TA245.7 873 3-1246 0.1246 Range Vegetation Response to Burning Thicketized Live Dak Savannah CONTENTS SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 METRIC UNITS-ENGLISH EQUIVALENTS . . . . . 4 SCIENTIFIC NAMES OF PLANTS MENTIONED IN TEXT . . . . . . . . . . . . . . . . . . . . . . . . 4 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 STUDY AREA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 MATERIALS AND METHODS . . . . . . . . . . . . . . . . . 6 RESULTSHHL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 5 Soil Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 6 Fuel Characteristics . . . . . . . . . . . . . . . . . . . '. . . . 6 Environmental Conditions During Burning . . 7 Green Grass Standing Crop . . . . . . . . . . . . . . . 7 Botanical Composition of Grass Stands . . . . . 9 Forb Standing Crop . . . . . . . . . . . . . . . . . . . . . . . 1O Botanical Composition of Forb Stands . . . . . . 11 Mulch Yield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Total Herbaceous Standing Crop . . . . . . . . . . . 11 Density of Running Live Oak . . . . . . . . . . . . . . 1" Height Change of Running Live Oak . . . . . . . 1b Standing Crop of Running Live Oak . . . . . . . . 14 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . 15 LITERATURE CITED . . . . . . . . . . . . . . . . . . . . . . . . . . 15 SUMMARY Response of live oak-dominated vegetation on thicketized uplands to burning in the fall 1974, spring 1975, and fall 1975 was evaluated through 1977 on the Aransas National Wildlife Refuge. Standing crop, species diversity, and botanical composition of her- baceous vegetation and density of woody plants were the major variables evaluated at various times after the burns. The areas were not grazed by domestic livestock during the study. Burning uplands in the fall generally increased grass standing crops, compared to burning in the spring, for at least one growing season. However, areas burned in the spring weremore productive than unburned areas based on evaluations for two growing seasons after the fires. Herbaceous species diversity increased following all burns but was greatest on areas burned in the fall. Burning tended to decrease the proportion of the grass species of poor grazing value in the stand, and to increase the pro- portion of the grass species of good-to-excellent for- age value by the second growing season after the fires. Gulfdune paspalum and Heller panicum in- creased the first growing season after burning but began to decline as the proportion of little bluestem increased during the second growing season on burned areas. Forb standing crop on uplands burned in the fall, averaged across sampling dates, was five times that of unburned areas and twice that of areas burned in the spring. Mulch accumulated rapidly after the burns on the upland sites in the absence of grazing by live- stock. Accumulated mulch on unburned areas appar- ently delayed emergence of forbs and grasses in the spring and severely reduced species diversity. Al- though live oak dominated the area, stem density de- clined somewhat as mulch accumulated after burn- 1n . The fires initially top killed the live oak, but den- sities of new sprouts the growing season after burn- ing exceeded preburn stem densities by as much as five times. Although live oak densities recovered rapidly following the burns, reductions in stem height and standing crop of live oak released her- baceous species. Following the initial burst of re- sprouting, live oak stem numbers usually returned to preburn densities by 2 years following the fires, ap- parently because of mulch accumulations and compe- tition from the herbaceous species which were re- leased by the fires. The vegetation also began to show signs of dete- rioration (heavy accumulations of mulch and reduced vigor of herbaceous plants) by 2 years following burn- ing. Thus, burning thicketized live oak savannahs every 2 or 3 years, depending on rainfall, may be necessary to maintain the vegetation at a successional stage most suitable for quality wildlife habitat and at optimum potential for livestock grazing use. KEYWORDS: Range vegetation and burning/live oak/standing crop/species diversity/woody plant density/botanical composition. 3 w! METRIC UNITS-ENGLISH EQUIVALENTS Metric Unit English Equivalent Centimeter . . . . . . . . . . . . . . . . 0.394 inch Hectare . . . . . . . . . . . . . . . . . . . . 2.47 acres Kilogram . . . . . . . . . . . . . . . . . . 2.205 pounds Kilogram per hectare . . . . . . . 0.893 pounds per acre Kilometer . . . . . . . . . . . . . . . . . . 0.62 statute mile Kilometer per hour . . . . . . . . . 0.62 miles per hour Liter . . . . . . . . . . . . . . . . . . . . . . 0.264 gallons Meter . . . s. . . . . . . . . . . . . . . . . . 