.4 4/40-12/40 0 Dahlonega (control 3/41-12/42 + .15 station) 12/42-3/43 0 3/43-10/43 -.061 10/43-2/44 -.15 2/44-11/44 -.24 11/44-12/45 -.27 1/47-11/53 -.31 11/53-9/54 -.37 10/54-9/71 -.40 10/71-9/73 -.37 Change in Name of River distance Reference discharge (cubic meters per second) streambed elevation * downstream gaging station and control station of station from dam Period change from initial (kilometers) gage height (meters) Rio Grande, New Mexico, Caballo Dam Year of dam closure 1938____________ Rio Grande below Caballo 1.3 - 28.3 2/38-10/38 0 Dam 10/38-12/39 0 1/40-9/40 -.061 1943 -.091 1944 -.12 1945 -.15 1946-48 -.46 3/55-12/55 -.40 1957 -.43 1958 -.40 1959-60 -.64 1961-62 -.70 1963-64 -.73 1965 -.70 1966 -.76 1967 -.70 1972 -.67 1974 -.76 1979 -.76 No suitable control station Table 14.—Changes in streambed elevations as estimated from streamflow-gaging-station rating tables—Continued Name of downstream gaging station and control station River distance of station from dam (kilometers) Reference discharge (cubic meters per second) Change in streambed elevation = Period change from initial gage height (meters) Frenchman Creek, Nebraska, Enders Dam Year of dam closure 1950 Frenchman Creek near 0.3 y i.3 2/46-9/48 0 Enders 10/48-1/50 -.061 1/50-9/51 -.15 10/51-9/54 -.12 10/54-1/59 -.18 1/59-9/60 -.21 10/60-4/62 -.27 4/62-9/62 -.30 10/62-9/63 -.34 10/63-4/67 -.37 4/67-5/68 -.43 5/68-9/72 -.46 9/72-10/78 -.49 No suitable control station —^Lowest discharge common to all rating tables. The flow exceeded 75 percent of the time. —The flow exceeded about 85 percent of the time. 4/ — The flow exceeded about 68 percent of the time. The flow exceeded about 83 percent of the time. —In adjacent drainage basin. —^The flow exceeded about 40 percent of the time. Marias River, Montana, Tiber Dam Year of dam closure 1955 Marias River near 3.2 2.8 10/55-9/79 0 Chester Marias River near 65 4.0 6/48-4/49 0 Shelby (control 4/49-4/50 -.030 station) 10/51-6/53 -.061 10/54-? -.15 10/57-9/59 -.21 10/59-9/61 -.24 10/61-9/63 -.27 10/63-6/64 -.21 6/64-3/76 -.18 ☆U.S. GOVERNMENT PRINTING OFFICE:1984— 776-041 / 40077 DAYS [5 % Huebnerite Veins near m Round Mountain, Nye County, Nevada GEOLOGICAL SURVEY ^PROFESSIONAL PAPER 1287 fjjGCUMENTS DEPARTMENT OCT 3 ^ '-934 LIBRARY ___ilNIVFRSITY OF CALIFORNIA li.fi. &E‘ OCT 16 •Huebnerite Veins near Round Mountain, Nye County, Nevada By DANIEL R. SHAWE, EUGENE E. FOORD, and NANCY M. CONKLIN GEOLOGICAL SURVEY PROFESSIONAL PAPER 1287 A study of the geology, mineralogy, and chemistry of huebnerite-bearing veins suggests their formation during two major episodes of mineralization UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1984UNITED STATES DEPARTMENT OF THE INTERIOR WILLIAM P. CLARK, Secretary GEOLOGICAL SURVEY Dallas L. Peck, Director Library of Congress Cataloging in Publication Data Shawe, Daniel R., 1925- Huebnerite veins near Round Mountain, Nye County, Nevada. (Geological Survey Professional Paper ; 1287) Bibliography: 40 p. Supt. of Docs. No.: I 19.16:1287 1. Huebnerite. 2. Mineralogy—Nevada—Nye County. I. Foord, Eugene E. II. Conklin, Nancy M. III. Title. IV. Series QE391.H84S53 1984 549'.74 For sale by the Branch of Distribution U.S. Geological Survey 604 South Pickett Street Alexandria, VA 22304 82-600307CONTENTS Abstract ......................... Introduction and acknowledgments Geologic setting of the veins . . . Distribution and form of the veins Alteration related to the veins . . Description of the vein minerals . Primary gangue minerals . . . Quartz................... Muscovite................ Allanite................. Fluorite................. Barite ................... Calcite................... Chalcedony................ Monazite ................. Primary ore minerals........... Huebnerite................ Scheelite ................ Tetrahedrite-tennantite . . Page 1 1 2 4 4 4 6 6 9 10 10 13 14 14 14 14 14 15 16 Description of the vein minerals—Continued Primary ore minerals—Continued Sulfides......................................... Secondary minerals.................................... Sulfides ....................................... Oxides, tungstates, carbonates, sulfates, phosphates, and silicates................................ Remobilized minerals ............................ Summary of textures and paragenesis................... Chemical composition of the huebnerite................... Iron-manganese composition........................... Tungsten-manganese composition ....................... Growth zones.......................................... Trace-element composition............................. Variations in tungsten mineralization.................... Geochemistry of the huebnerite veins..................... Discussion .............................................. Summary and conclusions.................................. References cited......................................... Page 17 19 19 20 21 21 23 23 25 25 27 28 29 30 39 40 ILLUSTRATIONS Page Figure 1. Geologic map of Round Mountain area..................................................................................... 3 2. Geologic map of tungsten area........................................................................................ 5 3. Diagrammatic sketch of huebnerite-bearing quartz vein ............................................................... 7 4. Sketch of thin section showing vein wall ............................................................................ 7 5. Drawing of huebnerite crystals....................................................................................... 8 6. Drawing of quartz with fluid inclusions.............................................................................. 9 7. Photomicrograph showing zoned huebnerite with fluorite......................................................... 12 8. Drawing of quartz-vein selvage...................................................................................... 12 9. Photomicrograph showing transitional vein margin.................................................................... 13 10. Photomicrograph showing late-stage veinlet.......................................................................... 13 11. Drawing of barite after pyrite ..................................................................................... 14 12-28. Photomicrographs showing: 12. Zoned huebnerite............................................................................................. 15 13. Zoned and fractured huebnerite, sample DRS-79-67 ............................................................ 16 14. Zoned and fractured huebnerite, sample DRS-79-68 ............................................................ 16 15. ’’Milled” huebnerite......................................................................................... 17 16. ’’Milled” and “mixed” huebnerite............................................................................. 17 17. Huebnerite, veined and replaced by scheelite................................................................. 18 18. Huebnerite veined with scheelite............................................................................. 19 19. Scheelite filling vug in quartz ............................................................................. 19 20. Tetrahedrite-tennantite filling vug, sample DRS-78-2A ...................................................... 20 21. Tetrahedrite-tennantite filling vug, sample DRS-78-2B 20 22. Copper minerals filling vug................................................................................. 21 23. Pyrite (to limonite) filling vugs........................................................................... 22 24. Vug in quartz vein....................................................................................... 23 25. Fractured and “milled” sphalerite .......................................................................... 24 26. Galena replaced by covellite................................................................................ 25 27. Sphalerite with exsolved chaleopyrite ...................................................................... 26 28. Tetrahedrite-tennantite..................................................................................... 26 illIV CONTENTS Page Figure 29. Scanning electron micrograph showing manganese oxide on huebnerite ................................................... 27 30-33. Photomicrographs showing: 30. Tetrahedrite-tennantite replaced by stibiconite ........................................................... 28 31. Secondary minerals in vein quartz........................................................................... 28 32. Vug in vein quartz.......................................................................................... 29 33. Late veinlet in quartz...................................................................................... 29 34. Paragenetic diagram of huebnerite veins......................................................................... 30 35-38. Graphs of: 35. Spectrographic vs. microprobe analyses of iron in huebnerite................................................ 34 36. Correlation between huebnerite iron and altitude............................................................ 35 37. Comparison of FeO and MnO in huebnerites.................................................................... 36 38. Comparison of MnO and W03 in huebnerites.................................................................... 37 TABLES Page Table 1. Description of huebnerite-bearing quartz-vein samples..................................................................... 6 2. Chemical analyses of muscovite.......................................................................................... 10 3. Spectrographic analyses, minerals from huebnerite-bearing quartz veins................................................. 11 4. Approximate lower limits of determination for elements analyzed by the six-step spectrographic method................. 11 5. Electron microprobe analyses of huebnerite.............................................................................. 31 6. Spectrographic analyses of huebnerite .................................................................................. 32 7. Spectrographic analyses of quartz-vein material......................................................................... 38 8. Other analyses of quartz-vein material.................................................................................. 38HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA By Daniel R. Shawe, Eugene E. Foord, and Nancy M. Conklin ABSTRACT Small huebnerite-bearing quartz veins occur in and near Cretaceous (about 95 m.y. old) granite east and south of Round Mountain. The veins are short, lenticular, and strike mostly northeast and northwest in several narrow east-trending belts. The quartz veins were formed about 80 m.y. ago near the end of an episode of doming and metamorphism of the granite and emplacement of aplite and pegmatite dikes in and near the granite. An initial hydrothermal stage involved deposition of muscovite, quartz, huebnerite, fluorite, and barite in the veins. Veins were then sheared, broken, and recrystallized. A second hydrothermal stage, possibly associated with emplacement of a rhyolite dike swarm and granodiorite stock about 35 m.y. ago, saw deposition of more muscovite, quartz, fluorite, and barite, and addition of scheelite, tetrahedrite-tennantite, several sulfide minerals, and chalcedony. Finally, as a result of near-surface weathering, secondary sulfide and numerous oxide, tungstate, carbonate, sulfate, phosphate, and silicate minerals formed in the veins. Depth of burial at the time of formation of the veins, based on geologic reconstruction, was about 3-3.5 km. The initial hydrothermal stage ended with deposition of quartz at a temperature of about 210°C and pressures of about 240-280 bars (hydrostatic conditions) from fluids with salinity of about 5 weight percent sodium chloride. Fluorite then was deposited at about 250°-280°C from solutions of similar salinity and containing a small amount of carbon dioxide. During shearing that followed initial mineralization, quartz was recrystallized at a temperature of 270°-290°C and in association with fluids of about 5 weight percent sodium chloride equivalent and containing carbon dioxide. Late-stage fluorite was deposited from fluids with similar salinity but devoid of carbon dioxide at a temperature of about 210°C. Huebnerite in the veins is of nearly end-member composition in the huebnerite-wolframite-ferberite series; iron content increases with altitude in the system. The present veins are interpreted to be near the bottom of the original system. A postulated upper part, now removed by erosion, may have formed in granite cupolas where deposits were richer and the tungstate much more iron rich. INTRODUCTION AND ACKNOWLEDGMENTS Small huebnerite-bearing quartz veins in granite just east of Round Mountain, Nye County, Nev., have been known since 1907 (Ferguson, 1921, p. 388-390). The veins have produced only a small quantity of tungsten since their discovery, and residual surface material near the veins also has produced a little tungsten, in 1915. According to Schilling (1963), the total production of tungsten from the Round Mountain district is between 1,000 and 10,000 units of W03 (1 unit equals 20 lbs). H. K. Stager (oral commun., Sept. 1983) indicated that recorded production is 500-1,000 units, mostly from residual surface material. Some of the tungsten veins have been prospected for uranium (Krai, 1951, p. 154). The Round Mountain district is better known as a gold district, having produced a little more than 0.5 million oz of gold through 1959 (Koschmann and Bergendahl, 1968, p. 194), both from placer deposits and from lode deposits in Tertiary rhyolite. A stockwork gold deposit in Tertiary rhyolite was reactivated in 1976, and during the first 2 years of renewed mining more than 50,000 oz of gold was produced (R. J. Leone, written commun., 1978). About 50,000 oz of gold per year has been produced since then (Russell Wood, oral commun., Dec. 1980) for a total production of about 200,000 oz during the present phase of mining. Production of the deposit and development of new reserves are continuing. Newly discovered reserves in a nearby ore body of 8.4 million oz gold and 15.7 million oz silver were announced recently (Wall Street Journal, Jan. 5, 1982). Despite the apparent economic insignificance of the huebnerite veins, their study is important for several reasons. First, the huebnerite is of nearly end-member composition, a rarity worldwide. Second, the veins appear to have been formed many millions of years following emplacement (in Cretaceous time) of a granite plu-ton with which they are associated. Commonly, such tungsten-bearing veins are interpreted to have formed as a final event related to magmatic emplacement of granite. Third, tourmaline in the granite pluton is not related genetically to the tungsten-bearing veins as is common the world over, but instead it is related to an Oligocene granodiorite stock that invades the granite and is substantially younger than the huebnerite veins. Finally, the huebnerite veins present evidence of two major episodes of mineralization. The first episode saw deposition of huebnerite along with quartz, muscovite, and fluorite closely following invasion in Late Cretaceous time of the granite by aplite and pegmatite dikes. The second episode of mineralization was introduction into the huebnerite veins of several sulfide minerals, l2 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA tetrahedrite-tennantite, barite, and chalcedony, and formation of scheelite, in a zone spatially associated with the Oligocene granodiorite stock, and at a time probably shortly following emplacement of the stock. This study is part of a broad investigation by the U.S. Geological Survey of the geology and mineral deposits of the Round Mountain and Manhattan 7V2-min-ute quadrangles, Nye County, Nev. We wish to thank Charles M. Taylor of C. M. Taylor Microprobe Co. for examining one of the huebnerite specimens (DRS-79-68) by electron microprobe to measure iron content of color-zoned crystals. The late Graham R. Hunt of the U.S. Geological Survey performed optical-spectral studies of several huebnerite specimens. Discussions with C. G. Cunningham and J. T. Nash of the Geological Survey proved helpful in our interpretation of the fluid inclusion data, and discussions with G. A. Desborough and B. F. Leonard, III, of the Geological Survey aided our microprobe and other mineralogic studies. Comments on the manuscript by G. P. Landis and W. H. Raymond improved our interpretations of some of the data. GEOLOGIC SETTING OF THE VEINS The huebnerite veins near Round Mountain occur in and near Cretaceous granite that forms a large pluton extending 23 km southeast across the Toquima Range from Round Mountain to a long-inactive silver camp at Belmont (fig. 1). The granite is emplaced in Paleozoic sedimentary and metamorphic rocks. A swarm of Oligocene rhyolite and andesite dikes and a granodiorite stock intrude granite about 3 km east and southeast of Round Mountain. Miocene silicic volcanic rocks over-lie both granite and sedimentary rocks locally in the area, and Quaternary alluvium fills the intermontane valleys that flank the Toquima Range (Ferguson, 1921; Shawe, 1977a). The Paleozoic rocks are marine sedimentary rocks: quartzite, silty argillite, and limestone of Cambrian (possibly in part latest Precambrian) age, and argillite, limestone, dolomite, chert, and quartzite of Ordovician age. Aggregate thickness of the Paleozoic rocks, in part probably the result of repetition by thrust faulting, is at least 1.5 km. Locally the argillite has been metamorphosed to phyllitic shale, and near the contact with granite it consists of knotted (chloritoid) schist and muscovite-biotite schist. In places the limestone is tre-molitized strongly, although not necessarily in proximity to the granite contact. In only a few places does limestone contain epidote and garnet, possibly as a result of contact metamorphism adjacent to granite. “The granite typically is a coarse-grained, granular-textured, light-gray rock that contains quartz, micro-cline, orthoclase, sodic plagioclase, biotite, and muscovite. Accessory minerals are monazite, apatite, [zircon,] and iron-titanium oxide minerals. Fluorite and tourmaline are present locally. Some rocks of the pluton are quartz monzonite and granodiorite rather than granite” (Shawe, 1977a). An apparently early phase of the granite pluton present 8 km south-southeast of Round Mountain and extending to Belmont is porphy-ritic and contains large microcline crystals mostly 2-8 cm long. Much of the granite is foliated, especially near the border of the pluton. Foliation that is manifested chiefly by alined biotite and muscovite flakes, and locally by lensoid quartz and feldspar grains, is conformable with the contact. The moderately outward dipping contacts of the granite pluton and the conformable foliation in the granite impart a domelike form to the pluton. The pluton is cut by numerous aplite and pegmatite dikes and by quartz veins. Aplite and pegmatite commonly occur together in individual dikes, and thus are coeval, although some dikes locally are cut by younger dikes of the same composition. Some foliated dikes are cut by unfoliated dikes. Quartz veins everywhere cut the aplite and pegmatite dikes. The veins include many that are barren and some that are sulfide-bearing in addition to those with huebnerite. In places iron-oxide stain in granite is intense; numerous limonite pseudomorphs after cubic pyrite suggest that most of the iron-stained rock is hydrothermally altered granite that has been weathered. The date of emplacement of the granite was 90-100 m.y. ago (Shawe, 1977a). Several potassium-argon ages on muscovite and biotite from foliated granite, pegmatite dikes, and quartz veins are about 80 m.y., indicating that doming and metamorphism of the granite, and emplacement of pegmatite and aplite dikes and quartz veins, took place in a single episode many millions of years after emplacement of the pluton. The Oligocene dike swarm east and south of Round Mountain trends northeast and is about 1.5 km wide and 11 km long. The Oligocene stock intrudes the swarm near its northeast end. The intrusives are dated as about 35 m.y. old (Shawe, 1977a). Mineralization that formed tourmaline along with andalusite and dumortier-ite in granite and tourmaline in rhyolite dikes, is spatially associated with, and apparently genetically related to, the granodiorite stock. Surprisingly the tourmaline mineralization is in no way related to the much earlier episode of huebnerite mineralization, but as described later, did involve a small degree of tungsten mineralization.GEOLOGIC SETTING OF THE VEINS 3 0 10 KILOMETERS 1 __________________________________I EXPLANATION KNOWN ZONE OF TUNGSTEN MINERALS QUATERNARY ALLUVIUM TERTIARY VOLCANIC ROCKS CRETACEOUS GRANITE CRETACEOUS PORPHYRITIC GRANITE PALEOZOIC SEDIMENTARY AND METAMORPHIC ROCKS ^ MINE Figure 1.