o O " — O |“JNVIN0SH1IIMS S3 I dVaa II LIBRARI ES^ SMITHSONIAN-1 INSTITUTION Z NOIlfUl z r“ z > SMITHSONIAN INSTITUTION^NOliniliSNI ~ NVINOSHJLmS^S 3 IHVa8n_LIBRAR 00 _ _ 2 * c n 2 co 2 X CO o Z , i^NviNOSHiiws^sa i ava a 11 li b rar i es^smithsonian institution Nounii 2 \ 01 _ co z: co CO CO O Xijvosv^ __ o -J z _j z -j SMITHSONIAN INSTITUTION NOIJinillSNI NVIN0SH1IINS S3iavaail LIBRAF m co = co NViNOSHims saiavaaii libraries Smithsonian institution Nounii 2 ^ _ CO 2 .... CO 2 CO < 5 . , ,/S ^ * E ' s X CO o c z v 9n d.c^x > :> ~ "X > CO Z co * 2 co * 2 5 SMITHSONIAN INSTITUTION N0I1D1I1SNI N.VINOSHilWS S3iavaail LIBRAF co x co — co o o v _ Xi!iX O l‘JNVINOSHimS Saiavaail^LIBRAR I E S^ SMITHSONIAN^ INSTITUTION NOIlfUl r~ 70 ~ WZ m x^vosv^x ~ rn co — co _ co — 5 SMITHSONIAN INSTITUTION NOIKUIISNI NVIN0SH1IWS S3 I a V a a 1 1 _ L 1 B R A F CO 2 I NVIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOIlfUl CO — in - “ co CO vinoshims saiavaan libraries Smithsonian institution NouniiiSNi z r~ z r“ z ' p CD A, J _ ^mAkWjr>/ f— 'Q. (7) — MITHSONIAN INSTITUTION NOUfllllSNI NVINOSH1IWS SSiavaail LIBRARIES Z \ CO z & .. ^ CO Z CO Z tO VIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSN CO “ CO ~ CO CO /5^\ w CO / *W~' S 2 ~\ > z CO * z co * z MITHSONIAN INSTITUTION NOIlfUllSNI NVIN0SH1IWS S3iavaail LIBRARIES to — CO CO O __ O v _ O VIN0SH1IINS S3 lavaan^LIBRARI E SZ SMITHSON IAN-1 1 NSTITUTION^ NOlinillSN v Z r~ Z ~ - O ✓'■'T n iTX O s'TTri O , MITHSONIAN INSTITUTION NOIinillSNI NVIN0SH1HNS S3iavaail LIBRARIES z * co z co z \ < V s E - u 2 h wW/#s- 2 ~l £ o x f/w^Ws o f$£y i 9 CO Sx CO to plL ;«4$| CO yZ' to g i. ? % ai ° ^ 1 CO ■'•*" Z CO z CO iviNOSHims S3 iava a ii libraries smithsonian institution NoiinmsN co ~ co • — CO HA ITUPAM I A A THE rEXAS JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP. — Any person or group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. Dues for members are $20.00 annually; student members, $12.00 annually; sustaining members, at least $30.00 in addition to annual dues; life members, at least $400.00 in one payment; patrons, at least $500.00 in one payment; corporate members, $250.00 annually; corporate life members, $2000.00 in one payment. Library subscription rate is $45.00 annually. Payments should be sent to Dr. Michael J. Carlo, P.O. Box 10986, Angelo State University, San Angelo, Texas 76909. The Texas Journal of Science is a quarterly publication of the Academy and is sent to all members and subscribers. Inquiries regarding back issues should be sent to Dr. William J. Clark, Drawer H6, College Station, Texas 77844. AFFILIATED ORGANIZATIONS Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers American Association for the Advancement of Science The Texas Journal of Science (USPS 616740) is published quarterly at Lubbock, Texas U.S.A. Second class postage paid at Post Office, Lubbock, TX 79401. Please send form 3579 and returned copies to Texas Tech Press, Box 4240, Lubbock, TX 79409. ISSN 0040-4403 THE TEXAS JOURNAL OF SCIENCE Volume 39, No. 1 February 1987 CONTENTS Reproductive and lipid patterns of a semiarid-adapted anuran, Bufo cognatus. By David R. Long . 3 A survey of the lead distribution in the soil of Corpus Christi, Texas. By George Harrison . 15 A study of the Cu(H20)6+/CuCl4 /ethanol system for solar energy storage. By L. Gene Spears, Jr., Larry G. Spears, and Joycelyn C. Spears . 23 Relative mobility of lead and copper in soils: an example from the Bonanza District, Saguache County, Colorado. By Joseph C. Cepeda . 29 Influence of environmental factors on oxygen consumption of Clibanarius vittatus (striped hermit crab). By V A. Wolfenberger . . . 37 Diurnal activity patterns of desert mule deer in relation to temperature. By Bruce D. Leopold and Paul R. Krausman . . . . . . 49 Limnological characteristics of Aquilla Lake, Texas, during impoundment. By Michael J. Maceina and Mary F. Cichra . 55 Pollen analysis of late-Holocene sediments from a central Texas bog. By Richard G. Holloway, L. Mark Raab, and Robert Stuckenrath . 71 Habitat and population destruction and recovery in the parthenogenetic whiptail lizard, Cnemidophorus laredoensis (Sauria: Teiidae), in southern Texas. By James M. Walker . . . 81 Long-term response of live oak thickets to prescribed burning. By Marlin D. Springer, Timothy E. Fulbright, and Samuel L. Beasom . . 89 Noteworthy records of mammals from the Texas Panhandle. By Robert R. Hollander, J. Knox Jones, Jr., Richard W. Manning, and Clyde Jones . 97 General Notes Range and habitat expansion of the introduced slug, Angustipes ameghini, in extreme southern Texas. By Raymond W. Neck . 103 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Editor: J. Knox Jones, Jr., Texas Tech University Assistant to the Editor: Marijane R. Davis, Texas Tech University Associate Editor for Botany: Randy Moore, Baylor University Associate Editor for Chemistry: Marvin W. Rowe, Texas A&M University Associate Editor for Computer Science: Ronald K. Chesser, Texas Tech University Associate Editor for Mathematics and Statistics: George R. Terrell, Rice University Associate Editor for Physics: Charles W. Myles, Texas Tech University Editorial Assistants: Robert R. Hollander, Texas Tech University Richard W. Manning, Texas Tech University Scholarly papers in any field of science, technology, or science education will be considered for publication in The Texas Journal of Science. Instructions to authors are published one or more times each year in the Journal on a space-available basis, and also are available from the Editor (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409, 806/742-2487, Tex-an 862-2487). REPRODUCTIVE AND LIPID PATTERNS OF A SEMI ARID-ADAPTED ANURAN, BUFO COGNATUS David R. Long Department of Biological Sciences, Texas Tech University, Lubbock, Texas 79409 Abstract. — Patterns of lipid storage and reproduction of Bufo cognatus on the Llano Estacado of northwestern Texas were examined over three years. Seasonal variation in fatbody mass occurred only in females and resulted from changes in the lipid content of the ovaries during dormancy. Histological examination of testes showed a temporal spermatogenic pattern, but no seasonal variation was evident in testis mass, volume, or seminiferous tubule diameter. Ovarian mass showed annual variation, but large follicles were present throughout the activity season. These patterns indicate reproductive readiness throughout the activity season. Liver lipid-free dry mass but not liver lipid mass increased by the end of the activity season suggesting a nonlipid component was the primary energy source for metabolism during dormancy. These patterns of reproductive activity and lipid utilization appear to be physiologically advantageous for an amphibian having moisture-dependent activity regimes but inhabiting a relatively warm and semiarid environment. Key words. Anura; Bufo cognatus ; fatbody; lipid; reproduction. Lipid is an efficient storage form of potential energy and the strategy employed by a species in allocating stored energy is an important consideration to workers attempting to understand the evolution of reproductive tactics in ectothermic vertebrates. Consequently, annual lipid cycles of species have been the focus of numerous studies (see Derickson, 1976, Fitzpatrick, 1976, and Pond, 1978, for reviews). In sexually mature aquatic and mesic anurans, body lipid is typically low during dormancy as a result of metabolic use, and is used during reproductive activity for vitellogenesis and possibly for steroid production. An increase in stored body lipid occurs after breeding in preparation for winter dormancy (Bush, 1963; Mizell, 1965; Brenner, 1969; Byrne and White, 1975; Chieffi et al., 1975; Pierantoni et al., 1983). Reproductive organs also show seasonal variation with periods of recrudescence and quiescence synchronized around the breeding season (Mizell, 1964; Byrne and White, 1975; Pierantoni et al., 1983). Arid and semiarid environments present osmotic problems for terrestrial amphibians and usually necessitate brief periods of activity that are coincident with high moisture levels. Because the duration of the activity season and the time available for foraging within the season are major determinants of how much lipid can be stored by an individual in a given year, a study of a species that survives in a semiarid habitat might demonstrate patterns of reproduction and lipid use that are modified to accommodate short and sporadic periods of annual activity. The semiarid environment and unpredictable summer rainfall of the Llano Estacado of northwestern Texas is relatively harsh for terrestrial The Texas Journal of Science, Vol. 39, No. 1, February, 1987 4 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 amphibians. There, the Great Plains toad, Bufo cognatus , is intermittently active in spring and summer during times at which moisture and temperature conditions are favorable for activity between dry periods. Bufo cognatus is common throughout western Texas and its habitat generally encompasses areas devoid of permanent bodies of water. For this reason, Bufo cognatus was used to assess the lipid storage/ mobilization strategy employed by an amphibian that has moisture-restricted activity periods. Initial data collected in 1983 indicated this species does not demonstrate a pattern of lipid mobilization typical of temporate zone ectothermic vertebrates (Long, 1985). Therefore, data were collected for three years to examine reproductive and lipid patterns of B. cognatus and to compare them to the previously described patterns of mesic anurans. The following hypotheses were tested: 1) Bufo cognatus will have low lipid stores on emergence from winter dormancy with a net increase in stored lipid prior to entering the next hibernation period, and 2) male and female gonadal states will change through the activity season with gamete development in the spring and gonadal quiescence at the end of breeding activity. Materials and Methods Bufo cognatus was collected throughout its activity season in 1983, 1984, and 1985 from an unwatered city park and surrounding roads of Lubbock, Texas. Toads were collected at nights, returned to the laboratory, and refrigerated at 7°C. In 1984 and 1985, male toads were pithed within 48 hours of collection and testes were placed in Bouin’s fixative. All individuals were frozen for later processing. In this paper, I examine reproductive and lipid patterns in mature individuals only. Snout-vent length (SVL) was measured for all toads prior to removal of fatbodies, liver, ovaries, and oviducts. The carcass, minus these organs, was oven-dried at 75° C to constant mass. The diameter of 10 of the largest ovarian follicles from each mature female was measured with calipers in 1984 and 1985 to estimate average mature follicle size. Females without developed ovaries or enlarged oviducts were classified as immatures. Ovaries, oviducts, and livers were weighed for wet mass and oven-dried to constant mass. Fatbodies were weighed in two groups per individual — carcass fatbodies (four pairs of subcutaneous fatbodies and a fatbody anterior to the heart) and gonadal fatbodies. Carcasses and ovaries were ground prior to lipid extraction. Mass was measured to the nearest 0.001 gram. Lipid extraction of carcasses (minus fatbodies), livers, ovaries, and oviducts was made using chloroform:methanol (2:1, volume/ volume — Folch et al., 1957) at 70°C in a Soxhlet apparatus for eight hours (livers, oviducts) or 10 hours (carcasses, ovaries). Extraction times were sufficient as indicated by the dry, nongreasy, powdery consistency of the crushed extracted tissue and the solvent was clear when extractions were terminated. Lipid mass of tissue was determined as dry mass minus lipid-free dry mass. Testes from toads collected in 1984 and 1985 were embedded in paraffin, sectioned at 7 jum, and stained with Delafield’s hematoxylin and eosin. Sexual maturity of males from 1984 and 1985 was confirmed by presence of spermatogenic activity. Sexual maturity of males collected in 1983 was based on comparison of SVL of mature toads collected in 1984 and 1985. During the 70 percent ethanol step of dehydration, maximum length and width of testes (minus Bidder’s organ) were measured and combined testis mass (including Bidder’s organ) was obtained. Mean testis volume was determined using the equation for determining the volume of a column (volume = 7rr2h, where r = mean testis width/ 2 and h = mean TOAD REPRODUCTIVE AND LIPID PATTERNS 5 testis length). Sectioned testes were qualitatively assessed for spermatogenic activity. In addition to mass and volume, quantitative assessment of testis activity included obtaining the mean diameter of 10 seminiferous tubules per individual. Toads were assigned to one of five groups based on reproductive patterns of individuals in the year they were collected. The five seasonal groups were determined as follows: 1) posthibernation, premating period (May to first breeding night); 2) the first mating night of each year (7 June 1983, 13 June 1984, 22 May 1985); 3) period of mating activity (after first breeding night through mid-July); 4) period when temperature was still favorable for mating but mating was not observed (late July through August); 5) postmating, prehibernation period (September). Toads collected the first mating night (group 2) were considered premating because they were collected early that night (2100 to 2330 hours), no females were ovipositing, and no eggs were found in the water. Analysis of covariance (ANCOVA) procedures, using SVL as the covariate, were used to test lipid and reproductive data for intergroup variation. Data to be used in ANCOVA were subjected to Hartley’s F-max test to test for homogeneity of group variances (Sokal and Rohlf, 1981) and group-covariate interactions were examined. Natural logarithmic transformation of fatbody mass, carcass lipid mass, liver dry mass, liver lipid mass, and liver lipid-free dry mass for each sex was performed to eliminate heteroscedasticity and group- covariate interaction prior to ANCOVA procedures. Liver lipid-free dry mass was used as the covariate when intergroup variation in liver lipid mass was examined. Following ANCOVA analysis, a priori contrast was made between early (groups 1 and 2) and late (groups 4 and 5) collection groups. This contrast was based on the assumption that if seasonal variation occurred, it would be most evident between early and late samples; thus verifying the variation that was anticipated to occur over dormancy. Linear regression analyses were used to test for relationships between continuous variables. The Statistical Analysis System (SAS) package (SAS Institute, Inc., 1982) was used for all statistical procedures. Statistical significance was determined at P < 0.05. Means are presented plus or minus two standard errors. Results Bufo cognatus initiated activity in May. Few toads were observed on any given night prior to the breeding season. In late May and early June, warm temperatures combined with rain or high humidity stimulated males to aggregate and chorus at local ponds and ephemeral playas. Females were not as numerous as males at breeding sites. By mid-July, breeding activity terminated and toads were again scarce. Mature individuals were no longer observed after late September. Mature males averaged 81.0 mm SVL (n, 91; SD, 11.3 mm; range, 52.0-108.0). Mean female SVL was 94.1 mm (n, 51; SD, 9.6 mm; range, 78.0-112.0). Males There was no variation (ANCOVA with covariate SVL) in log transformed fatbody mass (FBM) (F 4,84 = 0.31, P = 0.87) or log carcass lipid mass (CLM) (F 4,84 = 1.02, P = 0.40) among the five groups (Fig. 1). This lack of seasonal variation in fatbody mass is unusual for anurans. ANCOVA for log transformed liver values indicated intergroup variation for all analyses (Table 1). Contrasts between early and late samples indicated liver lipid-free dry mass (LFDM) was larger at the end of the 6 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure 1. Seasonal pattern of SVL adjusted lipid stores in male Bufo cognatus. Solid circles = fatbody mass; open circles = carcass lipid mass (minus fatbodies). Mean ± 2 SE. Sample sizes are indicated for each group. See Materials and Methods for group designation. activity season (F 1,84 = 7.77, P< 0.01). Although liver lipid mass, adjusted for liver LFDM, varied significantly among groups, the early-late contrast comparison was nonsignificant (F 1,84 = 0.01, P= 0.92) and no biological explanation for the intergroup variation could be discerned. Mean testis mass, volume, and seminiferous tubule diameter (when adjusted for SVL) remained constant through the activity season (Table 2). Seasonal trends were evident in the spermatogenic process (Fig. 2). Toads collected early in the year (group 1) had large numbers of spermatozoa arranged in bundles or had free spermatozoa in the lumen of the seminiferous tubules (Fig. 2A). Few spermatocytes or spermatids were present in these testes. Toads in groups 2 and 3 showed similar concentrations of bundled and free spermatozoa along with an increased number of primary spermatocytes (Fig. 2B). Fewer spermatozoa were present in group 4 (postreproductive group) (Fig. 2C). In groups 4 (Fig. 2C) and 5 (Fig. 2D), large numbers of germinal cysts containing all spermatogenic stages were present, and most testes exhibited small concentrations of bundled spermatozoa associated with Sertoli cells. The absence of quantitative variation in the testes might have occurred because of the constant replacement of germinal cysts which allowed mature spermatozoa to be available during much of the activity season. Females Seasonal variation in log FBM, but not log CLM, was revealed by ANCOVA with covariate SVL (FBM: F 4,45 = 2.55, P = 0.05 and CLM: F 4,45 = 1.63, P = 0.18) (Fig. 3). A pattern of increased fatbody mass by the end of the season was suggested in Figure 3 and comparisons between early and late groups confirmed a seasonal increase in FBM (F 1,45 8.08, P< 0.01) but not in CLM (F 1,45 = 0.84, P = 0.36). TOAD REPRODUCTIVE AND LIPID PATTERNS 7 Table 1. Seasonal group values for male Bufo cognatus liver parameters. Liver dry mass1 (g) Liver lipid mass2 (g) Liver LFDM1 (g) Group n x 2 SE x 2 SE x 2 SE 1 17 0.450 0.198 2 23 0.308 0.161 3 15 0.650 0.205 4 20 0.613 0.186 5 15 0.410 0.209 0.093 0.064 0.141 0.055 0.079 0.069 0.022 0.063 0.121 0.069 0.369 0.127 0.232 0.103 0.511 0.131 0.521 0.119 0.323 0.133 ANCOVA F 4,84 = 3.64 P< 0.01** 6.63 P< 0.01** 5.31 P<0.01** Significant differences at 0.05. ANCOVA based on log transformed values. ‘Adjusted for snout-vent length. 2Adjusted for liver lipid-free dry mass (LFDM). Log transformed liver values varied among groups (Table 3). Early as compared to late group contrasts indicated an increase in liver LFDM by the end of the activity season (F 1,45 = 8.84, P < 0.01). As with males, the early as compared to late group contrast was not significant for liver lipid mass (F 1,45 = 2.15, P — 0.15), although intergroup variation was evident with ANCOVA. The highest female liver lipid value was in group 3, possibly representing an increase in vitellogenic activity to replace ova voided during the breeding season. In support of this, the contrast between group 1 and group 3 was significant (F 1,45 = 6.58, P= 0.01). Ovary wet mass and oviduct dry mass varied in the activity season (Fig. 4). As anticipated, ovaries were largest in the beginning of the year and became significantly reduced in size after the start of the breeding season. Ovary mass began to increase by the end of the activity season. Log values Table 2. Seasonal group values for testis parameters of Bufo cognatus. All values adjusted for snout-vent length. These parameters were examined for specimens collected in 1984 and 1985, only. Testis mass (g) Testis volume (mm3) Tubule diameter (jum) Group n x 2 SE x 2 SE x 2 SE 1 5 0.072 0.022 2 6 0.073 0.020 3 6 0.081 0.023 4 20 0.064 0.011 5 3 0.085 0.034 80.4 26.0 84.6 23.5 96.7 26.9 83.5 12.8 97.4 38.3 297.0 43.1 249.7 39.0 281.4 45.4 270.3 22.2 258.3 60.8 ANCOVA F 4,34 = 0.89 P = 0.48ns 0.43 P = 0.78ns 0.80 /> = 0.53ns 8 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure 2 (A-D). Temporal spermatogenic pattern of Bufo cognatus, X 360. A) Postdormancy individual (group 1) with large concentrations of bundled and free spermatozoa (SZ); B) reproductive individual (groups 2 and 3) with large concentrations of spermatozoa (SZ) and primary spermatocytes (SC); C) postreproductive individual (group 4) with small concentrations of spermatozoa (SZ), but large numbers of primary spermatocytes (SC) and secondary spermatocytes transforming into spermatids (ST); D) predormancy individual (group 5) showing all spermatogenic stages. TOAD REPRODUCTIVE AND LIPID PATTERNS 9 GROUP Figure 3. Seasonal pattern of SVL adjusted lipid stores in female Bufo cognatus. Solid circles = fatbody mass; open circles = carcass lipid mass (minus fatbodies). Mean ± 2 SE. Sample sizes are indicated for each group. See Materials and Methods for group designation. of FBM and ovary wet mass were not correlated (r = 0.053, n = 49, P = 0.72). This was probably a result of follicle retention by females. Oviduct mass followed the same pattern as ovarian mass. Early as compared to late contrasts confirmed the significant reduction in mass of both structures (ovaries— F 1,43 = 26.54, P< 0.01; oviducts— F 1,42 = 15.97, P< 0.01). Diameter of the largest ovarian follicles did not show intergroup variation (F 4,31 = 1.36, P = 0.27), suggesting that mature follicles are available throughout the year. Average large follicle diameter was 1.03 mm ± 0.03 mm (number of females examined, 36). Ovaries contained 24.3 ± 5.3 percent lipid (n, 24) at their largest mass (groups 1 and 2). Ovary lipid mass increased with ovary LFDM (r2 = 0.891, n = 49, P< 0.01). Oviduct lipid mass increased with oviduct LFDM (r2 = 0.444, n = 49, P< 0.01). Early as compared to late contrast indicated a larger amount of lipid present in ovaries at the beginning of the season (when adjusted by ovary LFDM) (F 1,42 = 6.31, P — 0.02), implying the over-winter increase in ovarian mass also entails a significant increase in the ovary lipid content. The increase in adjusted ovary lipid mass from late to early groups was 0.351 gram. Variation in female FBM between late and early groups may be explained by lipid used in ovarian enlargement during dormancy. When ovarian lipid mass was added to female lipid stores, seasonal variation in female lipid mass was no longer evident (ANCOVA — F 4,45 = 1.77, P = 0.15; early as opposed to late contrast — F 1,45 = 0.15, P = 0.70). Male-Female Lipid Comparisons Although females possessed larger lipid stores than males (compare Fig. 1, 3), males actually had higher levels of FBM and CLM when SVL was 10 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 3. Seasonal group values for female Bufo cognatus liver parameters. Group n Liver dry mass1 (g) Liver lipid mass2 (g) Liver LFDM1 (g) X 2 SE X 2 SE X 2 SE 1 6 0.405 0.381 0.102 0.156 0.355 0.287 2 19 0.585 0.210 0.150 0.087 0.464 0.159 3 4 1.117 0.494 0.277 0.191 0.797 0.373 4 11 0.977 0.281 0.107 0.117 0.822 0.212 5 11 0.860 0.282 0.181 0.110 0.663 0.213 ANCOVA F 4,45 = = 3.24 2.98 2.78 P — 0.02* P = 0.03* P = 0.04* Significant differences at 0.05. ANCOVA based on log transformed values. 'Adjusted for snout-vent length. 2Adjusted for liver lipid-free dry mass (LFDM). used in ANCOVA to adjust for sexual dimorphism in body size (FBM — F 1,139 = 9.97, P< 0.01; CLM — F 1,139 = 51.66, P< 0.01). SVL adjusted mean values for males were 2.48 ± 0.51 grams (FBM) and 1.81 ± 0.13 grams (CLM), and for females were 1.05 ± 0.68 grams (FBM) and 0.94 ±0.18 grams (CLM). Lipid sequestered in the ovaries accounted for the lower levels of lipid for females. When FBM and CLM were summed and ovarian lipid mass added for females, no difference in total body fat content was evident between the sexes (F 1,138 = 2.44, P= 0.12). Discussion The a priori hypotheses presented in the introduction were falsified. Only mature female Bufo cognatus showed seasonal lipid store variation. Both sexes retained mature gametes throughout much of the activity season and no quanitified testis variation was evident. These patterns appear to be an adaptation to the environmental unpredictabilities present on the Llano Estacado of northwestern Texas. Individual variation in stored lipid, as evidenced by the large standard errors in Figures 1 and 3, is probably a result of variation in amount and types of available food. Dimmitt and Ruibal (1980) calculated that 11 to 12 feedings were necessary to provide metabolic requirements for one year in B. cognatus that were dormant 10 months of the year. The tendency to risk foraging during dry weather might vary among individuals and account for much of the within-group variation in the lipid reserve. Examining the reproductive pattern of this species seemed to provide some explanation for the patterns of lipid use. Males showed no seasonal variation in testis size, so no periodic demand for stored lipid was required for testicular recrudescence. Temporal patterns of spermatogenic activity were evident; however, spermatozoa and TOAD REPRODUCTIVE AND LIPID PATTERNS 11 GROUP Figure 4. Seasonal variation in ovary and oviduct mass of Bufo cognatus. Mean ± 2 SE. See Materials and Methods for group designation. spermatids were present in testes throughout most of the study. Maintaining even low concentrations of viable spermatozoa might be a reproductive adaptation allowing for mating preparedness throughout the activity season, particularly because mating time is contingent on appropriate temperature and moisture conditions (Bragg, 1937, 1940a, 1945). On the Llano Estacado, suitable environmental conditions are not predictable (NOAA, 1983); consequently, males appear to be prepared for mating much of the activity season. Females presented a similar situation. Ovaries were largest at the start of seasonal activity. During the breeding season, ovary mass dropped to its lowest point; however, females still retained mature follicles of ovulatory size. Ovary mass nearly doubled from this point to the end of the activity season, but was still significantly less than at the beginning. Ovary mass continued to increase during dormancy at the expense of the fatbody stores. Ingestion was probably the primary source of lipid for reproductive processes during the activity season; therefore, the demand on the lipid reserve during that time was minimal. Oviduct mass displayed a seasonal pattern similar to that of the ovaries because of the reduction of secretory products (jelly) associated with ovulation (Jorgensen and Vijayakumar, 1970). Because oviduct lipid mass increased with LFDM, the enlargement of these structures during dormancy also placed a burden on the lipid store. 12 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Van Beurden (1979) described patterns of gamete development similar to those of B. cognatus in an Australian desert anuran ( Cyclorana ). As in this study, retention of mature gametes throughout the year appeared to allow for breeding preparedness for times when favorable moisture conditions existed. Bragg (1940b) found ovulatory-size follicles in B. woodhousei late in its activity season, leading him to suspect that not all mature females oviposited in a given year. His observation might better be explained by the reproductive pattern demonstrated by Bufo cognatus. Females breed, but do not ovulate all mature follicles in a given year, and follicle maturation continues throughout the year. This allows continual breeding preparedness and spreads the demand for protein and lipid for ovarian enlargement over a broader period of time. Therefore, inadequate food ingestion during a dry period might not significantly affect ovarian development. Energy for metabolism during dormancy in apparently derived primarily from a nonlipid liver material (such as glycogen). Liver LFDM increased by the end of the activity season suggesting a nonlipid component was being stored and then used during dormancy. Liver lipid showed seasonal flux, but no net increase in lipid was evident by the end of the season. Male carcass lipid did not change over dormancy and the variation in females could be accounted for by ovarian lipid requirements. In conclusion, the amount of fat present in B. cognatus probably represents a net accumulation with age and remains seasonally constant when adjusting for individual variation in body size and accounting for ovarian lipid content. Any demands on the lipid reserve for metabolism during dormancy are negligible and a nonlipid material such as glycogen appears to be the main energy source at this time. Females use fatbody lipid for enlarging reproductive structures during dormancy, but most vitellogenic lipid during the active season is apparently obtained from ingestion. By retaining mature gametes throughout the activity season, both sexes are prepared for breeding anytime favorable climatic conditions exist. Acknowledgments I thank F. L. Rose, K. W. Selcer, and M. R. Willig for critical comments on this manuscript. I also appreciate the assistance provided by M. Scioli, K. Selcer, G. Henson, M. Bartlett and others in collecting toads during this study. The assistance of G. Henson in processing specimens is appreciated. This study was part of a dissertation submitted to the Graduate School of Texas Tech University in partial fulfillment of the requirements for the Doctor of Philosophy degree. Literature Cited Bragg, A. N. 1937. Observations on Bufo cognatus with special reference to the breeding habits and eggs. Amer. Midland Nat., 18:273-284. - . 