3.28 feet Square meter . . . . . . . . . . . . . . 10.758 square feet (Degrees centigrade X 1.8) + 32 . . . . . . . . . . . . . . Degrees fahrenheit Mention of a trademark name or a proprietary product does not constitute a guarantee or warranty of the product by the Texas Agricultural Experiment Station and does not imply its approval to the exclusion of other products that also may be suitable. 4 SCIENTIFIC NAMES OF PLANTS MENTIONED IN TEXT Common Name Grasses Big bluestem Brownseed paspalum Bushy bluestem Carolina jointtail Florida paspalum Gulf cordgrass Gulfdune paspalum Hairyawn muhly Heller panicum Little bluestem Pan American balsamscale Sheep panicum Silver bluestem Thin -paspalum Yellow Indiangrass Forbs American snoutbean Coulter conyza Hairy bonamia Hairypod pepperweed Plains wildindigo Prairie senna Sawtooth frogfruit Spadeleaf Texas croton Western ragweed Scientific Name Andropogon gerardii Paspalum plicatulum Andropogon glomeratus Manisuris cylindrica Paspalum floridanum Spartina spartinae Paspalum monostachyum Muhlenbergia capillaris Panicum oligosanthes var. helleri Schizachyrium scoparium var. frequens Elyonurus tripsacoides Panicum ovinum Bothriochloa saccharoides var. torreyana Paspalum setaceum Sorghastrum nutans Rhynchosia americana Conyza coulteri Stylisma villosa Lepidium lasiocarpum Baptisia leucophaea Cassia fasciculata Phyla incisa Centella asiatica Croton texensis Ambrosia psilostachya Woody Plants American beautyberry Live oak Post oak Redbay Saw greenbriar Yaupon Callicarpa americana Quercus virginiana Quercus stellata Persea borbonia Smilax bona-nox Ilex vomitoria Range Vegetation Response to Burning Thicketized Live ()ak Savannah C. I. SCIFRES AND D. M. KELLEW Woody plants, previously suppressed by natural fires, have increased on the Coastal Prairie appar- ently because of continuous overgrazing, indiscri- minant burning, and unsuccessful attempts at fann- ing (Blakely, 1947). As the woody plants increased in stature and spread to previously open areas, the pro- duction of range forage species of high value for graz- ing by cattle decreased and/or were replaced by less desirable species. Much of the Coastal Prairie was originally savannah, a grassland with isolated live oakl trees or scattered mottes which were of consid- erable value as shade for livestock and cover for wildlife. The closing of savannahs by increasing cover of woody plants was referred to as ”thicketiza- tion” by Dyksterhuis (1957). On much of the Coastal Prairie, live oak savannahs have become heavily thicketized, and the interspaces between the scat- tered large trees or mottes are now infested with dense stands of the shinnery form of live oak. The shinnery form, hereinafter referred to as ”running” live oak, spreads rapidly by a vigorous lateral root system, but individual stems are usually less than 1 meter tall. Thicketized savannahs may be improved for cat- tle grazing by use of herbicides (Scifres and Haas, 1974), but the chemicals damage or control the large trees as well as the underbrush. The optimum treat- ment, considering needs of livestock and wildlife, vsfould remove most of the underbrush but spare the large trees. Therefore, prescribed fire may have more potential than conventional brush control methods for restoring savannahs. Fire has influenced the development of most natural grassland communities (Vallentine, 1971) and is t. -latural method of keeping primitive forests open and in stable equilibrium (Biswell, Schultz and 1’ tific names of plants mentioned in text are listed on the pre- x 1g page. p Respectively, professor and former research assistant, The Texas Agricultural Experiment Station (Department of Range Science). Launchbaugh, 1955). Prior to settlement by white man and intensification of agriculture, fires periodi- cally suppressed woody vegetation on prairies and savannahs of the coastal region of Texas (Blakely, 1947). Since fire generally favors grasses at the ex- pense of woody vegetation (Daubenmire, 1967), its absence allows gradual recolonization by woody species on almost all types of grasslands (Saur, 1950; Cook, 1908). Burning can magnify drouth stress in arid areas but may be beneficial in temperate zones when moisture is not usually limiting. Wildlife habitat is usually improved by burning (Harne, 1938). Com- pared with mature browse, new growth following burning of woody species is more palatable and nutri- tious, characterized in part by higher crude protein content which may persist for several years (Dewitt and Derby, 1955). Prescribed burns on the Sabine and Laccassine Refuges in Louisiana were consid- ered simultaneously beneficial to water fowl, fur- bearing animals, white-tailed deer (Odecoilus virginiana), and cattle (Lynch, 1941). Increased cal- cium and