—Geologic map of the area near Round Mountain, showing zones of tungsten minerals east and south of Round Mountain.4 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA DISTRIBUTION AND FORM OF THE VEINS The huebnerite veins east of Round Mountain are found in an area that covers about 25-35 km2 (fig. 2). Despite this broad distribution, the veins occur in only a few clusters that are narrow east-trending zones no more than 200 m wide and 4 km long (fig. 2). The huebnerite veins have been found in a vertical interval of about 500 m, between altitudes of about 2,125 m and 2,625 m (6,800 ft and 8,400 ft). Geologic reconstruction of the granite pluton prior to erosion suggests that veins in the core of the pluton were 1.5-2 km below its apex. There is no evidence that individual veins extend for more than about a hundred meters vertically and most probably are much less than that. Veins within the east-trending clusters are narrow (mostly no more than 30 cm wide) and short (generally less than 30 m long). According to Brown (1911), numerous quartz stringers and veins that ranged in thickness from 2.5 cm to 1 m extended across the principal easttrending belt. Mostly tabular, the veins pinch and swell locally and may vary greatly in thickness in a short distance. They occur singly in most places within the east-trending clusters, although in a few places they appear to occur as very short, en echelon segments. Ferguson (1921, p. 389) stated that the veins locally “are close enough together to give a banded appearance to the granite.” Some of the thicker veins split along strike into several thin, nearly parallel veinlets. Most of the veins are in granite where they tend to strike northeast and dip moderately to steeply southeast (table 1). A few veins 5 km south of Round Mountain occur in muscovite-biotite schist close to the granite contact. These veins lie in the schist foliation and dip away (southwestward) from the granite contact at low to moderate angles. One vein is known in quartzite north of the granite contact and east of Round Mountain. At the time of tungsten mining in 1911, as reported by Brown (1911), workings consisted of three tunnels in granite. Tunnel No. 1 followed veins and stringers for 60 m across the belt, with a crosscut extending 60 m west that exposed “a large body of milling ore, varying in width from three to ten feet [1-3 m].” Tunnel No. 2 was “a drift on the strike of the belt for a distance of 400 feet [120 m], exposing large bodies of good milling ore, varying in width from four to twelve feet [1.2 to 3.7 m].” Tunnel No. 3 was also “a drift on the belt for a distance of 390 feet [120 m], cross-cutting many veins and stringers,” with two drifts from it following separate veins 30 m in length. These workings exposed “very good ore bodies***having a width of from five to fourteen feet [1.5 to 4.3 m].” The locations of these tunnels are not certain now, but probably they are the same as the workings known near the bottom of Shoshone Canyon at and near the Darrough prospect (fig. 2). ALTERATION RELATED TO THE VEINS The huebnerite-bearing quartz veins are bounded by a selvage of hydrothermally altered rock. This selvage is most evident in granite and is inconspicuous in schist. Granite adjacent to veins has been strongly altered so that virtually all feldspar crystals have been replaced by sericite. Biotite commonly has been converted to muscovite, and coarse muscovite crystals have developed locally in masses of sericite within replaced feldspar crystals. Quartz, pyrite, chalcopyrite, fluorite, and huebnerite have been added to wall rocks locally. According to Brown (1911) huebnerite crystals occurred in bunches in clay selvages adjacent to the larger veins and in the enclosing granite walls. Our studies were not adequate to determine the thickness of the altered granite selvage adjacent to huebnerite veins. The granite in some places has been widely altered—mainly sericitized and pyritized—probably by mineralizing events other than the tungsten episode, and this fact has prevented our distinguishing the altered rocks related only to tungsten mineralization. Schist wall rocks adjacent to huebnerite veins contain finely crystallized muscovite and biotite, but so does all the schist in the vicinity of the granite contact. Some of the schist wall rocks contain tiny cubes of pyrite (weathered to limo-nite) and minor huebnerite. We were unable to determine alteration products related solely to the emplacement of the huebnerite veins. Locally, within the area of huebnerite veins, the granite has been strongly greisenized with development of much muscovite and some fluorite. Greisen is here defined as altered granite consisting of dominant muscovite and containing significant fluorite; quartz, feldspar, and other minerals may be present. In places the greisen bounds quartz veins not known to contain huebnerite but probably of the same mineralizing episode, and in other places greisen forms small pipes not closely associated with quartz veins. DESCRIPTION OF THE VEIN MINERALS Minerals of the huebnerite veins comprise three groups: primary (hydrothermal) gangue minerals, primary (hydrothermal) ore minerals, and secondary (alteration) minerals formed as a result of weathering of the veins. Primary gangue minerals are quartz, muscovite, allanite, fluorite, barite, calcite, and chalcedony. Monazite may be a gangue mineral, but we have no direct evidence that it is. Primary ore minerals are huebnerite, scheelite, tetrahedrite-tennantite, pyrite,DESCRIPTION OF THE VEIN MINERALS 5 Figure 2.—Generalized geologic map of the tungsten area near Round Mountain, showing localities of samples described in this report. Open circle is locality where huebnerite was identified in vein material; only collected samples are labeled with sample numbers. Open circle with X is locality referred to in text. Thin solid line, contact, ftsm, Paleozoic sedimentary and metamorphic rocks; Kgp, Cretaceous porphyritic granite; Kg, Cretaceous granite; Tgd, Tertiary granodiorite stock; Tv, Tertiary volcanic rocks; Qa, Quaternary alluvium. Cross-lined areas, tungsten-mineralized zones (0.005-1.0 percent W or visible huebnerite); shaded areas, base-and precious-metal-mineralized zones (as much as 3.0 percent Cu, 10.0 percent Pb, 5.0 percent Zn, 0.00036 percent Au, and 0.15 percent Ag). Dashed lines are generalized topographic contours, in feet, labeled with iron content (in percent, determined by spectrog-raphic analysis) in huebnerite at sample localities (see table 6): 0.2 percent Fe approximates 6,800 ft, 0.3 approximates 7,000, 0.7 approximates 7,400, 1.0 approximates 7,600, 2.0 approximates 7,800, 3.0 approximates 8,400 (see fig. 36). (1 ft=0.3 m.)6 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Table 1.—Description of huebnerite-beariruj quartz-vein samples collected east and south of Round Mountain [Localities shown on figure 2; leader (—), not determined] Sample No. Sulfide or DRS- Strike and dip of vein Host rock Scheelite sulfosalt Barite 67-30 N. 15° W. 35° E. Granite no no no 67-33 N. 60° E. 55° SE. —do sparse no yes 73-17 N. 15° E. 70° E. —do no no no 73-19 N. 30° E. 50° SE. —do no no no 73-21 N. 40° E, 60° SE. —do no no no 73-27 — — —do no no no 73-168 — — —do no no no 74-64 N. 40° W. 30° NE. —do — — no 74-66 N. 18° E. 61° E. —do abundant yes no 74-70 N. 42° E. 51° SE. —do no no yes 74-72 — — —do no no no 74-76 N. 35° E. 16° SE. —do sparse no no 74-105 N. 19° E. 39° E. —do no no no 74-193 N. 30° E. 85° SE. —do sparse yes no 74-194 N. 10° E. 80° E. —do — yes no 74-200 — —do no no no 74-218 — — —do no no no 74-221 N. 45° E. 65° SE. —do abundant yes yes 74-252A,B N. 15° E. 79° W. —do sparse yes no 78-1 — — Quartzite no yes no 78-2 N. 35° E. 67° SE. Granite abundant yes yes 79-67A.B — — Schist sparse no no 79-68 N. 15° W. 40° W. —do abundant yes no galena, sphalerite, chalcopyrite, pyrrhotite, covellite, and blaubleibender covellite (a variety of covellite that remains blue in polarized light under the reflecting microscope). Secondary minerals are covellite, blaubleibender covellite, chalcocite, stromeyerite, acanthite, stibiconite, azurite, malachite, chrysocolla, cerussite, anglesite, brochantite, limonite, jarosite, plumbojaro-site, autunite, pyrolusite, psilomelane, other manganese oxides, tungstitef?), ferritungstite, jixianitef?), stolzite, calcite, and opaline silica (chalcedony). PRIMARY GANGUE MINERALS QUARTZ Alpha quartz is the principal component of the veins. It tends to fill the vein space as a solid mass of inter-grown crystals that individually are as much as 10 cm long and 2 cm wide. In places vugs remain, and a few veins have much open space in their cores into which euhedral, commonly slightly tapered quartz crystals penetrate (fig. 3). Crystals here tend to grow perpendicular to the vein walls and may develop forms resembling cockscomb structure. However, many quartz crystals are randomly oriented, and large open spaces within veins may be lined with doubly terminated quartz prisms that have grown transversely on crystals that penetrate the vugs (fig. 3). Most of the quartz is milky white but tends to be only slightly milky or clear and colorless where crystals penetrate vugs. Locally clear quartz crystals that penetrate vugs are smoky gray. Shears parallel to the walls of the veins are common, and shear surfaces generally are slickensided. Only rarely has quartz in parts of the veins been strongly broken and sheared, and then refilled with younger quartz. For example, in one place (sample DRS-74-193) vein quartz has been brecciated, the breccia filled with chalcedony, and the chalcedony-filled breccia veined with clear mosaic quartz (grains less than 1 mm in size) that shows mild strain but is virtually free of fluid inclusions. However, some veins are composite, such that part of the vein consists of quartz that filled an initial open-space fracture that was again opened and filled with quartz. The reopened fracture in some places is one wall of the original vein and in other places it is within the original vein. Massive quartz of the older and younger parts of a vein generally seems to merge owing to recrystallization, as described below. Massive, milky-white vein quartz may be extensively recrystallized, and virtually all of it is strained. As seen in thin sections, this material appears as dominantly anhedral intergrown crystals (mosaic structure; locally,DESCRIPTION OF THE VEIN MINERALS 7 Figure 3.—Diagrammatic sketch of huebnerite-bearing quartz vein (sample DRS-79-68) in quartz-mica schist. Sample is from prospect dump; attitude of schist foliation at prospect is N. 15° W., 40° W. Wavy dash pattern, schist wall rock; sparse stipple, massive milky quartz; clear colorless quartz crystals project into vugs (solid white); cross lining, muscovite selvages and seams; solid black, huebnerite; close stipple, fluorite. almost mortar structure) with uneven extinction, mostly 1-5 mm across (fig. 4), but locally also as strung-out aggregates of tiny crystals that average about 0.1 mm in size (fig. 5). Leonard, Mead, and Conklin (1968, p. 56) described similar quartz in a tungsten lode at the New Snowbird deposit, Idaho. In places the finer grained aggregates occur in zones that parallel vein walls, and they probably resulted from shearing. The recrystallized quartz in places contains very abundant, minute, somewhat cymoidal fractures 0.03-0.15 mm long that are arranged in narrow elongate zones parallel to the vein walls. The zones of minute fractures vary according to the crystallographic orientation of the quartz crystals. Where the c axis is parallel to the vein walls, the fractures are nearly parallel and on end. Where the c axis is oblique or normal to the vein walls, the fractures are arranged en echelon in the elongate zones, which themselves are parallel to the vein walls. Commonly, alternating zones of fractures have alternating orientations of the contained fractures, such that the direction of zone elongation bisects the acute angle formed by the alternating fracture orientations, producing a crude herringbone effect. Fluid inclusions are present in much of the quartz. As seen in thin sections, most of the inclusions occur in strings or trains that commonly transgress grain boundaries (fig. 6). Most sets of subparallel fluid inclusion trains are alined with vein walls, but other orientations are common (fig. 5). In places, sets of subparallel fluid inclusion trains seem restricted to small fields or cells that show an orientation different from those of adjacent fields, as though the fields became rotated unevenly following formation of the trains of inclusions. Fluid inclusions tend to be absent from quartz near free crystal faces that border vugs in the massive vein Figure 4.—Sketch of thin section of sample DRS-79-68B showing locally transgressing contact between huebnerite-bearing quartz vein (right) and quartz-mica schist wall rock (left). Wall rock schist (subparallel dashes) contains abundant tiny pyrite cubes weathered to limonite (black specks); some schist septa are almost wholly muscovite (unoriented dashes) locally with some pyrite (black blebs); other septa contain clear unstrained quartz (solid white) and minor huebnerite (high relief, close-stippled mineral) in muscovite. Quartz vein (no huebnerite present in this thin section) has muscovite (bladed) selvage and contains wisps and flakes of muscovite. Quartz (solid white) is strained, and near vein selvage shows well-developed mosaic structure. A thin, limonite-filled fracture (near top of thin section) extends from a schist septum across vein wall and into muscovite selvage of the vein. quartz (figs. 6, 19, 20). Quartz prisms that penetrate voids and are almost wholly surrounded by space appear to be almost totally devoid of fluid inclusions (figs. 18, 21). Thus, as seen in hand specimens, quartz crystals that penetrate vugs appear to be clear and colorless, whereas massive quartz is opaque and white as the result of diffusion of light caused by millions of microscopic fluid inclusions. The fluid inclusions vary in shape and size. Extremely small (<0.01 mm) inclusions tend to be equidimensional and to show negative quartz crystal forms. Medium-size (0.02-0.05 mm) inclusions are somewhat oblong and slightly irregular, with stubby spurs extending from some corners. Large (as much as 0.1 mm) inclusions tend to be quite ragged and irregular, some being weblike in form. All three forms of inclusions occur mostly in trains, although all also may occur randomly scattered. There seem to be no criteria evident by which any of the fluid inclusions could be identified as primary; virtually all may be secondary. However, J. T. Nash of the U.S. Geological Survey (written communs., January 1969 and March 1980) reported sparse primary fluid inclusions in clear, colorless quartz. Most of the fluid inclusions contain a single gas bubble whose diameter is about half the diameter of the inclusion, but sizes of bubbles relative to apparent volume of the fluid in the inclusions vary. Some inclu-8 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA mmDESCRIPTION OF THE VEIN MINERALS 9 Figure 6.—Quartz charged with fluid inclusions (black specks), thin section of sample DRS-78-2A. Small vug (center) is partly lined with botryoidal chalcedony (on upper left quartz prism face) and filled with fibrous chalcedony; core of vug is granular chalcedony. Note the dominant subparallel orientation of trains of fluid inclusions, some of which transgress boundaries of quartz crystals, and the lack of fluid inclusions near the vug. sions contain a second bubble, consisting of liquid carbon dioxide, within which the gas bubble is suspended. The diameter of this carbon dioxide bubble varies substantially among different inclusions, from slightly longer to perhaps twice as long as the diameter of the enclosed gas bubble. Fluid inclusions that contain only a single bubble are much more abundant than those that also contain liquid carbon dioxide. The carbon dioxide-bearing inclusions are irregularly distributed; most thin sections examined in this study contain very few. Locally, however, the carbon dioxide-bearing inclusions are quite abundant, but still unevenly distributed. One thin section of quartz vein material from sample DRS-67-30 shows abundant carbon dioxide-bearing fluid inclusions in one large area that is defined by a set of subparallel trains of fluid inclusions, whereas another part of the thin section that is defined by another set of trains of fluid inclusions is almost devoid of carbon dioxide-bearing inclusions. According to J. T. Nash (written communs., January 1969 and March 1980), massive milky quartz of the veins contains small fluid inclusions that have filling temperatures of 250°-270°C. Liquid carbon dioxide is present in many of these fluid inclusions. Nash estimated that original fluids contained about 2-10 mol percent (5-20 weight percent) carbon dioxide. Freezing temperatures of fluid inclusions in this quartz indicate a salinity of 5.2 weight percent sodium chloride equivalent. Fluid inclusions in clear quartz crystals that penetrate vugs give filling temperatures of about 190°C. The fluid inclusions in clear quartz contain no liquid carbon dioxide. Freezing temperatures of fluid inclusions in clear quartz indicate a salinity of 5.7 weight percent sodium chloride equivalent. MUSCOVITE Muscovite is a common component of the huebnerite veins, and although it makes up perhaps only 1-5 percent of the vein material, it is everywhere present (figs. 3, 4, 5). The muscovite is colorless, very pale silvery gray, or very pale yellowish green. It almost universally forms a thin selvage 1-5 mm thick on the vein walls. Muscovite flakes tend to grow perpendicular to the vein margins (figs. 4, 5). Although the c axes of the crystals commonly are nearly parallel to the margin, their orientation otherwise appears to be random so that the muscovite forms a feltlike layer viewed normal to the vein wall. In places the muscovite in the vein Figure 5 (facing page).