1940a. Observations on the ecology and natural history of Anura. I. Habits, habitat and breeding of Bufo cognatus Say. Amer. Nat., 74:322-349, 424-438. TOAD REPRODUCTIVE AND LIPID PATTERNS 13 - . 1940b. Observations on the ecology and natural history of Anura. II. Habits, habitat, and breeding of Bufo woodhousei woodhousei (Girard) in Oklahoma. Amer. Midland Nat., 24:306-321. - . 1945. The spadefoot toads in Oklahoma with a summary of our knowledge of the group. II. Amer. Nat., 79:52-72. Brenner, F. J. 1969. The role of temperature and fat deposition in hibernation and reproduction in two species of frogs. Herpetologica, 25: 105-1 13. Bush, F. M. 1963. Effects of light and temperature on the gross composition of the toad, Bufo fowleri. J. Exp. Zool., 153:1-13. Byrne, J. J., and R. J. White. 1975. Cyclic changes in liver and muscle glycogen tissue lipid and blood glucose in a naturally occurring population of Rana catesbeiana. Comp. Biochem. Physiol., 50A:709-715. Chieffi, G., R. K. Rastogi, L. Iela, and M. Milone. 1975. The function of fat bodies in relation to the hypothalamo-hypophyseal axis in the frog, Rana esculenta. Cell Tiss. Res., 161:157-165. Derickson, W. K. 1976. Lipid storage and utilization in reptiles. Amer. Zool., 16:711- 723. Dimmitt, M. A., and R. Ruibal. 1980. Exploitation of food resources by spadefoot toads ( Scaphiopus ). Copeia, 1980:854-862. Fitzpatrick, L. C. 1976. Life history patterns of storage and utilization of lipids for energy in amphibians. Amer. Zool., 16:725-732. Folch, J., M. Lees, and G. H. Sloane Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem., 226:497-509. Jorgensen, C. B., and S. Vijayakumar. 1970. Annual oviduct cycle and its control in the toad, Bufo bufo L. Gen. Comp. Endocrinol., 14:404-411. Long, D. R. 1985. Fat body complement of bufonid toads: an adaptation for inhabiting fluctuating habitats. Presented paper at the Southwestern Association of Naturalists Annual Meeting, Glendale, Arizona. Mizell, S. 1964. Seasonal differences in spermatogenesis and oogenesis in Rana pipiens. Nature, 202:875-876. - . 1965. Seasonal changes in energy reserves in the common frog, Rana pipiens. J. Cell. Comp. Physiol., 66:251-258. National Oceanic and Atmospheric Administration (NOAA). 1983. Local climatological data. Annual summary with comparative data. Lubbock, Texas. National Climatic Data Center, Asheville, North Carolina. Pierantoni, R., B. Varriale, C. Simeoli, L. Di Matteo, M. Milone, R. K. Rastogi, and G. Chieffi. 1983. Fat body and autumn recrudescence of the ovary in Rana esculenta. Comp. Biochem. Physiol., 76A:31-35. Pond, C. M. 1978. Morphological aspects and the ecological and mechanical consequences of fat deposition in wild vertebrates. Ann. Rev. Ecol. Syst., 9:519-570. SAS Institute, Inc. 1982. SAS user’s guide: statistics. Cary, North Carolina. 584 pp. Sokal, R. R., and F. J. Rohlf. 1981. Biometry. W. H. Freeman and Co., San Francisco, 2nd ed., 859 pp. Van Beurden, E. 1979. Gamete development in relation to season, moisture, energy reserve, and size in the Australian water-holding frog Cyclorana platycephalus. Herpetologica, 35:370-374. Present address of author: Department of Biology, Wilkes College, Wilkes-Barre, Pennsylvania 18766. A SURVEY OF THE LEAD DISTRIBUTION IN THE SOIL OF CORPUS CHRISTI, TEXAS George Harrison U.S. Geological Survey, Quissett Campus, Woods Hole, Massachusetts 02543 Abstract. — Four-hundred and eighty-five soil samples taken in June 1984 within the city limits of Corpus Christi, Texas, had a mean lead (Pb) value of 208 parts per million with a range of eight to 2969 ppm. A baseline of 13 ppm was established for the community based on remote or virtually untrafficked areas within the city limits. The mean Pb value for samples taken in parks was 55 ppm, for school playgrounds, 57 ppm. All other samples, which were taken from the edges of roadways and freeways, had a mean of 250 ppm. Samples from the northern (oldest) section of the city showed a Pb value 19.5 times above the baseline value, whereas samples from the southern (newer) section of the city showed a Pb value 4.5 times above baseline. The expected correlation of high traffic-flow and increased Pb values was observed except for two instances — lower than expected Pb values around highly trafficked, elevated freeway interchanges, and an inverse relationship between traffic-flow rates and Pb values on a portion of the freeway system. Key words : tetraethyl lead; leaded gasoline; vehicular emissions; lead contamination; lead baseline values. Natural lead (Pb) concentrations can vary from two to 200 parts per million (ppm) depending upon geographical location (Swaine, 1955). Average concentrations of Pb in the soil may range from 10 to 37 ppm (Nriagu, 1978). The mean value for Pb in the earth’s crust is 17 ppm and 20 is considered the mean value by some for North America (Nriagu, 1978). The dominant source of Pb in the urban environment is vehicular exhaust fumes that result from the combustion of gasoline containing tetraethyl lead (Motto et al., 1970; Albasel and Cottenie, 1985). The particulate Pb compounds in exhaust gas cover a wide range of particle size from diameters of 0.01 jum to several milimeters, and reflect vehicle speed. At higher speeds (60 miles per hour), a larger portion of the Pb exhausted is in the form of coarse particles — larger than 5 pm (Hirschler and Gilbert, 1964). Such subtleties help determine the degree of Pb distribution and concentration in an area. Leaded exhaust emissions produce triethyl lead salts (Et3Pb+X— ), which are the first degradation products of tetraethyl lead. These salts freely permeate the plasma membranes of mammalian and plant cells, both of which are quite sensitive to the triethyl lead ion (Et3Pb+)- The high toxicity of this ion, coupled with its ability to pass easily through membranes, makes this compound potentially hazardous to cells of living organisms (Stournaras et al., 1984). Studies of Pb in soils have shown the increased concentrations in heavily trafficked urban areas of the world (Motto et al., 1970; Ward et al., 1975; Nriagu, 1978; Harrison and Williams, 1982; Lau and Wong, The Texas Journal of Science, Vol. 39, No. 1, February, 1987 16 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 1982; Albasel and Cottenie, 1985). Lead variation with soil type, distance from roadways, concentrations in various flora and fauna, and in humans living near roadways also are well documented (Ward et al., 1975; Solomon and Hartford, 1976; Wong and Tam, 1978; Dissanayake et al., 1984; Garcia-Miragaya, 1984). However, detailed overviews of major cities are not available; many works have Pb values based on widely separated sampling sites or in localized segments of a city. Such specialized or localized examinations of Pb concentrations may not give an accurate overview of the actual distribution of Pb or may miss trends characteristic of a given community. Therefore, the intent of this research is to present a detailed overview of Pb distribution in a major city based on numerous closely-spaced sampling sites, and to look for anomalies or trends, or lack of. Such works as this could be of particular interest to urban planners, city traffic divisions, city-county health organizations, environmental scientists, and the general public. Methods I sampled 485 locations within the Corpus Christi, Texas, city limits from 19 to 29 June 1984. The numerous parks throughout the city made excellent sampling sites for use as residential Pb indicators and for comparison to roadside soil samples. School playgrounds were sampled where there were no convenient parks. Soil samples were collected from the middle of parks and school playgrounds, grassy medians, grassy roadside embankments and borders, and the grassy areas in and around elevated freeway interchanges. The edge of frontage roads had to be sampled from the Crosstown Expressway (Fig. 1, item J) to Airline (Fig. 1, near item E). These areas provided the only soil samples close to that portion of the freeway (S.P.I.D. — South Padre Island Drive). By sampling only vegetated soil, I felt this helped reduce the possibility of a biased sample from a site where a leaded substance might have been dumped recently. The only consideration given to the soil type was that it not be sandy; more than 99 percent of the samples were clay dominated, whereas two were caliche. This preference for clay soil reflects the dominant soil type for the Corpus Christi area. Sampling the same type of medium also increased the uniformity of samples and sampling technique. The top two centimeters of soil were collected with a TeflonR knife (a 3.5 by 28 centimeter TeflonR bar was carved to form a knife). Field samples averaged about 100 grams and were placed in plastic bags. The samples later were ground to a fine powder, homogenized, and subsampled. The average weight of each subsample was 3.5- gram dry weight. These then were placed in TeflonR beakers, covered with watch-glasses, and refluxed for 24 hours in concentrated HNO3 under heat lamps. The samples then were dried and the final solution consisted of the sample, 10 milliliters of concentrated HNO3, and 90 millileters of deionized water. Analyses were performed on a Perkin — Elmer 360 flameless atomic absorption spectrophotometer. A National Bureau of Standards soil sample was used for quality control. Their results were 714 ± 28 ppm of Pb, whereas mine were 694 ± 24 ppm Pb, a difference of three percent. Any use of trade names is for descriptive purposes only and does not imply endorsement by the U.S. Geological Survey. Results The mean Pb value for the city of Corpus Christi, based on all 485 samples, was 208 ppm with a range of eight to 2969 ppm. Samples taken LEAD IN SOIL AT CORPUS CHRISTI, TEXAS 17 in comparatively remote and untrafficked areas within the city limits yielded a baseline of 13 ppm Pb. In the rural area surrounding the city, a lower baseline is probable. Values that exceeded the proposed baseline of 13 ppm Pb are expressed as a concentration factor (CF) relative to the 13 ppm. Some of the values obtained are listed in Table 1. The highest concentration of Pb recorded (2969 ppm) was from an embankment of the Crosstown Expressway (Fig. 1, item J) at site 121 (item A). The park with the highest Pb value was Louisiana Parkway with a value of 318 ppm (Fig. 1, item B). It is bounded on either side by a two-lane roadway that is heavily trafficked. The school playground having the highest value (Fig. 1, item C) was Travis Elementary (258 ppm), which is removed from any major streets and is in an older section of the city. Travis is one of the newer schools in the city, having been built in 1980. A distinct difference existed in the average amount of Pb between the northern (older) and southern (newer) sections of Corpus Christi, with the freeway (S.P.I.D.) being the divider between the two. Exclusive of S.P.I.D. values (62 sites), the area north of S.P.I.D. had a mean Pb value of 254 ppm (CF 19.5). Specifically, northside parks showed a mean value of 66 ppm (CF 5.1), whereas the southside parks had a value of 25 ppm (CF 1.9). North of S.P.I.D., the soils near roadways showed a mean Pb value of 313 ppm (CF 24.1); to the south the value was 67 ppm (CF 5.2). As expected, the locations with higher traffic-flow rates generally had increased Pb values. Exceptions, however, were the elevated freeway exchanges (Fig. 1, items, F,G,H). They are elevated approximately 10 to 50 feet on a combination of earthen embankments and concrete pillars. Samples taken around the grassy areas of these elevated exchanges were surprisingly low in Pb content (mean of 89 ppm) considering the heavy traffic flow on these structures. Based on traffic-flow rates taken from 14 February through 8 April 1983 by the Texas Department of Public Safety, the vehicular-flow rates for a 24-hour period (v/24 hours) at various locations on Figure 1 were: item D, 97,000 v/24 hours; item F, 61,530 v/24 hours; item G, 29,500 v/24 hours (Subsequent flow rates are from the same source.) The Crosstown Expressway (Fig. 1, item J) and S.P.I.D. are the two main roadways by which people commute. The Pb distribution was most representative of this (Fig. 1). On S.P.I.D. at item E (Fig. 1), the average traffic flow was approximately 32,500 v/24 hours. The two sampling sites on this section of S.P.I.D. with Pb values greater than 1000 ppm had a mean Pb value of 1172 ppm. Most sites along S.P.I.D. had elevated Pb values. However, the western portion of S.P.I.D, curving northward to intersect 1-37, did not, even though the traffice flow was 1.7 times greater (55,340 v/24 hours) than the flow for the eastern portion of S.P.I.D. (from the Crosstown Expressway eastward). 18 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 1. Lead (Pb) values and ranges for the Corpus Christi, Texas, sampling sites. Samples Mean Range Standard deviation Concentration factor (CF) (13 ppm baseline = CF 1.0) All samples (485 sites) 208 ppm 8-2969 ppm 236 CF 16.2 Parks (94 sites) >1000 ppm (0%) 100-999 ppm (15%) 50-99 ppm (17%) <50 ppm (68%) 55 ppm 8-318 ppm 66 CF 4.2 Schools (12 sites) >1000 ppm (0%) 100-999 ppm (17%) 50-99 ppm (8%) <50 ppm (75%) 57 ppm 1 1-258 ppm 77 CF 4.4 All others (379 sites) >1000 ppm (6%) 100-999 ppm (43%) 50-99 ppm (15%) <50 ppm (36%) 250 ppm 8-2969 ppm 250 CF 19.4 Ocean Drive, a four-lane, heavily trafficked roadway, follows the northern edge of Corpus Christi. Five of the samples taken along it were in excess of 1000 ppm (Fig. 1; item I, mean of 1291 ppm; item K, 2597 ppm; item L, 1977 ppm). In each case, the sites are located at major intersections. For item I, the traffic-flow was approximately 37,000 v/24 hours and items K and L had an average of approximately 23,000 v/24 hours. Discussion Prior studies, as cited in the introduction, have established that Pb exhausted from motor vehicles burning tetraethyl lead is concentrated in and around roadways and is directly proportional to the traffic volume and type of traffic flow. Figure 1 of this study amply reaffirms this. What is of note in this study are: (1) the lower Pb values at the elevated freeway interchanges; (2) the inverse relationship of Pb to traffic flow between the east and west ends of S.P.I.D.; and (3) the differences in Pb values between the older, northern section of the city and the newer, southern section. A possible explanation for the lower Pb levels around the elevated freeway interchanges may be that particles exhausted at higher speeds are coarser than those exhausted at lower speeds (Hirschler and Gilbert, 1964). Consequently, these larger and presumably heavier coarse-particles would settle out immediately onto the roadway. LEAD IN SOIL AT CORPUS CHRISTI, TEXAS 19 Figure 1. Distribution and concentration values of lead within the Corpus Christi, Texas, study area. The inverse relationship between Pb concentrations and traffic-flow rates at either end of S.P.I.D. probably is attributable to the presence or absence of various traffic lights and closely spaced exit/ entry ramps on S.P.I.D. located at major intersections. These slow the flow of traffic considerably at this eastern portion of S.P.I.D. At the west end, traffic moves much more efficiently for there is only one traffic light and few exit/ entry ramps. Although the Crosstown Expressway has no traffic lights, it also is burdened with numerous exit/ entry ramps that slow vehicular traffic during peak periods of traffic flow. Consequently, there are high Pb levels along its length. Of the 20 sampling sites with Pb values in excess of 1000 ppm, 11 are associated with the Crosstown Expressway and S.P.I.D. (east end) freeway system. The main concentration of Pb exhausted from traffic generally is limited to a narrow zone within 100 meters of a roadway (Nriagu, 1978). Consequently, the Pb is capable of being concentrated in the older and more congested portions of a city. Kinard et al. (1976) noted that freeway vehicular emissions enveloping certain metropolitan areas, coupled with emissions from inner-city traffic flow, present a potential hazard to the health of inner-city residents. This may apply to vehicular emissions in general because Pb is not the only contaminant exhausted (for example, chlorine, bromine, sulfur, and carbon monoxide). Of special note are the higher mean Pb concentrations found north of S.P.I.D. as compared to those of the south. While the area north of S.P.I.D. carries the greater traffic burden, it is also the older section of 20 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure 2. Distribution of residential and nonresidential areas within the Corpus Christi, Texas, study area. town. Therefore, in addition to the concern for general exposure to exhausted Pb, length of exposure time to Pb emissions is an additional health question. The area to the south of S.P.I.D. is comparatively new. Development there did not begin until about 25 years ago and major expansion has occurred within the last 10 years. Until development of the southside, all major residential areas were confined to the north side of S.P.I.D. The major industrial locations have not changed and most commercial businesses blend in with the residential areas in the form of small strip-shopping centers, individual business structures, and major shopping malls. Older and larger commercial businesses (mainly industrial suppliers) are confined to the northwest segment of the city (Fig. 2). As Figure 2 shows, the residential area is extensive and well segregated from the major industrial and commercial enterprises. The section of the city bisected by the Crosstown Expressway (Fig. 1, item J) contains some of the oldest residential areas in the city. Second- and third-generation families live within this area and many schools are located there. This implies a consistant and long-term exposure of the residents to exhausted Pb. This potential hazard is demonstrated in two related studies dealing with house dust. House dust in inner-city homes of Boston averaged 2000 ppm Pb (Kreuger, 1972), and in Champaign- Urbana, Illinois, Solomon and Hartford (1976) found home-interior dusts averaged 600 ppm and nonresidential interior samples averaged 1400 ppm. “Such contamination for home interiors could pose serious health problems for small children or infants who enjoy placing fingers and toys LEAD IN SOIL AT CORPUS CHRISTI, TEXAS 21 into their mouths; there will inevitably be some fraction of this contaminated dust on the toys or fingers” (Solomon and Hartford, 1976). Even though the average Pb content of gasoline has declined, the total discharge in the environment still is considerable if the increase in consumption is accounted for (Rodriguez-Flores and Rodriguez- Castellon, 1982). Even with the mandatory no-lead gasolines beginning in 1986 for the United States, the lead problem probably will be with us for some time. The residence time of Pb in urban dust and soils is unknown and will vary according to the geographic, climatic, and topographic conditions of each city. In coastal communities such as Corpus Christi, much of this Pb is washed into the surrounding bays, estuaries, and seas via storm and street drains. Pb can be concentrated even further in sediments and organisms near the outfalls of these drains (Holmes, 1974; Harrison and Martin, 1982; Harrison, 1984). Conclusions The three main factors influencing Pb concentrations from vehicles are: (1) proximity to the source; (2) volume of traffic; and (3) the loitering time of this volume of traffic in a given area as the result of traffic signals, reduced speed limits, and congestion resulting from closely spaced exit/ entry ramps. As shown in this study, the impact of these factors may be magnified in the older, more congested sections of a major urban community. How great a factor the local wind conditions are in dispersing or concentrating the Pb was not studied. Although Corpus Christi has a high average wind velocity (12 to 13 miles per hour) and the wind is dominately from the south in the summer months, areas sampled on the far north side of the city did not demonstrate any unusual buildup nor was there a graduation of values across the city coincident with wind direction. Such studies as this are not statements on the toxicity of Pb, but only on its relative availability and actual distribution. However, high Pb values present a greater potential for toxicological repercussions than do low values. Even though leaded gasolines soon will not be permitted in the United States, as long as there are combustion engines and increasing numbers of people to use them, traffic emissions in general will continue to produce the spector of potential ill-health. Acknowledgments I wish to thank F. T. Manheim and J. C. Hathaway for their critical review of this paper; their comments were most helpful. I, also thank Susan Peterson for typing this manuscript. Happy Sesquicentennial Texas. 22 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Literature Cited Albasel, N., and A. Cottenie. 1985. Heavy metal contamination near major highways, industrial and urban areas in Belgian grassland. Water, Air, and Soil Pollut., 24:103- 109. Dissanayake, C. B., A. Senaratne, and S. V. R. Weerasooriya. 1984. Environmental significance of trace elements in human hair — a case study from Sri Lanka. Internat. J. Environ. Studies, 23:41-48. Garcia-Miragaya, J. 1984. Levels, chemical fractionation, and solubility of lead in roadside soils of Caracas, Venezuela. Soil Sci., 138:147-152. Harrison, G. 1984. A survey of the trace-metal content of Corbicula fluminea and associated sediments in the tidal Potomac River. U.S. Geol. Surv. Open-File Rep. 84- 558, 33 pp. Harrison, G., and E. A. Martin. 1982. Heavy-metal contamination of Crassostrea virginica and associated sediments of the corpus Christi Bay System, Texas. U.S. Geol. Surv. Open-File Rep. 82-1060, 16 pp. Harrison, R. M., and C. R. Williams. 1982. Airborne cadmium, lead and zinc at rural and urban sites in north-west England. Atmosph. Environ., 1 1:2669-2681. Hirschler, D. A., and L. F. Gilbert. 1964. Nature of lead in automobile exhaust gas. Archives Environ. Health, 8:297-313. Holmes, C. W. 1974. Map showing distribution of selected elements in surface-bottom sediment of Corpus Christi and Baffin Bays, Texas. U.S. Geol. Surv. Misc. Field Studies Map MF-571. Kinard, J. T., J. Tisdale, and E. Alexander. 1976. Assessment of lead distribution patterns in urban and rural environments. J. Environ. Sci. Health, 11:153-164. Kreuger, H. W. 1972. Lead content of dust from urban households. Project 72-4, Kreuger Enterprises, Inc., Cambridge, Mass. Lau, W. M., and H. M. Wong. 1982. An ecological survey of lead contents in roadside dusts and soils in Hong Kong. Environ. Res., 28:39-54. Motto, H. L., R. H. Daines, D. M. Chilko, and C. K. Motto. 1970. Lead in soils and plants: its relationship to traffic volume and proximity to highways. Environ. Sci. Tech., 4:231-237. Nriagu, J. O. 1978. The biogeochemistry of lead in the environment, part A: ecological cycles. Elsevier/ North-Holland Biomedical Press, Amsterdam, 298 pp. Rodriguez-Flores, M., and E. Rodriguez-Castellon. 1982. Lead and cadmium levels in soil and plants near highways and their correlation with traffic density. Environ. Pollut. ser. B, 4:281-290. Solomon, R. L., and J. W. Hartford. 1976. Lead and cadmium in dusts and soils in a small urban community. Environ. Sci. Tech., 10:773-777. Stournaras, C., G. Weber, H. -P. Zimmermann, K. H. Doenges, and H. Faulstich. 1984. High cytotoxicity and membrane permeability of Et3Pb+ in mammalian and plant cells. Cell Biochem. Function, 2:213-216. Swaine, D. J. 1955. The trace element content of soils. Commonwealth Bur. Soil Sci. Tech., Comm. no. 48, Herald Printing, York, England. Ward, N. I., R. D. Reeves, and R. R. Brooks. 1975. Lead in soil and vegetation along a New Zealand state highway with low traffic volumes. Environ. Pollut., 9:243-250. Wong, M. H., and F. Y. Tam. 1978. Lead contamination of soil and vegetables grown near highways in Hong Kong. J. Environ. Sci. Health, 13:13-22. A STUDY OF THE Cu(H20)6+/CuCl47 ETHANOL SYSTEM FOR SOLAR ENERGY STORAGE L. Gene Spears, Jr., Larry G. Spears, and Joycelyn C. Spears Department of Chemistry, Rice University, Houston, Texas 77005; Department of Natural Sciences, University of Houston- Downtown, Houston, Texas 77002; and Science Department, Cypress Creek High School, Houston, Texas 77070. Abstract. — In ethanol, CuCl2*6H20 can establish an equilibrium between the hydrated Cu(II) ion and the tetrahedral copper chloride anion. This reaction provides a potential means of storing solar energy via the unhydrated CuCli ". On cooling, heat is released based on the heat of hydration and the specific heat of the solution. To determine the potential capacity of this system for energy storage, the solubility of cupric chloride in ethanol at different temperatures, the AHhyd, and the specific heat of the solution were determined. Theoretical calculations then were made to compare the heat storage capacity of the CuCb* 6H20/ethanol solutions to that of water, for a solar energy collection system. Key words : solar energy storage; hydrated salts; solubility of copper chloride; heat of hydration; heat capacity. One of the major problems associated with the utilization of solar energy as a primary energy source for heating and cooling is the lack of a practical, efficient means of storing large amounts of energy after it has been collected. Sensible heat storage via heated liquids or rocks has the disadvantages of low energy density and requires large volumes for limited collector temperatures (Wentworth et al., 1981). Latent heat storage, which utilizes the heat associated with phase changes, normally from solid to liquid, also is utilized as a method for solar energy storage (Meinel and Meinel, 1979). Another method for storage utilizes a reversible endothermic reaction (Wentworth et al., 1981). In a previous study (Spears and Spears, 1984), it was shown that a solution of cobalt (II) chloride in isopropyl alcohol could be used for storing solar energy via the reversible reaction between the hydrated cobalt (II) ion and the chlorinated cobalt (II). In addition to the specific heat of the solution, this system utilizes the heat of hydration for the cobalt (II) ion for additional heat storage capacity. For this system, the value of the equilibrium constant and the solubility as a function of temperature determine the potential for heat storage. Based on the above research, solutions of the hydrated salts of Al2(S04)3, Cr(NC>3)2, CuCL, Fe (NC>3),3, FeCL, and NiCL in ethanol were selected for preliminary screening. Of these, CuCL appeared to be the most desirable and was chosen for further study. In making this choice, several factors were considered: 1) the solubility of the salt; 2) the sensitivity of the system to temperature change and the ability to reach equilibrium quickly; 3) the heat of hydration of the salt; and 4) the availability of a convenient The Texas Journal of Science, Vol. 39, No. 1, February, 1987 24 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. I, 1987 instrumental method of analysis to follow changes in the equilibrium system. This last factor prevented the consideration of many hydrated salts. Because water is a part of the equilibrium reaction, its concentration in the solvent is important. When CuC12-(H20)6 is dissolved in ethanol, the following equilibrium is established: Cu(H20)6++4Cf« C11CI4 +6H2O. In this system, the forward reaction is endothermic and the reverse reaction is exothermic. Thus, the forward reaction could be used to collect and store solar energy, whereas the reverse reaction could be used to release the stored energy. A series of experiments were performed using the CuC12*6H206 ethanol system to determine its potential use as an energy storage medium. Because CuCl4 had a distinct green color and Cu(H20)6+ is blue, visible spectrophotometry was used to follow the reaction. Using the recorded data and published information, calculations were made to compare the heat storage capacity of this system to that of water, for a solar energy collection system. Experimental Section The majority of the experimental research was concerned with following the reactions of CuCl2*6H20 dissolved in ethanol, by the use of a Model-2 Perkin Elmer recording spectrophotometer with a controlled temperature cell attachment, in the visible region of the spectrum. Because the green CuCd" complex has a major absorption peak at 281 nm and the Cu(H20)6+ complex does not absorb in this region, a Beer’s Law calibration plot was determined using a cell pathlength of 1.0 centimeter. A value of 288 M-1 cm _1 for the molar absorptivity constant was calculated from the resulting linear plot. Because the CuCll- and Cu(H20)6+ions were the only Cu-containing ions detected in the test solutions, the total concentration of dissolved Cu2+ was assumed to be equal to the CuCh concentration plus the Cu(H20)6+ concentration. The ethyl alcohol used in this study was 99.9 percent pure, with less than 0.005 percent water. The copper (II) chloride used had a purity of 99.5 percent. It was pulverized and heated, at 150°C for three hours, to dryness. All weighings and sample transfers involving the anhydrous copper chloride were done in a dry box. Results Determination of Product / Reactant Ratios Due to difficulty in measuring values for the free water concentration, values for the product/ reactant ratio rcucin [Cu(H20)?