TDOC Z TA245. 7 B873 no.131L CONTENTS Historical Introduction . . . . . . . . . . . . . . . . . . . . . . . Experimental Procedure . . . . . . . . . . . . . . . . . . . . . . . . 7 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .‘Q . ll "The Short-Season Effect" . . . . . . . . . . . . . . . . . . . . . . . 41 References Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 SUMMARY Experiments were conducted in Texas from 1975 through 1976 to investi- gate the place and value of rapid—fruiting cottons in production schemes directed at minimizing damage from the boll weevil, Anthonomus grandis Bohe- man. Narrow-row planting and the selection of certain genotypes produced more cotton bolls earlier in the season and increased yields. The benefits of genotype selection depended on the length of time during the bloom and boll set period that injurious numbers of weevils were delayed. Short-sea- son types were more productive than their long-season counterparts in cases where the weevil-free period was shorter; however, where weevils were delayed the slower-fruiting cottons produced equally well. Susceptibility to boll- lock injury increased daily as bolls developed from later flowers following initiation of blooming. Seventy percent or more of the lint harvested (500- 750 pounds of lint per acre) was produced from flowers of the first 3 weeks of blooming. It is during this critical time period that significant boll weevil damage to squares must be prevented if a satisfactory yield (500-750 pounds of lint per acre) is to be obtained without initiation of costly late-season insecticide applications. KEYWORDS: Boll weevil/ Anthonomus grandis Boheman/ control/ cotton/ geneticsl short-season genotypesl narrow—row plantingl age of cotton boll/ "short-season effect"/ Texas. THE ‘suoar-snzgsow EFFECT‘ IN COTTON AND ESCAPE FROM THE BOLL WEEVILI R. D. Parker, J. K. Walker, G. A. Niles and J. R. Mulkey* The tenure of Fredrick William Mally with the Agricultural and Mechanical College of Texas in College Station was a brief one of 3 1/2 years; but he left a cornerstone of research for viable cotton production that now extends over a significant amount of today's acreage. Entomology at the College began in 1899 with the arrival of Mally, who was engaged to develop a solution to the shocking destruction brought by a small, gray insect, the boll weevil, Anthonomus_grandis Boheman. Erupting out of Mexico, this destructive cotton pest had struck through a large part of Texas by 1899, leaving little doubt that any of the cotton acreage of the Old South would escape. Mally's stewardship at the College, his research and extension programs, the recomendations he submitted, the rejection of these same recomendations by growers, and the slowly evolving boll weevil management practice during the next 70 years or so are thick with irony. 1Texas Agricultural Experiment Station, Texas A&M University. The work reported herein was funded by the Rockefeller Foundation and by an IBP sponsored project entitled "The Principles, Strategies, and Tactics of Pest Population Regulation and Control in Major Crop Ecosystems." This support was made available through a grant from _ National Science Foundation (NSF GB 34718) and a cooperative agreement with Agricultural Research Service, U. S. Department of Agriculture. *Respectively, former graduate assistant (Department of Entomology), Texas A&M University, presently, entomologist, the Texas Agriculture Extension Service; associate professor (Department of Entomology), professor (Department of Soil and Crop Sciences), Texas A&M University, and associate professor, Texas A&M University (Agriculture and Research Extension Center at Uvalde). _2_ Mally saw that every advantage was being surrendered to the boll weevil by the then present manner of cotton production. Oddly enough, the cotton plant itself exaggerated the abilities of the weevil. Varieties of those years were extremely slow to fruit, and somehow, this characteristic was related to increased losses. Moreover, the lateness of these cottons served the interests of weevils surviving the winter; it was only the fall generations of the pest, feeding on cotton that remained growing in fields, that survived to the following spring. Mally's important recommendations, then, were these: plant early-maturing cottons (new developments at about this time) and destroy all cotton stalks promptly after harvest. Not entirely satisfied with Mally's guidance, growers made the scientist's position so unpleasant that he resigned in 1902.’ If the quick change to faster—fruiting cottons seemed awkward for planters of 1900, the absence of the mechanics necessary and of the willingness to harvest cotton without delay and to promptly destroy stalks, stamped the researcher's recommendations as unworkable and out of the question. And yet, the thrust of later entomological findings by scientists from a number of quarters pointed to the wisdom of Mally's suggestions; in the final analysis it would be these recommendations, imperfect as they were, that permitted cotton production in the presence of the boll weevil. Improvement in the system would await the organic insecticides after World War II. But what of this earliness in cotton —— an attribute that was observed to reduce losses by the weevil? How should it be measured? Not always tied to the newly introduced fast—fruiting varieties, _3._ earliness often seemed a function of the interaction of cotton with the soil on which it was grown. Alluvial, bottomland soil could grow a larger, more vigorously fruiting plant, but boll set might be slower, and ironically, weevil damage could be intensified. On the other hand, the less vigorous cottons of the upland soils fared better in their relationship with the weevil: more fruit could be set before weevils reached great populations. Numerous accounts of the conflict between the boll weevil and the cotton plant likened it to a race between the fruiting ability of the plant and the relentless increase of the insect's numbers. This "relentless" increase, occurring in those seasons when rainfall was abundant, often became even more dedicated. Where the cotton plant won the first lap or so of the race, bolls would be set and some production assured. The more bolls established before weevils increased to great numbers, the higher were the yields. Obviously, bolls were less vulnerable than squares. Earliness, when it operated to subdue weevil damage, likely was noted only after the fact; when at the conclusion of a growing season it became apparent that in a particular cotton field, in a given season, fruit set had been rapid and a reasonable yield had resulted despite the boll weevil. The reasons for or against these events occurring were at best only vaguely understood. A number of agricultural scientists examined the matter (Bennet 1908, Ewing 1918, Cook 1923, Ware 1925, Verhalen et al. 1975); yet the descriptions of earliness in cottons and its interface with the weevi1's numbers were not adequate to structuralize the interaction so that predictive extensions might follow. If studies provided generalities of this relationship, these fell short of the kind -4- of delineation needed to more perfectly understand, and address, the problem of the pest. The arrival of the organic insecticides after World War II seemed to relegate the idea of "earliness in cottons as a means to escape damage from the boll weevil" to a less prominent status. Whether cotton fruited rapidly or not, these economical and effective insecticides (applied in multi—applications) were the kind of solution that growers had long imagined. By 1964, over 19 million pounds of insecticides was being applied to cotton to guard against the boll weevil and other cotton insects. No longer was the boll weevil approached with the 1ess—than—perfect cultural steps of Fredrick Mally. The scheme worked well until a series of events, evidence of the biological skills of the insect, showed themselves. The development of insect resistance to pesticides, especially by the bollworm, Heliothis E23 (Boddie), and the tobacco budworm, Heliothis virescens (F.), cut through the enthusiastic promise of multi—treatment programs and had, by the mid—l970's rerouted the strategies of pest control in cotton in Texas. In 1976, a little over 2 million pounds of insecticides was applied to the crop. Hardly a reroute at all, most of these "new" Texas schemes embrace the basic Mally tenets of the early years of this century. Variously described, the new direction is more commonly referred to as "short—season cotton production." At the center of the scheme are cotton genotypes that, under a variety of soils and growing conditions, produce relatively high yields in a compressed fruiting period. Certain conditions enforce this fruiting habit: reduced nitrogen and the avoidance of late irrigation are important. It _5._ is widely held that the early squares are important and should not be lost to the cotton fleahopper, Pseudatomoscelis seriatus (Renter), or “__ to the_boll weevil. A number of studies have demonstrated that short-season cotton production performs well (Heilman et al. 