—Zoned huebnerite crystals (dense stip-ple)grown on vein wall (black), thin section of sample DRS-79-68A. Compound (twinned?) crystal at right displays simple growth zone pattern. Note that crystal touches the vein wall locally, and otherwise has grown upon quartz (solid white) or muscovite (bladed) selvage. Complexly zoned huebnerite crystal at left commenced growth at the left end of the compound crystal, near the vein wall. Note that growth zones are grossly different above and below a growth boundary that extends diagonally toward the upper left of the crystal; uniform extinction under the microscope indicates crystallographic continuity across the growth boundary of the two differently zoned segments. The complex growth zonation of the huebnerite suggests a possibly complex history of crystal growth, partial dissolution, and regrowth of a differently zoned segment. Fluorite (high-relief, patterned, labeled F) forms two small crystals, one between quartz and huebnerite crystals near the initiation point of the complexly zoned huebnerite crystal, and one near the middle of the upper edge of that crystal. Barite (high-relief, patterned, labeled B) forms an irregular patch on the upper right point of the complexly zoned huebnerite crystal. Muscovite (plumose and bladed) has grown on huebnerite crystals and in a vug (upper left). Quartz (solid white, and marked with numerous strings of fluid inclusions) is strained; at the left, shapes of original growth prisms remain; at the right, quartz has recrystallized and displays mosaic and mortar structure.10 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA selvage shows a slightly plumose habit (figs. 4, 5). Thin lensoid layers of muscovite 1-2 mm thick are found locally within quartz veins, oriented parallel to the vein walls (fig. 3). Originally these may have been selvages on veins that were reopened and again filled with quartz so that the selvage that clung to the original vein quartz then formed a layer enclosed within vein quartz. Muscovite flakes in some of these internal lenses have a pronounced “rake” (fig. 3) as though the flakes were dragged out of their normal perpendicular alinement by shearing within the vein. Irregular small patches and clusters of muscovite flakes are found inter-grown with quartz and huebnerite throughout the veins. As seen in thin sections, muscovite also occurs as crystals 0.1-1 mm long that fill thin shears in quartz that parallel the vein walls (fig. 4). Unlike the muscovite crystals in thicker lensoid layers that are oriented normal to the muscovite layers, the muscovite crystals in the thinner shears are parallel to the vein walls. These thin, muscovite-filled shears are closely associated with parallel streaks of finely crystallized quartz. A purified portion1 of muscovite selvage from sample DRS-79-68 was analyzed by J. E. Taggart of the U.S. Geological Survey using X-ray fluorescence techniques. Results of the analysis, in weight percent, are given in table 2. Muscovite separated from samples DRS-79-67 and DRS-79-68 was analyzed spectrographically; results are given in table 32. These samples were handpicked but not carefully purified mechanically, and the analytical results show evidence of minor impurities. The unusually high magnesium content of the musco-vites shown by both X-ray fluorescence and spectro-graphic analyses, and the abnormally high silica and abnormally low alumina shown by the X-ray fluorescence analysis, indicate that the muscovites are phengi-tic. In phengite the octahedral A1 is replaced by Mg and (or) Fe+2 and the tetrahedral A1 is replaced by Si. Data for a phengite of similar composition from a pegmatite at Amelia, Va. (Glass, 1935), are given for comparison in table 2. Substantial amounts of silver, copper, molybdenum, lead, and zinc in sample DRS-79-68 (table 3) are likely present in sulfide impurities, and abnormally high amounts of tungsten in samples DRS-79-67 and DRS-79-68 (table 3) are probably present in huebnerite or scheelite. Most of the other minor elements shown in ’Purification of the muscovite entailed trimming of the selvage with adiamond saw, crushing. grinding, and sieving. A 35-60 mesh fraction was treated with magnetic separation and heavy liquids; bromoform (D 2.85) floated quartz, and methylene iodide-bromoform mixture (D 3.10) sank heavy minerals. The sample was finally hand picked to assure a virtually pure concentrate (0.80 g). Approximate lower limits of determination for elements analyzed by the six-step spectrog-raphic method, data for which are presented in tables 3, 6, and 7, are given in table 4. Table 2.—Chemical analysis of muscovite from Round Mountain huebnerite vein compared with chemical analysis of phengite from Amelia, Va., pegmatite (Glass, 1935) [Leaders (—), not looked for] Muscovite from Round Mountain huebnerite vein Phengite from Amelia, Va. , pegmatite SiC>2 49.5 49.16 A12°3 30.0 30.81 Fe (as Fe20^) .17 — FeO — 1.43 MgO 2.94 2.22 CaO .05 — Na20 <.2 .48 k2o 11.10 10.90 tio2 .09 — p205 <•05 — MnO .60 — s <.02 — Loss on ignition* 3.91 — h2o+ — 4.73 h2o- — .15 Other — .19 Totals 98.36 100.07 *30 minutes at 900°C. the spectrographic analyses could be present in the muscovite structure. Amounts of lithium and barium in muscovite from sample DRS-79-68 are sufficiently high that, if added to the components analyzed by X-ray fluorescence, they would raise the total to nearly 100 percent. Muscovite collected from vein material at the Dar-rough tungsten prospect (sample DRS-67-122, same location as DRS-73-168) has a potassium-argon age of 79.2 ±2.2 m.y. (Marvin and Dobson, 1979, p. 23). ALLANITE A few small elongate prisms of a brown, pleochroic mineral showing inclined extinction, high relief, and moderate birefringence were identified optically as allanite in sample DRS-78-2B. The allanite crystals are enclosed in vein quartz (fig. 21) and are inferred to have been deposited during the initial main phase of vein formation. FLUORITE Fluorite is widely but irregularly distributed through the huebnerite veins. Where present it is generally sparse and may make up less than 1 percent of a vein;DESCRIPTION OF THE VEIN MINERALS 11 Table 3.—Semiquantitative spectrographic analyses, in weight percent, of some minerals from huebnerite-bearing quartz veins (sample localities shown on fig. 2) [Analyses by Nancy M. Conklin. N, not detected at limit of detection or at value shown (see table 4). L, detected, but below limit of determination or below value shown. Also looked for but not found: Au, Co, La, Pd, Pt, Te, U, Zr, Ce, Ge, Hf, In, Re, Ta, Th, Tl] Mineral Sample No.— Muscovite DRS-79-67 Muscovite DRS-79-68 Fluorite DRS-79-67 Barite DRS-74-221 Tetrahedrite- tennantite DRS-78-1 Tetrahedrite- tennantite DRS-78-2 Galena DRS-74-221 Si >10 >10 0.03 0.07 1 3 0.3 A1 >10 >10 .07 L .001 .015 .015 .015 Fe .7 1.5 .1 .07 .7 1 .003 Mg 3 3 .005 L .001 .001 .007 L .001 Ca 1.5 .05 >10 .007 1.5 3 .003 Na .3 .3 N .05 N .2 .1 N .05 K 7 10 N N N N N Ti .07 .07 N N N N N Mn .3 .3 .15 .003 .02 .005 .0007 Ag N .01 N .002 >2 >2 .3 As N N N N >10 >10 N B .003 .003 N N N .015 N Ba .07 .7 .0015 >10 .05 .7 .007 Be .007 .007 N .0003 N N N Bi N .0015 N N .15 .7 .3 Cd N N N N .15 .07 N Cr .0002 .0003 N L N N N Cu .0015 .07 N .0015 >10 >10 1 Ga .003 .007 N N N N N Li .07 .15 N N N N N Mo N .07 N N N N N Nb .0015 .0015 .005 N N N N Ni .0015 N N N N N N Pb .003 .7 N .01 .1 .15 >10 Sb N N N N >10 >10 .07 Sc .0007 .001 N N N N N Sn N .001 N N N N N Sr .03 .003 .7 .5 .007 .07 .0015 V .007 .007 N N .0015 .0015 N W .3 .02 .3 N .03 N N Y N N .015 N N N N Yb N N .0015 N N N N Zn N .03 N N 7 5 N Table 4.—Approximate lower limits of determination for elements analyzed by the 6-step spectrographic method (in percent) Si 0.002 Ga 0.0005 Sr 0.0005 A1 .001 Gd .005 Sm .01 Fe .001 Ge .001 Ta .02 Mg .002 Hf .01 Tb .03 Ca .002 Ho .002 Te .2 Na .05 In .001 Th .02 K .7 Ir .005 Tl .005 Ti .0002 La .003 Tm .002 P .2 Li .005 U .05 Mn .0001 Lu .003 V .0007 Ag .00005 Mo .0003 W .01 As . 1 Nb .001 Y .001 Au .002 Nd .007 Yb .0001 B .002 Ni .0005 Zn .02 Ba .00015 Os .005 Zr .001 Be .0001 Pb .001 Bi .001 Pd .0001 Cd .002 Pr .01 Ce .015 Pt .003 Co .0003 Re .003 Cr .0001 Rh .0002 Cu .0001 Ru .001 Dy .005 Sb .015 Er .005 Sc .0005 Eu .01 Sn .001 locally, however, such as where wall rocks are schist, it may constitute as much as 25 percent of short segments of a vein. It is mostly white but also pale green and pale to deep purple; Ferguson (1921, p. 389) described “complex delicate pink crystals” in many veins; color varieties occur both separately and together. Typically, fluorite is a late vein mineral, filling vugs. Locally, the tabular interior parts of veins that were open space following quartz deposition are lined or filled with fluorite (fig. 3). Crystals that line vugs commonly have cubic form. Also, fluorite occurs in minor amounts associated with both early and late vein minerals. Tiny crystals are seen under the microscope in or adjacent to huebnerite (fig. 5), and fluorite is found enclosed in some growth layers in zoned huebnerite (fig. 7). Locally it is found intergrown with muscovite in vein selvages (fig. 8), where it apparently formed during initial stages of vein formation. Where quartz has been recrystallized, traces of fluorite may occur together with muscovite and pyrite in small patches interstitial to quartz12 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 0.5 mm Figure 7.—Part of a zoned huebnerite crystal, plane-polarized light, thin section of sample DRS-79-68A. Crystal faces on the inner part of the huebnerite crystal (lower right) are coated with a thin, uneven layer of fluorite. Zoned and partly broken layers of huebnerite that contain blebs of quartz and fluorite overlie the inner fluorite layer. Quartz is molded against the corroded exterior of the huebnerite crystal (upper left). grains (fig. 9). Fluorite occurs also in trace amounts as a post vein mineral, within parts of veins that were strongly sheared following vein formation (fig. 8). Examination of sample DRS-79-67 with an ultraviolet lamp indicated the presence of both phosphorescent and nonphosphorescent fluorite, suggesting the possibility of two separate episodes of fluorite deposition that resulted in fluorites of slightly different trace-element composition. Also, some fluorite in the huebnerite veins shows bluish-white fluorescence under short-wave ultraviolet radiation, and some does not. A six-step semiquantitative emission spectrographic analysis of the pale fluorite in DRS-79-67 is presented in table 3. Minor impurities suggested by the data are pyrite (iron), muscovite (silicon and aluminum), and huebnerite (manganese and tungsten). Strontium, yttrium, and ytterbium likely are present in the fluorite structure. Figure 8.—Edge of huebnerite-bearing quartz vein (above) and sheared wall rock (below), thin section of sample DRS-79-68C. Sheared wall rock consists of muscovite (subparallel dashes), pyrite weathered to limonite (black cubes and blebs), and a trace of fluorite (high relief, patterned, labeled F); the sheared zone (2-5 mm thick) laterally (beyond the field of view) transgresses the margin of the vein and may be enclosed wholly in vein material or wholly in wall rock. Tetrahedrite-tennantite and its oxidation products occur in the sheared zone elsewhere. Vein selvage is muscovite (bladed and plumose), randomly oriented against the vein wall, and inward tending to be plumose, locally intergrown with fluorite that constitutes about 20 percent of the selvage. Huebnerite (high relief, dense stipple) and strained quartz (solid white, marked by fluid inclusions) fill the interior of the vein. Fluid inclusions are present in fluorite. According to J. T. Nash (written communs., January 1969 and March 1980), fluorite that fills vugs in quartz vein material contains fluid inclusions that have filling temperatures of 230°-260°C. Freezing temperatures of fluid inclusions in this fluorite indicate a salinity of 5.8-6.0 weight percent sodium chloride equivalent. Some of the fluorite that has fluid inclusions with a filling temperature of 260°C contains no visible carbon dioxide but decrepitates on heating, suggesting the presence of some carbon dioxide dissolved in the aqueous phase of the inclusions. Freezing temperatures of fluid inclusions in this fluorite indicate a salinity of 4.5 weight percent sodium chloride. Late-stage fluorite, also occurring as vug fillings, gives filling temperatures of about 190°C. No liquid carbon dioxide is evident in fluid inclusions in the late-stage fluorite.DESCRIPTION OF THE VEIN MINERALS 13 0.5 mm Figure 9.—Transitional vein margin, plane-polarized light, thin section of sample DRS-67-33D. Field of view includes quartz (upper right) that contains fluid inclusions and is similar in appearance to vein quartz, plagioclase (upper left) that shows limonite-filled cleavages, and randomly oriented bladed muscovite (lower left); all three minerals are intergrown in lower right. Pyrite cube (center) that is oxidized to limonite and contains small blebs of quartz is abutted by an elongate grain of fluorite on its upper left face. Massive vein quartz lies just below the field of view. BARITE Barite is present only locally and in very small amounts in the huebnerite veins. Rare, small, irregular patches that are aggregates of tabular crystals have grown on huebnerite (fig. 5), where barite probably was an early vein mineral. Barite occurs also as white crystals less than 1 mm across in places intergrown with quartz and elsewhere penetrating vugs where free surfaces show characteristic tabular orthorhombic form. Locally, tabular barite crystals as large as 3 cm across and 0.5 cm thick occur in vugs in vein quartz. A thin section of sample DRS-74-70 as seen under the microscope contains late-stage barite (fig. 10). In this sample “milled” granite lies in a matrix of vein quartz that contains huebnerite and is itself veined by thin (about 1 1 mm Figure 10.—Late-stage veinlet in huebnerite-bearing quartz vein, plane-polarized light, thin section of sample DRS-74-70. Quartz (right, and upper left) is strained, and is crowded with fluid inclusions, some of which contain liquid carbon dioxide; note subparallel trains of abundant fluid inclusions. Irregular late-stage veinlet (lower left to upper right) is chalcedony that contains barite pseudomorphs after cubic pyrite (light-gray, squarish crystals), muscovite (randomly oriented, bladed crystals), and scattered dustlike limonite (black). Note that some barite pseudomorphs and muscovite crystals occur in vein quartz outside the chalcedony vein-let. Beyond the field of view, clear quartz veinlets that contain barite cut chalcedony veinlets. mm or less), irregular veinlets of chalcedony that contain muscovite and barite. The barite appears to have replaced pyrite inasmuch as it occurs as aggregates of barite crystals pseudomorphous after a cubic mineral (fig. 10). The cores of some such pseudomorphs contain limonite, probably oxidized from original pyrite (fig. 11). Barite separated from sample DRS-74-221, collected from a huebnerite- and sulfide-bearing quartz vein (fig. 2), was analyzed spectrographically (table 3). Of the elements detected, barium, and probably calcium, lead, and strontium, are present in the barite structure; other elements may be present in mineral impurities within the analyzed barite.14 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 1 mm Figure 11.—Barite pseudomorphous after pyrite in quartz vein, thin section of sample DRS-74-70. Variously oriented barite crystals (white, patterned) replace periphery of pyrite cube (center), core remnant of which has been oxidized to limonite (black). Unidentified high-relief, moderately birefringent, pale-yellow (color not shown) mineral (labeled J) at lower edge of pseudomorph may be jarosite. Pseudomorph lies in strained vein quartz (white) that shows mosaic and mortar structure and contains abundant trains of fluid inclusions (dark specks, mostly alined subparallel). Muscovite (bladed) forms individual crystals and short stringers scattered in quartz. CALCITE Calcite occurs in minor amount and only locally as tiny (0.03 mm and less) crystals associated with scheel-ite, pyrite (oxidized), and muscovite, and it is considered to be hydrothermal in origin in those places. CHALCEDONY Chalcedony is found in late-stage veinlets in the huebnerite-bearing quartz veins; because it contains minerals such as muscovite, barite, and pyrite (weathered to limonite), it is interpreted to be a hydrothermal mineral in these occurrences. MONAZITE We have no direct evidence of the occurrence of monazite in the quartz veins, although Ferguson (1921, p. 389) reported recovery of monazite along with hueb-nerite from residual surface material near a vein 3 km southeast of Round Mountain. Ferguson (1921, p. 389-390) inferred that the monazite and huebnerite were derived from a pegmatitic quartz vein. Monazite is an accessory mineral of the granite locally, for example in Shoshone Canyon along the vein where sample DRS-73-168 was collected. PRIMARY ORE MINERALS HUEBNERITE Huebnerite is the principal primary ore mineral of the tungsten-bearing quartz veins. Small local concentrations may exceed 10 percent of the vein material (these parts have been prospected or mined), but the huebnerite content of much of the veins ranges from but a few percent down to substantially less than 1 percent. Huebnerite content varies as much within a vein as from vein to vein. The huebnerite is dark brownish black on fresh cleavage surfaces, showing deep-blood-red internal reflection in bright light. Cleavage is well developed on {010}, and cleavage (parting) less developed on {100}. Huebnerite occurs as stubby, tabular to prismatic crystals, generally subhedral to euhed-ral, either singly or as composite clusters. Single crystals are as much as 5 cm long, although commonly 1 cm or less in size. In granite huebnerite tends to occur in the cores of quartz veins where it is intergrown with quartz as subhedral crystals and as aggregates of crystals. In this environment tabular, somewhat tapered crystals are common. In schist huebnerite tends to occur as stubby, euhedral crystals growing from one wall of flat veins (fig. 3). Vein material with huebnerite growing from one wall was not observed in place in schist, but was noted in material from prospect dumps. Presumably the huebnerite is on the footwall of these flat veins, analogous to the example from Panasqueira, Portugal, illustrated by Kelly and Rye (1979, fig. 9). A thin (3 cm), northeast-striking, nearly vertical quartz vein in granite at the locality marked with an open circle and X on figure 2 displays prismatic crystals that grow perpendicularly into the vein from its northwest wall. Where huebnerite has grown from vein walls, crystals commonly fan or flare outward from their base in the direction of growth (fig. 3). The center of the base of such a crystal is in direct contact with the wall, and outward the crystal rests on a progressively thicker selvage of muscovite, which shows thickest development on the walls between separate huebnerite crystals (fig. 3). This relationship indicates simultaneous deposition of huebnerite and muscovite beginning with the onset of vein mineralization. Inclusion of flakes of muscovite within huebnerite also indicates synchronous crystallization of huebnerite and muscovite. Many of the subhedral to euhedral crystals of huebnerite show growth zones that, as seen in thin sectionsDESCRIPTION OF THE VEIN MINERALS 15 Figure 12.