+] SOLAR ENERGY STORAGE 25 Figure 1. Solubility of CuCl2*(H20)6 in ethyl alcohol/ H2O solutions at different temperatures. were used instead of Keq. These ratios were determined at various temperatures and water concentrations for 0.002 M CuCh solutions (Fig. 1). From these data, the percent CuCli present in each solution was determined (Table 1). Solubility of CuCh^EhO in Ethanol Saturated solutions of CuCh with varying water concentrations were heated to three different temperatures and their absorbances at 281 nm were recorded. Concentration values for CuClJ were determined and using the data in Table 1 the corresponding concentration values for Cu(H20)6+ were calculated. These two values were added together to obtain the total solubility as a function of temperature and water concentration (Fig. 2). The solubility value for 24° C and four percent water compares favorably with a reported value of 53g/100mL in 95 percent ethanol at 15°C (Perry and Chilton, 1973). 26 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 1. Percent composition of 0.002 M CuCE/ethanol solutions. %h2o T(°C) % CuCiF % Cu(H20)6+ 4 24 48.9 51.1 4 40 58.0 42.0 4 60 66.5 33.5 8 24 23.2 76.8 8 40 41.3 58.7 8 60 53.0 47.0 12 24 2.15 97.8 12 40 27.7 72.3 12 60 43.0 57.0 Determination of Specific Heat The specific heat of a solution describes the amount of heat it can absorb. A 0.1 M solution of copper (II) chloride in 88 percent ethanol was assumed to be representative for this system and a modified Nalgene Dewar flask/calorimeter was used for the specific heat measurements. Following the procedure described by Morss and Boikes (1978), an average value of 4.20 J g”1 °C_1 was obtained for the specific heat using 50 mL samples at 24° C and an initial calorimeter temperature of 40° C. Determination of the Heat of Hydration In addition to the specific heat, the CuCF/ethanol solution is capable of storing heat via the CuClJ- tetrahedral complex. On cooling, this complex can revert to its hydrated form, Cu(H20)6+, and in doing so releases heat due to hydration (AHhyd). By use of conventional methods (Shiflett, 1978), a value for AHhyd — 63.2 kJ mole”1 was determined. A theoretical value of -66.5 kJ mole"' can be calculated from published thermodynamic data (Perry and Chilton, 1973). Energy Storage Calculations The heat storage capacity of a solar heating system using the CuCh/ ethanol system as the energy storage medium was calculated based upon data collected during this study. The solar collector was assumed to concentrate incoming heat at a 4:1 ratio, and to operate at a mean temperature of 60° C. Using Figure 1, the concentration of copper chloride in a saturated solution of 88 percent ethanol was determined to be 2.27 M at 60° C. The heat content of solution is due both to the heat of hydration and the specific heat. In Figure 2, it is seen that for 88 percent ethanol at 60° C, the CuClJ /Cu(H20)6+ ratio is 0.768, and at 20° C it decreases to 0.0219. From these values it can be determined that at 60° C, 43.4 percent of the copper is unhydrated, and at 20° C, 2.1 percent is unhydrated. Thus on cooling from 60° C to 20° C, 41.3 percent of the CuClJ SOLAR ENERGY STORAGE 27 Figure 2. Effect of temperature on the CuCh /Cu(H20)6+ ratio at different H2O concentrations. undergoes hydration. The heat released from 1.0 L of 2.27 M CuCh in ethanol for this amount of hydration would be 63.2 kJ mole _1 X 2.27 mole L_1 X 0.412 = 59.3 kj Lf '. The specific heat of 1.0 L of the above solution is approximately 4.20 J g 1 °C‘ X 0.80 mL_1 X 1000 mL L_1 = 3.36 J°C"‘ L"\ where 0.80 g mL'1 equals the assumed density of the solution (this value would vary depending on the concentration of dissolved salt and temperature). If the temperature of the cool solution is 20° C and that of the heated solution is 60° C, the total heat absorbed per liter would be 59.3 kj _1 + 3.36 kj °C_1 L“‘ X (60°C-20°C) = 194 kJ L'1. 28 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 This calculation assumes that the value of AHhyd remains constant in the temperature interval 20-60° C. If water was used as the storage medium under the above conditions, it would have a total heat absorption capacity of 4.18 kJ °C_1 L"1 X (60° C-20° C) = 167 kJ L'1. Discussion In the above, it has been shown that the copper (II) chloride/ 88 percent ethanol system has a theoretical energy storage capacity approximately 16 percent greater than that of water assuming that AT=40°C. Figure 2 indicates that large values of AT would favor this type of system. Our data indicate that this equilibrium system adjusts rapidly when the temperature is changed. An additional advantage to this system is the lower freezing point than that of water. It is not proposed that this system be pursued on a commercial basis due to the obvious high cost for the solute. However there are other hydrated salts that are less expensive, more soluble, and have higher heats of hydration that copper (II) chloride that may be economically and thermodynamically suitable for this purpose. Based on this research, it is recommended that liquid phase equilibra systems involving hydrated salts be considered as a method for storing solar energy in systems where pumping of the liquid storage medium is required. Literature Cited Meinel, A. B., and M. R Meinel. 1979. Applied solar energy, an introduc¬ tion. Addison-Wesley, London., 651 pp. Morss, L. R., and R. S. Boikes 1978. Chemical principles in the laboratory. Harper & Row, New York, 298 pp. Perry, R. G., and C. H. Chilton (eds.). 1973. Chemical engineers handbook. Mc-Graw- Hill, New York, 1958 pp. Shiflett, R. B. 1978. Calorimetry and solar energy. J. Chem. Ed., 55:103. Spears, L. G., Jr., and L. G. Spears. 1984. Chemical storage of solar energy using an old color change demonstration. J. Chem. Ed., 61:252-254. Wentworth, W. E., B. W. Johnston, and W. M. Raldow. 1981. Chemical heat pumps using a dispersion of a metal salt ammoniate in an inert solvent. Solar Energy, 26:141- 146. RELATIVE MOBILITY OF LEAD AND COPPER IN SOILS: AN EXAMPLE FROM THE BONANZA DISTRICT, SAGUACHE COUNTY, COLORADO Joseph C. Cepeda Department of Geosciences, Killgore Research Center, West Texas State University, Canyon, Texas 79016 Abstract. — The relative mobility of Pb and Cu in the soil profile downslope from tailing piles in mining districts is a function of climate, relative solubility of metal sulfates, slope, pH, and soil type. In the Bonanza District, Saguache Co., Colorado, 50- to 85-year-old mine dumps with more than 4400 parts per million of lead and 970 of copper show a narrow, well-defined dispersion train downslope from the dump. Mobility of copper on this steep (29 degrees) slope is limited relative to that of lead. Copper contents in the soil approach background values (50 ppm) approximately 60 meters down the slope, but lead contents 80 meters down the slope are still four to five times the background value of 500 ppm. The dispersion haloes are probably produced by downslope movement of resistant metallic sulfides and possibly native copper particles on a steep slope in a semiarid climate. Key words : soils; Colorado; geochemical prospecting. The bulk of the exposed economic ore deposits on the earth’s surface already have been discovered by conventional geologic and field techniques. Additional reserves of metallic minerals remain hidden beneath a few centimeters to a few meters of soil, alluvium, or glacial debris. Geochemical prospecting techniques seek to locate these hidden bedrock deposits by identifying their anomalous, metal-rich haloes composed of metallic elements dispersed in the soil horizon in a general downslope or downstream direction from the deposit. An anomalous concentration is generally defined as more than three times the background value for that element. Background values can be determined by regional geochemical surveys such as the National Uranium Resource Evaluation (NURE) surveys or by values obtained from samples sideslope or upslope from mine dumps or ore deposits. Previously exploited mining districts contain the combination of factors necessary to produce an ore deposit, and in geochemical surveys of such areas it is desirable to be able to confidently distinguish between anomalies produced from an old mine dump and those produced by previously undiscovered mineralization. This type of study also has environmental implications — that is, defining the extent and degree of metallic contamination of soils and surface and subsurface waters. The Bonanza Mining District was selected for this study because it is mineralized by several elements, including lead and copper. Relative The Texas Journal of Science, Vol. 39, No. 1, February, 1987 30 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 values of Pb and Cu in soils downslope from the dump compared to values from samples taken from the mine dump indicate relative element mobility. The objective of this study was to evaluate and compare the geometry and size of the dispersion haloes for both Cu and Pb. Geology and Mineralogy of the Bonanza District The Bonanza Mining District, Saguache Co., Colorado, is located in the extreme northeastern part of the San Juan volcanic field in south central Colorado. Named for the town of Bonanza on Kerber Creek, it encompasses an area of about 100 square kilometers (30 square miles). The highest peaks in the district have elevations of 3660 meters (12,000 feet) to 4000 meters (13,200 feet). The elevation of the town of Bonanza is approximately 2850 meters (9400 feet) above sea level. Slopes are steep and many in the northern part of the district are heavily wooded with spruce, Douglas fir, Ponderosa pine, and aspen. In the southern part of the district, the vegetation is more sparse, particularly on south-facing slopes. Precipitation in the district is a function of altitude and occurs as thundershowers in summer and moderate to heavy snowfall in winter. Average annual precipitation at the elevation of the study area is about 45 centimeters (18 inches). Burbank (1932) described the distribution and structure of the Paleozoic sedimentary units and the Tertiary volcanic sequence and intrusive rocks in a comprehensive investigation of the geology of the district. The main deposits of the Bonanza District are veins containing Pb, Zn, Cu, and Ag. Mineralogically, these veins contain pyrite, sphalerite, galena, chalcopyrite, bornite, and silver-bearing tennantite in a gangue of calcite, quartz, rhodocrosite, and barite (Burbank, 1932). Burbank (1932) was the first to recognize a collapse feature in the district, which he suspected may have been formed by the eruption of magma from beneath the Bonanza area. Subsequent investigators (Karig, 1965) have recognized the Bonanza Caldera, one of 18 known or inferred calderas in the San Juan volcanic field. Most of the productive veins cut the Rawley Andesite, a sequence of intermediate composition volcanic rocks older than the 36 million-year- old Bonanza Caldera (Burbank, 1932; Varga and Smith, 1984). Ore- bearing veins also cut the Bonanza tuff of dacite to rhyolite composition (Varga and Smith, 1984). It was the eruption of the Bonanza tuff 36 million years ago that caused the formation of the Bonanza Caldera. The mining history of the district began in 1879-80, and the majority of the metal output occurred between 1904 and 1930, when the Rawley mill was closed and its equipment sold for salvage value (Burbank, 1932). Thus, the majority of the tailings in the district are 50 to 85 years old. LEAD AND COPPER IN SOILS 31 Yenter (1984) mapped the soil in the study area as a Bushvalley cobbly loam. This is a shallow, well-drained soil with moderate permeability and typical of mountain slopes and ridges. This soil type contains 15 to 35 percent clay and two to three percent organic matter. Soil pH ranges from 6.1 to 7.8. (Yenter, 1984). Sample Collection and Preparation Soil samples were collected at depths of 10, 20, and 30 centimeters at 19 points downslope from a mine tailing dump in June 1981. Four samples were collected from various parts of the mine dump itself, as shown in Figures 1 and 2. Individual samples were taken at each depth along three traverses spaced 36 meters (120 feet) apart. Sample spacing along the traverse lines was nine meters (30 feet). See Figures 1 and 2. Bedrock outcrops in the vicinity of the dump area consisted of andesitic volcanic breccias of the Rawley Andesite. The grassy slope downhill from the dump has a slope of 29 degrees to the southeast. The 30 to 60 mesh and less than 60 mesh fractions were processed, digested in a hot acid bath, and analyzed for Pb and Cu by atomic absorption spectrophotometry, using techniques described in Cepeda (1986). Results and Discussion Lead and copper contents in the soil in the less than 60 mesh fraction at a depth of 30 centimeters are shown in Figures 1 and 2. Data for other mesh fractions and depths are similar to that shown in Figures 1 and 2, although at a depth of 10 centimeters for both Cu and Pb, metal values are not as high and are distributed over a wider area, probably the result of dispersion by southwesterly winds. Contours derived from the data show a rapid decrease in copper content downslope from the dump. Background or near background values are measured approximately 45 meters (150 feet) from the base of the mine dump. Similar results were obtained from copper contaminated soil in the Old Cleora Mining District (Cepeda, 1986), where background values were attained 60 meters downslope from the dumps. However, at Old Cleora, the slope angle was a gentle six degrees in a sandy soil. Lead contents in the soil remain high all the way down the slope and are more than 50 percent of values in the dump samples at a downslope distance of 76 meters (250 feet), the limit of the slope, before it is intersected by a gravel road. The gravel road marks the base of the slope. At this point it is still well above the background value of approximately 500 parts per million. In contrast, unmineralized Rawley Andesite has a Cu content of 18 to 21 ppm and a Pb content of 21 to 32 ppm (Varga and Smith, 1984). The Bonanza tuff has similar Pb and Cu contents (Varga and Smith, 1984). The downslope decrease in Pb and Cu values with distance from the dump is depicted graphically in Figure 3. Soil contents of Pb and Cu in this figure have been normalized to the highest value of each metal in 32 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Mine Dump O33 O30 * 0 9Meters \ - 1 0 30Feet Scale O Sample Point 54 O 33 Figure 1. Copper concentration (in parts per million) in less than 60 mesh fraction, soil samples at a 30 centimeters depth. Bonanza, Colorado. Arrows indicate downslope direction on 29 degree slope. the dump. Thus the highest mine dump value is assigned a value of 1.0 and downslope soil metal contents are plotted as a fraction of this highest value. Pb and Cu background values are also shown on Figure 3. Previous investigators of element mobility in soils have come to mixed conclusions. Several investigations have noted the high mobility of copper in the soil in both a desert (Snoep and Zeegers, 1979) and a humid (Cameron, 1975) environment. Cameron (1975) suggested that the mobility of copper may be due to the high relative solubility of copper sulfate (14.3 grams per 100 milliliters) relative to the lead sulfate (0.004 gm/100 ml). Bogoch and Brenner (1977) noted that lead (and zinc) concentrations in soils adjacent to mineralization in a temperate climate are exceptionally high but downslope and downstream mobility is low. LEAD AND COPPER IN SOILS 33 511 493 686 463 0 9Meters i _ i 0 30Feet Scale O Sample Point 832 533 Figure 2. Lead concentration (in parts per million) in less than 60 mesh fraction, soil samples at a 30 centimeters depth, Bonanza, Colorado. Arrows indicate downslope direction on a 29 degree slope. Hoffman and Fletcher (1972) reported a low mobility of copper in an alkaline geochemical environmental in a semiarid climate. In an alkaline environment, the migration of dissolved copper, as copper sulfate, would be inhibited. Hoffman and Fletcher concluded that in such an environment copper gives consistent and meaningful geochemical patterns related to mineralization. This appears to be the case in the Bonanza District where copper exhibits limited mobility in the soil. The limits of Eh and pH in natural environments (Baas-Becking, 1960) are restricted to a pH between 2 and 10 and an Eh of between — 400mV and +800mV. The Eh characteristics of soils and meteoric waters range from about +100 to +700 mV. Measured pH for the Bushvalley soils range from 6.1 to 7.8. Copper species stable in this part of the Eh-pH diagram include chalcocite, native copper, cuprite, and malachite. The 34 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Downslope DistanceCin meters) Figure 3. Pb and Cu soil contents downslope from mine dump, normalized to highest mine dump value. Highest Pb and Cu values are 4473 and 977 parts per million, respectively. Pb and Cu background values are 500 and 50 parts per million, respectively. solubility of copper as Cu++ ion in aqueous solution increases with decreasing pH and is approximately 10”4 molar at a pH of 6 at 25° C and with P co2 =10”35 atm (Garrels and Christ, 1965). An increase in pH would promote the precipitation of Cu as malachite or cuprite. A decrease in Eh (more reducing conditions) would favor the precipitation of chalcocite or native copper. A recent study by Hostettler (1984), however, suggested that Eh measurements are difficult to make, and it is not only necessary to consider the effect of minerals but also gases, masses of water, and microorganisms. Conclusions The origin of the dispersion haloes in the Bonanza area, a semiarid climate, is probably due to downslope movement of resistant particles on a steep slope, although it is possible that Cu could be transported in solution. Microscopic examination of panned concentrates of the 30 to 60 mesh fraction of soil reveals metallic sulfide and native copper particles. But indication of relative mobility derived from this study should be utilized with caution because of possible variables such as climate, soil type, pH, Eh, slope, and type of mineralization. Acknowledgments Financial support for the laboratory portion of this project was supplied by a 1981 Organized Research Grant from the Killgore Research Center, West Texas State University, LEAD AND COPPER IN SOILS 35 Canyon, Texas. Ron Thomason, Head, Department of Plant Science, kindly allowed use of the department’s atomic absorption spectrophotometer and supplied standards for calibration. Trenton Richards assisted the principal investigator in the collection of samples at Bonanza, Colorado. Steve Ritter assisted in laboratory preparation of samples and analytical work. Robert Sawvell provided drafting materials and supplies. Literature Cited Baas-Becking, L. G. M. 1960. Limits of the natural environmental in terms of pH and oxidation reduction potentials. Geol., 68:243-284. Burbank, W. S. 1932. Geology and ore deposits of the Bonanza Mining District. U.S. Geol. Surv. Prof. Paper, 169:1-166. Bogoch, R., and I. B. Brenner. 1977. Distribution and dispersion of lead and zinc in anomalous soils and stream sediments, Mount Hermon area, Israel. J. Geochem. Expl., 8, 529-535. Cepeda, J. C. 1986. Copper in soil samples downslope from copper-tungsten mine tailings, Cleora District, Chaffee County Colorado. Texas J. Sci. 38:59-64. Cameron, E. M. 1975. Geochemical methods of exploration for massive sulphide mineralization in the Canadian Shield. Proc. 5th Internat. Geochem. Expl. Symp., Vancouver, Elsevier Publ. Co., pp. 21-49. Garrels, R. M., and C. L. Christ. 1965. Solutions, minerals and equilibria. Harper & Row, New York, 450 pp. Hoffman, S. J., and K. Fletcher. 1972. Distribution of copper at the Dansey-Rayfield River Property, south-central British Columbia, Canada. Jour. Geochem. Expl., 1:163- 180. Hostettler, J. D. 1984. Electrode electrons, aqueous electrons, and redox potentials in natural waters. Amer. J. Sci. 284:734-759. Karig, D. E. 1965. Geophysical evidence of a caldera at Bonanza, Colorado. U.S. Geol. Survey Prof. Paper, 525-B: 9-12. Snoep, J., and H. Zeegers. 1979 .Multi-metal soil geochemistry: a tool for identification and exploration of porphyry copper deposits. Two examples from Peru. J. Geochem. Expl. 11:103-130. Varga, R. J., and B. M. Smith. 1984. Evolution of the early Oligocene Bonanza Caldera, northwest San Juan Volcanic Field, Colorado. J. Geophys. Res., 89:8679-8694. Yenter, J. M. 1984. Soil survey of Saguache County area, Colorado. U.S. Dept. Agric., Soil Conserva. Serv., 203 pp. INFLUENCE OF ENVIRONMENTAL FACTORS ON OXYGEN CONSUMPTION OF CLIBANARIUS VITTATUS (STRIPED HERMIT CRAB) V. A. WOLFENBERGER Texas Chiropractic College, Pasadena, Texas 77505 Abstract. — The oxygen consumption of the hermit crab, Clibanarius vittatus , was measured under combinations of salinity — 10, 20, 30 and 40 parts per thousand, temperature — 16, 20 and 24° C, the presence or absence of selenium, and two sites of collection of animals. The general trend was for oxygen consumption by the crabs to increase with increasing temperature. The rate of oxygen consumption tended to decrease with increasing salinities. Selenium depressed the rate of oxygen consumption in animals from the bay environment, whereas erratic results were obtained in animals from the Gulf environment. Key words : oxygen consumption; hermit crab; selenium; salinity; temperature. Oxygen consumption of an organism long has been used as an indicator of stress because it is an indirect measurement of metabolic rate. Various environmental conditions might be expected to produce differing levels of stress in an organism subjected to them. Quantity of oxygen consumption, therefore, would be an indicator of the extent of stress an organism was experiencing. Four variables were investigated for their influence on the oxygen consumption of Clibanarius vittatus , the striped hermit crab. Temperature and salinity are obvious factors that might stress a euryhaline organism. Stress from a pollutant was the third factor. Incorporating two different populations provided crabs that would be expected to show differences in their abilities to adapt to the various conditions included in this study. Interactions or synergistic effects of environmental factors allow a more realistic evaluation of the stress an organism experiences than do single factor evaluations. The primary purpose of this study was to determine the effects of these environmental factors on the rate of oxygen consumption of the hermit crab, Clibanarius vittatus , and to compare these effects in the two populations of the species — one from a bay environment and one from a surf environment. Clibanarius vittatus is a euryhaline crustacean common in estuarine environments of the Gulf of Mexico. Above a salinity of 20 parts per thousand it maintains its body fluid concentration slightly hyperosmotic to the medium. In more dilute surroundings, it is hyperosmotic to the medium (Sharp and Neff, 1980). Changes in salinity are known to influence oxygen consumption of some crustaceans. However, a sufficient acclimation period occasionally allows return to near-normal oxygen The Texas Journal of Science, Vol. 39, No. 1, February, 1987 38 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 consumption rates. Shumway (1978) found that Pagurus bernhardus (with shells) increased oxygen consumption sharply when ambient salinity was reduced in 75 percent sea water, either abruptly or gradually, but returned to near-normal oxygen consumption after approximately three hours. Sarojini and Nagabhushanam (1968) found that the oxygen consumption of Diogenes bicristiamnus (an Indian hermit crab) decreased with decreasing salinity. Vernberg (1956) reported Clibanarius vittatus acclimated to normal sea water consumed 1.28 microliters of O2 per gram weight per minute. The nonestuarine Pagurus bernhardus exhibited withdrawal behavior, cessation of ventilation, and negligible oxygen uptake when exposed to 20.5 to 22.5 parts per thousand salinity (Davenport et al., 1980). The relationship between temperature and oxygen consumption of poikilothermic organisms such as Clibanarius vittatus is often a logarithmic one. Respiratory response to chronic temperature change is less than to acute temperature change; acclimation may occur, resulting in return to near-original rate of oxygen consumption. This return is termed compensation (Precht, 1973). It has not been determined whether Clibanarius vittatus compensates or not. The third ecological factor considered in this study is a pollutant. Some effects of pollutants on respiration in crustaceans have been reported. Exposure for 60 days to mercury at concentrations of 0.5 to 1.0 parts per trillion did not significantly alter the respiratory rate of postlarval Penaeus setiferus (Green, et al., 1976). Vernberg et al. (1979) found respiratory rates of Palaemonetes pugio exposed to cadmium under static conditions were lower than for nonexposed animals. The effects of interactions of a pollutant and environmental factors may offer a more realistic means of simulating environmental situations. The synergistic effects of mercury and temperature-salinity on the oxygen consumption of Uca pugilator larvae were studied by Vernberg et al. (1973). Mercury depressed the oxygen consumption of the larvae at 25 and 30° C and enhanced oxygen consumption at 20° C. Suboptimal temperature- salinity regimes generally depressed oxygen consumption. The effect of the added stress of mercury was temperature dependent. The pollution factor considered here is the presence or absence of selenium. Awareness of selenium as a pollutant is increasing. The growing sulfur and petroleum industries are only two of the potential sources of the element on the Texas coast. Geologically, selenium occurs widely and can enter the environment via the weathering of rocks and soils (NAS, 1976). The element also may enter the environment as a result of combustion of fossil fuels (Fleming et al., 1979; Bertine and Goldberg, 1971; Gutemann and Bach, 1976). It also can be a run-off contaminant from soil dressing containing selenium in selenium deficient areas (Gissel- Nielson and Gissel-Nielson, 1973). Inasmuch as selinium tends to OXYGEN CONSUMPTION IN HERMIT CRAB 39 accumulate in sediments (Sidelnikova, 1970), dredging operations also might increase available selenium in marine environments. Several studies have been done concerning the accumulation of selenium by marine organisms. Crustaceans that have been found to accumulate selenimum include the euphausid, Meganyctiphanes norvegica (Fowler and Benayoun, 1976a), a shrimp, Lysmata seticaudata (Fowler and Benayoun, 1976b), and the water flea, Daphnia pulex (Sandholm, 1973). The fourth parameter to be considered is that of habitat or population effect on physiological response. Whereas several studies have been done comparing populations of invertebrates, not all such investigations have used populations from proximate locations and of the same species. King (1965) reported differences in oxygen consumption of Callinectes sapidus from brackish water and from marine waters after acclimation to the same conditions. Loft (1956) found considerable differences in the respiratory rates of two populations of Palaemonetes varians as they were subjected to different salinities. The collection sites for the two populations of Clibanarius vittatus in this study differed (at least obviously) only in salinity; the bay environment, Steadman’s Island, had a salinity range from 12 to 40 parts per thousand, whereas the Gulf-exposed environment ranged from 24 to 35 parts per thousand. Methods This study considers the influence of temperature, salinity and selenium on the oxygen consumption of Clibanarius vittatus collected from two habitats near Port Aransas, Texas (Fig. 1). One site, Steadman’s Island, is a bay environment to the south of the causeway connecting Aransas Pass and Port Aransas. The second site is on the southwestern shore of Corpus Christi Pass through Mustang Island into Corpus Christi Bay. Animals at this site were collected gulfward from the point at which the Highway P52 bridge crosses the waterway. This area is exposed to Gulf waters, temperatures and salinities. The animals were transported dry in styrofoam chests to laboratory facilities in College Station, Texas. They were maintained in fiberglass tubs containing 30 parts per thousand artificial sea water at room temperature (approximately 20° C ± 2°C). They were fed Tetra Min ad libitum. No animals were subjected to experimental conditions until they had been in the laboratory a minimum of 10 days, and no animals were used that had been in captivity longer than 90 days. No records were kept of sex. The mean dry weights of the animals from the two collection sites were 1.180506 grams for animals from the Gulf environment and 1.466789 grams for the animals from the bay environment. An F-test (Number Cruncher