1977, Namken and Heilman 1973, Sprott et al. 1976); insecticides have been reduced and yields and profits increased. But the questions of why and under what circumstances the rapid—fruiting of cotton serves to limit weevil damage have remained unanswered. Walker and Niles(l97l) approached a solution to these mysteries in their model which characterizes the different effects of the boll weevil on both short— and long—season cottons where injurious infestation did not develop until the second generation, a period that starts about day 30 of the blooming period. In this model most blooming had occurred in the short-season cotton by day 30, and a substantial number of bolls (12 days or older) were established on plants. Bolls of these ages were presumed unsusceptible to weevil damage. On the other hand, the long-season cotton had bloomed less with a complement of younger bolls present than were on the short-season plant. Walker and Niles presented other data that inferred that a crop could escape major injury if overwintered weevil infestations were no larger than 20 females per acre. When numbers of the overwintered brood were increased beyond 60, damaging infestations appeared well before day 30, or before adequate boll set had been secured. §This study established that the time of emergence of the first and second generations of weevils could be predicted by knowledge of the flowering pattern of cotton. -6- This model was tested (Walker et al. 1976), and the results demonstrated that, where weevils did not attain damaging numbers until the second generation, substantial production was realized, especially where new short-season cottons were used and planted in narrow-row configurations. Walker et al. (1977) later established the probabilities of injury to bolls of different ages present at the time weevils became severe in the second generation. It was concluded that "when weevils are delayed until the second generation, the level of boll damage may be more closely involved with the age distributions of the bolls. A production scheme that increases the number of older bolls (perhaps 12 days of age) by the time of the beginning second generation emergence would be less sensitive to weevil attack." This experiment illustrated the daily decline in susceptibility of cotton bolls. Bolls in the 12-day—or—older category escaped major losses of locks to the weevil. The experiments reported herein were designed to explore and develop understandings of this "short—season effect" or the particular situation that permits effective cotton production where a combination of rapid—fruiting (and rapid boll set) and a delay in damaging numbers of the weevil until about day 30 of blooming exists. _7_. EXPERIMENTAL PROCEDURE Experiments were conducted during 1975 at a dryland Blackland site in Falls County in a 35-inch rainfall zone and during 1975-76 in Uvalde and Frio Counties, an area termed the Winter Garden. Rainfall is about 23 inches per annu in the Winter Garden; both irrigated and dryland culture are practiced. The objective was to evaluate, without late—seasonal insecticidal control, various components of short-season cotton production, i.e., genotype, planting configuration, irrigation (in the Winter Garden) and levels of insecticides for overwintered weevil control. Nitrogen was held to about 25 pounds per acre in all tests. During 1976 at the Uvalde location, a companion study was carried out where overwintered weevil insecticidal treatments were applied plus eight late-season applications. Boll weevils were the only major insects at the test locations, although a cotton plant disease did produce premature defoliation of certain'genotypes at the Uvalde site in 1976. The primary field design at these locations was a split plot with three or four replications. Main plots were two levels of overwintered weevil treatments (one or three applications in the“ Winter Garden and two or none in Falls County); the first was initiated at first one—third grown squares, and succeeding applications followed at four-to six—day intervals. Plots untreated or receiving a single application are designated Tl, those treated twice or three times, T2. Main plots were divided into two in—season irrigation levels (I1 and I2) (Winter Garden); these were divided into three planting configurations —- single-drill plantings on 38-inch beds (S1), double—drill plantings on 38-inch beds (S2), and -8- single—drill plantings on 25-1/3-inch beds (S3). The S2 and S3 configurations are generally termed "narrow-row plantings." The planting configurations were further