phosphorus contents (Green, 1935) are also benefits of burning herbaceous forage species. Fire increases surface soil temperatures in the spring which encourages early emergence of forage plants and may stimulate nitrification (Wright, 1974). Burn- ing improves utilization of range forages by livestock (Gordon and Scifres, 1977), establishes mineral soil seedbeds, and may suppress various diseases (Wright, 1974). The primary objective of this study was to de- (termine the response of thicketized Coastal Prairie vegetation to controlled burning in the spring or fall. STUDY AREA The Aransas National Wildlife Refuge is located on Blackjack Peninsula and bounded by San Antonio and Saint Charles Bays to the east and west, and by the Aransas Bay to the South. The refuge is located on a strip of deep whitish sand, 4 to 12 kilometers 5 " w’. wide, occurring along the coast from Baffin to San Antonio Bays and interrupted by Corpus Christi, Copano, and Saint Charles Bays. This sandy body, sometimes called Flour Bluff Ridge, is the remains of barrier islands that formed offshore in the Pleis- tocene and, driven inward by wave action, finally came to rest 100,000 or more years ago at the outer edge of the Beaumont Formation Delta (Jones, 1975). The major upland soils, Mustang and Galveston fine sands, are acid to moderately alkaline and usu- ally contain more than 90 percent sand, less than 10 percent clay, and less than 1 percent organic matter in the surface 30 centimeters. The soils are deep, moderately drained, rapidly permeable, and often saturated or inundated in late falland early spring. The climate of the region is classified as dry sub- humid (Jones, 1975). Mean annual rainfall is 91.4 cen- timeters. Relative humidity may vary from 80 to 90 percent during the mornings to less than 60 percent in the afternoons. Average maximum temperature during midsummer is 33 degrees centigrade with frequent highs of 40 degrees centigrade. The average minimum temperature is 24 degrees centigrade. The average high in January is 21 degrees centigrade, and freezing weather is likely to occur only three or four times each year. The average date of the first killing frost is December 20, and the last killing frost usually occurs by January 30. The study area, consisting of approximately 2,025 hectares and typified by live oak mottes and upland flats interspersed with sloughs and swales, is separated from the remainder of the Refuge by a deer-proof fence. The deep sandy loam uplands typi- cally support mid- and tallgrasses including big bluestem, little bluestem, yellow indiangrass, gulfdune paspalum, and Dichanthelium species. Major forb species are western ragweed, plains wild- indigo, Texas croton, hairy bonamia, and Cassia species. The major woody species is live oak, which occurs as scattered large trees, in mottes, and in shinnery form. Redbay, yaupon, American beautyberry, and saw greenbriar are commonly as- sociated with live oak. MATERIALS AND METHODS Areas of approximately 700 hectares were burned in October 1974, April 1975, or October 1975. The area burned in October 1974 was reburned in November 1976. Vegetation of an unburned area, lo- cated outside the fence, was monitored for compari- son with vegetation of the burned areas. Two or three permanent 183-meter lines were established on each burned area and on the unburned area for vege- tation sampling. Relative humidity, air temperature, and wind direction and velocity were recorded during each burn except during the spring burn. Standing fine fuel. (herbaceous vegetation) was clipped to a 2.5- centimeter stubble height from twenty-five 0.25- square-meter areas equidistantly spaced along each 6 sampling line and separated into green and dead. After removal of standing fine fuel, mulch was c4- lected from each sampling station. Standing fine and mulch samples were oven-dried and weighed. Twenty-five samples of fine fuel and mulch were col- lected for water-content determinations. At the same time, triplicate soil samples were taken from 0 to 8, 8 to 15, and 15 and 30 centimeters deep for gravimetric determination of soil water content. Soil textures from the same depths were determined by the hy- drometer method (Bouyoucos, 1962), and pH on a 1:4 slurry (Mortensen, 1965). Post burn standing crop of herbaceous vegetation was harvested at 4-month intervals from fifteen to fifty 0.25-square-meter sampling areas equidistantly located along each permanent line. Care was exercised not to harvest areas sampled previously. Total oven dry standir“ crop (green + dead), total green standing crop, and green standing crop by plant species were recorded. Live oak preburn and postburn densities, heights, and standing crop to a 2.5-centimeter stub- ble height were recorded for plants less