Statistical System, Version 4.2, 1985) indicated these two means were significantly different at the a = .05 level. Experimental conditions included four salinities, three temperatures, and the presence or absence of a pollutant, selenium. Preparation of animals from each site for experimentation included 96 hours acclimation in three liters of sea water at salinity of 10, 20, 30, or 40 parts per thousand with six to 10 animals per four-quart aquarium. An airstone was placed in each container. These aquaria were kept in an incubator at 16°C, 20° C, or 24° C (all temperatures are ± 1°C) for at least 96 hours prior to performing any experimental tests. These temperatures and salinities are within the ranges for the Texas coast. A 12-hour, light-dark cycle was maintained. Animals subjected to similar conditions were exposed to the pollutant selenium in the form of sodium selenite. To yield the 100 parts per million 40 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure 1. Map of the Port Aransas, Texas, area. The bay collection site is indicated by an arrow. The Gulf-exposed collection site is indicated by a star. exposure concentration of selenium, 0.65714 gram of the compound was added by dissolving the salt in the three liters of sea water 72 hours after acclimation had begun so that the exposure time to the pollutant was 24 hours. The oxygen consumption of whole animals subjected to experimental conditions was determined. Shells were scrubbed prior to testing. After the 96-hour acclimation period, an animal was placed in a quart container (three-inch inside diameter) containing 502 milliliters of artificial sea water at the acclimation temperature and salinity. A marble was added to facilitate agitation of the water. The sea water had been filtered through a .4 n Millipore filter to remove bacteria and other microscopic organisms, and it was then vigorously aerated to saturate it with oxygen. Immediately after addition of the animal, a two-milliliter sample of sea water was taken using a glass syringe, which then was sealed. The sea water was OXYGEN CONSUMPTION IN HERMIT CRAB 41 covered with eight milliliters of paraffin oil to prevent gas exchange with the atmosphere (Wofford, 1978). The syringes were submerged in an acclimation temperature water bath. The quart containers with animals were kept in an incubator at the acclimation temperature during the test period. Initial samples of the sea water were analyzed at the acclimation temperatures for oxygen content, using a Radiometer PHM 71 Mk 2 blood gas analyzer that had been modified to maintain constant lower temperatures. At the termination of the experimental period, each container was agitated to disperse diffusion gradients. A second sample was taken by placing the syringe through the paraffin oil; it was analyzed for oxygen content using the Radiometer. The animals were removed from their shells, weighed and dried to a constant weight. An individual animal was tested only once at only one temperature and salinity. Five animals were tested under each set of conditions. Partial pressures of oxygen were converted to milliliters of oxygen by the equation: ml O2/ 1 = solubility coefficient X mm Hg of sample barometric pressure X(1000) (Giese, 1968). The solubility coefficient in milliliters O2 per milliliters of water was obtained from Weiss (1970). Measurements of oxygen consumption were made weight specific. Statistical analysis included multiple analyses of variance (Ott, 1977) incorporating all four factors, as well as multiple analyses of variance performed separately for each site incorporating the other three factors. In instances in which a significant effect attributable to treatment was determined at the a = .05 level, multiple linear regression analyses were performed. Results The oxygen consumption rates of Clibanarius vittatus not exposed to selenium are shown in Figure 2. At 16°C, the oxygen consumption of animals from both sites peaked in 30 parts per thousand salinity at a mean of 6.61 milliliters per milligram per hour for animals from Corpus Christi Pass and of 7.20 for the same measurement for those from Steadman’s Island. The instances of lowest rate of oxygen consumption differed between the two sites; for the Pass, the lowest rate occurred in 20 parts per thousand, whereas it occurred in 40 parts per thousand for Steadman’s Island animals. At 20° C, the animals from Corpus Christi Pass had variable rates of oxygen consumption over the range of salinities tested. The rate of oxygen consumption of animals from Steadman’s Island at 20° C peaked in 20 parts per thousand salinity and was depressed to 30 parts per thousand. For both sites, the rate of oxygen consumption was usually lower at 20° C than at 16°C. For Corpus Christi Pass, the oxygen consumption rates at 24° C were higher than for either of the other two temperatures, except the rate at 16°C in 30 parts per thousand was higher than that at 24° C in the same salinity. For animals from Steadman’s Island, the rates for all measurements at 24° C were higher than for the other two temperatures at the same salinities. For both sites, the highest mean rate of oxygen consumption was at 24° C in 40 parts per thousand salinity. 42 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 0 0 20 50 40 SAL N TV "N PPT Figure 2. Oxygen consumption of whole Clibanarius vittatus from Corpus Christi Pass (top) and Steadman’s Island, not exposed to selenium. Temperatures are 16° C, A; 20° C, □; and 24° C, Oi standard errors are represented A, H and O, respectively. Each open figure represents the mean of five animals. OXYGEN CONSUMPTION IN HERMIT CRAB 43 Figure 3. Oxygen consumption of whole Clibanarius vittatus from Corpus Christi Pass (top) and Steadman’s Island, exposed to selenium. Temperatures are 16°C, A; 20° C, □; and 24° C, O; standard errors are represented by a, ■ and #, respectively. Each open figure represents the mean of five animals. 44 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 The pattern of the oxygen consumption rate for crabs exposed to 100 parts per million selenium for Corpus Christi Pass was depressed, but similar to the pattern for nonexposed animals from the Pass (Fig. 3); the point that did not lie in a similar relationship to the other was in 10 parts per thousand at 16°C, which was depressed to 0.87 milliliters per milligram per hour in the selenium-exposed animals; it was 3.34 in the nonexposed animals. However, rates of oxygen consumption at 20° C were elevated in the selenium-exposed animals in 10 parts per thousand and in 40 parts per thousand salinity, especially in the lower salinity. At 24° C, oxygen consumption increased greatly in 10 parts per thousand salinities, but these rates were still higher than for the nonexposed animals in similar circumstances. In 40 parts per thousand, however, the rate of oxygen consumption for the selenium-exposed animals decreased to 5.69 milliliters per milligram per hour, but rose in nonexposed animals to 10.22 (Fig. 2). Responses of animals from Steadman’s Island exposed to 100 parts per million selenium were generally lower than for the nonexposed animals. At 16°C, the rates of oxygen consumption did not vary greatly among the four salinities, but they were all lower than any of the rates for similarly treated nonexposed animals. At 20° C, the exposed hermit crabs displayed a gradual decrease in rate of oxygen consumption for 10 parts per thousand salinity; there was a sharp increase in consumption in 40 parts per thousand. The rate of oxygen consumption in selenium-exposed animals at 24° C was lowest in 20 parts per thousand salinity and increased at both higher and lower salinities, but was still less than the rates for nonexposed animals. In Clibanarius vittatus from Steadman’s Island, the rates of oxygen consumption tended to increase with temperature (Fig. 3). The rates of consumption for selenium-exposed animals were higher for Corpus Christi Pass than for Steadman’s Island. An analysis of variance (Table 1) indicated that temperature and site as single factors were influential on the oxygen consumption of the hermit crabs (a = .05), but the interactions of selenium and temperature and of selenium and site also were significant. Conclusions The rates of oxygen consumption of Clibanarius vittatus from both sites not exposed to selenium were generally higher at 24° C than at the other two temperatures, as would be expected (Kinne, 1964). However, there was also a trend for the rates at 16°C to be greater than those at 20° C. There was apparently an interaction of 16°C and 30 parts per thousand salinity that produced a peak of oxygen consumption for that temperature. OXYGEN CONSUMPTION IN HERMIT CRAB 45 Table 1. Analysis of variance of effects of temperature (TEMP), salinity (SAL), selenium (SE) and site on the oxygen consumption of Clibanarius vittatus. Source DF SS F PR>F Corrected total 195 4649.778 Model 30 1826.995 3.56' SE 1 19.985 1.17 0.2814 SAL 3 59.880 1.17 0.3241 SE-SAL 3 106.117 2.07 0.1049 TEMP 2 622.124 18.18 0.0001** SE-TEMP 2 127.244 3.72 0.0263** SAL-TEMP 6 202.164 1.97 0.0728 SE-SAL-TEMP 6 136.926 1.33 0.2448 SITE 1 104.525 6.11 0.0145** SE-SITE 1 372.109 21.75 0.0001** SAL-SITE 3 115.877 2.26 0.0823 TEMP-SITE 2 87.495 2.56 0.0806 Error 165 2822.783 R2 = 0.3929 ** Significant (a = .05). The lack of influence of salinity on the oxygen consumption of Clibanarius vittatus is similar to Shumway’s (1978) findings for Pagurus bernhardus in that oxygen consumption returned to near-normal rates after acclimation. The introduction of selenium into the pool of factors did not delete the 30 parts per thousand- 1 6° C peak for animals from Corpus Christi Pass. However, the overall rate at 16°C for the Pass animals was lower than for the nonexposed animals. Generally, rates of oxygen consumption for Corpus Christi Pass selenium-exposed animals increased over those that were not subjected to selenium. In general, oxygen consumption of estaurine invertebrates increases with decreasing salinity (Kinne, 1964). The overall response to salinity was for these animals to consume less oxygen in 30 parts per thousand than in either higher or lower salinities. This is not surprising because that level often is considered an ambient salinity. The most noticeable exceptions to this were at 16°C. Only animals from Steadman’s Island exposed to selenium did not consume less oxygen in 30 parts per thousand than in other salinities, indicating, perhaps, an interaction of 16°C and that concentration of selenium. The overall increase in rate of oxygen consumption of selenium-exposed hermit crabs from Corpus Christi Pass compared to nonexposed crabs from the Pass versus the overall decrease of oxygen consumption of selenium- exposed animals from Steadman’s Island compared to nonexposed animals from the Island was apparently an effect of site, or difference in animal size, inasmuch as the salinity-temperature-selenium regimes were the same for both sites. The two most obvious instances of increase in oxygen 46 THE TEXAS JOURNAL OF SCIENCE — VOL. 39, NO. 1, 1987 consumption at low salinity was of animals from Corpus Christi Pass that were exposed to selenium in 20° C and 24° C, perhaps indicating an interaction of low salinity and selenium for those animals. It is possible that the differences in size of the animals between the two sites had an influence on the oxygen consumed. The somewhat erratic results on the oxygen consumption of Clibanarius vittatus may be explained by the behavior of the animals. There was no way to assure uniform exposure to conditions. The animals had the option of retreating into their own mini-habitats, their shells, to escape the regime imposed on them. This is the most probable explanation of the results obtained for the hermit crabs exposed to 100 parts per million selenium. Shortly after the addition of the compound and its dissolution into the media, the animals became quite active. After approximately two hours, the usual situation upon opening the incubators was to find all animals withdrawn into their shells. Literature Cited Bertine, K. K., and E. E. Goldberg. 1971. Trace elements in clams, mussels and shrimp. Limnol. Oceanogr., 17:877-884. Davenport, J., P.M.C.F. Busschets, and D. F. Cawthorne. 1980. The influence of salinity on behavior and oxygen uptake of the hermit crab, Pagurus bernhardus L. J. Mar. Biol. Assoc. United Kingdom, 60:127-134. Fleming, W. F., W. H. Gutenman, and D. J. Lisk. 1979. Selenium in tissues of woodchucks inhabiting fly-ash landfills. Bull. Environ. Contamin. Tox., 21:1-3. Fowler, S. W., and G. Benayoun. 1976a. Selenium kinetics in marine zooplankton. Marine Science Commun., 2:43-67. - . 1976b. Influence of environmental factors on selenium flux in two marine invertebrates. Mar. Biol., 37:59-68. Giese, P. S. 1968. Cell physiology. W. B. Saunders Co., Philadelphia, 671 pp. Gissel-Nielsen G., and M. Gissel-Nielsen. 1973. Ecological effects of selenium application to field crops. Ambio, 2:1 14-1 17. Green, F. A. Jr., J. W. Anderson, S. R. Petrocelli, B. J. Presley, and R. Sims. 1976. Effect of mercury on the survival, respiration and growth of post larval white shrimp Penaeus setiferus. Mar. Biol., 37:76-81. Gutenmann, W., and C. A. Bache. 1976. Selenium in fly-ash. Science, 191:966-967. King, E. N. 1965. The oxygen consumption of intact crabs and excised gill as a function of decreased salinity. Comp. Biochem. Physiol., 15:93-102. Kinne, O. 1964. The effects of temperature and salinity on marine and brackish water animals: 2. Salinity and temperature-salinity combinations. Oceanogr. Mar. Biol., 2:281-339. Loft, B. 1956. The effects of salinity changes on the respiration rate of the prawn Palaemonetes varians (Leach). J. Exp. Biol., 33:730-736. National Academy of Science. 1976. Selenium. Washington, D.C., xi+203pp. Ott, L. 1977. An introduction to statistical methods and data analysis. Duxbury Press, Belmont, California, xi+730pp. Precht, H. 1973. Temperature and life. Springer- Verlag, New York, 779 pp. Sandholm, M. 1973. Uptake of selenium by aquatic organism. Limnol. Oceanogr., 18:496-499. OXYGEN CONSUMPTION IN HERMIT CRAB 47 Sarojini, R., and R. Nagabhushanam. 1968. Oxygen consumption of Diogenes bicristimanus in relation to environmental conditions. Broteria-Ciencias Naturais, 37: 1 73- 199. Sharp, M. S., and J. M. Neff. 1980. Steady state hemolymph osmotic and chloride ion regulation and percent body water in Clibanarius vittatus. Comp. Biochem. Physiol., 66A:455-460. Shumway, S. E. 1978. Osmotic balance and respiration in the hermit crab Pagurus bernhardus exposed to fluctuating salinities. J. Mar. Biol. Assoc. United Kindgom, 54:860. Sidelnikova, V. N. 1970. Several questions on the aqueous migration of selenium in deserts. In Ockeri goekhimii endozennykh gipergennykh professor (Outlines of geochemical, endogenic, and supergenic processes), 143 pp. A translation of a Russian manuscript. Vernberg, F. J. 1956. Study of the oxygen consumption of excised tissues of certain marine decapod Crustacea in relation to habitat. Physiol. Zool., 29:227-234. Vernberg, W. B., P. J. DeCoursey, M. Kelly and D. M. Johns. 1979. Effects of sublethal concentrations of cadmium on adult Palaemonetes pugio under static and flow through conditions. Bull. Environ. Contamin. Tox., 17:16-24. Vernberg, W. B., P. J. DeCoursey, and W. J. Padgett. 1973. Synergistic effects of environmental variables on larvae of Uca pugilator (Bose). Mar. Biol., 22:307-312. Weiss, R. F. 1970. The solubility of nitrogen, oxygen and argon in water and sea water. Deep-Sea Res., 17:721-735. Wofford, H. W. 1978. Physiological changes in the American oyster Crassostrea virginica during anaerobiosis. Master’s thesis, Texas A&M Univ., College Station, 81 pp. DIURNAL ACTIVITY PATTERNS OF DESERT MULE DEER IN RELATION TO TEMPERATURE Bruce D. Leopold and Paul R. Krausman School of Renewable Natural Resources, University of Arizona, Tucson, Arizona 85721 Abstract. — Desert mule deer ( Odocoileus hemionus crooki) activity in relation to temperature was studied in Big Bend National Park, Texas. No significant decrease in activity resulting from high temperatures occurred during spring and winter. Summer, and to a lesser extent late summer, activity significantly decreased when air temperature exceeded 33° C. Key words : activity patterns; desert mule deer; Odocoileus hemionus ; temperature. Mammals living in most desert regions evade or counter high temperatures through migration, behavioral (Vorhies, 1945; Bartholomew and Cade, 1957), or physiological adaptations (Taylor, 1969a, 1969b; Schmidt-Nielson, 1959). In North American deserts, however, large ungulates may not have developed physiological mechanisms to counter aridity and high temperatures. Knox et al. (1969) stated that the water kinetics of mule deer do not significantly differ from those of other ruminants. Additionally, preliminary investigations of desert mule deer kidneys collected in southwestern Arizona do not indicate any physiological adaptations to arid environments (Krausman, unpublished data). Desert bighorn sheep ( Ovis canadensis nelsoni) also lack physiological water conservation mechanisms (Horst, 1971). Thus, desert mule deer may have developed behavioral mechanisms to counter high temperatures and limited available water. Observations of desert mule deer activity were made in Big Bend National Park, Texas (BBNP), in 1980 and 1981. We analyzed deer activity relative to seasonal changes in temperature patterns to document behavioral responses to temperatures. We hypothesized that desert mule deer were less active during periods of high temperatures compared to periods of low temperature to avoid thermoregulatory stress. Study Area BBNP, Brewster County, is located 113 kilometers from Marathon, Texas. The park is characterized by hot summers, mild winters, and low rainfall. Temperatures often exceed 38° C in the desert regions in summer and are rarely as low as freezing in winter. Precipitation occurs primarily from May through October ranging from < 28 centimeters in the desert and the surrounding foothills to 41 centimeters in the mountainous regions. Two areas in BBNP, Paint Gap Hills and Panther Junction, were studied intensively. The Paint Gap Hills are an igneous intrusion nine kilometers from Panther Junction. Elevations range from 1067 to 1299 meters. Panther Junction is on the flats surrounding the foothills of the northeastern portion of the Chisos Mountains. Elevations range from 884 to 1250 meters. The Texas Journal of Science, Vol. 39, No. 1, February, 1987 50 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Materials and Methods Observations of Deer Independent deer observations were obtained while hiking the study area. Independent observations were of deer not influenced by another observation nor by the actions of the researcher. Only deer initially observed undisturbed and involved in an activity other than bedded were recorded. Fleeing deer were not included in the analysis. For each observation, we recorded air temperature two meters above ground, time, principal activity when initially observed, age (fawn, yearling, adult), and group size. In addition to the independent observations, activity of free-ranging deer at Panther Junction and Paint Gap were observed from dawn until dusk for 17 days in 1980 and 1981 (spring — seven days, summer — five days, late summer — two days, winter — three days). Total minutes individuals spent feeding, standing, and bedded were recorded. We calculated percent time active (feeding and standing) by pooling individuals. The observation period was from dawn until dusk or when deer moved out of sight during their evening feeding period. Partial observations (not lasting from dawn to dusk) were not included in the calculations. Only adult does were included in the analysis. Seasonal Activity Assessment Frequency histograms for the number of deer observed within temperature categories were constructed for each season (spring — February-April, summer — May-July, late summer — August-October, winter — November-January. Sample sizes for each season were relatively equal (spring, 121 days; summer, 107 days; late summer, 110 days; winter, 97 days) thus allowing seasonal comparisons of deer activity. Daily maximum and minimum temperature data were obtained from a weather station at Panther Junction (1140 meters). Contingency table analysis (Zar, 1984) was used to determine differences in frequencies of deer observed within temperature classes between seasons. Results and Discussion Temperature Patterns Monthly maximum and minimum temperatures at Panther Junction during the study periods ranged from 13.9°C to 37.0° C, and 2.8° C to 23.3° C, respectively. The maximum temperature for June and July was 37.2° C for 1980 and 33.6° C in 1981. The average maximum temperature for June and July from 1960-1979 was 33.9° C. Seasonal Activity Patterns In 1980-1981, 749 observations of deer were obtained; however, only 585 observations were of undisturbed deer and subsequently used in the analysis. Deer activity decreased when temperatures exceeded 30° C in summer and 25° C in late summer (Fig. 1). However, similar conclusions regarding activity could be stated for temperatures less than 6°C, but only in winter when temperatures actually fell that low. Diurual temperatures generally reached 33° C in summer and 31°C in late summer within 5 to 7 hours after sunrise. Deer activity decreased when maximum diurnal temperatures were attained. This decrease in activity may have resulted from ambient temperatures approaching or equalling deer body temperature (36-4 1°C) (Anderson, 1980). Other investigators found similar activity changes by deer in response to diurnal and annual temperature patterns (Clark, 1953; Dasmann and Taber, 1956; Miller, 1970). DIURNAL ACTIVITY OF MULE DEER 51 TEMPERATURE (°C) CLASSES Figure 1. Seasonal activity patterns of desert mule deer (DMD) in relation to temperature. (Sp, spring; S, summer; LS, late summer; W, Winter) in Big Bend National Park, Texas, from 1980 to 1981. Numbers on x axis indicate mid-point of temperature class; interval length, 5°C. Total number of observations, 585. Activity between seasons and temperature classes differed (P < 0.001, X2 = 255, df = 15). However, activity in spring and winter did not differ between temperature intervals ( P — 0.134, X2 = 0.843, df = 5). Colder temperatures during these seasons may have permitted deer to be more active during the day. White-tailed deer (O. virginianus) in southern Texas were active 82 percent of the time in winter as compared to 45 percent of the time in summer (Michael, 1970). During the winter in Arizona, desert mule deer spent 60 percent of the day feeding (Clark, 1953). Activity also differed between temperature classes for summer and late summer ( P < 0.001, X2 = 21.8, df = 4). The significant X2 was a result of differences in frequencies of deer sightings within temperature classes when temperature exceeded 26° C. Fewer deer were observed in the higher temperature classes during late summer (57 as opposed to 32 percent). In our study, the percentage of the day spent feeding by adult does increased from spring to winter. Does were active 18 percent of the day in spring (seven days of observation), increasing to 19.7 percent (five days) and 22.7 percent (two days) in summer and late summer, respectively. The highest percentage of the day spent feeding occurred in winter (35.5 percent, three days of observation). This activity may coincide with breeding activities when bucks are actively pursuing does. Twenty-five percent of the movements in winter in the Tucson Mountains, Arizona, was a result 52 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 of the rut (Clark, 1953). It is more difficult to explain why diurnal activity did not decrease in spring. Spring is the season of lowest rainfall, which may lower the nutritional qualities of forage and cause deer to increase foraging activity to meet their energy requirements. Also, group size was largest in spring (3.2) compared to other seasons (summer, 2.4; late summer, 2.1; winter, 3.0). Although the number of observations may be the same within temperature categories, the average group size varied seasonally. Michael (1970) found similar results. Dickinson (1978) stated that the only period of significant inactivity in spring was between five and eight hours after sunrise. Summary and Conclusions Contrasting the activity patterns of desert mule deer with those of deer studied in more temperate regions (Dasmann and Taber, 1956; Michael, 1970; Miller, 1970), we found that the frequency of deer activity decreased when ambient temperatures reached deer body temperatures. Our observations support our initial hypothesis that observed activity patterns are a behavioral mechanism to avoid temperatures that may result in thermoregulatory stress. However, we realize that measuring temperature without also considering thermal radiation and wind is simplistic in design. Further study is needed to assess these activity patterns in a more comprehensive manner. Acknowledgments We would like to thank the Rob and Bessie Welder Wildlife Foundation, the National Park Service, and the Big Bend Natural History Association for their support throughout this study. This study was funded by the Rob and Bessie Welder Wildlife Foundation, Sinton, Texas. This paper is Welder Wildlife Foundation Contribution no. 290. Literature Cited Anderson, A. E. 1980. Morphological and physiological characteristics. Pp. 36-40 in Mule and black-tailed deer of North America (O. C. Wallmo, ed.), Univ. Nebraska Press, Lincoln, xvii+605 pp. Bartholomew, G. A., and T. J. Cade. 1957. Temperature regulation, hibernation, and aestivation in the little pocket mouse ( Perognathus longimembus). J. Mamm., 38:60- 72. Clark, E. S. 1953. A study of the behavior and movements of the Tucson Mountains mule deer. M.S. thesis, Univ. Arizona, Tucson, 1 10 pp. Dasmann, R. F., and R. D. Taber. 1956. Behavior of Columbian black-tailed deer with reference to population ecology. J. Mamm., 37:143-164. Dickinson, T. G. 1978. Seasonal movements, home range, and home range use of desert mule deer in Pecos County, Texas. M.S. thesis, Sul Ross State Univ., Alpine, Texas, 156 pp. Horst, R. 1971. Observations on the kidney of the desert bighorn sheep. Desert Bighorn Counc. Trans., 15:24-37. DIURNAL ACTIVITY OF MULE DEER 53 Knox, K. L., J. G. Nagy, and R. D. Brown. 1969. Water turnover in mule deer. J. Wildlife Manag., 33:389-393. Michael, E. D. 1970. Activity patterns of white-tailed deer in southern Texas. Texas J. Sci., 21:417-428. Miller, F. L. 1970. Distribution of black-tailed deer ( Odocoileus heminous columbianus) in relation to environment. J. Mamm., 51:248-260. Schmidt-Nielson, K. 1959. The physiology of the camel. Sci. Amer., 201:140-151. Taylor, C. R. 1969a. The eland and the oryx. Sci. Amer. 220:88-95. - . 