—Zoned huebnerite crystal, plane-polarized light, thin section of sample DRS-79-68. Rectangular outline in A shows area of B. and less plainly in polished sections, are alternating lighter and darker brown zones that reflect the crystallographic form of the growing crystals (figs. 12-14). Some euhedral huebnerite crystals have been fractured, dismembered, or otherwise deformed during growth (fig. 14). Growth zones may be truncated or separated, and the resulting fracture space filled by later stage huebnerite or by quartz (figs. 13, 14). This dismemberment resulted from brecciation rather than replacement, as can be shown locally by the fact that segments of huebnerite can be fitted together if interstitial later huebnerite and quartz are removed. The “milling” of some huebnerite crystals as a result of excessive deformation of the vein material was accompanied by recrystallization of quartz, as evidenced by the fact that the dispersed “milled” fragments of huebnerite are embedded in fine-grained mosaic quartz (figs. 15, 16). In places where quartz shows pronounced mosaic structure and is sheared (for example, samples DRS-67-33 and DRS-74-221), huebnerite may appear “worm eaten” and is intergrown with granular quartz crystals. Locally this huebnerite may have lighter colored, mottled patches against quartz. Some huebnerite crystals show evidence of solution etching where they are adjacent to voids in the vein material, as well as where they are embedded in recrystallized quartz. Beddoe-Stephens and Fortey (1981) reported that, in the Car-rock Fell tungsten deposit in the English Lake district, wolframite that occurs in quartz veins near vugs tends to be veined and replaced by scheelite, arsenopyrite, pyrite, carbonate minerals, and less commonly by quartz. The density measured by Berman balance of three samples of huebnerite from Round Mountain quartz veins ranges from 7.24 to 7.26. Calculated density of pure MnW04 is 7.234. Chemical data for Round Mountain huebnerites are presented in a later section of this report. SCHEELITE Scheelite is a widespread but sparse component of the huebnerite veins. It is restricted, however, to a specific environment and is therefore present in some veins and virtually absent from others (table 1). It is associated with sulfides and sulfosalts within the areas16 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 5 mm Figure 13.—Zoned and fractured huebnerite crystal, plane-polarized light, thin section of sample DRS-79-67. White patches in lower half of view are areas plucked during grinding of thin section. Upper part of crystal lies along an irregular boundary against strained quartz (also white). Irregular corrosion of the outer part of the huebnerite crystal before crystallization of quartz is implied. Note converging growth zones in huebnerite crystal that suggest diminished and arrested growth of the crystal periodically along some faces, and that imply a directional influx of material that formed the growing crystal. Some growth zones have been truncated (upper left), but the mechanism of truncation is not evident. Note also fractured growth zones at top center where fractures are filled by light-colored huebnerite, and corroded huebnerite growth zones at right center filled by light-colored huebnerite. Scale is approximate. of base- and precious-metal mineralization of the tungsten-bearing veins, and it occurs in some veins devoid of sulfides and sulfosalts but close to the areas of base-and precious-metal mineralization. With one known exception, scheelite is absent from veins more distant from these zones of mineralization (compare table 1 and fig. 2). Within scheelite-bearing veins the mineral occurs in sparse to trace amounts. On a microscopic scale it is seen to occur closely associated with huebnerite, and there it ranges from a few percent to perhaps 50 percent of the tungsten-bearing mineral present in a thin section (sparse to abundant in table 1). White to very pale yellowish green scheelite occurs as irregular patches and blebs on the margins of and within huebnerite crystals, and as small veinlets or fracture fillings a fraction of a millimeter wide in huebnerite (figs. 17, 18). Locally, veinlets of scheelite as much as 1 mm wide fill fractures that extend through huebnerite and into adjacent vein quartz. Small specks, patches, and veinlets of scheelite occur in quartz near concentrations of huebnerite. The veinlets, some of which are aggregates of scheelite and muscovite (fig. 17), appear to fill late shears; the patches commonly are fillings of vugs in quartz (fig. 19). These relations of scheelite to Figure 14.—Zoned and fractured huebnerite crystal, plane-polarized light, thin section of sample DRS-79-68. White patches within crystal are areas plucked during grinding of thin section. Quartz (white, upper right) has grown on huebnerite. Note segment of zoned huebnerite (center) that was fractured and displaced upward (so that growth zones are offset) and then overgrown with continuous layers of huebnerite. Bleaching or filling of huebnerite has occurred along the margins of the fractured and displaced segment. Scale is approximate. other minerals are similar to those described for scheelite in a tungsten-bearing quartz lode at the New Snowbird deposit, Idaho, by Leonard, Mead, and Conklin (1968, p. C6). Microprobe examination of scheelite associated with huebnerite in sample DRS-74-193 revealed the presence of about 0.045 percent MnO, 0.35 percent FeO, 81.64 percent W03, 18.80 percent CaO, and 0.00 percent Mo03 (total, 100.83 percent; average of two analyses). Examination of the scheelite under ultraviolet light using a scheelite fluorescence analyzer manufactured by Ultra-violet Products, Inc., Los Angeles, Calif.1, showed that the scheelite fluoresces bright blue and contains virtually no molybdenum. TETRAHEDRITE-TENNANTITE Tetrahedrite-tennantite (referred to as tetrahedrite by Ferguson, 1921, p. 389-390) is present in some of the huebnerite veins where it is generally sparse but may locally make up a few percent of the vein material. These veins are found in zones of base- and precious-metal mineralization indicated on figure 2. Tetrahedrite-tennantite forms small grains as much as 5 mm across in milky white quartz. Under the microscope the grains are seen to fill vugs in quartz (figs. 20-22). ’Brand or manufacturers’ names used in this report are for descriptive purposes only and do not constitute endorsement by the U.S. Geological Survey.DESCRIPTION OF THE VEIN MINERALS 17 1 mm Figure 15.—“Milled” huebnerite crystal, plane-polarized light, thin section of sample DRS-79-68C. Original huebnerite crystal (dark gray) has been broken and is dispersed in strained and recrystallized quartz (white) that contains large numbers of fluid inclusions (tiny dark specks). Irregular patches of muscovite and fluorite (light gray) are intermingled with very fine grained mosaic quartz and fragments of huebnerite to the right of the large crystal of huebnerite. Tetrahedrite-tennantite separated from samples DRS-78-1 and DRS-78-2, collected from huebnerite-bearing veins (fig. 2), was identified optically in polished sections and by X-ray diffraction, and analyzed spectrographically. The spectrographic analyses are given in table 3. The substantial amounts of copper, antimony, arsenic, zinc, and silver indicated in the analyses show that the sulfosalt is zincian tetrahedrite-tennantite. Also, appreciable amounts of silicon, calcium, barium, and tungsten given in the analyses suggest the presence of some impurities in the analyzed sul-fosalts such as quartz, calcite, barite, and a tungsten mineral. The small amounts of aluminum given rule out the presence of much feldspar. Also, the low aluminum contents rule out a large number of sodium-bearing minerals to account for the sodium that is present; possibly a secondary iron-sodium sulfate such as natrojaro- 1 mm Figure 16.—“Milled” and “mixed” huebnerite fragments, plane-polarized light, thin section of sample DRS-79-68E. Extremely disrupted huebnerite (dark gray) consists of a large number of disoriented fragments of different sizes, color shades, and degrees of zoning, embedded in strained mosaic quartz (white) that contains large numbers of fluid inclusions (dark specks). site or copper-sodium sulfate such as krohnkite or nat-rochalcite could account for the presence of sodium. Elements such as iron, bismuth, cadmium, and lead may be present mostly in the sulfosalt structure, or otherwise in minor amounts of sulfide impurities. Minor amounts of tetrahedrite-tennantite are found along with pyrite, chalcopyrite, and fluorite dispersed as small crystals (less than 1 mm) in wall rock near huebnerite veins at the Darrough prospect (fig. 2). SULFIDES Sulfides are found in only trace to minor amounts (probably less than 1 percent) in most of the vein material from the tungsten-bearing veins. Local concentrations in vein material of hand-specimen size may reach several percent. Like tetrahedrite-tennantite, sulfides occur in the tungsten-bearing veins within the zones of base- and precious-metal mineralization shown on figure 2.18 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Figure 17.—Huebnerite, veined and partly replaced by scheelite, plane-polarized light, thin section of sample DRS-78-2B. Rectangular outline in A shows area of B. A, Quartz (white) that is locally sheared, strained, and universally crowded with fluid inclusions (dark specks) in variously oriented trains. Mosaic structure is developed only locally in quartz, and apparent original prismatic form of much of the quartz is still evident. Enclosed in quartz is a central patch of huebnerite (dark gray), irregularly veined and largely replaced by scheelite (medium gray). Quartz prism faces surround Pyrite is by far the most abundant and widespread sulfide. It and other sulfides such as sphalerite, galena, covellite (CuS), chalcopyrite, and pyrrhotite characteristically are late-stage minerals in the veins. Both covellite and blaubleibender covellite are present in the veins, but not everywhere together. The sulfides occur in shears in quartz, filling vugs in quartz in the fashion of tetrahedrite-tennantite (figs. 20-22), and in late-stage veinlets of chalcedony (fig. 10). Figure 23 shows pyrite (altered to limonite) filling vugs in quartz and partly replacing quartz prisms that penetrate vugs. Pyrite also forms crystals about 1 mm in diameter that, along with quartz prisms, line vug walls, and that appear to have grown contemporaneously with the late quartz (fig. 24). Pyrite occurs also as cubes disseminated locally in vein quartz (fig. 9) and widely in wall ,______1 mm_______, some of the patch and suggest that the scheelite may be in part a vug filling. Veinlet in quartz (lower part of view) contains scheelite and muscovite (bladed, light gray). B, Scheelite (light gray) appears to replace huebnerite (medium gray) irregularly; also it fills cracks where huebnerite has been fractured and separated, as for example in the upper right where a Z-shaped fracture separated slightly and slipped in a left-lateral sense to provide an opening that was filled with scheelite. rocks (figs. 4, 8). Sulfides also occur as irregular grains enclosed in mosaic quartz where they may have crystallized simultaneously with mosaic quartz but are probably younger than the initial stage of quartz mineralization. Sphalerite that is enclosed in mosaic vein quartz near vugs has been fractured, “milled,” and dispersed apparently simultaneously with recrystallization of quartz (fig. 25). Figure 26 shows examples of covellite that apparently replaced galena, and figure 27 shows chalcopyrite exsolved from sphalerite. Molybdenite has not been recognized within huebnerite veins, although spectrographic analyses of a few samples (DRS-79-68, table 3; DRS-67-33, table 6; DRS-74-194, DRS-74-221, DRS-74-252A, table 7) show minor amounts of molybdenum. Some late shearing in the veins postdated deposition of sulfide and sulfosalt minerals.DESCRIPTION OF THE VEIN MINERALS 19 1 mm Figure 18.—Huebnerite veined with scheelite, plane-polarized light, thin section of sample DRS-79-68A. Fractured and somewhat granulated huebnerite (dark gray) has been veined and otherwise filled with late-stage scheelite (medium gray). Huebnerite and scheelite are veined and surrounded by strained quartz (white) that displays mosaic structure. Quartz is crowded with fluid inclusions (dark specks). A spectrographic analysis of galena from DRS-74-221 is given in table 3. Silver, bismuth, and antimony probably are held in the galena structure, copper is present mostly in covellite, and silicon, aluminum, barium, and strontium are present in minor mineral impurities. SECONDARY MINERALS Secondary minerals in the huebnerite veins near Round Mountain are mostly oxidation products that resulted from near-surface weathering. They occur as alteration halos around primary vein minerals, particularly the ore minerals, and in fractures and vugs near the primary vein minerals. A few secondary minerals appear to be low-temperature precipitates that resulted from solution and redeposition of vein or wall rock components by near-surface ground waters. 1 mm Figure 19.—Vug in quartz (white) filled with scheelite (medium gray), plane-polarized light, thin section of sample DRS-79-68A. Quartz that bounds the vug consists of well-formed prisms, is only slightly strained, and contains few fluid inclusions. Elsewhere the quartz has well-developed mosaic structure, is strongly strained, and contains numerous fluid inclusions (dark specks). Huebnerite (dark gray, at top) is part of a grain that has been partly broken and filled with quartz locally. Muscovite (bladed, light gray) is randomly oriented in a clot in quartz 1 mm above the scheelite-filled vug, and elsewhere forms plumose bunches in and near huebnerite and in the vug. Some of the larger black blebs in quartz are limonite probably oxidized from pyrite. SULFIDES Sulfides have formed commonly as initial breakdown products of weathering of primary sulfide and sulfosalt minerals. As shown on figure 28, chalcocite (Cu2S), acanthite (Ag2S), and stromeyerite ((AgCu)2S), identified by optical properties, X-ray diffraction, and the scanning electron microscope, have formed as an initial breakdown of tetrahedrite-tennantite. In places covellite and blaubleibender covellite appear to be alteration products of primary copper-bearing sulfide minerals, and in addition appear to be alteration products of the initial alteration minerals chalcocite and stromeyerite (fig. 28).20 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA !______0.5 mm______| Figure 20.—Tetrahedrite-tennantite (black) filling vug in quartz (white), plane-polarized light, thin section of sample DRS-78-2A. Quartz contains abundant fluid inclusions (dark specks); inclusions tend to be least abundant where quartz prisms penetrate the vug. Irregular, light-gray veinlets near top of view are muscovite. Chrysocolla fills a minute vug at upper left. OXIDES, TUNGSTATES, CARBONATES, SULFATES, PHOSPHATES, AND SILICATES Oxide minerals have formed directly from weathering of some ore minerals such as huebnerite, or are further breakdown products of weathered sulfide and sulfosalt minerals. Manganese oxides, including pyrolusite (Mn02), psilomelane (colloidal Mn02), and probably other manganese oxides, are common alteration products of huebnerite (fig. 29). Some of the manganese oxide minerals contain significant barium, as indicated by scanning electron microscope studies. Weathered huebnerite has a dull-black appearance as a result of the presence of manganese oxide minerals. Black and dark-brown manganese oxide and manganese-iron oxide mixtures commonly line vugs or coat fractures in the quartz veins near the sites of primary ore minerals. Manganese oxide dendrites coat fracture surfaces in many places in and near the veins. i___1 mm_____j FIGURE 21.—Tetrahedrite-tennantite (black) filling vug in quartz (white), plane-polarized light, polished thin section of sample DRS-78-2B. White blebs within tetrahedrite-tennantite, some of which contain bubbles, are areas plucked during grinding of the thin section. Dark-gray areas within the large vug and in smaller nearby vugs are weathering products, mostly malachite. Quartz contains abundant fluid inclusions (dark specks) that are sparse or absent in small quartz prisms that penetrate the vug. Small, dark-gray, needlelike prisms at center right and above tetrahedrite-tennantite-filled vug are allanite. Light-gray vug filling at lower right edge of view is chalcedony. Stibiconite (Sb306(0H)) has formed both as an early (fig. 30) or late (fig. 28) weathering product of tetrahedrite-tennantite. Limonite is almost ubiquitous in and near the huebnerite veins. It has been derived mostly from the oxidation of pyrite, and it occurs commonly as pseudomorphs after pyrite (figs. 4, 8, 9, 10, 11, and 19), fracture fillings (figs. 4 and 33), and as a wash or in dustlike form permeating vein material and wall rocks. Tungstite, a hydrous oxide, occurs as a minor buff-yellow alteration product on the surfaces of and in fractures in huebnerite crystals. It is generally associated with other secondary tungsten minerals and (or) stibiconite. Chrysocolla is the only silicate weathering product identified in the huebnerite veins. It may have formedDESCRIPTION OF THE VEIN MINERALS 21 1 mm Figure 22.