than 1 meter tall from fifty 0.25-square-meter sampling areas equidistantly located along each sampling line. Den- sity, height, and standing crop of live oak regrowth were evaluated in the same manner at 4-month inter- vals following each burn. New sprouts, entire Stems, and leaves of live oak which were judged of adequate palatability and succulence to be browsed to ground level were separated from mature growth. Live oak canopy cover replacement was estimated from the permanent lines using a large, inclined 10-point frame (Arny and Schmidt, 1942; Levy and Madden, 1933). This approach facilitated evaluation of effects of fires on running live oak in the interspaces among the mottes. Statistical analysis was conducted among sea- sons of burning within sampling dates based on var- iation among sampling lines. Significant differences in variables between burned and unburned areas were accepted at the 0.05-level of probability based on paired comparisons (t-test). RESULTS Soil Characteristics The relative proportions of sand, silt, and clay were essentially the same at all depths from the up- land sites (Table 1). Sand content averaged 91 to 92 percent, and silt content was only 3 percent. Thus, the soils were highly permeable, but the surfaces dried rapidly following rains during the growing season. Organic matter content was less than 0.25 percent in the surface soil. —\ Fuel Characteristics Fuel loads were highly variable within burned areas, ranging from none (bare soil) to heavy stan \ grass and running live oak with understory IHLIICh accumulated to as deep as 6 centimeters. Preburn mulch loads averaged 2,600 to 3,000 kilograms per TABLE 1. SOIL CHEMICAL AND PHYSICAL CHARACTERIS- TICS TO 3O CM DEEP FROM UPLANDS ON THE ARANSAS NA- WNAL WILDLIFE REFUGE NEAR AUSTWELL, TEXAS Qrganic Textural components Depth matter (%) Textural (cm) pH (°/o) Sand Silt Clay class 0-8 5.6 0.21 92 3 5 Sand 8- 15 5. 7 0 .04 92 3 5 Sand 10-30 5.9 0.02 91 3 6 Sand hectare, regardless ofthe time of burning. Total standing fine fuel averaged 3,000 kilograms per hec- tare in the spring and about 6,000 kilograms per hec- tare in the fall. Running live oak, the major coarse fuel, yielded 2,200 to 3,000 kilograms per hectare at ‘ll burning dates. Other woody species did not con- tribute significantly to fuel loads except in small iso- lated areas. In general, the dormant grasses and mulch burned readily and ignited live oak stems 1 meter or less in height. However, very few of the large live oak trees (7 meters or taller) were damaged by the fires. Likewise, mottes of plants greater than 1 meter but less than 7 meters tall did not burn uni- formly or completely. Environmental Conditions During Burning The burns in 1974 and 1976 were conducted under partly cloudy skies with north winds averag- ing 12 kilometers per hour and gusting to 19 kilomet- ers per hour. Air temperature during the burns was approximately 29 degrees centigrade. Relative humidity averaged 69 to 75 percent during the burns in 1974 and 1976. Soil moisture content averaged 6 percent in the upper 15 centimeters and at the time of the burn in 1974, and surface (2 centimeters) soil temperature averaged 21 degrees centigrade. Soils were saturated and scattered pools of standing water occurred during the burn in 1976. The surface soil temperature immediately prior to the burn was 19 degrees centigrade. According to Refuge records, the spring burn (1975) was conducted on a bright day with air temperature of 26 degrees centigrade and a wind speed of 19 kilometers per hour. Green Grass Standing Crop Annual rainfall distribution on the Coastal Prairie is usually characterized by peaks in the spring to early summer and during the fall. Thus, these "periods were chosen for evaluation of vegetation re- covery following burning in this study. Although rainfall for 60 days following the burn in the fall 1974 totaled almost 53 centimeters, winter rainfall was low (about 12 centimeters total for December, January, a. .1 February) (Table 2). Moreover, only 1.1 cen- timeters total of rainfall were received during March and April. Therefore, green grass standing crop in "il 1975 was significantly lower on the area burned 1.. the fall of 1974 than on unburned, thicketized savannah (Figure 1). Rainfall totaled over 20 cen- timeters for May and June 1975 (Table 2), and by July TABLE 2. MONTHLY RAINFALL (CM) RECEIVED AT AUST- WELL, TEXAS, NEAR THE ARANSAS NATIONAL WILDLIFE REFUGE FROM 1974 THROUGH 1977 Year Month 1974 1975 1976 1977 Avg. January T1 4.4 1.2 7.3 3.2 February 2.8 1.9 0.2 3.6 2.1 March 7.6 0.3 1.5 3.0 3.1 April 1.3 0.8 11.7 11.5 6.3 May 6.1 11.8 16.4 21.8 14.0 June 7.9 8.4 18.8 11.1 11.6 July 0.8 6.7 27.5 2.5 9.4 August 14.0 16.9 3.1 5.0 9.8 September 16.5 10.6 9.1 10.8 11.8 October 