divided into various numbers of cotton genotypes. At the Blackland site, main plots contained the two levels of overwintered weevil treatments; subplots contained genotypes. Only single-drill, 38-inch planting was used at the dryland Blackland location. The genotype plots in these experiments were two rows wide for the 38-inch rows, and three rows wide for 25-1/3-inch rows; these were 50 feet long. The Falls County test was replicated four times; Frio County, three times; Uvalde (1975) four times; Uvalde (1976), three times. Twenty buffer rows were used to separate the two levels of insecticides; four buffer rows were planted between the irrigation levels. A second experiment was planted in 1976 at Uvalde, and here full- season insecticidal treatment was applied: three early—season treatments and eight late—season. Three genotypes were examined in a randomized block design. Plots were 50 feet long, two rows wide; there were three replications. The following is a description of the cotton genotypes examined. lX6—56 is a small, compact, fast—fruiting cotton. 6M—l0 is less compact than lX6—56 and is intermediate in its rate of fruiting. l209L and 1209 produce larger plants, have larger bolls than 1X6-56, and are slower to fruit. A07-79—R2 is a long—season, frego bract cotton. ORS-75C and 0HR—74C are okra leaf—frego bract stocks that fruit very rapidly. Frego bract cottons have been shown to possess some resistance to the boll weevil. _9_ Several commercial varieties were included in the tests. Lankart LX 571 is a widely planted cotton that, under dryland conditions and limited irrigation, fruits rather fast. TAMCOT SP-37, a variety we1l—adapted to South Texas, fruits quickly and yields well in dryland and irrigated cultures. Stoneville 213 is a long—season, relatively slow—fruiting variety that yields well under a high—input system.‘ Percentages of punctured squares and weevils per acre were determined from weekly examinations in each genotype subplot. Twenty—five squares were examined, and the row feet required to produce the 25 was measured; then, percentages of punctured squares and weevils per acre were calculated. Numbers of weevils per acre determined by fruit examination alone would, of course, underestimate the actual number. Before squaring, 150 row feet were examined on different dates to determine numbers of overwintered weevils. Adult weevil counts were not made at the Fall County test. Seasonal blooming was recorded in 6-1/4-foot permanently marked sampling areas in each of the genotype subplots. Blooms were tagged and dated, permitting later examination for weevil damage when bolls were harvested. As the cotton boll has four or five locks, sometimes three, damage will be presented as percent damage to locks that originated on a certain day of blooming or during a period of blooming. Collections also were made at Uvalde in 1976 of tagged TAMCOT SP—37 bolls that had shed; these were examined for evidence of injury from weevil larvae. Yields were determined from the 6-1/4—foot sampling areas where blooms were tagged. One to three harvests were made; these were ginned, and lint and seed weights were determined. ._lo_ Where statistical differences were noted in the analysis of variance, means were separated at the 5—percent level of probability with Duncan's Multiple Range Test. Table 1 gives various information pertinent to the studies. Seasonal history of cotton experiments in Falls, Frio, and Uvalde Counties, l975—7( Table 1. Date of Date of Production se Planting Date of insecticide treatment— irrigation rainfall C County date first bloom l 2 3 1 2 after plant Falls 4/16/75 e/19 e/osi/ 6/09 4.25 Frio 3/11/75 5/24 5/07 5/13 _ 5/19 6/10 6/18 14.40 Uvalde 3/13/75 6/04 5/16 5/22 5/28 6/20 ——E/ 6.01 Uvalde 3/23/76 6/06 5/18 5/24 5/28 6/ll 7/01 27.70 Uvalde-q/ 3/26/76 6/06 5/18 5/24 5/28 6/15 e/21 27.70 (full-season treatment) a/ -— Azinphosmethyl (0.25 lb AI/A). -—/ Rainfall negated the need for a second irrigation. c/ -— Dicrotophos (0.1 lb AI/A) was used for fleahopper control on this date in plots untrea overwintered weevils. d/ eight treatments were made beginning July 2. -— Azinphosmethyl at 0.25 lb AI/A and Pydrin at 0.125 lb AI/A were applied for each trea _11_ RESULTS Falls County, 1975. The boll weevil percent—punctured-square infestation was slow to develop at this location, and the percent—punctured squares never reached a high level; there was no statistical difference between plots treated twice for overwintered weevils or untreated. Percent—punctured—square count on day 28 of the blooming period was