1969b. Metabolism, respiration changes, and water balance of an antelope, the eland. Amer. J. Phys., 217:317-320. Vorhies, C. T. 1945. Water requirements of desert animals in the Southwest. Univ. Arizona Tech. Bull., 107: 487-525. Zar, J. H. 1974. Biostatistical analysis. Prentice-Hall Inc., Englewood Cliffs, New Jersey., 620 pp. LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE, TEXAS, DURING IMPOUNDMENT Michael J. Maceina and Mary F. Cichra Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas 77843 Abstract. — Various water chemistry concentrations, as well as phytoplankton, zooplankton, and benthic macroinvertebrate densities were determined for a 14-month period in Aquilla Lake during impoundment. Although large reductions in total phosphorus and chlorophyll a levels occurred, and water clarity increased as water volume expanded 10-fold during this period, the reservoir was considered to be eutrophic. Secondarily treated sewage effluent appeared to elevate phosphorus and conductivity concentrations in one arm of Aquilla Lake during the early impoundment phase. As the reservoir filled, however, these higher chemical concentrations were not evident. Regression analysis indicated that 60 percent of the variation in chlorophyll a levels was attributable to total phosphorus, total hardness, and to the change in lake volume. Total phosphorus appeared to be the most important parameter regulating algal biomass production. Phytoplankton density changes were not evident with water level increase, but community structure changed over time. Zooplankton abundance was high when compared to other Texas reservoirs. Substrate heterogeneity probably accounted for the differences in abundance of Chironomidae larvae and oligochaetes between the two reservoir arms. Key words : total phosphorous; chlorophyll a; phytoplankton; zooplankton; benthic macroinvertebrates. Typically, initial high productivity rates occur at all trophic levels in new reservoirs (Noble, 1980). Nutrient release from newly-flooded soils, detritus, and terrestrial vegetation, and subsequent autotrophic utilization contribute to eutrophy in recently formed impoundments. Over time, reservoir productivity usually declines, but these changes appear to be at least partially dependent on regional geology, hydraulic flushing rates, nutrient input, and sedimentation rates (Baxter, 1977; Canfield et al., 1982; Hoyer and Jones, 1983). However, only limited information pertaining to limnological conditions in newly impounded reservoirs in the western United States has been reported (Mullan and Applegate, 1965; Cooper et al., 1971; Funk and Gaufin, 1971). In Aquilla Lake, Texas, we measured certain physiochemical variables, as well as phytoplankton, zooplankton, and benthic macroinvertebrate densities, to describe limnological conditions in a new Texas reservoir. In addition, nutrient levels were measured in Hackberry Creek, a tributary of the reservoir that receives secondarily treated sewage effluent from the city of Hillsboro, Texas. Temporal variation in limnological variables was examined with respect to impoundment filling. Influence of sewage effluent on water quality also was determined. The Texas Journal of Science, Vol. 39, No. 1, February, 1987 56 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Study Area Aquilla Lake, a multipurpose U.S. Army Corps of Engineers reservoir in the Brazos River basin, is located in Hill County in east-central Texas (31°55', 97° 10'), approximately 80 kilometers south of Fort Worth (Fig. 1). Impoundment began on 29 April 1983 with water levels reaching conservation pool level (163.9 meters above sea level) on 21 March 1985. At this level, Aquilla Lake has a surface area of 1330 hectares and contains 6450 hectare- meters of water with maximum and mean depths of 15.2 and 4.9 meters, respectively. The dam site, located at the confluence of Aquilla and Hackberry creeks, forms an impoundment containing two large arms, each in a different vegetation region. The Aquilla Arm, located in the Eastern Cross Timbers region, has deep sandy soils with land cover primarily delineated as upland woodlands interspersed with pasture lands (Slack et al., 1986). The Hackberry Arm, located in the Blackland Prairie region, contains alkaline black clay soil with high organic content. Land use surrounding the Hackberry Arm is primarily agricultural cash crops. Secondarily treated sewage effluent is discharged from the city of Hillsboro (population, 7400) approximately five kilometers upstream from the reservoir. The total watershed area above the damsite is 798 square kilometers. Materials and Methods Limnological samples were collected every other month from six stations on Aquilla Lake between February 1984 and April 1985. Permanent sampling stations were established in February 1985 as the lake approached normal conservation pool (Fig. 1). Prior to this time, stations were located further downstream on the Aquilla and Hackberry Creek arms so that sampling was accomplished in lentic, and not lotic waters. Three stations each were located on the Hackberry and Aquilla Creek arms of the reservoir. Temperature and dissolved oxygen measurements were taken with a YSI Model 57 oxygen- temperature meter. Water transparency was measured to the nearest five centimeters with a 20-centimeter Secchi disk. Water level readings were obtained from the U.S. Army Corps of Engineers office in Whitney, Texas. Reservoir area and volume determinations were obtained from data provided by the U.S. Army Corps of Engineers (1978). Surface (0.5 meter) water samples were collected in acid-washed, one-liter Nalgene sampling bottles and placed on ice for analysis the next day in the laboratory. Lake pH was determined with a Cole-Palmer pH meter calibrated with standard buffers of pH 7.0 and 10.0. Conductivity was measured with an Electronic Switchgear Mark IV Conductivity Measuring Bridge standardized to 25° C. Total hardness, total alkalinity, nitrate-nitrite, ammonia, and organic nitrogen and total phosphorus concentrations were determined utilizing the procedures presented by Hach Chemical Company (1979). Total nitrogen was considered to be the sum of nitrate-nitrite, ammonia, and organic nitrogen. Total nitrogen to total phosphorus ratios were computed using weight concentrations. In September 1984, the effluent discharge rate (cubic meters per second) from the City of Hillsboro sewage treatment plant into Hackberry Creek was estimated and water samples were collected approximately two kilometers downstream from the discharge. Flow in Hackberry Creek during dry conditions in August and September was due entirely to plant discharge. During this time, discharge from Aquilla Creek was nil with intermittent, stagnant pools found along the creek bottom. Total phosphorus and nitrogen levels were determined in the laboratory and daily and annual nutrient loading rates from effluent discharge were derived for Aquilla Lake. Unpublished water chemistry data were provided by the U.S. Geological Survey. To determine chlorophyll a values, a known volume of water was filtered through a 12.5- centimeter Gelman GF/A glassfiber filter that retained particles larger than 0.3 /j. meters. Samples were frozen in a dessicator dish with analysis and calculations performed according to standard methods (A.P.H.A., 1976). Chlorophyll a values were not corrected for pheophytin. To determine the composition of the phytoplankton community, a single surface LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 57 Figure 1. Location and map of Aquilla Lake. Numeric values represent limnological sampling stations. Normal conservation pool elevation is 163.9 meters above sea level. (0.5 meter) 200-milliliter water sample was collected at each station and preserved in buffered Lugol’s solution. In the laboratory, a 10-milliliter aliquot from each sample was settled for 24 hours, and examined on an inverted microscope at X400. Phytoplankton were identified and enumerated in randomly selected strips until the count of the most abundant taxon approximated 100 (Lind, 1974). Keys used for identification were Prescott (1978) and Whitford and Schumacher (1984). Zooplankton were collected using a 153-ju meter Wisconsin plankton net. Triplicate vertical tows were taken from the reservoir bottom to the surface. When water depth exceeded five meters, tow distance was 5.0 meters. Samples were preserved in five percent formalin-rose bengal solution. In the laboratory, a 5.0-milliliter sample was placed on a Ward circular wheel, and zooplankton were identified, and counted using a binocular dissecting scope at X30-X60. Cladocerans were identified to genus when possible using the keys provided by Pennak (1978). Calanoids, cyclopoids, copepod nauplii, and rotifers were enumerated as such. Zooplankton counts were converted to concentrations (numbers per cubic meter) assuming 100 percent net efficiency. Triplicate benthic samples were collected at all stations with a 232 square centimeter Ponar dredge and preserved in a 10 percent formalin-rose bengal solution. Samples were washed in a U.S. Standard no. 30 seive, and organisms removed and placed in 70 percent ethanol. Organisms were counted under a binocular dissecting microscope at X20 and identified to the most practical taxonomic group using the keys of Pennak (1978) and Merritt and Cummins (1978). Data were analyzed using the statistical procedures published by the SAS Instutute Inc. (1985). Unless otherwise stated, statistical significance is defined as P < 0.05. Mean values for certain data ranged over several orders of magnitude and the variances were proportional to the means. Logio transformation of these data was performed and geometric means were computed. Unless otherwise stated, however, mean values represent arithmetric means. Duncan’s multiple range tests and t-tests were used to test differences among and between mean values. Results and Discussions Physio chemical Parameters Temperature and oxygen profiles indicated that Aquilla Lake is a warm monomictic reservoir (Wetzel, 1975). Oxygen stratification was evident 58 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 1. Mean, standard deviation, minimum, and maximum values for water chemistry parameters, chlorophyll a, and secchi disk readings collected from Aquilla Lake, February 1984 to April 1985. Parameter Mean Standard deviation Minimum Maximum pH 8.2 0.5 7.5 9.1 Conductivity (/zS) 600 140 440 980 Total hardness (mg.L-1 as CaCCL) 140 10 120 170 Total alkalinity (mg.L-1 as CaCCL) 140 20 100 180 Total nitrogen (mg.L 1 as N) 2.29 1.78 0.28 7.13 Total phosphorus (mg.L-1) 0.16 0.26 0.01 0.91 Chlorophyll a (mg.m3) 31 19 1.9 107 Secchi (cm) 50 15 20 85 during the summer 1984, however, maximum thermal variation in the water column was only 1.5°C. Thermal stratification probably did not occur in summer months due to wind mixing in this shallow lake. Between June and August 1984, mean depth was only 2.7 meters. During the study period, water quality parameters were highly variable (Table 1). Based on pH, total hardness, total alkalinity, and conductivity values, Aquilla Lake is moderately hard-water, well-buffered reservoir with a high mineral content. Mean total phosphorus (TP), total nitrogen (TN), and chlorophyll a (Chk) concentrations indicate that Aquilla Lake is highly eutrophic according to criteria established by Forsberg and Ryding (1980). These authors classified lakes as eutrophic when TP, TN, and Chi# concentrations exceeded 0.025 milligrams per liter, 0.60 milligrams per liter, and seven milligrams per cubic meter, respectively. The Trophic State Index (TSI) values described by Carlson (1977) for TP, Chla, and secchi disk readings were 77, 64, and 70, respectively, indicating that Aquilla Lake is eutrophic. TSI values are scaled from 0 to 100, where 0 represents extreme oligotrophy and 100 is hypereutrophy. Temporal water quality changes were apparent during the study period (Fig. 2). Lake pH levels declined from 8.8 to 9.0 in early 1984 to 7.6 to 8.3 by late 1984-early 1985. Conductivity readings generally declined from a high of 930 juS in February 1984 to 560 /jlS in April 1985. Total hardness fluctuated somewhat during the 14 month period, however, large changes were not observed. TN increased from a mean concentration of 0.50 milligrams per liter August 1984 to 4.56 milligrams per liter in December 1984. Mean TP concentrations declined from 0.78 milligrams per liter in February 1984 to 0.03 milligrams per liter in April 1985. Changes in many water chemistry parameters appeared to be influenced by increases in water level. Reservoir area increased from 309 hectares to 1359 hectares and volume from 664 hectare-meters to 6730 hectare- meters during this study (Fig. 2). Conductivity, TP, and total alkalinity were negatively correlated with the increase in reservoir area (Table 2). LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 59 Figure 2. Water chemistry concentrations, chlorophylla a, and secchi disk depth changes in Aquilla Lake compared to changes in lake volume and area. Water level increases probably caused reductions in conductivity and TP concentrations due to dilution. Nitrate-nitrite was positively correlated with water level and volume. Newly flooded terrestrial vegetation possibly elevated nitrate-nitrite concentrations as decomposition occurred. Funk and Gaufin (1971) reported elevated nitrate-nitrogen levels after inflow of runoff water in a new Wyoming reservoir. Leaching from the watershed also may account for variation in chemcial concentrations over time. Mean conductivity and TP concentrations were higher in the Hackberry Arm than in the Aquilla Arm in February and April 1984 (Table 3). After April 1984, when water levels increased, conductivity and TP levels were relatively similar between arms although some significant differences were observed in December and February 1984-85. Higher TP and conductivity concentrations measured in the Hackberry Arm in early 1984 when Aquilla Lake water level was low, may be due in part to sewage effluent. Based on a discharge rate of 0.0375 cubic meters per second and a TP concentration of 0.96 grams per cubic meter, as measured in September 1984, the annual phosphorus loading rate into Aquilla Lake from the Hillsboro sewage treatment plant was computed to be 1130 kilograms. This value assumes a discharge rate of 1.2 X 106 cubic meters per year, which 60 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 2. Pearson product-moment correlation coefficients relating water quality parameters, chlorophyll a concentration, and secchi disk readings to change in lake area and volume in Aquilla Lake. Parameter Aquilla Lake Area (ha) Volume (ha-m) Conductivity -0.33* Total phosphorous -0.39** -0.34* Total alkalinity -0.51** -0.50** Nitrate-nitrite 0.76** 0.75** Chlorophyll a -0.49** -0.46** Secchi 0.66** 0.65** *P<0.05, **P<0.01 was similar to rates reported by the U.S. Geological Survey (1983). Assuming no sedimentary loss of phosphorus and a hydraulic flushing rate of one year, effluent input would elevate TP concentrations in Aquilla Lake approximately 0.02 milligrams per liter. This TP level is comparable to reservoir concentrations measured in April 1985. Water chemistry data collected between February 1984 and April 1985 on Hackberry Creek indicated average TP concentration was approximately 17 times higher below the effluent discharge site than above (U.S. Geological Survey, unpublished data). Hence, Hackberry Creek probably will serve as a major source of phosphorus in Aquilla Lake, although some sedimentary phosphorus loss can be expected (Vollenweider, 1969). In addition, conductivity, sodium, postassium, chloride, total dissolved solids, and TN levels were higher in Hackberry Creek below the effluent discharge site when compared to concentrations collected above the site (U.S. Geological Survey, unpublished data). Thus, higher conductivity readings measured in the Hackberry Arm in early 1984 may have been due to sewage discharge. In September 1984, TN concentration was 12.5 grams per cubic meter below the effluent discharge site on Hackberry Creek with a computed annual loading rate of 14,800 kilograms of nitrogen. TN differences between arms, however, were not evident. This suggests that nitrogen assimilation or loss occurred before reaching the first upper Hackberry Arm station, which is 1 1 kilometers downstream from the discharge site. Differences in soil types, as well as land use practices between Hackberry and Aquilla creeks also may have influenced nutrient and chemical concentration in the reservoir. The rapid rise in water levels probably negated some of the water quality differences that might have been observed between arms. During this study period, water clarity in Aquilla Lake generally increased from a mean Secchi disk reading of 25 to 80 centimeters (Fig. 2). Pearson product-moment correlation analysis indicated a significant positive relationship (r = 0.66) between Secchi disk readings and reservoir area during the study period (Table 2). Only a weak ( P =0.09) negative LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 61 relationship (r =—0.24) was observed between Secchi disk readings and Chi a levels. Generally, higher ChD concentrations reduce water clarity in lakes (Canfield and Bachmann, 1981). Increases in water clarity associated with rising water levels probably were caused by dilution and a reduction of clay turbidity due to addition of organic material. In addition, the increase in mean depth may have resulted in less sediment disturbance. During the study period, mean depth increased from 2.0 to 4.9 meters. Wind related turbidity is usually greater in shallow lakes. Thus, higher water levels might have reduced the effects of wind on water clarity. Phytoplankton Chk? concentrations, a measure of phytoplankton biomass, averaged 31 milligrams per cubic meter during the study period (Table 1). However, concentrations were highly variable as individual station values ranged from 1.9 to 107 milligrams per cubic meter. Peak Chla levels were observed in February 1984, averaging 62 milligrams per cubic meter, and declined to 25 to 30 milligrams per cubic meter in summer-autumn of 1984 (Fig. 2) . Reservoir volume increased 147 percent between December 1984 and February 1985 while mean Ch \a values declined from 29 to four milligrams per cubic meter; however, mean concentration increased to 25 milligrams per cubic meter by April 1985. Ch\a was positively correlated (r =0.68, P <0.01) to TP and also positively, but weakly correlated (r =0.27, P =0.05) to total hardness concentrations. LogioChD levels were negatively correlated (r =—0.45, P <0.01) to water volume changes in the reservoir. The multivariate regression equation was computed: Chla = - 69.0 + 49.1 TP + 0.66 THARD - 0.055 CHAVOL where Chkz = chlorophyll a concentration in mg.m’3 TP = total phosphorous in mg.L-1 THARD = total hardness in mg.L”1 CHAVOL = the percent change in lake volume. This equation was highly significant (P <0.01) as these three variables explained 60 percent (r2 =0.60) of the Chi a variation during the study period. The variables, TP, THARD, and CHAVOL, explained 46, 11, and three percent of Chla variation, respectively. These data indicate that TP was the most important parameter affecting Ch \a concentrations. Analogous with greater TP concentrations, significantly higher ChD levels were observed in the Hackberry Arm in February and April 1984 (Table 3) . Turbidity, dissolved solid concentrations, and other nutrients, such as TN, also effect Chla levels in lakes (Baxter, 1977; Hoyer and Jones, 1983). The mean TN:TP ratio was highly variable during the study period ranging 62 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 3. Mean water quality concentrations in Hackberry and Aquilla arms at Aquilla Lake. Parameter Arm 1984 1985 Feb Apr Jun Aug Oct Dec Feb Apr Conductivity Hackberry 970 740 530 560 600 460 590 570 (MS) * * * Aquilla 890 550 540 540 570 440 520 560 Total phosphorous Hackberry 0.88 0.13 0.12 0.07 0.07 0.06 0.06 0.03 (mg.L-1) * * * * Aquilla 0.69 0.06 0.02 0.04 0.06 0.05 0.09 0.02 Chlorophyll a Hackberry 80 52 27 27 25 32 5.0 26 (mg.irf3) * * Aquilla 43 44 23 33 24 26 3.9 25 *t-test indicates higher mean value is significantly (P<0.05) greater than corresponding lower mean value for a particular sampling date. from 1.3 to 91 (Fig. 3). When TN:TP ratios are less than 10, TN becomes limiting to algal production (Sakamoto, 1966). This was evident between February 1984 and August 1984. When TN:TP ratios are between 10 and 17, either nutrient can be limiting. When the ratio exceeds 17, TP is generally considered to limit phytoplankton production. The TN:TP ratio was greater than 17 between October 1984 and April 1985 suggesting TP limitation. Regression analysis indicated TP to be the most important parameter regulating Chla levels. Therefore, as Aquilla Lake reached conservation pool level, phosphorus appeared to be limiting algal biomass production. Stauffer (1985) also reported higher phytoplankton production in lakes that had greater calcium carbonate concentrations. Our regression model suggests total hardness was a moderately important parameter regulating Chkz levels in Aquilla Lake. Rapid increase in lake volume affected Chkz directly by dilution of the phytoplankton community and indirectly by reducing TP levels in the lake. Total phytoplankton abundance was highly variable as geometric mean densities ranged from 1.73 X 103 cells per milliliter in June 1984 to 7.04 X 103 cells per milliliter in October 1984 (Table 4). Chkz concentration was only moderately correlated (r =0.32, P =0.03) to total phytoplankton density. Cellular Chkz content can be highly variable among different phytoplankton taxa (Canfield et. al., 1985), which probably accounted for this weak relationship. Phytoplankton community structure changed over time as chlorophytes dominated the flora in February 1984. At this time, Chlamydomonas and Ankistrodesmus numerically comprised 70 percent and 24 percent of the plankton flora, respectively. Cyclotella chains were the predominant algal form in April 1984 (43 percent), December 1984 (36 percent), and February 1985 (73 percent). In April 1985, Cyclotella LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 63 0_J—i i i i i i i i FEB APR JUN AUG OCT DEC FEB APR 1984 1985 Figure 3. Mean Oligochaeta and Chironomidae larvae densities between Aquilla and Hackberry arms. The “a” indicates the higher mean value is significantly (/^O.OS) greater than the lower mean value for a particular sampling date. singles comprised 80 percent of the phytoplankton community. Nitzchia and Melosira diatoms also were numerically abundant during peak Chrysophyta densities. The cyanophytes, Raphidiopsis and Oscillatoria , were the predominant algae collected in August 1984 (31 and 40 percent, respectively) and in October 1984 (42 and 30 percent, respectively). Changes in total phytoplankton density were not evident with water level increase, although seasonal as well as water level increase appeared to alter phytoplankton community structure. Invertebrates Five major genera of Cladocera ( Daphnia , Bosmina, Ceriodaphnia, Diaphanosoma, and Chydorus) were collected from Aquilla Lake during the study period. Other cladoceran genera that were identified, but the low abundance of which prevented statistical analysis, included Alona, Alonella , and Pleuroxus. Chronological differences in Daphnia abundance 64 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 4. Geometric mean density (number per milliliter) of major phytoplankton classes in Lake Aquilla. Geometric mean values in columns followed by the same letter are not significantly (P>0.05) different. Taxon Month Chlorophyta Cyanophyta Chrysophyta Total phytoplankton 1984 Feb 4369a 23c 60c 4635b Apr 1002b 35c 2790a 4439b June 280d 236b 79 lb 1727d Aug 2 1 7d 3954a 656b 5254ab Oct 819b 54 1 9a 5 1 3b 7036a Dec 46 lc 123b 2325a 3 1 32c 1985 Feb 327cd ld 2382a 2802c Apr 249d 0d 4747a 5248ab were evident as greatest mean density occurred in April 1984 (Table 5). Daphnia declined during the summer months, then increased between October 1984 and April 1985. Significantly lower numbers of Daphnia were collected in April 1985 than in April 1984, however, no differences were observed between February sampling dates. Bosmina displayed similar temporal differences as peak abundance occurred in February and April 1985. Bosmina density was significantly higher in February 1985 when compared to February 1984. Differences were not observed between April sampling dates. Maximum Diaphanosoma abundance occurred in June 1984, then declined during winter and spring 1985. Ceriodaphnia density was extremely low between February and August 1984, but increased in late 1984. Ceriodaphnia peaked in April 1985 and this organism became the dominant cladoceran in the lake. Chydorus was relatively abundant in early 1984, however, few of these organisms were collected after that time. The rapid increase in water levels that occurred in early 1984 and early 1985 did not appear to reduce total cladoceran density. The density of Cladocera was significantly higher in April 1984 as compared to April 1985 (Table 5); however, Cladocera abundance in February 1985 was greater than that observed in February 1984. Chydorus , which is a small, shallow- water, littoral cladoceran (Pennak, 1978), appeared to be negatively affected by increase in water level. During the study period, mean depth increased from 2.0 to 4.9 meters and pelagic sampling stations were deeper by the end of the study. Thus, a reduction in the open-water Chydorus populations might be expected. The zooplankton community was not sampled in shallow water where higher Chydorus densities possibly occurred. Maximum calanoid, cyclopoid, and nauplii Copepoda density occurred in April 1984 (Table 6). Abundance declined in summer and autumn of LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 65 Table 5. Geometric mean density (number per cubic meter) of Cladocera zooplankton collected in Aquilla Lake. Geometric mean values in columns followed by the same letter are not significantly (P>0.05) different. Taxon Month Daphnia Bosmina Diaphano- soma Cerioda- phnia Chydorus Total Cladocera 1984 Feb 5500b 200b 0e 0e 130b 7700e Apr 80,220a 18,300a 0e 4d 5700a lll,000a Jun 23d 190° lQ,700a 3d 2cd 12,300d Aug 120c 4C 4700b 4d 0d 5500f Oct 4300b 220b 4100bc 2300b lcd 13,300d Dec 6400b 12,300a 2200bc 2800b od 25,300c 1985 Feb 3600b 21,200a 9d 310c 4C 26,500c Apr 6600b 20,900a 1800c 30,800a 0d 62,900b 1984, but generally increased between December 1984 and April 1985. Nauplii density declined significantly in April 1985 when compared to April 1984, but a difference was not apparent for February sampling dates. Similarly, calanoid and cyclopoid numbers were significantly lower in April 1985 than in April 1984. Greater numbers of calanoids were collected in February 1985 when compared to 1984, but cyclopoids were less abundant in February 1985 when compared to mean density one year earlier. Total copepod density was two-fold and five-fold lower in February and April 1985 when compared to corresponding months in 1984. Rotifers, which included the genera Asplanchna, Polyarthra, Hexarthra, Keratella, Conochilus, and Brachionus, declined dramatically during the study period (Table 6). Peak rotifer abundance occurred in April 1984. A 25- to 50- fold reduction in rotifer density was observed when compared to data collected in early 1984. Similarly, total zooplankton abundance was highest in April 1984. Lowest densities were recorded in summer 1984 and gradually increased from October 1984 to April 1985 (Table 6). In Aquilla Lake, the annual mean density of zooplankton collected from February to December was 125 organisms per liter. This density was considerably higher than that reported for six Texas reservoirs, where average annual densities ranged from 22 to 60 organisms per liter as reviewed by Cichra et al. (1985). High levels of allochthonous and autochthonous organic material probably were present during reservoir filling and impoundment. This material, in addition to relatively high ChD concentrations, provided ideal conditions for zooplankton population expansion in Aquilla Lake. Although water levels increased in early 1985, chronological reductions in copepods and rotifers may not have been directly due to this factor. Between February and April 1984 reservoir area and volume increased 65 66 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 6. Geometric mean density (number per cubic meter) of Copepoda, Rotifera, and total zooplankton collected in Aquilla Lake. Geometric mean values in columns followed by the same letter are not significantly (PX).05) different. Taxon Month Copepods nauplii Calanoida Cyclopoida Copepoda Rotifera Total zooplankton 1984 Feb 1300bcd 1900e 42,000b 46,300b 5 1 ,200b 107,000b Apr 14,900a 36,900a 174,000a 232,000a lll,000a 459,000a Jun 140e 2100e 2200e 5300f 1900d 20,200e Aug 130cd 5500d 2400e 9300e 4700c 23,200e Oct 460d 8600c 4100d 14,500d 2100d 30,700d Dec 2800b 8100c 1 l,300c 22,900c 4200c 53,200c 1985 Feb 1700bc 10,700bc 10,300c 24,600c 2000d 55,200c Apr 2100bc 19,000b 16,500c 43,400b 2100d 1 13,000b and 125 percent, respectively. Maximum zooplankton abundance occurred in April 1984 indicating the change in reservoir area did not reduce zooplankton density. Between December 1984 and February 1985, lake area and volume increased 73 and 147 percent. Although many other variables may be involved in regulating zooplankton abundance and composition, reduction in Chk* levels possibly was an important factor in the decline in zooplankton. In February and April 1984, Chla averaged 55 milligrams per cubic meter. One year later, average Chi a concentration for the same time period was 15 milligrams per cubic meter. Correlation analysis indicated cyclopoids (r =0.44), rotifers (r =0.55) and total zooplankton abundance (r =0.40) all were positively, but weakly, correlated with Chltf levels. Higher Ch \a levels during this time period would be advantageous to herbivorous zooplankton. Noonan (1979) reported a positive relationship (r =0.51) between total zooplankton abundance and Chi a in natural and artificial midwestern temperate lakes. Canfield and Watkins (1984) also found a positive correlation (r =0.66) between Chkz and zooplankton in natural Florida lakes. Short term variation in zooplankton density, as well as selective grazing by fish and invertebrate planktivores, also may have affeted the zooplankton community in Aquilla Lake. Spatial differences in zooplankton abundance were not evident during the study period. Predominant benthic macroinvertebrates collected were Oligochaeta and Diptera larvae, including Chironomidae and Chaoborus (Table 7). Comparison of geometric mean densities for pooled data between arms indicated chironomids to be three times more abundant in the Aquilla Arm. Oligochaeta density was four times higher in the Hackberry Arm. Nondipteran immature insects, including Ephemeroptera, Trichoptera, and Odonata, were numerically more abundant in the Aquilla Arm (Table 7). LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 67 Table 7. Geometric mean density (number per square meter) of dominant macroinvertebrates collected from the Hackberry (stations 1-3) and Aquilla (stations 4- 6) arms, February 1984 to April 1985. Nondipteran Insecta include immature Ephemeroptera, Trichoptera, and Odonata. Taxon Arm Hackberry Aquilla Chironomidae larvae 185 * 540 Chaoborus larvae 6 * 13 Nondipteran Insecta 1 * 3 Oligochaeta 771 * 170 Total benthic invertebrates 1462 * 1157 *t-test indicates higher mean value is significantly (P<0.05) greater than corresponding lower mean for a particular taxon. Seasonal and spatial density differences were evident as chironomids were found in greater numbers in the Aquilla Arm for six out of eight sampling dates (Fig. 3). Substrate differences probably account for benthic macroinvertebrate heterogeneity between arms in Aquilla Lake. Correlation between bottom oxygen concentrations and benthic organism density was not evident. The predominantly clay substrate in the Hackberry Arm provides suitable burrowing habitat for aquatic oligochaetes (Wetzel, 1975). Nondipteran Insecta in this study, with the exception of Hexagenia , all inhabit the bottom substrate surface. The harder bottom type in the Aquilla Arm may be more suitable to these organisms. Although oligochaete and chironomid densities appeared generally to decline during the study period (Fig. 3), significant differences were not evident. Therefore, impact of rising water levels on these organism was not apparent. Conclusion Although reductions in TP and Chi a concentrations occurred as Aquilla Lake filled, the reservoir was considered eutrophic during the entire impoundment period. Spatial differences for some chemical and Chi a levels were evident early in the impoundment phase. Secondarily treated sewage effluent from Hackberry Creek appeared to cause TP and conductivity concentrations to be higher in the Hackberry Arm during the early phase of impoundment. After this time, however, rising water levels probably negated water quality differences between arms. The hydraulic flushing rate in Aquilla Lake is predicted to be greater than one year; therefore, phosphorus from sewage effluent will be a major nutrient source for the reservoir. High algal chlorophyll productivity will probably be maintained as TP was the most important parameter regulating Chi a levels. Zooplankton densities in Aquilla Lake were higher when compared to other Texas reservoirs. We expect zooplankton abundance to decline as 68 THE TEXAS JOURNAL OF SCIENCE VOL. 39, NO. 1, 1987 planktivorous fish populations become established and reduction in newly- flooded organic material occurs. Because Chla was moderately correlated with zooplankton, anticipated elevated levels of planktonic algae will contribute to zooplankton production. Spatial heterogeneity in the profundal benthos was possibly attributable to substrate differences, and not to water quality or reservoir level increases. Acknowledgments Financial support for this study was provided by the Fort Worth District Corps of Engineers, contract no. DACW-29-C-0410. Bob Betsill, Sheila Bell, Mike Hoy, and Scott Smith assisted in field collections. William Clark and Bob Betsill offered valuable suggestions to the manuscript, and Sally Kim helped in the preparation of the paper. This manuscript represents contribution no. TA 21738 of the Texas Agricultural Experiment Station, Texas A&M University. Literature Cited American Public Health Association. 