8.8 7.2 14.1 8.1 9.6 November 43.8 3.9 11.1 13.8 18.2 December 5.4 5.2 20.3 0.4 7.8 TOTAL 115.0 78.1 135.0 98.9 106.9 ‘Trace. grass standing crop on areas burned the previous fall tended to exceed that of the unburned area (Figure 1). Soil water content was also higher on the burned than on the unburned areas in July 1975 (Table 3)., Cumulative rainfall for the period July through November exceeded 45 centimeters (Table 2). Con- sequently, standing crop of green grass was no dif- ferent between burned and unburned areas in De- cember although there was a tendency for higher TABLE 3. SOIL-WATER CONTENT (°/o) AT THREE DEPTHS (CM) AT SEVERAL DATES IN 1975, 1976, AND 1977 ON THE UNBURNED AND BURNED STUDY AREAS OF THE ARANSAS NATIONAL WILDLIFE REFUGE NEAR AUSTWELL, TEXAS Collection dates 5°11 1975 1976 1977 depth (cm) July December February April July February I Unburned 0-8 10 16 14 9 27 38 8-15 10 13 ' 11 8 24 30 15-30 10 17 12 8 25 30 Avgl 10 16 12 8 25 32 Fall burn, 1974 0-8 23 13 5 10 19 14 8-15 15 10 8 a 10 16 18 15-30 15 11 7 7 22 24 Avg 17 11 7 9 20 20 Spring burn, 1975 0-8 18 12 9 8 14 36 8-15 14 10 9 7 10 30 15-30 14 10 8 7 14 26 Avg 15 11 9 7 13 30 Fall burn, 1975 0-8 10 7 16 34 8-15 12 10 16 29 15-30 11 10 18 26 Avg 11 9 17 30 ‘Weighted average across soil depths. ZRebumed in November 1976. 2700 UNBURNED BURNED I800 " /\- / . . /_ 2400 '- __ , ___ , FALL, I974 AND 197s /\ _ __ __ SPRING, I975 2'00 . . . . . . .. FAl-l-J975 l ‘k \ \_\ '5 a '3 o. € a? a. O ‘5 I500 - o z g I200 ' . as: o’ " 1: I500 - \ l- =E=E 555i =5=£ \ x 50o - § o \ _ y -'== s :\\ ==:= NE FEB. I976 APRIL I976 JULY I976 E VALUATION DATES contribution of standing dead material by the coarse perennial grasses. Vegetation burned in the fall of 1974 followed the same trend as on the area burned in the spring until it was reburned in the fall 1976. These data reinforce the interpretation of botani- cal composition changes that burning could be re- peated at about 3-year intervals, rainfall permitting, to maintain herbaceous stands at a successionally op- timum level. Grazing use would eventually reduce the standing dead herbaceous vegetation on un- burned areas and would be expected to delay its ac- cumulation on burned areas. Density of Running Live Oak Since live oak spreads by vegetative reproduc- tion from lateral roots, most of the study area is typified by scattered mottes of live oak connected by dense stands of oak stems less than 1 meter tall. Re- sponse of running live oak in the interspaces among the mottes was emphasized in this study since it cov- ered the majority of the area. Figure 3. Mulch (kg/ha) at various dates’ after burning thicketized savannah on the Aransas Na- tional Wildlife Refuge near Austwell, Texas. As- terisked bars indicate that mean values are signifi- . 95-percent level. NE indicates area not evaluated. FEB. I977 Initial top kill of live oak after burning the inter- spaces exceeded 95 percent, regardless of the date of burning. However, stem densities at 6 to 12 months after burning exceeded preburn stem densities (Fig- ure 5). Burning apparently breaks apical dominance which allows rapid development of new live oak shoots from the lateral roots. Live oak stem density by February 1976 significantly increased after burn- ing in the fall of 1974 and following burning in the spring of 1975, compared to that of unburned areas. By April 1976, stem density on areas burned in the fall of 1975 had increased significantly, compared to preburn densities. By April 1976, the rate of increase in live oak stem densities began to stabilize on all areas except the most recent burn installed in the fall of 1975 (Fig- ure 5). During July 1976, a decrease in density of a green live oak stems was recorded on areas burned during the fall of 1974 or spring of 1975. Apparently, the mulch accumulations, competition for moisture, and the effects of shading from herbaceous species, Figure 4. Total herbaceous standing crop (kg/ha) nah on the Aransas National Wildlife Refuge nea. Austwell, Texas. Asterisked means within a date of evaluation are significantly different from that of the unburned area at the 95-percent level. 7 _____ UNBURNED _ _,_ , FALL, I974 AND 191s __ 5 _ _ _ sPRlNe, r975 f: . _ . _ _ _ . FALL,l975 O 5 \- Z d 4 " 8 I / ,9: I 3 I- 1!‘ Z ZN‘ 5 12-2 2-m- ‘TEX? ‘T T. g u } . . g i‘) . i141 \ B2“ :3 / / \ I - / - / O l l l l l . April July Dec Fab. April July Feb. l I975 I976 E VALUATION DATES 12 cantly different from that of unburned areas at the at various dates after burning thicketized savan- -» 0 ‘ 3 b . ’ . ‘T? w’: a!" ‘T’ ,7'\. ' . a 2- / -~ \ r\ v'c \ _v Q O 0 ‘Pi QnIII i" é g | p f ‘kali’ ° T'- H o ° * Lu o O ~ _ ,- > e ‘m --- - 3 w \ / g ..| - \ ..