—Copper minerals (shades of gray) filling vug in quartz (white), plane-polarized light, polished thin section of sample DRS-78-2B. Island remnants of original vug filling are irregular blebs of tetrahedrite-tennantite; these are surrounded by secondary copper and antimony oxides (in the photomicrograph these minerals are indistinguishable and together appear as nearly black patches); surrounding the tetrahedrite-tennantite and associated oxides and filling the arm of the vug in the lower three-fourths of the view are green secondary copper minerals (mostly malachite?) and minor brown secondary antimony(?) and arsenic(?) minerals (dark gray); the other arms of the vug (upper left and upper right) contain azurite (medium gray) that shows concentric growth layers. Beyond the area of the view, azurite is bounded by crystal faces in open space of the vug. Note that the quartz prisms that penetrate the vug are nearly devoid of fluid inclusions (dark specks) that abound in the remaining quartz. Small vug in lower right is lined with chalcedony. early in the weathering of copper-bearing sulfides inasmuch as it lines vugs that were later filled with malachite (fig. 31). Chrysocolla also was deposited before or contemporaneously with the precipitation of chalcedony that fills vugs (fig. 32). It fills late fractures, and itself is veined by very late chalcedony (fig. 33). The tungstates, ferritungstite ((W,Fe+3)204(0H)2-1/2-H20—Machin and Susse, 1975—yellow tungstic ochre, sample DRS-74-221), jixianite(?) (Pb(W,Fe + 3)2- (0,0H)7, brownish-red tungstic ochre), and stolzite (PbW04, orange-yellow tungstic ochre), are minor alteration products found associated with the primary tungsten minerals. Because ferritungstite, jixianite, and stibiconite are all reported to have pyrochlore-type structure, conventional X-ray diffraction techniques cannot readily distinguish them with certainty, particularly when they are known to occur as mixtures. Jixianite is still an incompletely described mineral and discrepancies exist in the published descriptions (Jian-chang, 1979; Hogarth and Chao, 1979, p. 1330). Several carbonates—azurite, malachite, cerussite, and calcite—and sulfates—anglesite, brochantite, and jarosite-plumbojarosite—are common alteration products in the vicinity of primary vein sulfides. According to Krai (1951, p. 154), the phosphate autunite occurs in trace to small amounts in the tungsten veins east of Round Mountain; the mineral presumably is a weathering product of an unidentified primary uranium mineral. REMOBILIZED MINERALS Two minerals appear to have been in part simply remobilized from earlier vein minerals and deposited in very late stages. The silica of chalcedony probably was derived mainly from preexisting vein quartz; it appears to have been deposited in several late stages of mineralization (see fig. 34), the earlier ones probably low-temperature hydrothermal and the later ones possibly during weathering by cold ground waters. Opaline silica that fluoresces bright yellowish green under both shortwave and longwave ultraviolet radiation, probably owing to the presence of the uranyl ion, commonly coats fracture surfaces. Calcite that fluoresces pale salmon to orange pink and that may be slightly phosphorescent sparsely coats fractures in the veins along with other secondary minerals, particularly chalcedony. The calcite may have been reworked from earlier primary vein calcite or from calcium-bearing wall rocks. SUMMARY OF TEXTURES AND PARAGENESIS The huebnerite-bearing quartz veins near Round Mountain were filled with gangue and ore minerals to form nearly solid lenticular-tabular masses that locally contain vugs. Most vein walls are sharp and bounded by muscovite selvages, but in places veins merge with country rocks as if formed in part by replacement, and locally near veins some vein minerals impregnate the country rocks. Figure 34 shows the interpreted paragenesis of mineral deposition, and events of deformation and recrystallization. The paragenesis is based on the previously described detailed interrelations of minerals—local sequences of mineral deposition, and correlations of specific mineral deposition from place to place.22 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Figure 23.—Pyrite (altered to limonite) filling vugs in quartz and partly replacing quartz prisms that penetrate vugs, plane-polarized light, thin sections of samples DRS-74-252A, (A) and DRS-74-252A2 (B). A, Quartz prisms (white) that penetrate vugs contain The paragenetic diagram (fig. 34) shows that an initial hydrothermal phase witnessed deposition of major amounts of quartz and lesser amounts of muscovite, huebnerite, fluorite, barite, and pyrite. A period of shearing and recrystallization followed in which much quartz was reconstituted; some new muscovite, huebnerite, fluorite, and pyrite were added at this stage, and muscovite and huebnerite were in part mechanically redistributed. Some scheelite and tetrahedrite-tennantite may have formed in the later part of this stage. A second hydrothermal stage saw deposition along shears and in the earlier vugs of small additional amounts of muscovite, quartz, fluorite, and barite; formation of minor calcite, scheelite, tetrahedrite-tennantite, and several sulfides; and finally precipitation of chalcedony. In places more than one younger episode of hydrothermal remineralization is evident. Near-surface weathering of the veins resulted in formation of small amounts of calcite, covellite, chalcedony, chalcocite, stibiconite, chrysocolla, malachite, azurite, manganese oxide, and jarosite. bleblike grains of limonite after pyrite (black). B, Granular pyrite in upper right (orbicular forms) has altered to concentric layers of limonite, jarosite, and chalcedony. Some uncertainty exists regarding part of the relations shown on figure 34. Muscovite is associated with chalcedony in some late veinlets that cut earlier vein minerals; the muscovite appears to have grown in place although it is possible that the muscovite crystals were mechanically emplaced in the chalcedony veinlets. More than one major episode of shearing may have affected the veins: an earlier episode that took place when quartz was recrystallized at the time carbon dioxidebearing solutions moved through the veins, and a later (perhaps much later) episode that preceded and (or) accompanied scheelite, tetrahedrite-tennantite, and sulfide mineralization. The paragenetic diagram (fig. 34) has been simplified inasmuch as several minerals of the veins are not shown. The positions of some minerals in the paragenetic sequence are uncertain, although most of the deleted minerals can be placed in their appropriate positions. Allanite probably was contemporaneous with early vein quartz. Chalcopyrite probably belongs with sphalerite; blaubleibender covellite with covellite; acanthite andDESCRIPTION OF THE VEIN MINERALS 23 i 1 mm_____| Figure 24.—Vug in quartz vein, under crossed nicols, thin section of sample DRS-74-252B2. Pyrite crystals (black) and quartz prisms (white and shades of gray) line vug and are coated with chalcedony (fine grained, mottled gray). The core of the vug is filled with granular quartz (white, black, and shades of gray). stromeyerite with chalcocite; and cerussite, anglesite, brochantite, limonite, jarosite-plumbojarosite, autunite, tungstite, ferritungstite, jixianite(?), and stolzite with stibiconite, chrysocolla, malachite, azurite, manganese oxide minerals, or jarosite. CHEMICAL COMPOSITION OF THE HUEBNERITE Chemical data were acquired to establish the iron-manganese compositions, Ihe compositional variations, if any, in visible growth zones, and the trace-element compositions of the huebnerites, for the purpose of better defining the environment of their deposition. Samples of huebnerite from 20 localities (shown on fig. 2) were analyzed using electron microprobe and emission spectrographic techniques. Synthetic huebnerite standards were made for use in the electron microprobe determinations. Stoichiometric proportions of Mn02, W02, and Fe metal were fused to a melt and allowed to crystallize. The resulting material was entirely crystalline, and only slight iron-manganese variation was found in the 95 mol percent MnW04 standard. Natural huebnerite was prepared for analysis by hand picking crushed material selected from the vein samples. Examination of cleavage fragments under the microscope revealed no other mineral phases nor fluid inclusions, although it was evident that oxidation and replacement by iron-manganese oxides along cleavage planes, fractures, and other irregular surfaces have occurred in various degrees. Conditions for microprobe analysis were 15 kilovolts operating voltage, 30 nanoamperes sample current, 10 micrometer beam diameter, and counting times of about 10 seconds on samples and standards using integrated beam current. Data were reduced by a computer program utilizing a linear least-squares best-fit line. No elements other than iron, manganese, and tungsten were detected above background. Results of the microprobe analyses of 19 samples are given in table 5. Polished thin sections of four huebnerite-bearing samples (DRS-73-17, DRS-74-221, DRS-79-67, and DRS-79-68) were prepared for microprobe study of growth zones. The polished thin sections of huebnerite were examined under the microscope, in both transmitted and reflected light, and under the SEM, and no other mineral phases that might define the growth zones were detected. Six-step semiquantitative emission spectrographic analyses of 23 samples of huebnerite are presented in table 6. Details of the six-step method, its precision, and comparison with other analytical methods were discussed by Myers, Havens, and Dunton (1961). In spite of the fact that the crushed huebnerite samples were carefully hand picked to remove other visible mineral phases before analysis, it is believed from the analytical data, as discussed later, that minor impurities were contained in some of the analyzed huebnerites. IRON-MANGANESE COMPOSITION The analytical data (tables 5 and 6) show that the mean iron contents of the huebnerites range from about 0.1 to 3.0 weight percent (recorded as Fe for spectrographic analyses and as FeO for microprobe analyses; the high value of 7.0 percent Fe (spectrographic) in sample DRS-74-221A is discounted, being attributed provisionally to pyrite). A comparison of the analytical data for iron (fig. 35) shows that iron values determined spectrographically are somewhat higher on average than those determined by microprobe. Variations in the iron content of the Round Mountain huebnerites suggest the possibility of compositional zoning in the vein system. Indeed, the iron content (deter-24 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 1 mm Figure 25.—Sphalerite that has been fractured, “milled,” and dispersed in recrystallized mosaic quartz, under crossed nicols, thin section of sample DRS-74-252B2. A, Strongly broken and dispersed sphalerite (black) in mosaic quartz (white and shades of gray). Broken sphalerite at left center has been filled with chalcedony (fine grained, mottled gray). Note vug at lower right filled with chal- mined by spectrographic analyses, which we believe to be more representative of the bulk compositions than are the microprobe analyses) appears to be a sensitive indicator of altitude within the vein system, increasing from values of about 0.2 percent Fe at an altitude of 2,125 m (6,800 ft) to about 3.0 percent Fe at an altitude of 2,625 m (8,400 ft) (figs. 2 and 36). It should be noted, however, that although the altitude correlation of microprobe-determined FeO is similar to that of spec-trographically determined Fe (recalculated to FeO), sample DRS-74-221B is anomalous in its FeO (recalcu-lated)-FeO ratios (see fig. 35). As discussed later, the vertical compositional zoning of the vein huebnerites is thought to reflect change in the pressure-temperature-composition conditions upward through the vein system. A comparison of the FeO and MnO contents of the huebnerites as determined by microprobe analysis (fig. 37) reveals two fields of composition. A field of huebner- i______1 mm________| cedony (fine grained to fibrous, mottled gray). B, Fractured sphalerite (black) filled with chalcedony (fine grained, mottled gray). Note that quartz (white and shades of gray) penetrates fractures a short distance, suggesting that quartz recrystallized and molded itself against sphalerite as the fractures in sphalerite opened. Chalcedony deposition likely was the latest event. ites with highest MnO to FeO ratios includes only huebnerites with which sulfides are associated in the veins. A field of huebnerites with lowest MnO to FeO ratios includes only huebnerites with which no sulfides are associated in the veins. In this comparison, sample DRS-74-221B does not appear anomalous. Assuming that the huebnerites and sulfides were deposited simultaneously in the vein system, we could explain the low content of iron in the sulfide-associated huebnerites by its preferential deposition in pyrite. Assuming that the sulfides were deposited in the veins later than the huebnerites, as the geological and mineralogical evidence indicates, we could explain the low content of iron in the sulfide-associated huebnerites by movement of iron out of huebnerite to be deposited as pyrite at the time of sulfide mineralization. We note the “reworked” character of the huebnerites that developed following their initial deposition but that predated (or accompanied?) deposition of sulfides.DESCRIPTION OF THE VEIN MINERALS 25 Figure 26.—Galena partly replaced by covellite, reflected light, polished sections of sample DRS-74-221. A, Galena (white) partly replaced by covellite (shades of gray) in irregular patches and along crystallographic planes. B, Galena (white) intergrown with or partly replaced by covellite (gray). Dark-gray areas near top and at right are quartz (high-relief fragments) and mounting medium. TUNGSTEN-MANGANESE COMPOSITION The variations in tungsten contents of 75.9 to 78.9 percent W03, as determined by microprobe analysis (table 5), compared to an ideal composition for huebner-ite of 76.6 percent W03, led to the calculation of the molecular proportions of the components based on the microprobe data (structural formulas given in table 5). The results indicate compositions near stoichiometry and suggest that the compositional variations resulted largely from an instrumental factor, most likely variation in sample current, and are not real variations in tungsten composition. We also plotted the MnO versus W03 compositions of the analyzed huebnerites (fig. 38) to evaluate further the compositional variations. Here again two distinct fields are shown, one of huebnerites with associated sulfides in the veins and one of huebnerites without associated sulfides in the veins. An exception is sample DRS-73-17 (queried on fig. 38), a huebnerite without associated sulfides that falls in the field of huebnerites with associated sulfides. The tendency for positive correlation of MnO and W03 within the separate fields suggests instrumental variation, probably in sample current during the time the analyses were made. GROWTH ZONES Because of the extremely low total iron contents, microprobe analyses of visible growth zones in huebnerite were inconclusive. Sample DRS-79-68 was examined by Dr. Charles M. Taylor (oral commun., 1979) and darker colored zones were found to contain very slightly more iron than lighter colored zones (0.2 versus 0.15 weight percent Fe). However, our studies failed to show convincing variations in iron content between darker and lighter parts of the huebnerite crystals, possibly because of the very thin character of the dark growth layers compared to the microprobe beam diameter employed. Nevertheless, local variation in the iron contents of the huebnerites (table 5) is evident despite our inability to correlate such with color variation.26 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 0.1 mm Figure 27.—Sphalerite (medium gray) that contains exsolved blebs of chalcopyrite (light gray), reflected light, polished section of sample DRS-74-252B. Iron-manganese zonation within huebnerite single crystals has been described by a number of authors (for example, Takla, 1976; Bird and Gair, 1976). The growth layers described by Bird and Gair (1976) for Hamme, N.C., huebnerites are similar in appearance to those observed in Round Mountain huebnerites, except that in the Round Mountain examples, growth layers appear to be much thinner. Patchy and random iron-manganese distribution in wolframite was reported by Moore and Howie (1978). Clark (1970) pointed out that many wolframites contain small but significant amounts of ferric iron and that this feature has received little attention in most recent studies. In crystals where the total iron is only 1.0 weight percent or less, small variations in amounts of ferrous (Fe+2) and ferric (Fe+3) iron might produce color zonation (Fe+2 producing lighter colored zones and Fe+3 producing darker colored zones). Moore and Howie (1978) reported appreciable excess ferrous oxide in some analyzed Cornish wolframites. Study of the stability relations of the wolframite series by Hsu (1976) has shown that wolframite of any composition 0.25 mm Figure 28.—Tetrahedrite-tennantite (light gray to white) partly replaced by a narrow halo of chalcocite, acanthite, and stromeyerite (mottled light to medium gray) and by intervening patches of gray isotropic mineral (stibiconite?) and covellite (mottled medium gray), reflected light, polished section of sample DRS-74-252A. Dark-gray areas are quartz (except mounting medium at lower right). can remain stable under the oxidation states prevailing in hydrothermal environments. The stability of ferber-ite and huebnerite does not differ appreciably in the presence of both oxygen and sulfur. Apparently neither f02 nor /S2 exerts any noticeable influence on the composition of wolframite. Optical spectral studies (Graham R. Hunt, oral commun., 1980) were not able to distinguish Fe+2 from Fe+3 in several Round Mountain heubnerites (DRS-73-17, DRS-74-66, DRS-79-68) because of very low total iron content. Moore and Howie (1979) in a study of cassiterite from Cornwall suggested that alternating color zones in that mineral resulted from other than chemical variation. They found that small variations in iron content did not correlate with color zones. Banerjee (1969) and Banerjee, Johnson, and Krs (1970) could not show conclusively by Mossbauer spectral studies of the cassiterite that color variation was related to differences in the Fe+2:Fe+3 ratio. We note, however, that the extremelyDESCRIPTION OF THE VEIN MINERALS 27 , 60 /xm | | 6 /im , Figure 29.