1976. Standard methods for the examination of water and wastewater. 14th ed., American Public Health Association, Washington, D.C., 1193 pp. Baxter, R. M. 1977. Environmental effects of dams and impoundments. Ann. Rev. Ecol. Syst., 8:255-283. Canfield, D. E. Jr., and R. W. Bachmann. 1981. Predictions of total phosphorus concentrations, chlorophyll a and Secchi depth in natural and artificial lakes. Canadian J. Fish. Aquat. Sci., 38:414-423. Canfield, D. E. Jr., J. R. Jones, and R. W. Bachmann. 1982. Sedimentary losses of phosphorus in some natural and artificial Iowa lakes. Hydrobiol., 87:65-76. Canfield, D. E. Jr., and C. E. Watkins. 1984. Relationships between zooplankton abundance and chlorophyll a concentrations in Florida Lakes. J. Freshwater Ecol., 2:335- 344. Canfield, D. E. Jr., S. B. Linda, and L. M. Hodges. 1985. Chlorophyll-biomass-nutrient relationships for natural assemblages of Florida phytoplankton. Water Res. Bull., 21:381- 391. Carlson, R. E. 1977. A trophic state index for lakes. Limnol. Oceanogr., 22:361-369. Cichra, M. F., J. M. Campbell, P. W. Bettoli, and W. J. Clark. 1985. Zooplankton community structure and dynamics in Lake Conroe, Texas. Texas J. Sci., 36:235-246. Cooper, W. A., R. S. Hestand, and C. E. Newton. 1971. Chemical limnology of a developing reservoir (Lake Meredith) in the Texas Panhandle. Texas J. Sci., 23:241- 251. Forsberg, C., and S. Ryding. 1980. Eutrophication parameters and trophic state indices in 30 Swedish waste-receiving lakes. Arch. Hydrobiol., 80:189-207. Funk, W. H., and A. R. Gaufin. 1971. Phytoplankton productivity in a Wyoming cooling- water reservoir Pp. 167-178, in Reservoir fisheries and limnology (G. E. Hall, ed.), Spec. Publ. 8, Amer. Fish. Soc., Washington, D.C., 511 pp. Hach Chemical Company. 1979. Water analysis handbook. Hach Chemical Company, Loveland, Colorado. Hoyer, M. V., and J. R. Jones. 1983. Factors affecting the relation between phosphorus and chlorophyll a in midwestern reservoirs. Canadian J. Fish. Aquat. Sci., 40:192-199. Lind, O. T. 1974. Handbook of common methods in limnology. C. V. Mosby Press, St. Louis, 154 pp. LIMNOLOGICAL CHARACTERISTICS OF AQUILLA LAKE 69 Merritt, R. W., and K. W. Cummins. 1978. An introduction to the aquatic insects of North America. Kendall/ Hunt Publishing Co., Dubuque, Iowa, 441 pp. Mullan, J. W., and R. L. Applegate. 1965. The physical-chemical limnology of a new reservoir (Beaver) and a fourteen-year-old reservoir (Bull Shoals) located on the White River, Arkansas and Missouri. Proc. Ann. Conf. Southeastern Assoc. Game Fish Comm., 19:413-421. Noble, R. L. 1980. Management of lakes, reservoir, and ponds. Pp. 265-296, in Fisheries management (R. T. Lackey and L. A. Nielsen, eds.), John Wiley & Sons, New York, 422 pp. Noonan, T. A. 1979. Crustacean zooplankton and chlorophyll a relationships in some Iowa lakes and reservoirs. Unpublished M.S. thesis, Iowa State University, Ames, 80 pp. Pennak, R. W. 1978. Freshwater invertebrates of the United States. John Wiley & Sons, New York, 2nd ed., 803 pp. Prescott, G. W. 1978. How to know the freshwater algae. W. C. Brown Co., Dubuque, Iowa, 3rd ed., 293 pp. Sakamoto, M. 1966. Primary production by phytoplankton community in some Japanese lakes and its dependence on lake depth. Arch. Hydrobiol., 62:1-28. SAS Institute, Inc. 1985. SAS user’s guide: statistics, version 5 ed. SAS Institute, Inc., Cary, North Carolina, 1209 pp. Slack, R. D., M. J. Maceina, and M. D. Hoy. 1986. Post-impoundment environmental study of Aquilla Lake. Final report submitted to U.S. Army Corps of Engineers, Fort Worth, 238 pp. Stauffer, R. E. 1985. Relationships between phosphorus loading and trophic state in calcareous lakes of southeast Wisconsin. Limnol. Oceanogr., 30:123-145. U.S. Army Corps of Engineers. 1978. Aquilla Lake design memorandum no. 9 master plan. U.S. Army Engineers District, Fort Worth Corps of Engineers, Fort Worth. U.S. Geological Survey. 1983. Water resources data for Texas. U.S. Geological Survey, Austin, Texas, 2:1-407. Vollenweider, R. A. 1969. Possibilities and limits of elementary models concerning the budget of substances in lakes. Arch. Hydrobiol., 66:1-36. Wetzel, R. G. 1975. Limnology. W. B. Saunders Co., Philadelphia, 743 pp. Whitford, L. A., and G. J. Schumacher. 1984. A manual of freshwater algae, revised edition. Sparks Press, Raleigh, North Carolina, 337 pp. POLLEN ANALYSIS OF LATE-HOLOCENE SEDIMENTS FROM A CENTRAL TEXAS BOG Richard G. Holloway, L. Mark Raab, and Robert Stuckenrath Palynology Laboratory, Anthropology Department, Texas A&M University, College Station Texas, 77843, Anthropology Department, California State University, Northridge California, 91330, and Radiocarbon Laboratory, University of Pittsburgh, Pittsburgh, Pennsylvania Abstract. — Pollen analysis of a short sedimentary column from Weakly Bog, Leon Co., Texas, has provided important new evidence concerning the late-Holocene vegetational changes of central Texas. Utilization of pollen influx values have provided evidence for a shift from oak-woodland to savannah-like plant communities between 1500 and 2000 years ago, which is interpreted as the establishment of the present Post Oak Savannah. Documentation of this shift in vegetation is correlated with existing models of gradually increasing aridity and drying that have been proposed elsewhere for eastern Texas and the southern plains. Key words: pollen analysis; vegetational changes; late-Holocene; Texas. Today, the region of east-central Texas is ecologically important because it is located along the prairie-forest ecotone. It was equally important in the past because it stood between the forests of eastern Texas and the dry scrub plant communities of west-central Texas. As early as the 1940’s, Potzger and Tharp (1943, 1947, 1954) reported the recovery of fossil pollen from late-glacial and Holocene sediments from Sphagnum- bog localities in Lee, Robertson, and Milam counties, Texas. This pioneering pollen work was followed by the pollen analysis of Sofje Bog, located in Gonzales County, Texas, by Graham and Heimsch (1960) and their re-examination of some sediments from Gause Bog that had been examined previously by Potzger and Tharp (1954). This re-examination was an attempt to verify the reported occurrence of boreal conifers in late-glacial Texas sediments from this region as originally reported by Potzger and Tharp (1943). Larson et al. (1972) reported on the pollen analysis of Hershop Bog, also located in Gonzales County, Texas. That deposit provided an almost continuous pollen record dating to 12,000 years B.P. and indicated a distinct vegetational change occurred about 10,000 yrs. B.P. from an oak-parkland to an oak-savannah dominated assemblage. Unfortunately, the most recent 2000 years of sediment were missing from this record. According to Larson et al. (1972) the bog suface vegetation had been destroyed by an attempt to drain the bog, and concomitant lowering of the water table may have destroyed the pollen assemblage for this crucial, most recent time period (Holloway, 1981). This same phenomenon was encountered by Bryant (1977) in his examination of the pollent record from Boriack Bog, Lee Co., Texas. Again, due to draining and later over-grazing of the bog area (Bryant and The Texas Journal of Science, Vol. 39, No. 1, February, 1987 72 THE TEXAS JOURNAL OF SCIENCE — VOL. 39, NO. 1, 1987 Holloway, 1985), the upper 2000 years of the pollen record were not reported in the pollen assemblage from this core. The pollen records from earlier periods are good, yet no well-documented pollen assemblages representing the last 2000 to 2500 years of deposition are extant from this region of Texas. The pollen record from Weakly Bog, Leon Co., Texas, thus is significant because it helps close the gap in our knowledge of the paleoenvironmental history of this area by providing new information on this crucial time period in the vegetational history of central Texas during the past 2400 years. Earlier pollen analytical studies generally reported palynological data in the form of percentage frequencies. Recently, with the advent of more precise dating methods, a better quantitative method for reporting data has become available. This method, which involves calculation of pollen influx values, was initially developed by Benninghoff (1962) and later refined by Davis (1969). Although initially designed to deal with lacustrine sediments, this method has been utilized in analyzing bog sediments. The present study is important inasmuch as it represents the first time that this statistical method has been utilized on bog sediments from Texas and allows an improved interpretation of the paleovegetational changes that have occurred within the state. Methods and Materials Based on evidence provided by the Espey Huston and Associates Company, Weakly Bog was located in northern Leon County, Texas, approximately five miles northwest of Jewitt. The bog area is narrow and extends over a distance of one and a half miles in an area presently delimited by the Post Oak Savannah vegetational zone (Gould, 1975). The peat deposits were relatively shallow but in some areas extended to 1.5 meters. The area is underlain by deposits of the Carrizos Sands aquifer. A continuous column of sediment 139 centimeters in length was obtained from this deposit with a modified Livingston Piston Sampler. The core segments were extruded in the field and transferred to the Palynology Laboratory at Texas A&M University in three-inch PVC tubing. The samples were stored at 3°C until subsampled. The core sections were initially measured and described, and then split lengthwise. A subsample of one cubic centimeter was removed at 10 centimeter intervals throughout the core as well as at sedimentary contacts. Two tablets, each containing 12,330 ±210 Eucalyptus pollen grains, were added to each cubic centimeter subsample following Davis (1969) for later calculation of pollen concentration and pollen influx values. The subsamples then were stored at 3°C until the fossil pollen grains were chemically extracted. Pollen was extracted following the procedure of Faegri and Iversen (1975). Samples initially were heated for one hour in a 10 percent solution of NaOH in order to soften the organic material and remove humates. A 10 percent solution of HC1 was used inorder to remove carbonates and the samples then were screened through 200 jum mesh screen to remove the larger particles. Siliates were removed overnight by a solution of 70 percent HF. The remaining silicates and other inorganic materials were removed by a heavy density separation using ZnCL (S.G. 1.99-2.00). Extraneous organic matter was removed by acetolysis, after which the samples were dehydrated and transferred to 1000 cs silicon oil with Butanol. For ease in identification, safranin stain was added during dehydration. POLLEN FROM A CENTRAL TEXAS BOG 73 The polliniferous residue was examined using 400X magnification and minimum pollen counts of 200 grains were used throughout (Barkley, 1934). The entire slide was scanned using transect intervals of one millimeter in order to avoid the problem of nonrandom distribution of palynomorphs on the microscope slide (Brooks and Thomas, 1967). Critical identification of some grains was accomplished with 1000X magnification (oil) or Hoffman modulation phase contrast microscopy. Fungal spores and spores of cryptogamic plants were tabulated but not included within the pollen sum. Indeterminate grains, which were those too badly degraded to identify, were included within the pollen sum. Pollen concentration (PC) values were calculated by the following formula: _ number contaminant added (number fossil grains counted) number contaminant grains counted (sediment volume) expressed in grains per cubic centimeter. This figure then was divided by the average rate of sedimentation as determined from the radiocarbon dates (Table 1), which produced the pollen influx values (Davis, 1969). Because of the observed sedimentary discontinuity (Table 1) at 58 centimeters, pollen influx values were calculated as two discrete units. The upper unit (58 centimeters) was deposited in 1550 years, whereas the lower unit (71 centimeters) took only 250 years for deposition. Results Bog Stratigraphy The basal portion of the core was composed of light-colored, fine-grained sands affiliated with the Carrizos Sands formation, common in this region of Texas. Little organic matter was present in the lower deposits (129- 139 centimeters) quite unlike the situation reported elsewhere in East Texas bogs (Bryant, 1977). Due to the low organic content in the basal section of the core, the bottom section could not be radiocarbon dated, nor was pollen extracted in statistically valid quantities. All radiocarbon analyses were performed by the Smithsonian Institution Radiocarbon Laboratory and the results are presented in Table 1 . The upper 50 centimeters of the core were composed of a high fibrous, organic deposit with increasing silicate content with depth. Between 50 and 93 centimeters, the column was composed of a high silicate organic deposit with less fibrous material than in the above section, possibly the result of decomposition. This section was interrupted by two district levels. The first (50-58 centimeters) was composed of high silicates with small sand inclusions. The second level (72-76 centimeters) was a small fibrous lens. From 93 to 129 centimeters, the column was composed of silicates with an increasing coarseness and percentage of sands and silts (Fig. 1). Pollen Results The relative pollen frequencies clearly are dominated by Quercus , Apiaceae (Umbel family), and Poaceae pollen and secondarily by pollen of Cyperaceae (Fig. 1). Pollen frequency and pollen influx values, calculated by SAS (SAS, Institute Inc., 1982), can be obtained by contacting the senior author. Pinus pollen, while consistently present, is represented only 74 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 1. Radiation dates from Weakly Bog, Leon County, Texas. Smithsonian sample number Depth Radiocarbon years B.R SI-5860 20-30 cm modern SI-5861 50-58 cm 1550 + 55 SI-6152 58-65 cm 2125 + 65 SI-5861 80-90 cm 2260 ± 60 SI-6044 120-129 cm 2375 + 65 SI-5862 129-139 cm no date intermittently in excess of two percent. The presence and abundance of Apiaceae pollen (four to 19 percent) is somewhat surprising, however, and probably reflects localized conditions. The one segment of the core that does suggest some degree of variation is between 50 and 80 centimeters and is pronounced at the 75-centimeter level. As noted in the sedimentary column (Fig. 1), the 75-centimeter level was marked by a mat layer of grasses, which is interesting in that both Poaceae and Cyperaceae pollen decrease at this level. An abrupt decrease in the pollen of Apiaceae also is noted. Several taxa appear to increase at this level but not as abruptly. Pollen of Pinus, Betula, and Celtis , increase slightly and there is an apparent increase in the number of taxa at this level. The lower portion of the core (61-124 centimeters) is characterized by extremely high pollen influx values (Fig. 2). While the pollen influx in this lower portion is dominated by Quercus pollen, all taxa increase in influx values. Arboreal pollen peaks between 75 and 104 centimeters. Figure WEAKLY BOG, LEON COUNTY, TEXAS Figure 1. Relative pollen frequencies from Weakly Bog, Leon County, Texas. POLLEN FROM A CENTRAL TEXAS BOG 75 Figure 2. Pollen influx values of selected taxa from Weakly Bog, Leon County, Texas. 2 presents pollen influx values for selected taxa and shows that Cyperaceae, Ambrosineae, high spine Asteraceae, Poaceae, and Pinus pollen all decrease in the lower 20 centimeters of the core, peaking between 74 and 104 centimeters. Indeterminate pollen (Fig. 2) shows a steady increase toward the base as expected. Pollen influx values decrease dramatically from 52 centimeters to the surface. In part, this is correlated with the mathematical manipulations of the data based on radiocarbon analysis, but the event noted is thought to be significant. Pollen influx values for all taxa are decreased and this may represent more stable conditions in this area for the past 1500 years. Discussion Results obtained from pollen analysis of sediments from Weakly Bog show two distinct pollen assemblages, which can be interpreted as representing two major plant communities. The lower section (58-129 76 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 centimeters) shows high pollen influx values for all taxa. As indicated by the radiocarbon dates (Table 1), this section was deposited rapidly. Thus the high influx values in conjunction with rapid sedimentation agrue strongly for a dominant woodland plant community that was much denser than at present. This woodland was clearly dominated by Quercus , which accounts for 25 percent or more (Fig. 1) of the pollen rain and clearly shows its dominant role in the community. Pinus pollen is present in the assemblages yet only in minute quantities. As Davis (1969) has previously explained, when Quercus and Pinus are both present in the vegetation, Quercus pollen is overshadowed by the much more prolific pollen production of Pinus. Thus because of the low frequency of Pinus , even though this taxon is present in the pollen assemblage, it is interpreted as having been absent from the local community. Alternatively, the change in sedimentation may be a function of compaction. If the lower portion of the core was compacted, this would explain the high pollen influx values. As Table 2 shows, there is no significant change in pollen concentration values throughout the length of the column as would be expected with compaction. The fact that a significant change in pollen influx values does occur at the sedimentary contact reinforces the interpretation of a shift in the distribution of the plant taxa. Because of the short time span involved with the deposition of these sediments, and the lack of concomitant changes in pollen concentration values, we conclude that no direct evidence of compaction is present. Thus it is apparent that a dramatic change in pollen influx values and sedimentation rates occurred approximately mid-way through this deposit. Pollen influx values for nonarboreal, herbaceous taxa, especially Poaceae, in these levels (Fig. 2) are quite high. This is indicative of open areas clearly dominated by grasses and composites and is entirely consistent with the interpretation of an oak-hickory woodland. Based on the radiocarbon data (Table 1), there is a 500 to 600 radiocarbon year erosional break occurring at the 55-centimeter level in this core. Radiocarbon dates (Table 1) suggest a 500 to 600 radiocarbon year hiatus in the sedimentation. It is only during the past 1500 years however, that pollen influx values and sedimentation rates both have decreased dramatically. The dominant taxa of the upper assemblages are not appreciably different (Fig. 1) from the earlier period, yet an abrupt change in the structure of the community does appear to have occurred. The decrease in pollen influx values may signify an expansion of hervaceous understory components and a more open, less arboreal structure to the paleovegetational community (Fig. 2). More grassland areas were present in the vicinity of Weakly Bog and, although composed of the same taxa,, the woodland was more open, resembling more a savannah. We interpret this as representing the establishment of the Post Oak Savannah in this POLLEN FROM A CENTRAL TEXAS BOG 77 Table 2. Pollen concentration values, Weakly Bog, Leon County, Texas. Depth (cm) Pollen concentration Total pollen influx 1 292,680 10,198 8 434,762 15,288 15 444,023 15,471 16 565,713 19,711 26 672,419 23,429 34 427,954 14,911 35 798,408 27,819 46 495,422 17,262 51 334,581 11,658 61 351,258 62,501 71 778,509 138,525 75 648,894 115,462 80 579,263 103,072 90 725,678 251,971 94 473,208 164,308 104 400,609 139,100 114 332,194 115,345 124 242,966 84,363 134 4,749 1,649 139 2,258 784 region of Texas. This interpretation implies a somewhat drier climate than was present earlier. Herbaceous components generally produce less pollen per plant; thus, with the decrease of arboreal taxa that would have occurred during the thinning of the forest, the pollen influx was significantly reduced. An indication of this is seen in the pollen frequency diagram (Fig. 1), which shows only slight decreases in Quercus pollen and concomitant increases in herbaceous pollen, although generally there is little or no change in pollen frequencies throughout the column. The upper 25 centimeters of the core were dated as being essentially modern in age. The presence of taxa such as Ostrya/ Carpinus , Moraceae, Anacardiaceae, Myrica , and Myriophyllum in this section suggest current conditions. According to the radiocarbon analysis, the shift from mesic conditions to drier conditions as reflected in the decreased pollen influx values occurred at some point between 1500 and 2100 years B.P. Hall (1982), based on pollen and molluscan faunal data, recognized a gradual shift from wetter to drier conditions from a series of archaeological sites in Oklahoma. He postulated that this shift occurred about 1000 B.P., which is somewhat earlier than the date for this phenomenon suggested by Wendland and Bryson (1974) in their proposed climatic model. Pollen analytical studies from Ferndale Bog in southeastern Oklahoma have likewise indicated a decrease in pollen concentration values somewhat later than 1554 ± 70 years B.P. (Albert, 1981). Unfortunately, these data 78 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 did not include pollen influx calculations and thus cannot be compared directly with the results from Weakly Bog. Analyses of a sediment column from Wood County, Texas (Holloway, unpublished data), is also in agreement with the record from Weakly Bog. Although preservation there was not as good as that from Weakly Bog, indications of a lowered water table and concurrent drying, which began approximately 1500 years B.P., were recognized. On the basis of the available comparative data, the vegetational record from Weakly Bog agrees with Hall’s (1982) interpretation of a gradual shift to drier conditions during the late Holocene. This shift occurred approximately 500 years earlier in central Texas than in northern Oklahoma, but supports the model of a gradual warming trend toward more xeric conditions as reported elsewhere (Bryant, 1977; Bryant and Shafer, 1977; Hall, 1982; Bryant and Holloway, 1985). Conclusions Pollen analysis of 139-centimeter section bog deposit recovered from Leon County, Texas, reveals a history of vegetational succession during the last 2400 years. During the earlier part of this record, the regional vegetation was composed of an oak-hickory woodland association, dominated by Quercus but with large prairie openings. Pine was not a member of the local plant community. Sometime between 2100 and 1500 years B.P. there was a shift to increasing aridity with the resultant reduction in arboreal cover. The canopy level was reduced yet the composition of the plant community remained virtually unchanged. Grassland areas increased in dominance with the corresponding decrease in pollen influx values. The fossil pollen record from Weakly Bog is extremely important in that it provides new data on recent alterations to plant community structure in this region of eastern Texas. The utilization of pollen influx values, heretofore not a common practice in this region, is shown to be indispensible in recognizing these vegetational shifts. Acknowledgements We would like to express our sincere appreciation to Mr. Woody Frossard for his indispensible contributions to this project. We are also indebted to the Tarrant County Water Control District for funding this project. Logistical support during field collection was kindly supplied by the Archaeology Research Program at Southern Methodist University. Thanks also are given to the Anthropology Department at Texas A&M University for providing facilities during the laboratory portion of this study. Literature Cited Albert, L. E. 1981. Ferndale bog and natural lake: five thousand years of environmental change in southeastern Oklahoma. Oklahoma Archaeol. Surv., 7:1-127. POLLEN FROM A CENTRAL TEXAS BOG 79 Barkley, F. A. 1934. The statistical theory for pollen analysis. Ecology, 15:283-289. Benninghoff, W. S. 1962. Calculation of pollen and spore density in sediments by addition of exotic pollen in known amounts. Pollen et Spores, 4:432 Brooks, D., and K. W. Thomas. 1967. The distribution of pollen grains on microscope slides part 1. The non-randomness of the distribution. Pollen et Spores, 9:621-629. Bryant, V. M. 1977. A 16,000 year pollen record of vegetational change in central Texas. Palynology, 1:143-157. Bryant, V. M., and R. G. Holloway. 1985. A late-Quaternary paleoenvironmental record for Texas: an overview of the pollen evidence. Pp. 39-70, in Pollen record of late- Quaternary North American sediments (V. M. Bryant and R. G. Holloway, eds.) Stratigraphic Palynologists, Dallas, Texas, 423 pp. Bryant, V. M., and H. J. Schafer. 1977. The late Quaternary paleoenvironment of Texas: a model for the archaeologist. Bull. Texas Archaeol. Soc., 48:3-25. Davis, M. B. 1969. Palynology and environmental history during the Quaternary period. Amer. Sci., 57:317-332. Faegri, K., and J. Iversen. 1975. Textbook of pollen analysis. Hafner Press, New York, 295 pp. Gould, F. W. 1975. Texas plants — a checklist and ecological summary. Bull. Texas Agric. Exp. Sta., MP-585(revised): 1-121. Grahan, A. and C. Heimsch. 1960. Pollen studies of some Texas peat deposits. Ecology, 41:785-790. Hall, S. A. 1982 Late Holocene paleoecology of the southern plains. Quaternary Res., 17:391-407. Holloway, R. G. 1981. Preservation and experimental diagenesis of the pollen exine. Ph.D. dissertation, Texas A&M University, College Station, 317 pp. Larson, D. A., V. M. Bryant, and T. Patty. 1972 Pollen analysis of a central Texas bog. Amer. Midland Nat., 88:358-367. Potzger, J. E., and B. C. Tharp. 1943. Pollen record of Canadian spruce and fir from a Texas bog. Science, 98:584-585. - . 1947. Pollen profile from a Texas bog. Ecology, 28:274-280. - . 1954. Pollen study of two bogs in Texas. Ecology, 35:462-466. SAS Institute Inc. 1982 SAS user’s guide. SAS Institute Inc., Cary, North Carolina, 921 pp. Wendland, W. M., and R. A. Bryson. 