less than 20 percent (Table 2). Ten genotypes were planted at this test site; however, we selected only three for critical examination — the commercial variety Lankart LX 571, 1X6-56 and the frego bract genotype ORH—74C. There were no significant differences in the fruiting between genotypes due to insecticidal treatment; the means presented are an average of the two treatment levels. There were no significant differences in yields due to genotype (Table 2). Over 80 percent of the bolls of all cottons harvested were from flowers set by July 8 (Table 2); that is, from flowers produced during the first 20 days of flowering. Lock damage from weevils was low in locks developing from flowers of the first 18 days of blooming and gradually increased after this (Figure 1). The percent—lock damage for the season averaged across treatments and genotypes was about 9 percent. Lankart LX 571 produced fewer bolls than the other cottons, but this cotton possesses a larger boll. Figure 2 shows the rapid accumulation of bolls 10 days or older that occurred for Lankart LX 571. By about day 30 of the blooming period, 90 percent of the bolls to be harvested (about 90 percent of 539 pounds per acre) were l0 days old or older. In a brief period of 10 days, from day 19 of blooming to day 29, the major part of the crop -12- Table 2. Percent-punctured squares, yields of cotton, and numbers of bolls harvested (lO00s/A) from different flowering periods, Falls County, l975§/ % Punctured squareshj 6/17 6/26 7/10 7/17 Genotype T1 T2 T1 T2 T1 T2 T1 T2 ORH—74C 0 0 2 2 5 12 17 15 Lankart LX 571 0 O 0 1 12 15 18 17 1X6-56 0 O 0 0 10 17 14 16 Pounds lint/A ORH—74C 459 Lankart LX 571 539 1X6—56 431 Numbers of harvested cotton bolls from different flowering periods and (Z) of these of total / Genotype e/19-23 e/24-2s 6/29-7/03 7/04-085 7/09-13 Total ORH—74C 8 (6%) 14(10%) 49(34%) 52(35%) 23(16%) 146 a Lankart LX 571 10 (8%) 21(19%) 40(35Z) 32(28Z) 11(1OZ) 115 b 1X6—56' 17(12Z) 18(13%) 40(30%) 40(27%) 25(18%) 137 a Eheans are compared vertically; those sharing a common letter are not significantly different (P=0.05). b/ —-Data not analyzed. 2/ Day 20 of blooming. _]_3_ Figure 1. Percent of locks damaged by boll weevils from flowers occurring on different days of the blooming period, Falls County, 1975. FALLS COUNTY, I975 6.8% of total locks produced in first I8 days were weevil-damaged 3O F e 25 - % wEEvlL 2o - e ' DAMAGED LOCKS ,5_ _ - e |0 " ' . Q e 0 ‘.00 0' 5- e Q0 Q Q ' ' L l3 5 7 9 ll l3 I517 I92! 2325 DAY OF BLOONHNG -14- Figure 2. Accumulated bolls (10 days or older) and percent of these of total harvested, at different days of the flowering period, Falls County, 1975. f | | | I | | I I I I I I I I I I FALLS CQUNTY , I975 . 3:1 E x < u, mo _ 904/. OF HARVESTE - n: ‘l: BOLLS Io DAYS OR E 5 OLDER THIS DAY s‘ew- — O f! _ A KART LX 57l Q no: E L N 062.3% J O. Z O g9 6o _ _. w $5 i j o o ‘Q 3 2 g In 4o - - ° Q 1.1:" o 22w. 6 I- llJ Z n: g 2O _ _ 5' l‘ <~ a é 5 4 ‘- '- I I I I I I I I I I I I I I I I I I a 5 7 9 ll l3 I5 I7 l9 2| 23 25 2? 29 3| as DAYS AFTER FIRST BLOOMS _15_ in bolls attained l0-day—old—or—older status. This compares favorably with the short—season cotton of the Walker—Niles model; had weevils in the present Blacklands test become severe around day 30 of blooming, lock damage according to the model should not have been severe. The rapid-fruiting performance of Lankart LX 571 explains its wide use on certain soil types in Texas and under certain production schemes. In many situations a delay in boll weevil buildup allows effective production with this cotton before weevils become limiting; a large number of bolls of sufficient age to escape major weevil damage can be set. The slight boll weevil infestation did not adequately test the Walker—Niles model at the Blacklands site in 1975; fortunately, more severe numbers were encountered the same year in the Winter Garden test. Frio County, 1975. As a consequence of an extremely poor cleanup of cotton stalks in 1974, overwintered weevils infested this experiment in large numbers. The small test field essentially was the only cotton planted in Frio County in 1975, and overwintered weevils concentrated in it. Also, rain prevailed throughout the growing season, a situation highly favorable for weevil development. Percent—punctured squares was less where three early insecticidal applications were made (compared to a single application), but by day 20 of the blooming period in the T2 plots the infestation had soared to 50 percent (Figure 3). Significantly (P=.O5) fewer squares were punctured per acre through the first l0 days of blooming in the T2. There was evidence that, in addition to infesting the T2 plots between applications of insecticide, overwintered weevils were continuing to -15- Percent punctured squares at different points in time of Figure 3. the squaring period at three locations, 1975-1976 (T1 = one treatment; T2 = three treatments). —o— TREATED 1 11m»: FOR OVERWINTERED wzevnts --o-- TREATED s TIMES so - . <- -- - 8 FRIO CO.; UVALDE; UVALDE; 5 ‘ro- 191s -- 197s -. 