—Manganese oxide weathering product of huebnerite, sample DRS-79-68. A, oxide (pyrolusite?) coating a substrate of huebnerite; B, closer view of the manganese oxide mineral. Scales are 60 and 6 p.m (micrometers), respectively. low concentration of iron in the cassiterite would make detection of variation in the Fe+2:Fe + 3 ratio difficult. TRACE-ELEMENT COMPOSITION In light of the fact that the Round Mountain huebner-ites were cleaned by hand picking, rather than by thorough mechanical and chemical means, the mineral residence of many of the minor elements detected is uncertain. Spectrographic analyses of huebnerite from Japan that was carefully purified by electromagnetic and heavy liquid separations and by acid treatment (Lee, 1955, p. 2, 49) revealed traces of titanium, magnesium, silicon, calcium, aluminum, zinc, copper, and silver. These trace elements may have been present in the huebnerite structure, in consideration of Lee’s careful and effective methods of removing mineral impurities from the material that was analyzed. The minor amounts of elements detected by spectrographic analyses of the Round Mountain huebnerites thus may be held, in part, in the huebnerite structure rather than in mineral impurities. The character of our data, however, suggests that much of the minor elements detected are present as “impurities,” that is, separate mineral phases; possibly some extraneous minor elements reside in fluid inclusions. Spectrographic data given in table 6 suggest the identity of some “impurities” present in the analyzed huebnerites. Four samples contained greater than 0.5 percent Si, probably as quartz. Unusually high amounts (0.03-0.07 percent) of calcium and magnesium were detected in some huebnerites, and these elements in part may be present in carbonate minerals; some calcium may be present in scheelite (see, for example, Grubb, 1967). Certain elements, such as lead, zinc, copper, bismuth, silver, molybdenum, and cadmium, occur in amounts perhaps sufficient to indicate the presence of sulfides and sulfosalts, or weathering (secondary) products derived from them. The huebnerites that presumably contain sulfides or sulfosalts are widely and apparently randomly scattered throughout the tungsten-mineralized area. However, those with the highest amounts of sulfide- or sulfosalt-related metals were collected within the zones of base- and precious-metal mineralization (for example, samples DRS-74-221 and DRS-78-2; compare table 6 and fig. 2), although some huebnerite samples collected within the base- and precious-metal mineralization zones (for example, samples DRS-79-67 and DRS-79-68) contain no unusually high amounts of metals that could be attributed to sulfides or sulfosalts. Nickel and cobalt were detected and chromium was found in amounts perhaps anomalously high in a few huebnerite samples. The mineral residence of these elements is uncertain. Barium is present in significant quantities (0.03-0.1 percent) in some samples, probably reflecting admixed barite or barium-manganese oxides. Small amounts of strontium are present in the samples that contain unusually high amounts of barium. Small amounts of the yttrium-group rare-earth elements along with lanthanum, cerium, neodymium (in one sample), beryllium, niobium, scandium, and titanium were found in a few huebnerites; the mineral residence of these elements also is uncertain. Because of interference between the spectra of tungsten and niobium, niobium was determined on an emission line with a sensitivity less than that of the line normally used. Judged from the total of elements present other than the major elements comprising huebnerite (see table 6), most of the huebnerite samples analyzed were reasonably free of “impurities.”28 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA 0.5 mm Figure 30.—Tetrahedrite-tennantite (light gray) partly replaced by stibiconite (medium gray) along irregular fractures, reflected light, polished section of sample DRS-78-2. Dark-gray grain in upper right is quartz. VARIATIONS IN TUNGSTEN MINERALIZATION One east-trending zone of tungsten mineralization that extends 2.5 km eastward from a point 3 km due east of Round Mountain (fig. 2) appears to be virtually devoid of well-defined huebnerite-bearing quartz veins. Instead, the zone contains local areas of tungsten-mineralized rock that are tourmalinized granite, iron-and quartz-mineralized fractures in granite, sheared and iron-mineralized argillite, or iron- and quartz-mineralized fractures in Oligocene rhyolite. The tungsten content of mineralized samples collected at these occurrences is low and ranges from 0.005 to 0.01 percent; identity of the tungsten mineral or minerals here is not known. Despite their low tungsten content, the occurrences contain substantially more tungsten than other nearby mineralized rocks. The tungsten-bearing deposits in this east-trending zone contain, in weight i_______0.5 mm_________| Figure 31.—Secondary minerals in vein quartz (white), plane-polarized light, thin section of sample DRS-78-2A. Irregular vein-lets are filled with growth layers of chrysocolla (medium gray). Vug bounded by sharp quartz prism faces (left center) is lined with layers of chrysocolla (medium gray), in part colloform, and filled with malachite (black). Small, sharp-cornered vug below (lower left) is lined with colloform chalcedony (nearly white) and filled with chrysocolla. Fluid inclusions (dark specks) form variously oriented trains in vein quartz. percent, as much as 0.02 Ag, 0.3 As, 0.007 Bi, 3.0 Cu, 0.002 Mo, 0.2 Pb, 0.5 Sb, and 1.0 Zn. Because some of these mineral occurrences are in Oligocene rhyolite or tourmaline-mineralized granite near the Oligocene stock east of Round Mountain, it is clear that there was a Tertiary tungsten-mineralizing event substantially younger than the emplacement of the huebnerite veins in Late Cretaceous time. In view of the small amount of tungsten involved in the younger (Tertiary) mineralization, the tungsten conceivably was remobilized from the earlier formed (Cretaceous) veins. The paragenetic data that show significant deposition of scheelite during a second hydrothermal stage, possibly derived by remobilization of huebnerite, indicate the availability of tungsten during hydrothermal activityGEOCHEMISTRY OF THE HUEBNERITE VEINS 29 0.5 mm Figure 32.—Vug in vein quartz (white with numerous trains of fluid inclusions—black specks) lined with chrysocolla (gray) and filled with chalcedony (white, faintly layered), plane-polarized light, polished thin section of sample DRS-78-2B. Alternating layers of chrysocolla and chalcedony form a colloform mass on quartz prism face (below center). Stibiconite (black, lower right) is a selvage on tetrahedrite-tennantite (outside field of view). that took place at some time following initial vein formation. GEOCHEMISTRY OF THE HUEBNERITE VEINS Semiquantitative spectrographic and other analyses of five samples of tungsten-bearing quartz vein material, given in tables 7 and 8, provide information on variations in mineralization of the veins that are indicated by the varied distributions of minerals described previously. The samples are part of a group of several hundred collected throughout the area of figure 2 for the purpose of determining patterns of metal distribution (Shawe, 1977b). Those samples were collected with an effort to obtain the most intensely mineralized material within a local area, and consisted of visibly 0.5 mm Figure 33.—Late veinlet of secondary minerals cutting quartz (white) and huebnerite (dark gray), plane-polarized light, thin section of sample DRS-78-2A. Principal veinlet, extending across view, is filled with chrysocolla (light gray) that within the huebnerite crystal is partly replaced or filled with chalcedony (white). Thinner veinlets of chrysocolla follow cracks in huebnerite and along huebnerite-quartz contact. A thin, late veinlet of chalcedony follows the right margin of the principal chrysocolla veinlet; in the huebnerite crystal the chalcedony veinlet locally merges with chalcedony that replaced or filled chrysocolla. Outside the field of view the chalcedony veinlet is seen to transgress the chrysocolla veinlet to its left margin, indicating that it postdates the chrysocolla veinlet. Tiny grains of iron oxide (limonite) or sulfate (jarosite)—dark specks—are present within the late chalcedony veinlet. mineralized rocks such as strongly iron-stained rocks, quartz and iron oxide-coated fracture surfaces, thin quartz and iron oxide veinlets, and quartz vein material containing sulfide and other ore minerals. Three of the samples listed in tables 7 and 8 (DRS-74-194, DRS-74-221, and DRS-74-252A) were collected within zones of base- and precious-metal mineralization as determined by the geochemical survey, and two samples (DRS-74-64 and DRS-74-200) were collected outside of those zones. Highest amounts of a number of metals, including mercury, gold, silver, arsenic, bismuth, copper, molyb-30 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA MINERAL INITIAL HYDROTHERMAL SHEARING AND YOUNGER HYDROTHERMAL WEATHERING STAGE RECRYSTALLIZATION STAGE (SECONDARY) STAGE Muscovite Quartz Huebnerite Fluorite Barite Calcite Scheelite Tetrahedrite- tennantite Pyrite Sphalerite Galena Covellite Chalcedony Chalcocite Stibiconite Chrysocolla Malachite Azurite Manganese oxide Jarosite Figure 34.—Paragenesis of huebnerite veing. Thicknesses of bars suggest relative abundance of minerals in each stage of mineralization. denum, lead, antimony, and zinc, were detected in the three samples collected within zones of base- and precious-metal mineralization. On the other hand, tungsten and a few elements such as manganese, iron, barium, fluorine, and niobium appear to be randomly concentrated through all five samples. Two elements, germanium and selenium, are more abundant in veins outside the zones of base- and precious-metal mineralization than within them. DISCUSSION Huebnerite has been reported from a large number of occurrences worldwide, but nearly end-member huebnerite, as in Japan (Lee, 1955, p. 49), Egypt (Takla, 1976; Takla and others, 1977), Mexico (Ram-dohr, 1966, p. 1068), and western Transbaikal (Shnuraeva, 1972), is uncommon. In the United States substantial amounts of huebnerite have been minedDISCUSSION 31 Table 5.—Electron microprobe analyses, in weight percent, of 19 samples of huebnerite from quartz-vein samples collected east and south of Round Mountain (localities shown on fig. 2) [Analyses by E. E. Foord; variation in total and some of the variation in component values resulted from variation in conditions of the microprobe analyses. Leaders (-), not applicable] Sample No. DRS- FeO Min FeO Max FeO MnO wo3 Total No. of analyses Structural formulas (basis of 4 oxygens) 67-30 1.6 1.23 1.97 21.8 77.6 101.0 6 (Mn.92Fe.07^0.99W1.00°4 67-33 1.9 1.38 2.41 21.7 77.8 101.4 4 (Mn.91Fe-08)0>99W1>0004 73-17 .5 .22 .94 22.9 76.6 100.0 5 (Mn. 98Fe.02>1.00W1.00°4 73-19 1.4 1.05 1.57 21.9 77.2 100.5 5 (Mn>93Fe>06)0>99WK0004 73-21 1.5 .84 2.07 21.8 77.0 100.3 5 (Mn.93Fe.06^0.99W1.00°4 73-27 1.0 .22 1.80 22.9 78.9 102.8 7 (Mn.95Fe.04)0.99w1.00°4 74-66 .4 .00 1.08 23.0 75.9 99.3 5 (Mn.99Fe.02)1.01W1>0004 74-70 .6 .04 .92 22.5 77.5 100.6 5 (Mn.96Fe.03)0.99w1.01°4 74-72 .4 .00 .87 22.7 78.4 101.5 4 (Mn.96Fe.02>0. 98w1.01°4 74-76 2.1 .84 3.57 21.1 76.3 99.5 9 (Mn>91Fe#09)uooWuoo04 74-105 2.9 .17 5.94 21.6 76.7 101.2 6 (Mn.92Fe.12)1.04w0.9904 74-193 .5 .00 .48 23.3 77.7 101.5 3 (Mn.98Fe.o2)i.ooW1.0004 74-200 .4 .32 .39 22.6 78.2 101.2 2 (Mn>95Fe>02)0>97W1>0104 74-218 .2 .19 .29 22.9 76.9 100.0 3 (Mn.98Fe.0p0.99W1.00°4 74-221 .3 .00 .43 22.9 76.7 99.9 6 (Mn.98Fe.01>0.99wl.00°4 78-1 .1 — — 23.0 77.2 100.3 1 (Mn.98Fe.004)0.98W1.00°4 78-2 .1 .10 .12 23.1 76.7 99.9 2 (Mn.99Fe.004^0.99W1.00°4 79-67 .3 .30 .35 23.0 76.9 100.2 2 (Mn.98Fe.01)0.99w1.00°4 79-68 .3 .24 .30 23.3 77.4 101.0 3 (Mn.98Fe.01>0.99^1.00°4 from veins at the Hamme (Tungsten Queen) Mine, N.C. (White, 1945; Bird and Gair, 1976), and small amounts have been mined from or recognized in veins at Butte, Mont., in the San Juan Mountains, in Boulder, Chaffee, Gunnison, Park, and Summit Counties, at Leadville, and at Cripple Creek, Colo. (Eckel, 1961), in Lincoln County, N. Mex., in Valley County (Leonard and others, 1968) and in Lemhi County, Idaho, , and near Round Mountain, and from pegmatites in the Black Hills, S. Dak. (see, for example, Palache and others, 1951, p. 1064, 1068-1070). Elsewhere in Nevada huebnerite occurs near Austin (first reported occurrence of the mineral, in 1865) and near Osceola (Palache and others, 1951, p. 1064, 1070) in quartz veins similar to those near Round Mountain. Huebnerite also was found in quartz gangue of epithermal veins of the Tonopah district, associated with sulfides and sulfosalts (Mel-hase, 1935). Despite the description of the Tonopah veins as epithermal, they were probably deposited at temperatures (350°C-200°C; Bonham and Garside, 1979, p. 107) similar to those of the Round Mountain huebnerite veins. Of the deposits in the United States, only those at the Hamme Mine and near Round Mountain are reported to contain end-member huebnerite (si weight percent FeO) (Bird and Gair, 1976; Shawe and others, this report). The huebnerite veins near Round Mountain are similar to those at the Hamme Mine in their mineralogy, rock associations, and conditions of formation. At the Hamme deposit quartz veins that contain huebnerite, fluorite, sulfides, and tetrahedrite occur in a north-northeast-trending linear belt nearly 6 km long, mostly32 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Table 6.—Semiquantitative spectrographic analyses, in weight percent, of 23 samples of huebnerite from quartz-vein samples collected east and south of Round Mountain (localities shoum on fig. 2) [Analyses by Nancy M. Conklin. N, not detected at limit of detection or at value shown; L, detected, but below limit of determination or below value shown; G, greater than 10 percent. Also looked for but not found: Na, K, P, As, B, Pd, Pt, Sn, Te, U, V, Zr, Ga, Ge, Hf, In, Li, Re, Ta, Th, Tl. (—), not looked for. Approximate lower limits (in percent) of determination for elements analyzed by six-step spectrographic method as shown in table 4, except for Ag (0.0002), Ba (0.0003), Be (0.0002), Cd (0.005), Co (0.0005), Cr (0.00015), Dy (0.003), La (0.005), Mo (0.0005), Nb (0.02), Ni (0.0002), Sr (0.001), Y (0.002), Yb (0.0005), Zn (0.03)] Sample No. DRS-67-30 DRS-67-33 DRS-73-17 DRS-73-19 DRS-73-21 DRS-73-27 DRS-73-168 DRS-74-66 DRS-74-70 DRS-74-72 DRS-74-76A DRS-74-76B Si 0.03 0.7 0.7 0.07 0.15 0.05 0.1 0.07 0.15 0.03 0.05 0.15 A1 N .015 N .015 N .015 N .015 N .015 N .015 N .015 N .015 N .015 N .015 N .015 N .015 Fe 1 2 .7 1. 5 1. 5 1.5 .7 .7 1 .7* 1.5 1.5 Mg .03 .015 .003 .007 .015 .007 .007 .015 .015 .015 .015 .03 Ca .015 .015 .03 .03 .07 .01 .03 .015 .03 .02 .015 — Ti .0015 .007 .007 .0015 .005 .0015 .003 .001 .0015 L .001 .003 .0015 Mn G G G G G G G G G G G G Ag N .002 N N N N N N N N .0007 .0007 Ba .015 .015 .05 .07 .1 .015 N .0003 .005 .007 .0015 .005 .0015 Be N .0003 N N N N N N N N N N Bi .007 .003 N N N N N N N N N N Cd N N N N N N N N N N N N Ce N .015 N N N N N N N N N N Co N N N N .002 N N N N N N N Cr N .0005 N .0005 N .0005 N .0005 N .0005 .0007 N .0003 .0015 .0003 .00015 .0005 N .0003 Cu .007 .015 .015 .015 .02 .001 .0015 .0015 .0007 .0002 .007 .007 Dy .01 .005 .005 .003 .005 .003 .007 N .003 N .003 N .003 N .003 .003 La N .007 N N N N N N N N N N Mo N .001 N N N N N N N N N N Nb .02 .02 .03 N .02 N .02 N .02 .03 .03 N .02 N .02 N .02 N .02 Nd N .007 N N N N N N N N N N Ni N N N N N N N N N N N N Pb .03 .07 .15 N .01 .15 N .01 N .01 .015 N .02 N Sb N N N N N N N N N N N N Sc .003 .0015 .0007 .0015 .003 .0015 .0015 .003 .0015 .002 .0015 .003 Sr N N .002 .002 .0015 N N N N N N N W G G G G G G G G G G G G Y .02 .007 .007 .007 .007 .003 .007 .003 .003 .003 .003 .007 Yb .01 .005 .005 .003 .007 .002 .007 .002 .003 .003 .003 .005 Zn N N N N N N N N N N N N in a granitic (granodioritic-tonalitic) pluton but extending a short distance into phyllite wall rocks. The huebnerite contains about 0.2-1.2 percent FeO and shows growth zoning similar to, although coarser than, that of the Round Mountain huebnerites. Microprobe analyses of the zoned huebnerites showed that dark-colored growth layers commonly contain 0.1-0.4 percent more FeO than do the light-colored growth zones. De-positional temperatures of the Hamme deposit probably were less than 350°C (summarized from Bird and Gair, 1976). The significance of variations in manganese and iron contents of wolframites, commonly stated as the hueb-nerite-ferberite molecular ratio (H:F), as an indication of conditions of deposition is controversial. Early work in western Europe and Africa by Oelsner (1944, 1952, 1954), Leutwein (1952), Bolduan (1954), Varlamoff (1958), and de Magnee and Aderca (1960) indicated that huebnerite has been deposited at higher temperatures than ferberite. More recently, in the Soviet Union and elsewhere, Churikov (1959), Ganeev and Sechina (1960), and Taylor and Hosking (1970) presented data to suggest that the opposite has been true. Clark’s (1970) data on Cornwall deposits indicated that wolframite in pegmatitic veins contained more iron than that in associated “normal” (presumably lower temperature) lodes. Nevertheless, Hollister (1970) pointed out that Bolivian tin districts and Peruvian and North American molybdenum porphyry districts are zoned outward from huebnerite-rich cores to ferberite-rich peripheral zones. Landis and Rye (1974) noted zonation outward from wolframite to ferberite in the Pasto Bueno tungsten-base metal district, Peru. Other studies in Tasmania, Europe, and the United States by Edwards and Lyon (1957), Baumann and Starke (1964), Bird and Gair (1976), Amosse (1978a), and Moore and Howie (1978), indicated more complex relations between H:F and environment. The studies of Groves and Baker (1972) in northern Tasmania showed that regional variations in H:F probably reflect regional compositional differences among ore fluids, these differences being much greater than variations in the ratio within local ore areas. This conclusion is supported by the experimental data of Hsu (1976). Amosse (1978b) has suggested, by thermodynamic considerations applied to wolframite from the Borralha Mine, Portugal, that H:F cannot be used as a geothermometer, but rather indicates direction toward the source of mineralization. Hsu (1976) alsoDISCUSSION 33 Table 6.