___;___ UNBURNED Z 35 BURNED p w V ._. . _._ . - FALL, |974 AND |975 Figure 5. Density (thousands of stems/ha) of live (9 "2 7 _ __ _ __ sPRlNs, I975 oak on thicketized live oak savannah burned at Z _ _ _ _ _ _ . _ _ FALL,|975 various dates on the Aransas National Wildlife g _3 _ Refuge near Austwell, Texas. Asterisked means o indicate significant difference from unburned areas at the 95-percent level within an evaluation -4 l l l I date. Preburn April Feb. April July Feb. I975 I976 I977 EVALUATION DATES especially mid- and tall grasses, caused the decrease in live oak densities. Several factors must have in- teracted to cause reduction in the densities of live oak stems on the areas burned in the fall of 1974 or spring of 1975, rather than solely the dry summer growing conditions during 1976. Stem densities on the area burned during the fall of 1975 increased during that time period. Since the area burned in the fall of1974 was reburned in the fall of 1976, no additional live oak counts were taken. Height Change of Running Live Oak Live oak heights in the interspaces among the mottes on the unburned area were relatively stable during the study period with minor differences at- tributed to sampling error (Table 7). Regrowth rate of live oak burned in the spring of 1975 was somewhat more rapid than following burning in the fall of 1975. However, live oak stems on burned areas were shorter at all evaluation dates than those on un- TABLE 7. MEAN HEIGHTS (CM) OF LIVE OAK STEMS 1 METER IN HEIGHT OR LESS ON THICKETIZED LIVE OAK SAVANNAH BURNED AT VARIOUS DATES ON THE ARANSAS NATIONAL FWILDLIFE REFUGE NEAR AUSTWELL, TEXAS Evaluation datesl 1976 1977 Burn date February April July February Unburned 45 52 41 39 . .111 19742 16 19 23 3 Spring 1975 23 29 3O 24 fall 1975 4 12 16 17 . means within dates from burned areas were significantly different at the 95% level from those of the respective unburned area. ZReburned in November 1976. 3Regrowth was not sampled after the reburn in November 1976. burned areas, regardless of date of burning. There- fore, live oak regrowth would not be expected to achieve preburn heights within 24 months after burning. Percentages of regrowth stems in height classes illustrate the progression of stem elongation with time after burning (Table 8). The percentage of live oak stems in the taller classes on the area burned in the spring of 1975 were roughly equivalent to that recorded on areas burned in the fall of 1974, regard- less of evaluation date. These data indicate that live oak regrowth rate is most rapid during the spring. Consequently, time of burning, fall or spring in this study, will probably have little effect on regrowth heights after the first growing season following burn- TABLE 8. PERCENTAGE OF LIVE OAK STEMS BY HEIGHT CLASS IN FEBRUARY, APRIL, AND JULY OF 1976 AFTER BURN- ING ON SEVERAL DATES ON THE ARANSAS NATIONAL WILDLIFE REFUGE NEAR AUSTWELL, TEXAS Evaluation Height class (cm) date Burn date (1976) 0-5 6-10 11-20 21-40 Z41 Unburned February 0 1 6 30 63 April 0 0 4 27 69 July 0 3 11 28 59 Fall 19741 February 9 3 42 38 8 April 1 13 42 35 0 July 1 5 30 47 17 Spring 1975 February 4 12 31 32 21 April 1 13 21 42 23 July 0 5 20 47 28 Fall 1975 February 72 24 4 0 0 April 13 33 40 14 0 July 11 16 52 20 1 lReburned in November 1976. 13 w; IIIIIIHIIMIIHHHIIIIIIIIII Aiua as 37 ing. The percentage of stems in the height classes generally decreased with time after burning indicating a well-defined, maximum regrowth height for running live oak. Standing Crop of Running Live Oak Although density of stems was relatively uni- form, standing crop of live oak was highly variable over the study area. Therefore, the primary indicator of the effects of burning was the percentage of stand- ing crop made up of new growth. In April 1976, only 2 percent of unburned live oak was new growth, and 11 percent of the total weight was contributed by new growth in July 1976 (Table 9). In contrast, new growth afforded 16 and 26 percent of the live oak standing crop in April and July 1976, respectively, after burning in the fall of 1974. Although the total standing crop on areas burned in the spring of 1975 was greater at both evaluations in 1976 than on areas burned in the fall of 1974, the percentages of the live oak standing crop represented by new growth were similar. The most recently-burned area (fall of 1975) supported the least total standing crop but the greatest percentage of live oak standing crop as new growth on each sampling date. The area burned in the fall of 1975 was heavily utilized by wildlife, espe- cially white-tailed deer which browsed as many as 5O percent of newly emerged live oak sprouts during the 1976 growing season. Springer (1977) reported that browse was the most important food, as a percentage of the annual diet, of white-tailed deer on the