1974. Dating climatic episodes of the Holocene. Quaternary Res., 4:9-24. HABITAT AND POPULATION DESTRUCTION AND RECOVERY IN THE PARTHENOGENETIC WHIPTAIL LIZARD, CNEMIDOPHORUS LAREDOENSIS (SAURIA: TEIIDAE), IN SOUTHERN TEXAS James M. Walker Department of Zoology, University of Arkansas, Fayetteville, Arkansas 72701 Abstract. — Between June 1983 and June 1986, several local populations of the parthenogenetic whiptail lizard, Cnemidophorus laredoensis, were studied at sites situated between 1.6 and 2.2 km. S Chacon Arroyo, Laredo, Webb Co., Texas. During the investigation, three of the sites were subjected to catastrophic alteration by human intervention; the vegetational structure and most of the lizards at each site were destroyed by earth-grading operations. Habitat restructuring began within a few weeks, and within 14 months each of the altered sites had developed a new assemblage of grasses and weeds growing in sandy to gravelly soil. At one intensively studied site, C. laredoensis exhibited a rapid population recovery within a year, with more lizards being observed in the new open- structured, grass-weed assemblage than the original modified thorn shrub vegetation. As each individual is capable of reproduction in C. laredoensis , parthenogenesis confers an advantage over the bisexual C. gularis in the rapid recovery of depressed population levels as conditions return to normal in catastrophically altered habitats, or in the maintenance of low population levels under suboptimal conditions. Key words : whiptail lizards; Sauria; Teiidae; Cnemidophorus ; habitat destruction and recovery. The parthenogenetic whiptail lizard, Cnemidophorus laredoensis , was described from two localities in southern Laredo, Webb Co., Texas, by McKinney et al. (1973). Observations on the catastrophic destruction and recovery of the habitat and whiptail population at one of these sites, and at two others nearby, have provided some significant insights into the ecological relationships between C. laredoensis and one of its apparent bisexual progenitors, C. gularis , and into an understanding of the ecological significance of parthenogenesis in the genus Cnemidophorus. Data presented by McKinney et al. (1973), Bickham et al. (1976), and Wright et al. (1983) indicate that C. laredoensis (Fig. 1) is diploid, and originated through hybridization between C. gularis and C. sexlineatus. Desription of Study Sites Prior to Catastrophic Alteration I first visited the site of origin of a part of the type series of C. laredoensis on the east side of U.S. Highway 83 at 1.6 km. S Chacon Arroyo (designated W-3), Laredo, in July 1983. The species also was located on the west side of U.S. 83 (designated W-4), directly across from W-3, and on the east side of U.S. 83 at 2.2 km. S Chacon Arroyo (designated W-5*), Laredo, in July 1983. No individuals of C. gularis were observed or collected at W-3 and W-4 from 1983 through 1986; both C. laredoensis and C. gularis The Texas Journal of Science, Vol. 39, No. 1, February, 1987 82 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure I. Dorsal view of an adult female of Cnemidophorus laredoensis from W-3 at 1.6 km. S Chacon Arroyo, Laredo, Webb Co., Texas. were present at W-5* in 1983 and 1984. All known localities for C. laredoensis (sensu McKinney et al., 1973) are listed by Walker (1987); localities for the species are also cited by McCrystal et al. (1985), Dryden (1985), and Walker et al. (1986). In July 1983, W-3 consisted of a relatively isolated, degraded patch of modified thorn shrub habitat on both sides of unpaved Saltillo Street in the southern suburbs of Laredo. The dominant vegetational components at W-3 included buffelgrass, an introduced Indo-African species (Gould, 1975), scattered clumps of Opuntia , and a variety of shrubs (for example, mesquite, huisache, paloverde) growing in sandy to gravelly soil. The entire area was littered with dumped debris. The south-facing side of Saltillo Street, with sand deposits to a depth of half a meter, scattered clumps of buffelgrass, and scattered debris (Fig. 2), was one of several focal microhabitats of lizard activity. Numerous burrows were present along both sides of the street and many were occupied by individuals of C. laredoensis. When first observed in July 1983, W-4 consisted of a large flat area that had been extensively disrupted in the construction of a large pipeline and scraped clean of vegetation. Several C. laredoensis were observed along the southeastern margin in patches of buffelgrass and in buffelgrass- mesquite along the northwestern margin. Closely spaced loads of dumped waste earth rendered the southwestern part of the flat unsuitable for whiptails. Between July 1983 and July 1984, W-5* remained available as a nearby site where C. laredoensis and C. gularis could be observed or collected in syntopy. W-5* consisted of a grass-weed association of about three hectares separated from an extensive thorn shrub tract of about 25 hectares by a sandy roadbed. Most individuals of C. laredoensis were observed in the transition zone of the roadbed and thorn shrub vegetation, whereas C. gularis usually was found at the edge of, or within, the thorn shrub. Large amounts of debris were scattered between the road and surrounding habitat at W-5*. HABITAT OF CNEMIDOPHORUS LAREDOENSIS 83 Figure 2. View of habitat at W-3 looking east on unpaved Saltillo Street, 1.6 km. S Chacon Arroyo, Laredo, Webb Co., Texas, as it appeared when first visited in July 1983. Description of Study Sites After Catastrophic Alteration A visit to W-3 in June 1984 revealed that the entire site had been bulldozed (Fig. 3). The thorn shrub association (dominated by mesquite) had been completely destroyed and the natural and man-made debris pushed into large mounds. The developing plant growth, dating to May 1984 according to a resident of the area, was limited to recently germinated grasses and weeds, rooting cactus pads, and root sprouts from damaged shrubs. Presumably, the scattered individuals of C. laredoensis that remained at W-3 were those that escaped the devastation by fleeing or by residing along the margins of the tract, rather than animals that immigrated from surrounding sites. Although both sides of Saltillo Street had been extensively disrupted from grading operations, several active burrows were located, and lizards removed from two. By July 1984, less than three months after the site was cleared, W-3 supported a low-growing grass-weed assemblage with numerous open spaces. Two additional adults of C. laredoensis were removed from burrows in July 1984 to confirm that whiptails were still present at W-3; however, it seemed apparent that the habitat and lizard population had been irreparably damaged. Quite the opposite proved to be the case. By the next visit to the site in July 84 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Figure 3. View of habitat at W-3 looking east on unpaved Saltillo Street, 1.6 km. S Chacon Arroyo, Laredo, Webb Co., Texas, as it appeared in July 1984 about two months after earth-grading operations catastrophically altered the area. 1985, or about 14 months after the destruction of the habitat, a rapidly developing growth of grasses and weeds measuring half a meter or less in height was present (Fig. 4). Both sides of Saltillo Street had numerous burrows, from which a number of C. laredoensis were removed. Not only were individuals of C. laredoensis more widely distributed through the weedy growth in areas where they had been limited to certain microhabitats within and around the margins of the thorn shrub tract in 1983, but the number of lizards observed at W-3 indicated that an explosive recovery had occurred since the catastrophic alteration of the site. Whereas only about 30 lizards were observed by three persons in two days in July 1983, more than 50 C. laredoensis were observed at W-3 by four persons in only two hours in July 1985. Although W-4 and W-5* were located in the same general area of southern Laredo as W-3, each site produced different combinations of pioneering weedy growth following their alteration. The part of W-4 that had been scraped clean of vegetation in July 1983 was found to have stands of tall sunflower, a common weed in well-drained waste areas with compacted soil in the Rio Grande Valley, when inspected in June and July 1985. Also, much of the northern third of the flat was no longer suitable for whiptails owing to its recent use as a dumpsite for waste soil. Where lizards had been limited to patches of bunchgrasses along the margins of the flat in July 1983, they were now more generally distributed through the tall sunflowers. HABITAT OF CNEMIDOPHORUS LAREDOENSIS 85 Figure 4. View of habitat at W-3 looking east on unpaved Saltillo Street, 1.6 km. S Chacon Arroyo, Laredo, Webb Co., Texas, as it appeared in July 1985 about 14 months after catastrophic alteration. The most surprising development during the study period involved W- 5*; a site inspection in July 1985 revealed that the entire 25-hectare thorn shrub tract located south and west of the sandy roadbed had been destroyed by earth-grading operations. W-5*, which had been last observed in July 1984, had become a vast and nearly impenetrable expanse of Russian thistle. Observations were mostly confined to the transition zone between the road and surrounding habitat in July 1985; C. laredoensis was still present at W-5*, but C. gularis was neither observed nor collected at W-5* in July 1985. Discussion At the beginning of the investigation of C. laredoensis in 1983, the species had been reported from only two sites located 1.6 kilometers apart in Laredo, Texas, and had not been reported in Mexico. The species is now known to occur within a narrow zone on both sides of the Rio Grande over a distance of about 250 kilometers by road in parts of Webb, Zapata, Starr, and Hidalgo counties, Texas, and the Mexican state of Tamaulipas, and well away from the river in parts of Dimmit, LaSalle, and Starr counties (Walker, 1987). Observations on sites W-3, W-4, and W-5* in Laredo, as well as on numerous other sites in the Rio Grande Valley and peripheral 86 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 areas, have revealed that the local distribution of C. laredoensis is closely tied to the production and maintenance of disturbed areas with grass-weed associations, bunchgrass-mesquite disclimax communities, or ecotones near thorn shrub tracts from the activities of man or nature. Significantly, the three catastrophically altered Laredo sites, which differed somewhat in soil characteristics, drainage, topography, and exposure, produced three of the four plant assemblages with which C. laredoensis is most often associated throughout its range. A four-year study revealed that C. laredoensis is syntopic with C. gularis at a minimum of 19 of 25 localities in Texas and six of 1 1 in Tamaulipas. Instances of syntopy between the two taxa usually involved transtional areas between tracts of relatively undisturbed thorn shrub or remnants of such formations and grass-weed habitats or in bunchgrass/ weed-mesquite formations. A common feature at all sites inhabited by C. laredoensis is the presence of sandy or loamy soil. C. gularis is adaptable to all types of soil (for example, sandy, loamy, gravelly, or rocky substrates). Whereas C. gularis occurs in weedy habitats with almost the same frequency as C. laredoensis particularly where disturbed areas abut climax formations, the natural habitat of the species includes Chihuahuan thorn shrub, barretal, and upland thorn shrub climax formations that border or intrude into the Rio Grande entrenchment. C. gularis may occur in small numbers or temporarily disappear from local sites that have been subjected to recent catastrophic alteration or fragmentation into small patches. Although several authors have noted that parthenogenetic species of Cnemidophorus frequently are associated with ecotonal or altered habitats (Axtell, 1966; Wright and Lowe, 1968; Christiansen, 1971; Christiansen et al., 1971; Schall, 1976, 1978; Cuellar, 1977, 1979; Walker, 1987), the biological basis for this apparent pattern of ecological and geographical distribution has not been explained. Moreover, no actual observations on habitat and population destruction and recovery for a parthenogenetic form have been reported. This study indicates that the types of grass-weed habitats frequently occupied by C. laredoensis may begin development shortly after such catastrophic events as cultivation, earth-grading, and fire, and that an optimal habitat may develop in less than a year (Figs. 2-4). Land use patterns along the Rio Grande, which may preserve such grass- weed associations for many years or lead to the development of bunchgrass/ weed-mesquite disclimax communities, often result in sufficient habitat stability for C. gularis to become established or remain in syntopy with C. laredoensis . The observations on W-3, and other similar areas in the Rio Grande Valley, indicate that parthenogenesis is at greatest advantage in C. laredoensis in the aftermath of habitat changes that sharply reduce the size of a population. As each individual is an effective reproductive “deme,” C. laredoensis would be expected to persist in low numbers under suboptimal conditions more successfully than C. gularis because of its HABITAT OF CNEMIDOPHORUS LAREDOENSIS 87 inherently higher reproductive rate and the lack of a requirement for a threshold population size to maintain the mating structure. Neonates of C. laredoensis , which are 100 percent females, appear in July and August and grow rapidly to a snout-vent length of 55 to 60 mm before becoming inactive for the year in late November or early December. Egg production in females hatched in July and August commences the following April, and one or two clutches of two to four eggs are deposited by each female between mid-May and mid-July. With every individual in the population capable of producing eggs it is not surprising that the survivors of C. laredoensis at W-3 were able to generate a rapid population recovery following catastrophic alteration of the habitat. Field data from 36 sites inhabited by C. laredoensis indicate that the parthenogen reaches the highest population densities at sites where C. gularis is absent or occurs in small numbers. Whether this pattern results from interactions between lizard populations or that the same habitat conditions favoring C. laredoensis are inhibitory on C. gularis requires additional study; however, preliminary data indicate that the latter point is more important. Although C. gularis was present about 200 meters east of W-3 and was common at W-5* (about 0.6 kilometers south of W-3), the species has not appeared in samples from W-3 for at least 15 years (McKinney et al., 1973; C. O. McKinney, personal communication; data from this study). W-3 is locally separated from thorn shrub climax formations, the natural habitat of C. gularis , by streets, buildings and grounds, paved and unpaved parking lots, and expanses of grasses-weeds that impede immigration of lizards to the site. Additional conditions that may contribute to the absence of C. gularis at W-3 are: history of severe habitat disturbances (motor vehicles, dumping, human and animal degradation); evidence of periodic catastrophic alteration; and patchy structure of the habitat. In summary, the parthenogen C. laredoensis is revealed as a pioneering vertebrate in several types of grass-weed successional stages and bunchgrass/weed-mesquite disclimax communities that today mainly result from the activities of man. The species exhibits an impressive ability to rebound from depressed population levels as a result of a parthenogenetic mode of reproduction, a biological attribute that also allows the Laredo whiptail to persist at low population levels in severely degraded habitats, and in some areas overgrown with grasses or weeds or both. The absence of the species in most areas removed from the ecological influence of the Rio Grande, and its presence in a narrow riparian zone on both sides of the river, result from its inability to become established in areas devoid of sandy burrowing sites and perhaps to the pervasive presence of C. gularis in climax formations. The success of C. laredoensis in becoming widely distributed in communities, towns, and cities on both sides of the Rio Grande between Laredo-Nuevo Laredo and Progreso Lakes-Nuevo Progreso (that is, in 88 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 areas subject to frequent habitat changes) mitigates against the species being considered endangered or threatened. Acknowledgments Dean John C. Guilds, Fulbright College of Arts and Sciences, deserves major credit for providing the initial incentive grant (Walker-ZOOL-RIG-83) in support of this work in 1983 and a subsequent grant (Walker-ZOOL-RIG-86) in 1986. Former Chairman Duncan W. Martin, Department of Zoology, provided grants in support of this work in 1984 and 1985, without which my studies in the Rio Grande Valley could not have continued. I am especially grateful to James E. Cordes, Department of Zoology, for his tireless assistance in this study. Field research in Texas was conducted under the authority of Texas Parks and Wildlife permit no. 61. Literature Cited Axtell, R. W. 1966. Geographic distribution of the unisexual whiptail Cnemidophorus neomexicanus (Sauria: Teiidae) — Present and past. Herpetologica, 22: 241-253. Bickham, J. M., C. O. McKinney, and M. F. Mathews. 1976. Karyotypes of the parthenogenetic whiptail lizard Cnemidophorus laredoensis and its presumed parental species. Herpetologica, 32:395-399. Christiansen, J. L. 1971. Reproduction of Cnemidophorus inornatus and C. neomexicanus (Sauria: Teiidae) in northern New Mexico. Amer. Mus. Novit., 2442: 1-48. Christiansen, J. L., W. G. Degenhardt, and J. E. White. 1971. Habitat preferences of Cnemidophorus inornatus and C. neomexicanus with reference to conditions contributing to their hybridization. Copeia, 1971: 357-359. Cuellar, O. 1977. Animal parthenogenesis. Science, 197: 837-843. - . 1979. On the ecology of coexistence in parthenogenetic and bisexual lizards of the genus Cnemidophorus. Amer. Zool., 19: 773-786. Dryden, L. S. 1985. Cnemidophorus laredoensis (Laredo striped whiptail). SSAR Herp. Review, 16: 60. Gould, F. W. 1975. The grasses of Texas. Texas A&M Univ. Press, College Station, 563 pp. McCrystal, H. K., R. H. Dean, and J. R. Dixon. 1985. Range extension for the whiptail lizard Cnemidophorus laredoensis (Teiidae). Texas Jour. Sci., 36: 283-284. McKinney, C. O., F. R. Kay, and R. A. Anderson. 1973. A new all female species of the genus Cnemidophorus. Herpetologica, 29: 361-366. Schall, J. J. 1976. Comparative ecology of sympatric parthenogenetic and bisexual species of Cnemidophorus. Unpublished Ph.D. dissertation, Univ. Texas, Austin, 258 pp. - . 1978. Reproductive strategies in sympatric whiptail lizards ( Cnemidophorus ): two parthenogenetic and three bisexual species. Copeia, 1978: 108-116. Walker, J. M. 1987. Distribution and habitat of the parthenogenetic whiptail lizard, Cnemidophorus laredoensis (Sauria: Teiidae). Amer. Midland Nat., in press. Walker, J. M., S. E. Trauth, J. E. Cordes, and J. M. Britton. 1986. Cnemidophorus laredoensis (Laredo striped whiptail). SSAR Herp. Review, 17: 27-28. Wright, J. W., and C. H. Lowe. 1968. Weeds, polyploids, parthenogenesis, and the geographical and ecological distribution of all-female species of Cnemidophorus. Copeia, 1968: 128-138. Wright, J. W., C. Spolsky, and W. M. Brown. 1983. The origin of the parthenogenetic lizard Cnemidophorus laredoensis inferred from mitochondrial DNA analysis. Herpetologica, 39: 410-416. LONG-TERM RESPONSE OF LIVE OAK THICKETS TO PRESCRIBED BURNING Marlin D. Springer, Timothy E. Fulbright, and Samuel L. Beasom 12508 Rosemont NE, Albuquerque, New Mexico 87112, College of Agriculture and Home Economics, Texas A&I University, Kingsville, Texas 78363 and Caesar Kleberg Wildlife Research Institute, Texas A&I University, Kingsville, Texas 78363 Abstract. — A 10-year prescribed burning study was initiated in 1974 on the Aransas National Wildlife Refuge, Aransas Co., Texas, to test the hypothesis that fire will reclaim grassland that has been invaded by live oak ( Quercus virginiana Mill.). Three 700-hectare areas were burned in autumn 1974, spring 1975, and autumn 1975, with reburns at two- year intervals. Effects of repeated burns on live oak density, height, and acorn production in live oak thickets were determined. Burning increased species diversity in the forb community and favored fair to excellent forage grasses. Initial burns increased density of live oak stems by approximately 130 to 240 percent and repeated burns over the subsequent 10-year period had no further effect. Prescribed burns reduced height of live oak stems by 30 to 60 percent for up to three years. Live oak mottes generally did not burn during the first burns, but were more susceptible to subsequent burns. Burning of large trees in live oak mottes resulted in an increase in live oak stems of nearly 200 percent, Key words : brush management; community succession; fire; Quercus virginiana. Live oak ( Quercus virginiana Mill.) savannah, with scattered individual trees or mottes, occupies about 500,000 hectares of the Texas Coastal Prairie (Smith and Rechenthin, 1964). Historically much of the region was maintained in that condition by repeated burns of natural or aboriginal causes (Blakey, 1947; Stewart, 1963; Vallentine, 1980). In addition to occurring in mottes or as individual trees, live oak also may occur in a low, running form referred to as “running” or “shinnery” live oak (Scifres, 1980). In recent history, overgrazing and lack of fire have been implicated as contributing factors in the closing of savannahs by dense stands of shinnery live oak (Blakey, 1947; Dyksterhuis, 1957). Dyksterhuis (1957) referred to this phenomenon as the “thicketization” of savannahs. Much effort has gone into finding ways to reverse the increasing density of oak stems and return the Coastal Prairie to savannah. Root plowing, mowing, and roller chopping have been used but are expensive (Aransas Wildlife Refuge records, 1941-74). Herbicides are effective (Scifres and Haas, 1974; Meyer and Bovey, 1980) but damage large oaks as well as shinnery. Large trees and mottes provide shade for livestock and habitat for wildlife (Scifres, 1980). Treatments used to remove low brush should leave them intact. Because fire has influenced the development and maintenance of grasslands (including savannahs), a prescribed burning program was The Texas Journal of Science, Vol. 39, No. 1, February, 1987 90 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 initiated to burn live oak thickets on the Aransas National Wildlife Refuge, Aransas Co., Texas, in 1974. Early effects of prescribed burning on small mammals (Brown, 1977), white-tailed deer ( Odocoileus virginianus Rafinesque) (Springer, 1977), and vegetation (Springer, 1978; Scifres and Kelley, 1979; Kelley, 1980) were reported two years postburn. Original study plots have been burned at approximately two-year intervals since the initial burns of 1974-75. This study provided an evaluation of the long-term effects of prescribed burning on the vegetation. Materials and Methods The Aransas National Wildlife Refuge is a 21,862-hectare peninsula predominated by live oak thickets. It is a part of the Coastal Prairies and Marshes Region of Texas (Gould, 1975) and is situated primarily in Aransas County. The ecology and history of the area were described by Halloran and Howard (1965) and Jones (1975). An area of approximately 700 hectares was burned in autumn 1974 with reburns in autumn 1976, 1978, 1980, and 1983. A second area of equal size was burned in autumn 1975 with reburns in 1977, 1979, and 1981. A third area was burned in spring 1975. An unburned reference area that was sampled as a control in the first years of the study (1974-76) was inadvertently burned in 1978 and was reburned in 1980 and 1982. All sampling in burned areas was done along one, 183-meter line in each area. Lines were permanently marked with metal posts in 1974. A new unburned area of live oak thickets was chosen in 1984 and permanently marked for future reference. For assessment of the effect of multiple burns on live oak thickets, composition of the herbaceous stratum and density, height, and acorn production of live oak were estimated on burned and unburned areas in August 1984. For comparison, some data from 1975-76 are presented. Herbaceous composition in live oak thickets was estimated by frequency of occurrence with the step-point method (Evans and Merton, 1957). The nearest plant to a notch in the observer’s boot was recorded with each step along the permanent sampling line. Data from 100 points were recorded in each area. Density of live oak stems in live oak thickets was estimated by counting all stems in 71 plots that were 0.25 meter square and equidistantly spaced along the sampling lines. Average height of live oak was calculated from measurements of the four live stems nearest the corners of the 7 1 sample plots on each transect. Acorn production was estimated in live oak thickets on burned and unburned areas by counting acorns on all plants in the 71 sample plots in each area. To assess the impact of the burning program on mottes, the woody community of five burned mottes was sampled and compared to the woody community of five unburned mottes. Five plots one meter square were equidistantly spaced along a sample line bisecting each motte. Height and density of woody plants were recorded. The study was designed for descriptive purposes. Data were not subjected to statistical analyses because of the lack of replication and use of systematic sampling. Inferences about the results are restricted to the study area. Results and Discussion Species Composition in Live Oak Thickets Herbaceous communities of the burned areas in 1984 tended to be more diverse than unburned areas (Table 1); this was also true in 1975-76 (Scifres and Kelley, 1979). There were 16 species of grasses and forbs recorded RESPONSE OF LIVE OAK TO BURNING 91 Table 1. Species composition (percent occurrence) of the herbaceous plant community in live oak thickets of a live oak savannah, Aransas National Wildlife Refuge, Aransas Co., Texas, August 1984. Plants identified along the sample line but not recorded as a hit were recorded as trace (t). Species Unburned First burned autumn 1975a GRASSES Seacoast bluestem [ Schizachyrium scoparium var. littoralis (Nash) Gould] 20 27 Gulfdune paspalum ( Paspalum monostachyum Vasey) 14 t Purple lovegrass [ Eragrostis spectabilis (Pursh) Steud.] 11 t Switchgrass ( Panicum virgatum L.) 8 3 Brownseed paspalum ( Paspalum plicatulum Michx.) 5 8 Dicanthelium sp.b 3 11 Carpetgrass ( Axonopus affinis Chase.) 3 t Thin paspalum ( Paspalum setaceum Michx.) 2 10 Big bluestem ( Andropogon gerardii Vitman) 1 7 Bushy bluestem [. Andropogon glomeratus (Walt.) B.S.P.] t 6 Panicum sp.c t 0 Carolina jointtail [Coelorachis cylindrica (Michx.) Nash] 0 t Tumble lovegrass ( Eragrostis sessilispica Buckl.) 0 t Muhlenbergia sp. 0 t Unknown grass 3 3 SEDGES AND RUSHES 20 18 FORBS Spadeleaf [Centella asiatica (L.) Urban] 6 0 Sawtooth frogfruit ( Phyla incisa Small.) 1 t Verbena sp. 1 t Unknown forb 1 1 Wild bean [Strophostyles leiosperma (T. & G.) Piper] 0 1 Partridge pea ( Cassia fasiculata Michx.) 