—Continued Sample No. DRS-74-105 DRS-74-193 DRS-74-200 DRS-74-218 DRS-74-221A DRS-74-22IB DRS-78-1 DRS-78-2 DRS-79-67A DRS-79-67B DRS-79-68 Si 0.05 0.07 0.15 1 0.3 0.15 0.05 3 0.2 0.07 0.15 A1 N .015 N .015 N .015 N .015 N .015 N .0015 N .015 N .015 N .015 N .015 N .015 Fe 1.5 .3 1.5 .7 7 3 .15 .3 .3 .3 .2 Bg .007 .005 .005 .005 .0015 .002 .007 .003 .007 .002 .005 Ca .005 .015 .05 .005 .03 — .005 .07 .05 .01 .03 Ti L .001 .007 .002 .005 .003 .003 N .007 .03 .03 .007 Mn G G G G G G G G G G G Ag N N N N .005 .0007 N .0007 .005 N N N Ba .03 .0015 .05 .0007 .03 .007 .001 .015 .01 .007 .007 Be N N N N N N N N N N N Bi .0015 N N N L N N N N N N Cd N N N N N N N .015 N N N Ce N N N N N N N N N N N Co N N .0015 N N N N N N N N Cr .0007 .0002 .0002 .00015 .0002 N .0003 N .0005 N .0005 .0003 .0002 .0003 Cu .0015 .03 .02 .0007 .07 .03 .02 .3 .0007 .0015 .007 Dy N .003 N .003 .003 N .003 N .003 N .003 N .003 N .003 N .003 N .003 N .003 La N N N N N N N N N N N Mo N N N N N N N N N N N Nb N .02 .03 N .02 N .02 .02 .02 N .02 .05 .03 .03 .07 Nd N N N N N N N N N N N Ni N N .0003 .0002 .0003 N N N .0007 .0015 .002 Pb .003 .05 .007 .03 .2 .05 N .03 N N N .003 Sb N N N N .15 N N N N N N Sc .001 .003 .0005 .007 .007 .007 .005 .007 .003 .003 .0015 Sr .002 N .001 N .001 N N .003 .001 N .002 W G G G G G G G G G G G Y .003 .002 .003 N .002 N .002 N .002 .002 .002 N .002 N .002 N .002 Yb .003 .002 .003 .0007 .0007 .0007 .002 .002 .0007 .0007 .0005 Zn N N N N .03 .07 N N N N N concluded that H:F is of little value as a clue to temperature, pressure, and oxygen and sulfur fugacity during ore formation. Wiendl (1968, p. 265-279) showed that pH of mineralizing fluids generally is more important than temperature in determining the manganese-iron content of wolframite; a lateral increase in iron results from neutralization of ore fluids by wall rocks. Homer’s (1979) theoretical and experimental data indicated that huebnerite should form at temperatures above 200°C from neutral to weakly alkaline solutions, whereas ferb-erite should form through a wide temperature range but from acidic solutions. Studies by Vinogradova, Barabanov, and Sorokin (1980) suggested that ferruginous wolframites precipitated at low pH were richer in scandium and niobium than crystals precipitated from more alkaline (manganiferous) media. Niobium was not detected in less ferruginous wolframites formed in a less acid medium. Moore and Howie (1978) indicated that the uneven distribution of manganese and iron may be controlled not only by the concentration of these elements in ore fluids but by rapid changes in the composition of ore fluids, by competing mineral phases (for example, sulfides) if present, and by subsequent remobilization of manganese and iron. Voyevodin (1981) emphasized the influence of varied complex factors on composition of wolframites. Oscillatory zoning of iron and manganese in wolframite crystals may be controlled by local concentration gradients at the crystal-liquid interface, or by periodic supercooling of the liquid, or by “constitutional” supercooling as proposed for oscillatory zoning in plagioclase feldspars by Sibley and others (1976), or to the thermodynamic changes proposed by Amosse (1980). Major compositional changes in broad growth zones may be due to changes in external conditions. Lawrence (1961) attempted to correlate the habit of wolframite with temperature of crystallization. He concluded that stout prismatic crystals grew at higher temperatures, whereas platy crystals grew at lower temperatures; iron-composition differences (0.5 weight percent between the two types) were slight, however. Landis and Rye (1974) showed at Pasto Bueno, Peru, that early-stage wolframite (75 weight percent MnW04) formed stubby bladed crystals and was deposited at a temperature of about 235°C and late-stage ferberite (<5 weight percent MnW04) formed elongate crystals and was deposited at a temperature of about 180°C. We note in the huebnerite veins near Round Mountain a tendency for stubby crystals to form where wall rock is schist, which also is at the lowest altitude of vein exposure and where huebnerite contains only about 0.2 weight percent FeO. In granite and at higher altitudes, huebnerite may occur more commonly as tapered tabular crystals, and the mineral in this part of the vein system contains 0.7-3.0 weight percent FeO. We do34 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA FiGUEE 35.—Comparison of iron (recalculated as FeO) in huebnerite samples determined by six-step spectrographic analyses, and by microprobe analyses. Numbers are sample numbers. not have adequate data to indicate whether or not there was a temperature difference between deposition of the two forms of huebnerite. An assessment of the geologic setting of the huebnerite veins near Round Mountain permits an estimate of depth of burial of the veins at the time of their formation and an estimate of some of the parameters of the hydrothermal fluids that deposited the veins. A reconstruction of the domed granite, based on attitudes of granite contacts and of foliation within the granite, suggests that the apex of the pluton was about 1.5-2 km above the present erosional level. Perhaps another 1.5 km of Paleozoic sedimentary rocks lay above the dome at the time of its formation, interpreted to be the time of tungsten mineralization. Erosional stripping of some or all of the Paleozoic rocks during doming cannot be evaluated easily, and we discount it in this assessment. Thus the depth of formation of the present huebnerite veins was likely about 3-3.5 km beneath the surface. Fluid pressure at the time of vein formation is difficult to assess. Hydrostatic rather than lithostatic pressure may have prevailed, because of the evidence of some brittle deformation at the time of vein formation, and because locally much open space in the veins implies through-going solution flow. However, at times during the course of vein deposition, deformation took place under high confining pressure. Such deformation is evidenced by the recrystallization of prismatic quartzDISCUSSION 35 Figure 36.—Positive correlation between spectrographically determined iron composition of huebnerite samples and altitude in the vein system. crystals to form massive mosaic vein quartz that contains carbon dioxide-bearing fluid inclusions, and by the fact that along some vein walls partial replacement of wall rocks rather than open-space filling took place. Uncertainties as to the amount of carbon dioxide present in vein fluids, the temperature of the vein fluids at the level of and above the present huebnerite veins, and the height of the column of fluids in the vein system above the present veins also make an evaluation of fluid density uncertain. We use the assumptions that hydrostatic pressure prevailed and that the column of fluids extended upward to the surface. Like Landis and Rye (1974, p. 1043), we have arbitrarily selected an average density of 0.8 g/cm3 (80 bars/km) as possible and reasonable. Maximum fluid pressures at depths of 3-3.5 km thus would have been about 240-280 bars. If lithostatic pressure is assumed (granite density =2.65 g/cm3), maximum pressure would have been about 640-740 bars. We believe that clear quartz crystals that penetrate vugs, on which J. T. Nash (written commun., March 1980) determined filling temperatures of about 190°C and which contain no liquid carbon dioxide, were formed near the end of the initial episode of mineralization during which huebnerite was deposited. Temperatures at the beginning of this stage may have been much higher. Following this stage of vein formation, the veins were deformed, quartz was recrystallized except where quartz prisms penetrated vugs, and fluid inclusions, some bearing carbon dioxide, were formed in the recrystallized quartz. Nash determined filling temperatures of 250°-270°C for these fluid inclusions. Fluorite that fills vugs penetrated by clear quartz crystals contains fluid inclusions that have filling temperatures of 230°-260°C, according to Nash, and he indicated that some of the fluid inclusions with filling temperature of 260°C contain carbon dioxide dissolved in the aqueous phase. Younger fluorite contains fluid inclusions that have filling temperatures of about 190°C, and these inclusions contain no liquid carbon dioxide. These temperatures, corrected for estimated maximum pressures at about 5 weight percent equivalent NaCl (Lemmlein and Klevtsov, 1961, fig. 4), are about 210°C for early quartz and late fluorite, 270°-290°C for recrystallized quartz, and 250°-280°C for early fluorite, at pressures of about 240-280 bars. At pressures of about 640-740 bars, they are about 245°C for early quartz and late fluorite, 310°-330°C for recrystallized quartz, and 290°-320°C for early fluorite. A fluid under pressure of 240-280 bars (hydrostatic conditions) and with about 5 weight percent NaCl should boil at about 380°^480°C, and a fluid under pressure of 640-740 bars (lithostatic conditions, considered improbable) should boil at about 515°-545°C (Takenouchi and Kennedy, 1965, fig. 5). Perhaps the uneven distribution of carbon dioxide in fluid inclusions in recrystallized vein quartz suggests the possibility of boiling during the formation of these secondary inclusions. However, the pressure-salinity data that indicate boiling in the range 380°-480°C, well above the estimated temperature of formation (270°-290°C) of the carbon dioxide-bearing fluid inclusions in recrystallized quartz, argue against boiling. The variation in carbon dioxide contents of these secondary fluid inclusions suggests instead a variation with time in carbon dioxide contents of the depositing fluids through the period of quartz recrystallization. J. T. Nash’s estimate (written commun., March 1980) of 2-10 mol percent (5-20 weight percent) C02 in fluids that were trapped in inclusions in the recrystallized (milky) quartz seems reasonable. According to Takenouchi and Kennedy (1965, fig. 2), in the temperature range of 270°-290°C and pressure range of 240-280 bars, at which the recrystallized quartz formed, 6 weight percent NaCl solution could contain a maximum of 6-7 weight percent C02. At 210°C and the same pressures and salinity, the maximum carbon dioxide content36 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Figure 37.—Comparison of FeO and MnO in huebnerites analyzed by microprobe. Numbers are sample numbers. would be 5 weight percent; Nash detected no carbon dioxide in quartz and fluorite that were deposited at 210°C. The huebnerite veins near Round Mountain probably represent the lower part of a mineralized system, the bulk of which has been eroded away. Wolframite-bearing vein systems that are spatially associated with granitic rocks commonly appear to have maximum development—that is, show the greatest concentration of rich and large veins—at and near the crests of cupolas. Examples are Panasqueira, Portugal (Kelly and Rye, 1979), Borralha, Portugal (Amosse, 1978b); several localities in Thailand (Shawe, personal observations, 1974); Pasto Bueno, Peru (Landis and Rye, 1974); Las Guijas, Arizona (Sheikh, 1970); several localities in Tasmania (Groves and Baker, 1972); and Cligga Head, Cornwall, England (Moore and Jackson, 1977). If the Round Mountain vein system had reached as high as cupolas near the apex of the granitic pluton, it would have had a vertical extent of 1.5-2 km. However, G. P. Landis (written commun., 1980) suggested to us that a chemically integrated continuum of wolframite deposition possibly could not be maintained through such a vertical extent. The fact that nearly end-member huebnerite was deposited in the Round Mountain veins also suggests that they are near the base of a tungstenDISCUSSION 37 Figure 38.—Comparison of MnO and W03 in huebnerites analyzed by microprobe. Numbers are sample numbers; query indicates that the position of sample DRS-73-17 in the field of huebnerites with associated sulfides is anomalous. vein system. Once huebnerite started to precipitate, as hydrothermal fluids rose through fractures, solution compositions changed so that huebnerites deposited higher in the system were more iron rich. Thus, where end-member huebnerite is present in a vein system that shows systematic increase upward in the iron content of the huebnerites; we interpret that the end-member huebnerite indicates the lowest part of the vein system. Assuming that huebnerite compositions changed toward that of ferberite with increasing altitude in the vein system, the diagram of iron versus altitude of the Round Mountain veins (fig. 36) suggests that at a level38 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Table 7.—Semiquantitative spectrographic analyses, in weight percent, of five samples of tungsten-bearing quartz vein material (sample localities shown on fig. 2) [Analyses by Harriet G. Neiman. N, not detected at limit of detection. G, greater than 10 percent. Also looked for but not found: Au, B, Co, La, Ni, Pd, Pt, Sc, Sn, Te, U, Y, P, Ce, Ge, Hf, In, Li, Re, Ta, Th, Tl, Eu] Sample No. DRS-74-64 DRS-74-194 DRS-74-200 DRS-74-221 DRS-74-252A Si02 86 100 100 88 98 ai2o3 4.3 1.8 .65 <.5 <.5 Fe2°3 .88 .54 .47 .53 .66 CaO <.l .13 .25 .56 <.l k2o 1.4 .42 .23 .24 .20 Ti02 .10 <.05 <.05 <.05 <.05 P2°5 <1 <1 <1 <1 <1 MnO .66 <.05 .13 .44 <.05 S <.08 <.08 <.08 .42 .13 Cl <.2 <.2 <.2 <.2 <.2 F .08 .06 .07 .05 <.04 Hg .000023 .00015 .000009 .01 .002 Au <.000005 .000011 <.000005 .000117 <.000005 Ag <.002 .0094 <.002 .175 .0105 Sn <.00001 <.00001 <.00001 .00007 <.00001 Sb <.00001 .0142 .0011 .0432 .0168 Ge .0021 .00044 .00085 .00021 .00018 As <.00001 .0038 .00008 .0111 .0035 Se .00044 .00009 .00013 <.00001 .00001 1.5-2 km higher in the system—that is, the interpreted level of cupola formation at the apex of the granite plu-ton—the composition of the tungstate mineral would be about 18 percent Fe, or virtually end-member ferberite. The uncertainty of the slope of the curve as controlled by the data points of figure 36, as well as the uncertainty of slope above the segment of the vein system represented in the diagram, make that value of iron content rather uncertain. Nevertheless, the extrapolation suggests that the wolframite would contain substantially more iron than that present in the huebner-ites of the Round Mountain veins, and the composition of wolframite would be closer to compositions of wolframites in the cited cupola environments elsewhere in the world. Trace element components of wolframites have been reported by a number of authors (Lee, 1955; Churikov, 1959; Ganeev and Sechina, 1960; Zuyev and others, 1966; Mineev, 1968; Takla and others, 1977; Vinogradova and Barabanov, 1978; Moore and Howie, 1978; Vinogradova, Barabanov, and Sorokin, 1980). With the exception of Lee (1955, p. 5) and Ganeev and Sechina (1960) none of these workers presented details on methods or effectiveness of purification of the analyzed minerals. Takla and others (1977) and Zuyev and others (1966) reported inclusions of other minerals in the wolframites they analyzed for trace elements. We believe Table 8.—Analyses, in weight percent, for several components of five samples of tungsten-bearing quartz vein material (sample localities shown on fig. 2) [Si02, A1203, Fe203, CaO, K20, Ti02, P206, MnO, S, and Cl by X-ray fluorescence method by J. S. Wahlberg and J. W. Baker; F. by specific ion electrode method by Johnnie Gardner and Patricia Guest; Hg by wet oxidation plus atomic absorption method by J. A. Thomas; Au by fire assay plus atomic absorption method, and Ag by difference, by A. W. Haubert, L. B. Riley, and J. G. Crock; Sn, Sb, Ge, As, and Se by X-ray fluorescence by J. S. Wahlberg, J. 0. Johnson, and J. W. Baker] Sample No. DRS-74-64 DRS-74-194 DRS-74-200 DRS-74-221 DRS-74-252A Si G G G G G A1 1.5 0.3 0.2 0.2 0.15 Fe .2 . 1 .07 .1 .15 Mg .03 .02 .015 .015 .005 Ca .05 .03 .2 .3 .002 Na .5 N N N N K 1 N N N N Ti .03 .005 .001 .001 .0005 Mn .5 .05 .15 .2 .0015 Ag .0002. .005 N .1 .01 As N N N .07 N Ba .15 .01 .01 .02 .005 Be .00015 N N N N Bi .0007 .002 N .003 .002 Cd N N N .003 N Cr N N N .0001 N Cu .0007 .07 .0015 .15 .15 Ga .0005 N N N N Mo N .0015 N .0007 .003 Nb .003 .001 .002 .005 N Pb .005 .007 N .3 .03 Sb N .07 N 1 . 1 Sr .005 N N .0015 N V .0005 N N NN W 1 .07 .1 1 .007 Yb .0001 N N N N Zn N .05 N .02 .1 Zr .003 N N N N that the presence of small amounts of other minerals enclosed in the analyzed wolframites was likely in many of the reported studies of trace elements in wolframites, and in some cases therefore speculations on the residence and absolute amounts of trace elements in the wolframite structure are not valid. The association with wolframite of minor amounts of minerals of certain compositions, however, as suggested by the analytical data, undoubtedly has geochemical significance. Noting, then, the uncertainty of residence of many of the trace elements associated with the Round Mountain huebner-ites, we can point out some perhaps meaningful associations. Yttrium-group rare-earth elements dominate over cerium-group rare-earth elements, a relation also described for wolframites from the Soviet Union (Mineev, 1968; Vinogradova and Barabanov, 1978). We detected yttrium and ytterbium values as high as 0.007 percent each in Round Mountain huebnerites, whereas Ganeev and Sechina (1960) reported as much as 0.08 percent Y and Vinogradova and Barabanov (1978) reported as much as 0.036 percent Y in wolframites of the Soviet Union. Scandium was found in Round Mountain huebnerites in amounts as high as 0.07 percent,SUMMARY AND CONCLUSIONS 39 whereas Ganeev and Sechina (1960) found as much as 0.032 percent Sc and Vinogradova and Barabanov (1978) found as much as 0.11 percent Sc in wolframites from the Soviet Union. A microprobe study of five zoned wolframite crystals from vein-type deposits (Vinogradova, Barabanov, and Sorokin, 1980) showed a range of 0.17 to 1.04 percent Sc203. Beddoe-Stephens and Fortey (1981) detected as much as 0.03 percent Sc203 in wolframite from the Carrock Fell tungsten deposit in the English Lake district. Ganeev and Sechina (I960) suggested close correlations between scandium and iron and between yttrium and manganese in the Soviet Union wolframites, but we see no such correlations in the Round Mountain huebnerites, possibly owing to their narrow iron-compositional range. Niobium was undetected above 0.02 percent in most Round Mountain huebnerites, but it ranges as high as 0.07 percent in those in which it was found. Takla and others (1977) reported 0-0.05 percent Nb in Egyptian huebnerites, Ganeev and Sechina (1960) reported 0-0.1 percent Nb in Soviet Union wolframites, and Moore and Howie (1978) reported as much as 0.5 percent Nb in Cornwall wolframites. Vinogradova, Barabanov, and Sorokin (1980) reported as much as 0.56 weight percent Nb205 in zoned wolframites from the U.S.S.R. Zuyev and others (1966) showed by means of selective chemical extraction and electron microprobe studies that practically 100 percent of the tantalum and niobium in epithermal wolframite from quartz veins is due to the isomorphous entry of these elements into the mineral. Examination of hypothermal wolframite (from albitized and greisenized granite) showed that only part of the tantalum and niobium is an isomorphous admixture, the remainder occurring in inclusions of columbite and microlite. An average of 0.19 percent Ta205 was found to enter isomorphously into the wolframite structure. Beddoe-Stephens and Fortey (1981) detected (by electron microprobe) as much as 0.98 percent Nb205 in wolframite from quartz veins in the Carrock Fell tungsten deposit in the English Lake district; minute amounts of columbite are associated with the wolframite. Where other minerals have replaced the Carrock Fell wolframite, no concentration of niobium at mineral contacts was detected, suggesting that the niobium content of the wolframite was primary. Moreover, microprobe analyses showed that niobium enrichment occurred in zones in wolframite close to primary crystal faces, suggesting original growth zoning. A niobium-rich wolframite (20.25 weight percent Nb205) was reported by Saari, von Knorring, and Sahama (1968) from the Nuaparra pegmatite, Zambezia, Mozambique. The wolframite also contained 5.35 percent Ta205, 2.68 percent Ti02, 1.52 percent Sn02 plus Sb203, and significantly, 8.43 percent Fe203 compared to 7.43 percent FeO. We have interpreted that much of the calcium detected in analyses of Round Mountain huebnerites is actually resident in scheelite or calcite impurities. Phase equilibria studies by Chang (1967) have shown only a limited solid solution between CaW04 and MnW04 and even less between CaW04 and FeW04. Zinc was detected in huebnerite from only one locality (DRS-74-221), where the huebnerite-bearing quartz vein also contains appreciable sphalerite. Although a complete series of solid solutions forms in the system ZnW04-MnW04 above 840°C (Chang, 1968), it seems unlikely that a phase of this system is present in the Round Mountain material. The magmatic-metamorphic episode that was accompanied by tungsten mineralization near Round Mountain about 80 m.y. ago may have been widespread in the region. For example, we interpret that the Jurassic Snake Creek pluton in the southern Snake Range in eastern Nevada was metamorphosed and invaded by aplites and pegmatites and mineralized by