Aran- sas Refuge. Moreover, live oak was the most com- monly occurring browse in their diet, especially dur- ing the winter months. DISCUSSION Prescribed burning generally improved vegeta- tion of thicketized live oak savannah for use by live- stock and wildlife. Few of the large oak (7 meters or taller) trees were damaged by the fires. Likewise, scattered mottes of plants 1 to 7 meters tall were not effectively burned, apparently because of lack of a continuous distribution of fine standing fuel. Coarse understory fuel (shrubs and vines) carried fires at ground level through some mottes. Occasional crown fires resulted from a combination of high fuel loads in surrounding vegetation, adequate combustible climbing plants in the crowns of these trees or mottes, and moderate to high (16- to 48-kilometers- per-hour) gusts of wind. However, grasses, forbs, and live oak plants less than 1 meter tall in the inter- spaces among the mottes burned completely and uni- formly except on a few isolated small areas, usually swales, where the fuel was discontinuous. Accumulations of mulch and dead standing plants on the unburned area apparently acted as a buffer against variations in the microenvironment with changes in weather. Burning removed accumu- lated mulch and standing dead herbaceous material, leaving mostly ash and mineral soil which promoted early warming of the soil in the spring. Thus, growth 14 5 b b E 7 TKBBLSE? TOTAL STANDING CROP o1= LIVE OAK AND NEW GROWTH (KG/HA) AND STANDING CROP INCREASE (9/6) REPRESENTED BY NEW GROWTH IN APRIL AND JULY, 1‘ ON AREAS BURNED AT VARIOUS DATES ON THE ARANS NATIONAL WILDLIFE REFUGE NEAR AUSTWELL, TEXAS‘ New growthz Total standing crop2 kg/ha °/o of total3 Burn date April July April July April July Unburned 880 1970 6O 215 2 11 Fall 1974 300* 1020* 50 260 16 26 Spring 1975 1200* 2700* 220* 570* 18 21 Fall 1975 230* 390* 110 120 47 30 lOnly live oak less than 1 m in height at the onset of the study were included in these measurements. ZAsterisked means within a column are significantly different from means of unburned areas at the 95-percent level. 3Statistical analysis not conducted. of forbs and grasses in the spring was slower on the unburned area than on burned areas, except when the burns were followed by below average winter and spring rainfall. Nutritional stresses incurred by wildlife and livestock in the winter might be relieved earlier in the spring on burned areas than on un- burned areas, especially in years of average or higher rainfall. Forb production was stimulated by burning, re- gardless of time of treatment, but was most respon- sive to fires in the fall. Difference in forb production between spring and fall burning was attributed to direct fire damage from spring fires. Maximum forb standing crop occurred from mid-spring to mid- summer, regardless of burning treatment. Burning in the fall or spring also increased numbers of forb species and promoted the occurrence of legumes, especially plains wildindigo and American snout- bean. More diverse forb communities provide a broader range of nutritious forage for selection by wildlife and domestic grazing animals. Burning in . the spring, as opposed to burning in the fall, would not be advisable when maximum numbers of forb species and maximum forb production early in the year are desirable for improvement of wildlife hab- itat. ' Standing crop of green grass followed predicta- ble seasonal fluctuations on the burned areas with reductions during the winter months and increases during the spring and summer. However, burned areas generally produced more green grass than did unburned areas during the summer months. This was attributed to mulch accumulations and the standing dead plants which reduced space and light for initiation of new growth on unburned areas. Ear; lier initiation of grass growth the growing seast . after burning on burned than unburned areas was attributed to earlier warming of the blackened soil surfaces. However, standing crop in the spring v greater on unburned areas than on burned rangelana when rainfall was below average during the winter and early spring. This was especially true in April 1976 after dry conditions the previous winter. Ap- "rently, the mulch cover insulated the soil surfaces .nburned areas against moisture loss. In addition, regrowth on burned areas may have exerted a greater demand for soil water than growth on unburned areas. Spring burning generally resulted in less grass production than did fall burning, apparently because of fire damage to newly emerging grasses in the spring. More species of grasses occurred for two grow- ing seasons on burned than on unburned areas. In general, grass species desirable for livestock grazing occurred more frequently and composed more of the