197s _ D o g I I O 0 -u- -r- -4 u, | mresmnon , mrssnraou n: I l4’ REACHES 50% I REACHES 50% o f i’ 5O_ o: PUNCTURED __ u! INFESTATION ‘P PUNCTURED , _ o | SQUARES on MREACHES 50% SQUARES on | g I omr 2o OF PUNCTURED DAY 4s OF m '1 stoomms l souamzs on stoomms l . 40 — , m T2 -- ' DAY 2e or -- m r2 ' - R | I stoomms . : I I m 1'2 | a so - ¢ -- o i‘ -- 3 - 1' , I -‘ : I 1 2o - /”o -- l, -- o '1 ' a | o f I j I I0 - | -- g ,0 ~- o - o-o’ 0"»? f l‘ 1 . 1 1 1 1 1 PQ1 1 1 1 1 1 - —- 10 2o so 4o so so 1o 1o 2o so 4o so so 1o 1o 2o so 4o so so 1o DAYS AFTER THE APPEARANCE OF THE FIRST l/ 3 GROWN SQUARES _17_ colonize after the final treatment. For about 28 days after appearance of the first squares, the infestation was considerably less in T2 than in Tl. Numbers of overwintered and sumer generation weevils per acre appear in Table 3; the infestation was less in the T2 plots. The reductions in weevil numbers and square damage made a great difference in blooming between the two levels of treatment (Table 3). Significant increase of blooms resulted from an additional in—season irrigation; substantial increases were attributable to S2 and S3 planting and to genotype. The effect of these variables on numbers of harvested bolls is shown in Table 3. Trends are similar to those recorded for blooms. The variable of three overwintered weevil treatment had the most powerful effect in increasing yields (Table 3), but irrigation, S2 and S3 spacings, and genotypes also produced positive, additive increases. The benefits of the quicker-fruiting TAMCOT SP—37 and lX6-56 are apparent. Exploding populations of weevils appeared too early in the blooming period (even in the T2 plots) to allow for the accumulation of bolls of sufficient age to escape major damage. This is apparent in Table 3. However, percent damage to locks in bolls from different flowering intervals is less in T2 plots. It is striking that yields in excess of 500 pounds of lint per acre were realized, considering the magnitude of lock injury. Harvested, undamaged locks as influenced by insecticidal treatment appear in Figure 4. Only a small percent of the harvest came from blooms after the first 18 days of flowering. This experiment underlines the importance of the transition in Table 3. _13_ Number of weevils per acre,accumulated blooms from dif- ferent periods, total bolls harvested, and yields in pounds of lint per acre; total locks harvestediand per- cent of these damaged by boll weevils. Frio County, 197s 5/3/ N . g/ umber of weevlls/A T1 T2 May 21 367 o May 27 459 183 June l6 6093 ll23 June 23 3172 3900 Accumulated blooms (l0O0s/A) May 24- June June June 4 5-ll 12-23 Tl 36.6 e 107.6 e 225.4 e T2 60.2 f 179.3 f 358.9 f Il 44.7 e 136.6 e 284.2 e I2 52.0 f 150.3 f 300.2 f S1 36.4 e 107.0 e 228.2 e S2 58.6 f 165.6 f 321.2 f S3 50.3 i 157.7 i 327.2 i _]_9._ Table 3 (continued) 6M—10 44.1 e 126.2 e 255.2 e lX6—56 57.5 f 160.6 g 309.2 f TAMCOT SP—37 43.6 e 143.5 f 312.1 f Total bolls (1000s/A) harvested T1 121.0 e T2 217.9 f I1 159.2 e I2 179.7 f S1 141.9 e S2 175.6 f S3 190.7 f 6M—10 145.0 e lX6—56 179.6 f TAMCOT SP—37 183.8 f Pounds lint/A S1 S2 S3 Mean T1 132 163 168 154 e T2 455 541 535 511 f Mean 294 a 352 b 352 b -20- Table 3 (continued) Genotype 6M-10 1216-56 TAMCOT SP—37 Meat: I1 275 309 334 306 e I2 309 407 361 359 f Mean 292 a 358 b 348 b Total locka/produced Damaged boll locks Blooming (l000s/A— (Z) interval Tl T2 Tl T2 May 25-31 50 42 29.0 18.8 June 01-06 103 265 57.4 b 31.6 a June 07-12 186 304 87.8 b 61.8 a June 13-18 106 137 90.6 78.3 June 19-24 42 118 78.8 82.8 5/ followed by e, f, g are compared vertically. a common letter are not significantly different (P=0.05). b/ g/g/ First blooms appeared by May 24. Not statistically analyzed. Means followed by a, b are compared horizontally; those Means sharing ._2]__ Figure 4. Numbers of locks escaping weevil injury from different flowering periods, Frio County, 1975 (T1 = one treatment; T2 = three treatments). O-—-O I TREATMENT '50 ’ I-—I a TREATMENTS ' G a: u <\: _ - m I 8 O c I00 FRIO COUNTY, I975 m x o O _.| _ _ .1 .1 O m ‘q so- - 52 ¢ 2 ‘i O