tungstenbearing quartz veins at various times following emplacement; one such episode took place about 71-78 m.y. ago (data from Lee and Van Loenen, 1971; Lee and others, 1970). At Tungsten, site of an important scheelite-producing district in northwestern Nevada, granitic rocks were metamorphosed about 74-79 m.y. ago and tungsten mineralization took place about 72-76 m.y. ago (Tingley, 1975). Near the site of large production of scheelite ores at Pine Creek, Calif., about 200 km southwest of Round Mountain, plutonic rocks yielded K-Ar (biotite) ages of 70-80 m.y. in contrast to surrounding plutonic rocks that yielded K-Ar (hornblende) ages of 80-100 m.y. (Evernden and Kist-ler, 1970). Possibly these relations indicate metamorphism of the plutonic rocks in the vicinity of Pine Creek at the time of tungsten mineralization. At the Old Woman Mountains in southeastern California, tungsten mineralization took place in granitic rocks that yield a Rb-Sr (whole rock) age of 80 m.y. (Miller, 1977). SUMMARY AND CONCLUSIONS Huebnerite-bearing quartz veins were deposited about 80 m.y. ago near Round Mountain, mostly in Cretaceous granite (emplaced about 95 m.y. ago), when the granite was invaded by pegmatite and aplite dikes, was domed, and was metamorphosed. Depth of burial at the time of vein formation was probably about 3-3.5 km, based on geologic reconstruction. Through the initial stage of vein deposition, pressures were probably about 240-280 bars or only slightly higher, and salinity of the fluids persisted at about 5 weight percent NaCl. Muscovite, quartz, and huebnerite were the principal minerals of early-stage deposition in the veins. Quartz crystals40 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA deposited near the end of this initial phase were precipitated from solutions at a temperature of about 210°C and contained little if any carbon dioxide. Following the precipitation of quartz, fluorite was deposited at temperatures of about 250°-280°C, from solutions that contained small amounts of carbon dioxide (about 5 weight percent?). Also, after initial precipitation of vein quartz, the veins were deformed, and the quartz was recrystallized and permeated by solutions at a temperature of about 270°-290°C that contained carbon dioxide (5-20 weight percent?). Probably following this deformation, additional fluorite was deposited at a temperature of about 210°C from solutions that contained little or no carbon dioxide. Foose, Slack, and Casadevall (1980) documented post-mineralization deformation at the Tungsten Queen (Hamme) deposit, North Carolina. Casadevall and Rye (1980) presented evidence that metamorphic temperatures exceeded earlier primary-mineralization temperatures at the Tungsten Queen deposit. During the initial stage of vein deposition at Round Mountain, some pyrite was deposited along with the other vein minerals. Possibly other sulfides and sul-fosalts were deposited in the veins or in altered rocks adjacent to the veins at that time, but the amounts were likely quite small. After the initial hydrothermal stage of vein deposition, the veins were sheared and reopened, possibly at the time of widespread recrystallization of vein quartz, but more likely at a later time. A second stage of hydro-thermal mineralization then took place, involving minor deposition of muscovite, quartz, and fluorite, but characterized mainly by deposition of sulfides, tetrahed-rite-tennantite, scheelite (altered from initial-stage huebnerite), and near the end of the stage, barite and chalcedony. Although we have no direct evidence of the age of this mineralization, its dominant spatial association with 35-m.y.-old intrusive rocks (southwest end of rhyolite dike swarm and vicinity of granodiorite stock) leads us to believe that the younger hydrothermal stage of mineralization was related to a geologic episode that saw emplacement of those igneous rocks. Moreover, sulfide and minor tungsten mineralization took place in 35-m.y.-old rocks in the vicinity of the granodiorite stock. The mineralization of Tertiary rocks and the presence of chalcedony in late phases of the younger hydrothermal stage suggest that the younger hydro-thermal stage took place at a shallower level at lower temperatures and pressures than did the initial hydro-thermal stage. We note that Leonard, Mead, and Conklin (1968, p. C19-C20) interpreted the New Snowbird, Idaho, tungsten-bearing quartz lode, somewhat similar to the Round Mountain huebnerite veins, to be the product of a single (although complex) episode of mineralization, even though they admitted that it might consist of two principal stages: early higher temperature tungsten mineralization and late lower temperature sulfide mineralization, as Schneider-Scherbina (1962) had interpreted for so-called telescoped mineralization of the Bolivian tin-silver deposits. Following uplift and erosion of the Round Mountain huebnerite vein system in late Tertiary and Quaternary times the veins were weathered, with the development of numerous secondary minerals derived from the breakdown of primary vein minerals. Possibly the huebnerite veins near Round Mountain represent the lower part of a vein system that when formed extended to a much higher level and resulted in mineralization of cupolas at the apex of the granite pluton. By analogy with wolframite deposits known elsewhere in the world that are associated with granite cupolas, the richest part of the Round Mountain vein system may have formed at that level, and thus was destroyed by erosion long ago. If the zonation of the iron content of the huebnerites in the Round Mountain vein system, increasing with altitude, is valid, we are led to believe that any such eroded deposits in cupolas must have contained wolframites with iron compositions much higher than those in the Round Mountain veins. Cupolas of granite plutons in the Basin and Range province in Nevada and adjacent States that have had deformation and metamorphic histories similar to the granite near Round Mountain, and that have not been unroofed by erosion, seem to be likely sites for undiscovered wolframite vein deposits, perhaps of large size. Such postulated deposits, as well as known ones, may have formed dominantly during a major regional event of tungsten mineralization about 80 m.y. ago. REFERENCES CITED Amosse, Jean, 1978a, Variation in wolframite composition according to temperature, at Borralha, Portugal, and Enguyales, France: Economic Geology, v. 73, no. 6, p. 1170-1175. ------1978b, Physicochemical study of the huebnerite-ferberite (MnW04-FeW04) zonal distribution in wolframite (MnxFe(1_X)-W04) deposits; Application to the Borralha Mine (Portugal): Physics and Chemistry of Minerals, v. 3, p. 331-341. ------1980, Physicochemical interpretation of the mechanism of the oscillatory zoning in wolframite crystals: International Mineralog-ical Association, Collected abstracts, 12th General Meeting, Orleans, France (July 4-6, 1980), p. 29. Banerjee, K. S., 1969, Origin of weak ferromagnetism and remanence in natural cassiterite crystals: Journal of Geophysical Research, v. 74, p. 3789-3795. Banerjee, K. S., Johnson, C. E., and Krs, M., 1970, Mossbauer study to find the origin of weak ferromagnetism in cassiterite: Nature, v. 225, p. 173-175.REFERENCES CITED 41 Baumann, L., and Starke, R., 1964, Beitrag zur verteilung der H/F-koeffizienten innerhalb der wolframitlagerstatte Pechtelsgrun auf grund neuer rontgenographischer untersuehungen: Berg- akademie 16, p. 79-82. Beddoe-Stephens, B., and Fortey, N. J., 1981, Columbite from the Carrock Fell tungsten deposit: Mineralogical Magazine, v. 44, p. 217-223. Bird, M. L., and Gair, J. E., 1976, Compositional variations in wolframite from the Hamme (Tungsten Queen) Mine, North Carolina: U.S. Geological Survey Journal of Research, v. 4, no. 5, p. 583-588. Bolduan, Helmut, 1954, Genetische untersuchung der wolframitlagerstatte Pechtelgriin (Vogtland) und besonderer be-rucksichtigung der verteilung des H/F-koeffizienten und der spurenelemente niob und tantal in wolframit: Freiberger Forschungshefte, ser. C, no. 10, p. 46-61. Bonham, H. F., Jr., and Garside, L. J., 1979, Geology of the To-nopah, Lone Mountain, Klondike, and northern Mud Lake quadrangles, Nevada: Nevada Bureau of Mines and Geology Bulletin 92, 142 p. Brown, G. C., 1911, Round Mountain tungsten mine: The Salt Lake Mining Review, August 15, 1911, p. 16. Casadevall, Tom, and Rye, R. 0., 1980, The Tungsten Queen deposit, Hamme district, Vance County, North Carolina—A stable isotope study of a metamorphosed quartz-huebnerite vein: Economic Geology, v. 75, no. 4, p. 523-537. Chang, L. L. Y., 1967, Solid solutions of scheelite with other RnW04-type tungstates: American Mineralogist, v. 52, nos. 3-4, p. 427-435. ------1968, Subsolidus phase relations in the system ZnW04- ZnMo04-MnW04-MnMo04: Mineralogical Magazine, v. 36, p. 992-9%. Churikov, V. S., 1959, Certain features of the chemical composition of wolframites—Materials relating to the geology, petrography, mineralogy and geochemistry of ore deposits: Academy of Sciences of the U.S.S.R., p. 235-250. [In Russian.] Clark, A. H., 1970, Manganese-iron ratios in wolframite, South Crofty Mine, Cornwall—A discussion: Economic Geology, v. 65, p. 889-892. de Magnee, I., and Aderca, B., 1%0, Contribution a la connaissance du tungsten belt Ruandais (aspects geologiques, geochemique, metallogenique): Academie Royale Sciences Outre-Mer, c. 1, Sci. Nat. Mem., In. 8°, n.s.t., 11, f. 7, 56 p. Eckel, E. B., 1961, Minerals of Colorado, a 100-year record: U.S. Geological Survey Bulletin 1114, 399 p. Edwards, A. B., and Lyon, P. J., 1957, Mineralization at Aberfoyle tin mine, Rossarden, Tasmania: Proceedings of the Australian Institute of Mining and Metallurgy, v. 181, p. 93-145. Evemden, J. F., and Kistler, R. W., 1970, Chronology of emplacement of Mesozoic batholithic complexes in California and western Nevada: U.S. Geological Survey Professional Paper 623, 42 p. Ferguson, H. G., 1921, The Round Mountain district, Nevada: U.S. Geological Survey Bulletin 725-1, p. 383-406. Foose, M. P., Slack, J. F., and Casadevall, Tom, 1980, Textural and structural evidence for predeformation hydrothermal origin of the Tungsten Queen deposit, Hamme district, North Carolina: Economic Geology, v. 75, no. 4, p. 515-522. Ganeev, I. G., and Sechina, N. P., 1%0, The geochemical peculiarities of wolframites: Geochemistry, v. 6, p. 617-623. Glass, J. J., 1935, The pegmatite minerals from near Amelia, Virginia: American Mineralogist, v. 20, p. 754-756. Groves, D. I., and Baker, W. E., 1972, The regional variation in compositions of wolframites from Tasmania: Economic Geology, v. 67, p. 362-368. Grubb, P. L. C., 1967, Solid solution relationships between wolframite and scheelite: American Mineralogist, v. 52, p. 418-426. Hogarth, D. D., and Chao, G. Y., 1979, Jixianite, in New mineral names: American Mineralogist, v. 64, p. 1330. Hollister, V. F., 1970, Manganese-iron ratios in wolframite, South Crofty Mine, Cornwall—a discussion: Economic Geology, v. 65, no. 5, p. 592. Homer, C., 1979, Solubility and hydrolysis of FeW04 and MnW04 in the 25°-300°C range, and the zonation of wolframite: Chemical Geology, v. 27, p. 85-97. Hsu, L. C., 1976, The stability relations of the wolframite series: American Mineralogist, v. 61, p. 944-955. Jianchang, Liu, 1979, Jixianite, Pb(W,Fe+3)2(0,0H)7—a new tungsten mineral: Acta Geologica Sinica, v. 53, p. 46-49. Kelly, W. C., and Rye, R. O., 1979, Geologic, fluid inclusion, and stable isotope studies of the tin-tungsten deposits of Panasqueira, Portugal: Economic Geology, v. 74, no. 8, p. 1721-1822. Koschmann, A. H., and Bergendahl, M. H., 1968, Principal gold-producing districts of the United States: U.S. Geological Survey Professional Paper 610, 283 p. Krai, V. E., 1951, Mineral resources of Nye County, Nevada: Nevada University Bulletin, v. 45, no. 3, Geology and Mining Series 50, 223 p. Landis, G. P., and Rye, R. O., 1974, Geologic, fluid inclusion, and stable isotope studies of the Pasto Bueno tungsten-base metal ore deposit, northern Peru: Economic Geology, v. 69, no. 7, p. 1025-1059. Lawrence, L. J., 1961, Crystal habit of wolframite as an indication of relative temperature of formation: Neues Jahrbuch fur Mineralogie. Monatshefte, p. 241-247. Lee, D. E., 1955, Mineralogy of some Japanese manganese ores: Stanford University Publications, University Series, Geological Sciences, v. 5, 64 p. Lee, D. E., and Van Loenen, R. E., 1971, Hybrid granitoid rocks of the southern Snake Range, Nevada: U.S. Geological Survey Professional Paper 668, 48 p. Lee, D. E., Marvin, R. F., Stem, T. W., and Peterman, Z. E., 1970, Modification of potassium-argon ages by Tertiary thrusting in the Snake Range, White Pine County, Nevada, in Geological Survey research 1970: U.S. Geological Survey Professional Paper 700-D, p. D92-D102. Lemmlein, G. G., and Klevtsov, P. V., 1961, Relations among the principal thermodynamic parameters in a part of the system H20-NaCl: Geokhimiya, 1961, no. 2, p. 133-142 (in Russian); translated in Geochemistry, no. 2, p. 148-158, 1961. Leonard, B. F., Mead, C. W., and Conklin, Nancy, 1968, Silver-rich disseminated sulfides from a tungsten-bearing quartz lode, Big Creek district, central Idaho: U.S. Geological Survey Professional Paper 594-C, p. C1-C24. Leutwein, F., 1952, Die chemische zusammensetzung des wolframit und ihre lagerstattenkundliche bedeutung: Acta Geologica Akademiae Scientiarum Hungaricae, v. 1, p. 133-141. Machin, M. P., and Siisse, P., 1975, On the crystal structure of fer-ritungstite, (W,Fe+3)204(0H)2-V2H20: Fortschritte der Mineralogie, v. 53, p. 50 [abs.]. Marvin, R. F., and Dobson, S. W., 1979, Radiometric ages— Compilation B, U.S. Geological Survey: Isochron/West, no. 26, p. 3-32. Melhase, John, 1935, Nevada, the mineral collector’s Mecca: Mineralogist, v. 3, no. 4, p. 9-10, 37-38. Miller, C. F., 1977, Muscovite granite with high initial Sr^/Sr86, Old Woman Mountains, California; a product of upper crustal anatexis?: Geological Society of America Abstracts with Programs, v. 9, no. 4, p. 466.42 HUEBNERITE VEINS NEAR ROUND MOUNTAIN, NYE COUNTY, NEVADA Mineev, D. A., 1968, Average composition of lanthanoids in minerals: Geochemistry, no. 7, p. 825-835. [In Russian, English abs.] Moore, F., and Howie, R. A., 1978, On the application of the huebner-ite:ferberite ratio as a geothermometer: Mineralium Deposita, v. 13, p. 391-397. ------1979, Geochemistry of some Cornubian cassiterites: Mineralium Deposita, v. 14, p. 103-107. Moore, J. M., and Jackson, N., 1977, Structure and mineralization in the Cligga granite stock, Cornwall: Journal Geological Society of London, v. 133, p. 467-480. Myers, A. T., Havens, R. G., and Dunton, P. J., 1961, A spec-trochemical method for the semiquantitative analysis of rocks and ores: U.S. Geological Survey Bulletin 1081-1, p. 207-229. Oelsner, Oscar, 1944, Uber Erzgebirgische wolframit: Freiberger Geologischen Gesellschaft Bergbau, v. 20, p. 44-49. ------1952, Die pegmatitisch-pneumatolytischen lagerstatten des Erzgebirges mit ausnahme der kontakt-lagerstatten: Freiberger Forschungshefte, ser. C, no. 4, 80 p. ------1954, Bemerkungen liber die anwendbarkeit des H/F-koef- fizienten zur deutung der genese von wolframiten: Freiberger Forschungshefte, ser. C, no. 10, p. 62-67. Palache, Charles, Berman, Harry, and Frondel, Clifford, 1951, The system of mineralogy, 7th ed., vol. II: New York, John Wiley, 1124 p. Ramdohr, Paul, 1966, The ore minerals and their intergrowths, English translation of 3d ed.: Oxford, Pergamon Press. Saari, E., von Knorring, O., and Sahama, Th. G., 1968, Niobian wolframite from the Nuaparra pegmatite, Zambezia, Mozambique: Lithos, v. 1, no. 2, p. 164-168. Schilling, J. H., 1963, Tungsten mines in Nevada: Nevada Bureau of Mines Map 18. Schneider-Scherbina, Alexander, 1962, Uber metallogenetische epochen Boliviens und den hybriden charakter der sogenannten Zinn-Silber-Formation: Geologische Jahrbuch, v. 81, p. 157-170 [1964], Shawe, D. R., 1977a, Preliminary generalized geologic map of the Round Mountain quadrangle, Nye County, Nevada (with text): U.S. Geological Survey Miscellaneous Field Studies Map MF-833. ------1977b, Geochemical and generalized geologic maps showing distribution of iron, copper, lead, zinc, molybdenum, silver, antimony, arsenic, tungsten, barium, potassium, and boron in the Round Mountain quadrangle, Nye County, Nevada (12 separate maps): U.S. Geological Survey Miscellaneous Field Studies Maps MF-835A-I. Sheikh, A. M., 1970, Geology and ore deposits of Las Guijas tungsten district, Pima County, Arizona: Economic Geology, v. 65, no. 7, p. 875-882. Shnuraeva, L. Ya., 1972, The typomorphism of heubnerite of the Bom-Gorkonsk tungsten deposit (western Transbaikal): Zapiski Vsesoyuznevo Mineralogichestvo Obshchestvo 101, p. 118-122. [In Russian.] Sibley, D. F., Vogel, T. A., Walker, B. M., and Byerly, G., 1976, The origin of oscillatory zoning in plagioclase—A diffusion and growth controlled model: American Journal of Science, v. 276, p. 275-284. Takenouchi, Sukune, and Kennedy, G. C., 1965, The solubility of carbon dioxide in NaCl solutions at high temperatures and pressures: American Journal of Science, v. 263, p. 445-454. Takla, M. A., 1976, Electron microprobe of zoned wolframite from Elba, Egypt: Neues Jahrbuch fur Mineralogie Monatshefte, v. 11, p. 477-483. Takla, M. A., Sharkawi, M. A., and Fawzi, E., 1977, The geochemistry of Egyptian wolframites: Chemie der Erde, v. 36, p. 10-19. Taylor, R. G., and Hosking, K. F. G., 1970, Manganese-iron ratios in wolframite, South Crofty Mine, Cornwall: Economic Geology, v. 65, p. 47-53. Tingley, J. V., 1975, K-Ar dates on granodiorite and related scheelite-bearing quartz veins at Tungsten, Pershing County, Nevada: Isochron/West, no. 12, p. 3-4. Varlamoff, N., 1958, Les gisements de tungestene au Congo beige au Ruanda-Urundi, materiaux pour l’etude de leur geologie et de leur classification: Academie Royale Sciences, Colon. Brussels, nov. ser., 8, fasc. 2. Vinogradova, L. G., and Barabanov, V. F., 1978, On the question about the chemistry of wolframite: Zapiski Vsesoyuznevo Mineralogichestvo Obshchestvo 107, no. 5, p. 585-590. [In Russian.] Vinogradova, L. G., Barabanov, V. F., and Sorokin, N. D., 1980, On the distribution of iron, manganese, niobium, and scandium in wolframites: Zapiski Vesesoyuznovo Mineralogischestvo Obshchestvo, v. 109, p. 351-358. [In Russian.] Voyevodin, V. N., 1981, The relation between chemical composition of wolframites and the geological conditions of their formation: International Geology Review, v. 23, no. 5, p. 561-570. White, W. A., 1945, Tungsten deposit near Townsville, North Carolina: American Mineralogist, v. 30, p. 97-110. Wiendl, Ulrich, 1968, Zur geochemie und lagerstatten kunde des wolframs: Technischen Universitat Clausthal, Disertation zur erlan-gung des grades eines Doktor-Ingenieurs, 295 p. (Genehmigt von der Fakultat fur Bergbau. Hiittenwesen und Maschinenwesen). Zuyuv, V. N., Zubkov, L. B., Zubynina, K. B., Utkina, T. F., and Chistov, L. B., 1966, New data on the modes of occurrence of tantalum and niobium in wolframite: Doklady of the Academy of Sciences of the U.S.S.R., v. 166, no. 106, p. 105-107. * U.S. GOVERNMENT PRINTING OFFICE: 1984-776-041/4030 REGION NO. 8 ■