forage by weight on burned than on unburned areas. However, improvement in botanical composition of the grass stand on these sandy sites was somewhat ower than observed in other studies on the Coastal Prairie after burning clay sites. Maximum potential improvement in the species composition of grass stands on these sandy sites may not be achieved until several burns have been applied. Grasses on burned areas were more succulent and more available to grazing animals after the dead standing herbaceous material was removed. Preburn densities of live oak averaged approxi- mately 152,000 stems per hectare, with a range of 68,000 to 276,000 stems per hectare, on the study area. Live oak densities were considerably lower on the unburned areas the growing season following the fires than on burned areas because the fires stimu- lated resprouting by breaking apical dominance. Also, the removal of the mulch and standing her- baceous crop may have released some live oak re- growth. The apparent intolerance of live oak sprouts to shade was verified by observations within the oak mottes. The mottes with dense canopy cover (greater than 75 percent cover) contained few live oak sprouts, if any, in the understory. Mottes with open canopies (less than 50 percent cover) and few understory plants supported substantially more live oak sprouts than those with heavy canopies. Live oak regrowth stem densities increased for 2 years after burning and then declined to preburn den- sities. No causative agent for the mortality of the regrowth live oak stems was ascertained. Although a second burn might initiate the cycle again, sub- sequent burns should ultimately reduce live oak re- growth stem densities by increasing the competitive stress from herbaceous vegetation. ACKNOWLEDGMENTS The authors are grateful to Frank Johnson, Man- aaer, Aransas National Wildlife Refuge, for allowing of land and facilities for this research. Iirn Mutz ‘and Marlin Springer helped with much of the data collection. Iulia Scifres is gratefully acknowledged for drawing all figures and typing the manuscript. LITERATURE CITED Arny, A. C., and A. R. Schmidt. 1942. A study of the inclined quadrat method botanical analysis of pasture mixtures. Am. Soc. Agron. I 34:238-247. Bouyoucos, G. I. 1962. Hydrometer method improved for making particle size analysis of soils. Agron. I. 54:464-465. Biswell, H. H., A. M. Schultz, and I. L. Launchbaugh. 1955. Brush control in ponderosa pine. Calif. Agr. 9:3-14. Blakely, H. L. 1947. The role of brush in habitat improvement on the Aransas National Wildlife Refuge. Trans. N. Am. Wildl. Conf. 12:179-185. ' Cook, O. G. 1908. Change of vegetation on the South Texas prairies. USDA, Bur. Plan Ind. Cir. 14. 5 p. Daubenmire, R. 1967. Plants and Environment. A Textbook of Plant Autecology. Iohn Wiley and Sons, Inc., New York. 422 p. Dewitt, I. B., and I. V. Derby, Ir. 1955. Changes in nutritive value of browse plants following forest fires. I. Wildl. Manage. 19:65-70. Dyksterhuis, D. I. 1957. The savannah concept and its use. Ecol- ogy. 38:435-442. Gordon, R. A., and C. I. Scifres. 1977. Burning for improvement of Macartney rose infested Coastal Prairie. Texas Agr. Exp. Sta. B-1183. 15 p. Green, S. W. 1935. Effect of annual grass fires on organic matter and other constituents of virgin longleaf pine soils. I. Agr. Res. 50:809-822. I-Iarne, E. E. 1938. Some wildlife forest relationships. Trans. N. Am. Wildl. Conf. 3:376-380. Iones, F. B. 1975. Flora of the Texas Coastal Bend. Mission Press, Corpus Christi, Texas. 262 p. Levy, E. B., and E. A. Madden. 1933. The point method of pasture analysis. New Zeal. I. Agr. 46:276-279. Lynch, I. I. 1941. The place of burning in management of Gulf coast wildlife refuges. I. Wildl. Manage. 5:454-457. Mortensen, I. L. 1965. Partial extraction of organic matter. In C. A. Black (ed) Methods of Soil Analysis. p. 1405. (Part II). Amer. Soc. Agron., Madison, Wis. Saur, C. O. 1950. Grassland climax, fire and man. I. Range Man- age. 27:5-11. Scifres, C. I., and R. H. Haas. 1974. Vegetation changes in a post oak savannah following woody plant control. Texas Agr. Exp. Sta. MP-1136. 12 p. Springer, M. D. 1977. The'influence of prescribed burning on nu- trition of white-tailed deer on the coastal plain of Texas. Ph.D. Dissertation. Texas A&M Univ., College Station, Texas. 92 p. Vallentine, I. F. 1971. Range Development and Improvements. Brigham Young Univ. Press. Provo, Utah. 516 pp. Wright, H. A. 1974. Range burning. I. Range Manage. 27:5-11. 15 All programs and infonnation of The Texas Agricultural Experiment Station are available to everyone without regard to race, color, religion, sex, age, or national origin. The Texas Agricultural Experiment Station, Neville P. Clarke, Director, College Station, Texas 2.5M —- 11-79