0 1 Scarlet pea ( Indigofera miniata Ort.) 0 1 Yankeeweed ( Eupatorium compositifolium Walt.) 0 1 Snoutbean [Rhynchosia americana (Mill.) Metz.] 0 t Loosestrife ( Lythrum lanceolatum Ell.) 0 t Dayflower ( Commelina erecta L.) 0 t Wild indigo ( Baptisia leucophaea Nutt.) 0 t "Burned in 1975, 1977, 1979, and 1981. bIncludes D. augustifolium (Ell.) Gould, D. lanuginosum (Ell.) Gould, D. oligosanthes (Schult.) Gould, and D. sphaerocarpon (Ell.) Gould. Tncludes P brachyanthum Steud. and P hians Ell. on the unburned area in August 1984 as compared to 25 on the burned area. In 1975-76 certain species increased, including seacoast bluestem (Schizachyrium scoparium var. littoralis) after burning (Kelley, 1980). The percent occurrence of seacoast bluestem after four burns (1984) suggested that burning may favor this species as well as big bluestem ( Andropogon gerardii ), brownseed paspalum ( Paspalum plicatulum ), bushy bluestem (A. 92 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 glomeratus), thin paspalum ( P \ setaceum ), and some species of Dichanthelium (Table 1). A difference in forb composition between the burned and unburned areas was evident in 1984. There were 11 species of forbs recorded along the sampling line in the burned area as compared to four in the unburned area (Table 1). Five of the species in the burned area were legumes that were absent in the unburned community. A more diverse legume-dominated forb community caused by burning also was indicated in 1975-76 (Scifres and Kelley, 1979). Live Oak Density in Thickets Two burned areas received four and five burns, respectively, over the approximately 10-year period, and density of live oak stems remained higher than preburn densities (Table 2). In 1975, live oak density in thickets of the area first burned in autumn 1974 and the area first burned in spring 1975 (one growing season postburn) had increased 150 and 240 percent, respectively, over preburn densities (Kelley, 1980). By the second growing season (1976) live oak mean density appeared to be declining on both areas, and stem density on the area first burned in autumn 1975 in its first year postburn was approximately 130 percent greater than preburn density. During this second growing season, many live oak plants (regrowth) on burned areas showed signs of severe stress. It was hypothesized that the repeated stress of burning and increased competition of the more vigorous herbaceous community would eventually reduce the number of live oak stems in burned thickets. Data taken in August 1984, however, failed to support that hypothesis. Of the two burned areas monitored in 1984, the area first burned in autumn 1975, in the third growing season since its last burn, reflected the same pattern seen in earlier years. Live oak density was about 80 percent higher than preburn levels, but 25 percent lower than the first year postburn level of 1976. Density in the area first burned in autumn 1974, in the first year postburn, was almost 300 percent greater than prior to burning. There was no change in live oak density in the unburned area in the first three years of the study. Live Oak Stem Height in Thickets Top kill of live oak by burning live oak thickets initially reduced stem height to ground level. Stem height of live oak regrowth remained less than stem height on the unburned area even after three growing seasons (Table 3). Stem growth rate was greatest in the first year postburn, with progressively lower rates in the second and third years. At the rate of growth suggested by interpretation of data in the third year postburn, it was projected that it would require at least eight years for live oak in the burned thickets to reach a height approximately equal to unburned areas. RESPONSE OF LIVE OAK TO BURNING 93 Table 2. Mean density (stems per square meter) of living live oak in live oak thickets of a live oak savannah, Aransas National Wildlife Refuge, Aransas Co., Texas. Numbers in parentheses indicate number of growing seasons since the last burn. Values for 1974, 1975, 1976 are adapted from Scifres and Kelley (1979) and Kelley (1980). Date of first burn Sample date 1974 (preburn) 1975 (after first burn) 1976 (after first burn) 1984 (after multiple burns) Unburneda 10.0 12.8 8.0 — A-74b 6.8 17.0(1) 15.4 (2) 26.2(1) S-75c 11.2 37.7(1) 30.9 (2) — A-75d 27.6 — 62.7(1) 49.4 (3) aUnburned reference area was inadvertently burned in 1978. bBurned in the autumn of 1974, 1976, 1978, 1980, and 1983. cBurned in the spring of 1975. dBurned in the autumn of 1975, 1977, 1979, and 1981. Acorn Production in Live Oak Thickets Although top-killed live oak is capable of flowering and producing acorns on regrowth stems the first year postburn, it appeared that most energy was used in vegetative regrowth. In 1984, one growing season after the last of five burns, acorn production on live oak regrowth was less than acorn production on live oak in unburned thickets (Table 4). Burning- related reduction in acorn production was significant both as production per unit area and production per plant. Production of acorns per unit area on burned thickets in their second and third year postburn tended to be lower than on unburned areas. When compared as production per plant, production was lower on burned areas. This same trend was evident in 1975-76 (Springer, 1978). Table 3. Mean heights (cm) of living live oak stems in live oak thickets of a live oak savannah, Aransas National Wildlife Refuge, Aransas Co., Texas. Numbers in parentheses indicate number of growing seasons since the last burn. Values for 1975 and 1976 are adapted from Scifres and Kelley (1979) and Kelley (1980). Date of first burn Sample date 1975 (after first burn) 1976 (after first burn) 1984 (after multiple burns) Unburned 48.8 53.6 42.7 A-74a 23.4(1) 28.8 (2) 15.6(1) A-75b — 16.4(1) 29.1 (3) A-78c — — 24.1 (2) aBurned in the autumn of 1974, 1976, 1978, 1980, and 1983. bBurned in the autumn of 1975, 1977, 1979, and 1981. cBurned in the autumn of 1978, 1980, and 1982. 94 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Table 4. Mean production of acorns in live oak thickets of a live oak savannah, Aransas National Wildlife Refuge, Aransas Co., Texas, August 1984. Treatment Acorns per square meter Acorns per stem Unburned 32.2 1.7 Burned areas3 First growing season postburn (after five burns) 0.3 0.01 Second growing seasons postburn (after three burns) 17.0 0.3 Third growing seasons postburn (after four burns) 15.6 0.3 aAll burned areas have had 3-5 burns over a 10-year period. Woody Composition of Mottes Although large mottes generally did not burn in the initial burns of 1974- 75, observation in later years indicated that with each successive burn, mottes became more susceptible to burning. A comparison of aerial photography from 1974 and 1984 showed that approximately 20 percent of the area occupied by the original mottes had burned. After mottes burned, there was an increase in the number of live oak stems (throughout the original area encompassed by the motte) (Table 5). Shade tolerant species, typical of the understory of unburned mottes, remained in approximately the same densities on burned mottes. Conclusions Prescribed burning over a 10-year period, with burns at approximately two-year intervals, was not effective in reducing live oak density in thickets in live oak savannah on the Aransas National Wildlife Refuge. Although fire is widely accepted as one of the main factors in the formation and maintenance of grasslands, results of this study indicate that fire cannot be used to restore grasslands on the Refuge. It is effective in reducing height of live oak and probably in increasing availability of the area for wildlife and livestock. This effect, coupled with a more diverse postburn herbaceous community, suggests that the prescribed burning program can Table 5. Mean density (stems per square meter) of major woody species (live plants) in the mottes of a live oak savannah, Aransas National Wildlife Refuge, Aransas Co., Texas, August 1984. Species Unburned Burned Live oak ( Quercus virginiana Mill.) 0.9 2.6 Yaupon ( Ilex vomitoria Ait.) 1.2 1.0 Beauty berry ( Callicarpa americana L.) 0.4 0.4 Catbriar ( Similax bona-nox L.) 0.6 1.2 Mustang grape ( Vitis mustangensis Buckl.) 0.3 0.1 RESPONSE OF LIVE OAK TO BURNING 95 have practical benefits if burns are repeated at intervals frequent enough to maintain suppression of live oak height. Burning of mottes removed the large tree canopy cover and promoted formation of thickets. It may be desirable to preserve mottes because of their value as shade for livestock and habitat for wildlife. Fire lines could be constructed around the mottes to protect them from burning. Acknowledgments The authors wish to thank the Cesar Kleberg Wildlife Research Institute and the U.S. Fish and Wildlife Service for funding this study. Literature Cited Blakey, H. L. 1947. The role of brush control in habitat improvement on the Aransas National Wildlife Refuge. Trans. N. Amer. Wildlife Conf., 12:179-185. Brown, W. A. 1977. The influence of prescribed burning on small mammal populations of the Texas Gulf Coast Prairie. MS thesis, Texas A&M Univ., College Station, 95 pp. Dyksterhuis, E. J. 1957. The savannah concept and its use. Ecology, 38:435-442. Evans, R. A., and L. R. Merton. 1957. The step-point method of sampling — a practical tool in range research. J. Range Manag., 10:208-212. Gould, F. W. 1975. Texas plants — a checklist and ecological summary. Bull. Texas Agric. Exp. Sta., MP-585 (revised): 1-121. Halloran, A. F., and J. A. Howard. 1965. Aransas Refuge wildlife introductions. J. Wildlife Manag., 20:460-461. Jones, F. B. 1975. Flora of the Texas Coastal Bend. Mission Press, Corpus Christi, Texas, 267 pp. Kelley, D. M. 1980. Vegetation response to burning thicketized live oak savannah on the Aransas National Wildlife Refuge. MS thesis, Texas A&M Univ., College Station, 72 pp. Meyer, R. E., and R. W. Bovey. 1980. Control of live oak ( Quercus virginiana ) and understory vegetation with soil-applied herbicides. Weed Sci., 28:51-58. Scifres, C. J. 1980. Brush management-principles and practices for Texas and the Southwest. Texas A&M Univ. Press, College Station, 360 pp. Scifres, C. J., and R. H. Haas. 1974. Vegetation changes in a post oak savannah following woody plant control. Bull. Texas Agric. Exp. Sta., MP-1136: 1-11. Scifres, C. J., and D. M. Kelley. 1979. Range vegetation response to burning thicketized live oak savannah. Texas Agric. Exp. Sta. B-1246: 1-15. Smith, H. N., and C. A. Rechenthin. 1964. Grassland restoration part 1. The Texas brush problem. USDA-SCS 4-181 14, 16 pp. Springer, M. D. 1977. The influence of prescribed burning on nutrition in white-tailed deer on the Coastal Plain of Texas. Ph.D. dissertation, Texas A&M Univ., College Station, 71 pp. - . 1978. The effects of prescribed burning on browse, forbs, and mast in a Texas live oak savannah. Proc. Ann. Conf. Southeast Assoc. Fish and Wildlife Agencies, 31:188-198. Stewart, O. C. 1963. Barrier to understanding the influence of use of fire by aborigines on vegetation. Proc. Tall Timbers Fire Ecol. Conf., 2:1 17-126. Vallentine, J. F. 1980. Range development and improvements. Brigham Young Univ. Press, Provo, Utah, 2nd ed., 545 pp. NOTEWORTHY RECORDS OF MAMMALS FROM THE TEXAS PANHANDLE Robert R. Hollander, J. Knox Jones, Jr., Richard W. Manning, and Clyde Jones Departments of Biological Sciences and Museum Science, and The Museum, Texas Tech University, Lubbock, Texas 79409 Abstract. — Distributional and natural history data are presented for 12 species of mammals — one shrew, three bats, the armadillo, and seven rodents — from the Texas Panhandle. Key words : distribution; mammals; Panhandle, Texas. The Panhandle area of northern Texas is one of the least well known in the state from a zoological point of view, perhaps because of its sometimes inhospitable climate and relative lack of species diversity, and because it is far removed from most centers of biological study. All but the eastern edge of the Panhandle lies in the High Plains physiographic region. Some of the salient environmental features of the High Plains and the Rolling Plains adjacent to the east were described by Gould (1975). Environmental variables of the region were summarized by Owen and Schmidly (1986). In 1984, we began a survey of mammals in the Panhandle, concentrating our efforts primarily in the northern counties — those areas lying to the north of the Canadian River. Specimens thus far accumulated, along with some material already housed in The Museum of Texas Tech University (TTU), represent 12 species that are noteworthy in terms of distribution, taxonomy, and natural history. These data are recorded here. Notiosorex crawfordi crawfordi (Coues, 1877) The only previous report of the desert shrew from the northern part of the Texas Panhandle is of a specimen taken in the flood plain of Bugbee Creek, 9 mi. E Stinnett, Hutchinson County (Blair, 1954:241). We have at hand a nonpregnant female (TTU 6849) taken 14 mi. E and 7 mi. S Dumas, Moore County, on 27 June 1968 by W. H. Conley. Lasiurus borealis borealis (Muller, 1776) The only published record of the red bat from the Panhandle of which we are aware is the specimen reported by Blair (1954:242) from 9 mi. E Stinnett, Hutchinson County. We took a lactating female (TTU 42648) on the Tyson Ranch, 8 mi. NW Higgins, Lipscomb County, on 26 July 1985. It was the only bat captured in a mist net set over a spring in a small, open canyon. The Texas Journal of Science, Vol. 39, No. 1, February, 1987 98 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Plecotus townsendii pallescens (Miller, 1897) This big-eared bat, which probably occurs in suitable places throughout the Panhandle, has been reported previously only from Armstrong County (Davis 1974:69). We have examined a female (TTU 2038) that was obtained in Borger, Hutchinson County, on 11 May 1965, and three specimens from Collingsworth County — two males (TTU 43684-85) taken in hibernation in a large gypsum sinkhole, 4 mi. N and 13 mi. E Lutie, on 8 March 1986, and a female (TTU 43686) netted over Elm Creek, 3 mi. N and 12 mi. E Lutie on 13 May 1986. The female carried a single fetus (23 mm in crown-rump length). Antrozous pallidus pallidus (Le Conte, 1856) The pallid bat is known from the Panhandle on the basis of a specimen from Tascosa, Oldham County (Bailey, 1905:214), and small series from Briscoe and Deaf Smith counties (Martin and Schmidly, 1982:31-32). In the summer of 1985, we netted a series of adult A. pallidus (TTU 42650- 92), all but one males, in the steep canyon of Romero Creek on the Griffin Ranch, 17-18 mi. N and 1 mi. E Adrian, Oldham County, and another series (TTU 42693-725) on the Fain Ranch, 16 mi. N Amarillo, Potter County, which included adults and young-of-the-year of both sexes. These were taken along with Eptesicus fuscus at the entrance to a barn used as a night roost. Additionally, on 13 May 1986, a female (TTU 43687), pregnant with twins (crown-rump lengths 17 and 19 mm), was taken in a mist net set over Elm Creek, 3 mi. N and 12 mi. E Lutie, Collingsworth County. Myotis velifer, Pipistrellus subflavus, Plecotus townsendii (see above), and Tadarida brasiliensis were netted at the same place. Dasypus novemcinctus mexicanus Peters, 1864 The armadillo is recorded from the Panhandle only from Armstrong County (Davis, 1974:269). We found one (TTU 42726) of unknown sex dead on U.S. Highway 60 at a place 10 mi. SW Canadian, Hemphill County, an extension of known range of approximately 80 miles to the northeast. Additionally, it was reported to one of us (Hollander) by a reliable observer that one had been killed 7 mi. NW Higgins, Lipscomb County, on State Highway 213, on about 4 July 1985. Probably the armadillo occurs throughout at least the eastern part of the Panhandle region. Sciurus niger rufiventer E. Geoffroy St. -Hilaire, 1803 Fox squirrels have been reported previously from Hemphill and Hutchinson counties in the northern part of the Panhandle but the distribution there is not well documented. We have two additional records from the north side of the Canadian River in Hemphill County — 3 mi. E. Canadian (TTU 42731) and 13 mi. E Canadian (TTU 42730). MAMMALS FROM THE TEXAS PANHANDLE 99 Additionally, the collection at Texas Tech University contains six specimens from Hansford County, three from 3 mi. S and 6 mi. W Gruver (TTU 270-72) and three from 10 mi. S and 3 mi. W Gruver (TTU 269, 273- 74). In the summer of 1985, Hollander observed several fox squirrels in cottonwoods along Pitcher Creek, a tributary of the Canadian River, 16 mi. N Amarillo, Potter County, and at least two others at Boys Ranch, on the north bank of the Canadian in Oldham County. This squirrel undoubtedly occurs in suitable habitat all across the Panhandle along the Canadian and its immediate tributaries, and also in the drainage basin of the North Canadian or Beaver River in the extreme northeastern part. Geomys bursarius jugossicularis Hopper, 1940 This subspecies has been recorded from one locality in the northwestern part of the Panhandle — 15 mi. E Texline, Dallam County (Hall and Kelson, 1952:364). Tentatively, we assign specimens from 1 mi. N and 3 mi. W Dalhart, Dallam County (TTU 42749-50), and 1 mi. S Dalhart, in Hartley County (TTU 42751-56), to G. b. jugossicularis because they are noticeably paler than pocket gophers collected to the east of those localities (Collingsworth, Hemphill, Lipscomb, and Moore counties) that represent G. b. major. In a study of pocket gophers of the genus Geomys in Texas, Davis (1940) did not examine material from the northwestern Panhandle and, in fact, did not map Geomys as occurring in that area. Similarly, Honeycutt and Schmidly (1979) did not treat gophers from that part of the Panhandle, possibly because of the paucity of specimens then available. Hart (1978) recognized G. b. jugossicularis , placing it in his “ bursarius group” of subspecies, whereas Heaney and Timm (1983) regarded jugossicularis as a synonym of lutescens , which they reckoned as a species distinct from bursarius. Whatever the ultimate fate taxonomically of pocket gophers from the northwestern Panhandle, they do appear to represent a paler race than specimens from farther to the east and south. Castor canadensis Kuhl, 1820 Both Davis (1974:190) and Schmidly (1984:5) mapped the beaver as occurring only in the extreme eastern part of the Panhandle. We observed indisputable evidence of this rodent (complete with photographs) near the mouth of Romero Creek, on the north side of the Canadian River, 17 mi. N and 1 mi. W Adrian, Oldham County. Furthermore, local residents informed us that beavers inhabited Pitcher Creek, on the Canadian 16 mi. N Amarillo, Potter County, within the past five years, and this species also was reported to us as resident below the dam of Lake Meredith in Hutchinson County. Evidently, C. canadensis occurs sporadically along 100 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 the Canadian and its major tributaries. The subspecific identity of the Panhandle population is questionable. Reithrodontomys megalotis aztecus J. A. Allen, 1893 The western harvest mouse is generally uncommon in the Panhandle and evidently only locally distributed there. We are aware of only three previously reported specimens — one from Hutchinson County and two from Hansford County (Jones and Mursaloglu, 1961:21). We have collected 22 specimens to date, all taken in grassy-weedy habitats with good ground cover, from the following localities: 4 mi. N and 1 mi. E Dumas (TTU 43050-51), and 3 mi. S Dumas (TTU 43052), Moore County; 10 mi. N Stratford (TTU 43899-901), and 8 mi. S and 2 mi. E Stratford (TTU 43902- 17), Sherman County. The percentage of tail length to length of head and body in our mice is less than typical for R. m. aztecus (83.7 percent in 20 adult specimens as opposed to an average of more than 90 in specimens of aztecus examined by Jones and Mursaloglu, op. cit.), suggesting the need for detailed systematic study of this species in the Panhandle and adjacent areas. Baiomys taylori taylori (Thomas, 1887) Two adult pygmy mice, one of each sex, were trapped along a grassy- weedy fencerow, adjacent to mesquite pastureland, 9 mi. E Lutie, Collingsworth County, on 14 May 1986. These not only provide the first records of this rodent from the Panhandle, but represent the northernmost point of known occurrence in North America. Furthermore, the locality of capture lies but a few miles west of the Oklahoma line, a state whence B. taylori is known from a single specimen recently reported from just north of the Red River in Cotton County (Stangl and Dalquest, 1986:123). The male (TTU 43780) had testes measuring 3X2 mm; the female (TTU 43779) carried three fetuses that were 5 mm in crown-rump length. Neotoma albigula warreni Merriam 1908 Davis (1974:220) did not map the white-throated woodrat as occurring in the northern part of the Panhandle, although Cutter (1959:448-449) reported specimens from Hansford and Hutchinson counties, referring them to the subspecies warreni on the basis of coloration, size, and several cranial features. Subsequently, Rogers and Schmidly (1981:180) recorded specimens of N. a. warreni from Moore and Potter counties. We collected nine rats of this species in rocky habitats at the following localities: 5 mi. W Boys Ranch (TTU 42824-25), 6 mi. W Boys Ranch (TTU 42823), and 18 mi. N and 1 mi. W Adrian (TTU 42818-22), all in Oldham County, and 4.8 mi. NW Sanford, Hutchinson County (TTU 42817). We have compared our material with specimens of N. a. albigula from several places near the type locality in Arizona and with specimens of N. MAMMALS FROM THE TEXAS PANHANDLE 101 a. warreni from Cimmaron County, Oklahoma. They definitely agree with the former in color, but with warreni in cranial details as described by Rogers and Schmidly ( loc . cit.). Erethizon dorsatum bruneri Swenk, 1914. The porcupine probably ranges sparingly throughout the Panhandle. The only record from north of the Canadian River, however, is of one found dead in a cottonwood grove 9 mi. E Stinnett, Hutchinson County (Blair, 1954:254), which was overlooked by Davis (1974). We found one killed on FM 767 at a locality 6.4 mi. W Channing, Hartley County, on 23 August 1985. Only the left dentary and tail (TTU 43103) of this animal could be salvaged. We follow Hall (1981) in referring our specimen to E. d. bruneri. Systematics of E. dorsatum in Texas and adjacent regions are deserving of serious study. Acknowledgments We are indebted to several Texans who assisted our efforts in the field, but most especially to the John Henard family of Collingsworth County, Sam Mason of the Griffin Ranch in Oldham County, Robert Morris of the Fain Ranch in Potter County, Donald Tyson of the Tyson Ranch in Lipscomb County, officials of Cal Farley’s Boys Ranch in Oldham County, and Craig S. Hood, Texas Tech University. Field work was supported in part by funds made available by the Graduate School and Office of Research Services, Texas Tech University. Literature Cited Bailey, V. 1905. Biological survey of Texas. N. Amer. Fauna, 25:1-222. Blair, W. F. 1954. Mammals of the mesquite plains biotic district in Texas and Oklahoma, and speciation in the central grasslands. Texas J. Sci., 6:235-264. Cutter, W. L. 1959. The Warren woodrat in Texas. J. Mamm,, 40:448-449. Davis, W. B. 1940. Distribution and variation of pocket gophers (genus Geomys) in the southwestern United States. Bull. Texas Agric. Exp. Sta., 590:1-38. - . 1974. The mammals of Texas. Bull. Texas Parks and Wildlife Dept., 41:1-294. Gould, F. W. 1975. Texas plants — a checklist and ecological summary. Bull. Texas Agric. Exp. Sta., MP-585 (revised): 1-1 21. Hall, E. R. 1981. The mammals of North America. John Wiley & Sons, New York, 2: vi+601-1 181+90. Hall, E. R., and K. R. Kelson. 1952. Comments on the taxonomy and geographic distribution of some North American rodents. Univ. Kansas Publ., Mus. Nat., Hist., 5:343-371. Hart, E. B. 1978. Karyology and evolution of the plains pocket gopher, Geomys bursarius. Occas. Papers Mus. Nat. Hist., Univ. Kansas, 71:1-19. Heaney, L. R., and R. M. Timm. 1983. Relationships of pocket gophers of the genus Geomys from the Central and Northern Great Plains. Misc. Publ. Mus. Nat., Hist., Univ. Kansas, 74:1-59. Honeycutt, R. L., and D. J. Schmidly. 1979. Chromosomal and morphological variation in the plains pocket gopher, Geomys bursarius, in Texas and adjacent states. Occas. Papers Mus., Texas Tech Univ., 58:1-54. 102 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 Jones, J. K., Jr., and B. Mursaloglu. 1961. Geographic variation in the harvest mouse, Reithrodontomys megalotis, on the Central Great Plains and in adjacent regions. Univ. Kansas Publ., Mus. Nat. Hist., 14:9-27. Martin, C. O., and D. J. Schmidly. 1982. Taxomonic review of the pallid bat, Antrozous pallidus Le Conte. Spec. Publ. Mus., Texas Tech Univ., 18:1-48. Owen, J. G., and D. J. Schmidly. 1986. Environmental variables of biological importance in Texas. Texas J. Sci., 38:99-1 19. Rogers, D. S., and D. J. Schmidly. 1981. Geographic variation in the white-throated woodrat ( Neotoma albigula ) from New Mexico, Texas, and northern Mexico. Southwestern Nat., 26:167-181. Schmidly, D. J. 1984. The furbearers of Texas. Bull. Texas Parks and Wildlife Dept., 1 1 1 :viii+l-55. Stangl, F. B., Jr., and W. W. Dalquest. 1986. Two noteworthy records of Oklahoma mammals. Southwestern Nat., 31:123-124. GENERAL NOTES RANGE AND HABITAT EXPANSION OF THE INTRODUCED SLUG, ANGUSTIPES AMEGHINI, IN EXTREME SOUTHERN TEXAS Raymond W. Neck Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, Texas 78744 Of the veronicellid slugs that have become established in the Brownsville, Cameron Co., Texas, area (Neck, 1976, 1981, 1985), the most abundant is Angustipes ameghini (Gambetta, 1923). Native only to Paraguay (Gambetta, 1923), A. ameghini also is known from other localities on the Gulf Coastal Plain of the southern United States (Dundee, 1974; Burch and Van Devender, 1980). Initial Brownsville records were from urban residential yards (Neck, 1976). Stange (1978) reported that A. ameghini in Florida was “an urban slug found in vacant lots and cemeteries.” The purpose of this note is to report expansion of range to an adjacent county and recent dispersal to native brush tracts within Cameron County. Living individuals of A. ameghini were collected at two localities in Hidalgo County, Texas: 1) Edinburg — urban residential yard, 317 Enfield Road, slugs found under bricks and flower pots in garden area, four specimens collected on 28 September 1983, had been seen by occupants for “about three years” (Audrey and Wayne Holiman, personal communication); and 2) Santa Ana National Wildlife Refuge — landscaped area around headquarters building and nearby transplant garden-greenhouse area, five specimens collected on 13 and 14 October 1984 by Joe Ideker, seen only recently (dried bodies discovered during personal survey on 27 September 1984). Additional localities from urban habitats other than yards in Brownsville and native brush habitats near Brownsville in Cameron County are as follows: 1) Town Resaca at Boca Chica Blvd., Brownsville, five slugs found under downed branches in second-growth brush at edge of water, 22 June 1984, five additional slugs found 200 meters north in more open area with wood on ground, 21 October 1984; 2) Esperanza Ranch-area with remnant native brush habitat along Resaca de la Palma at eastern edge of Brownsville, three slugs found under downed mesquite trunk, 22 October 1984; and 3) West Palm Grove — 3.1 kilometers southeast of the Esperanza Ranch site, remnant palm-thorn scrub woodland, two slugs found underneath palm leaf litter, 22 October 1984. These additional Cameron County records are the first valid records of A. ameghini that are not from residential yards. The report from Rabb Palm Grove (Neck, 1976) actually refers to Sarasinula plebeia Fischer, 1868, a species that has not been observed at this locality since the original collection on 23 December 1970. The significance of the Town Resaca records is the ability of A. ameghini to survive in feral populations without direct supplemental water, although such habitat areas along resaca shorelines are quite restricted. The occurrence of A. ameghini at the Esperanza Ranch site indicates that brush communities associated with resacas also may be able to support this species. The West Palm Grove locality is not associated with a resaca and is a much more xeric area than the two localities previously mentioned. The initial Hidalgo County records and additional Cameron County records provide some insight into probable dispersal routes utilized by A. ameghini as well as likely limits of habitat suitability in the lower Rio Grande Valley. The Santa Ana population is apparently the result of slugs present in commercial nursery stock of local native plants that were brought onto the refuge. The Edinburg population further demonstrates that A. ameghini is being dispersed The Texas Journal of Science, Vol. 39, No. 1, February, 1987 104 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 1, 1987 within extreme southern Texas. The Town Resaca populations were founded by overland dispersing slugs over short distances from nearby residences. The more isolated nature of the Esperanza Ranch indicates that drainage ditches and resacas carrying urban runoff water may function as dispersal routes of A. ameghini. Field surveys have revealed no hint of the origin of the West Palm Grove population as this area is isolated from urban drainage by the levee system that limits the floodwaters of the Rio Grande. The environmental factor likely to be most important in limiting geographic and ecologic occurrence of A. ameghini in southern Texas is moisture rather than extreme winter temperatures. Populations of A. ameghini tend to be more restricted in Edinburg than Brownsville. Edinburg is both drier and cooler than Brownsville, although the differences are small. Feral populations in native brush communities no doubt will exhibit lower densities and suffer higher extinction rates than those in urban residential lawns. Significantly, A. ameghini is the most drought-tolerant of the three introduced veronicellids that have established populations in Brownsville. Literature Cited Burch, J. B., and A. S. Van Devender. 1980. Identification of eastern North American land snails. The Prosobranchia, Opisthobranchia and Pulmonata (Actophila). Trans. POETS Soc., 1:33-80. Dundee, D. S. 1974. Catelog of introduced molluscs of eastern North American (north of Mexico). Sterkiana, 55:1-37. Gambetta, L. 1923. Alcuni vaginuldi sub-americanai. Boll. Musei. Zool. Anat. Comp. R. Univ. Torino, 38:1-10. Neck, R. W. 1976. Adventive land snails in the Brownsville, Texas area. Southwestern Nat., 21:133-135. - . 1981. Noteworthy gastropod records from Texas. Texas Conchologist, 17:69-72. — . 1985. Tropical veronicellid, Laevicaulis alte (Ferussac), established in southern Texas. The Nautilus, 99:19-20. Stange, L. A. 1978. The slugs of Florida (Gastropoda: Pulmonata). Florida Dept. Agric. Consumer Serv. Ento. Cir., 197:1-4. DATES OF DISTRIBUTION OF VOLUME 38 TEXAS JOURNAL OF SCIENCE 38(1) 4 March 1986 38 (2) 12 May 1986 38 (3) 1 August 1986 38 (4) 25 November 1986 THE TEXAS ACADEMY OF SCIENCE, 1986-87 Officers President: President-Elect: Vice-President: Immediate Past President and Executive Secretary: Treasurer: Editor: A AS Council Representative: Directors 1984 E. D. McCune, Stephen F. Austin State University James A. Neal, U.S. Fish and Wildlife Service 1985 George B. McClung, San Angelo Barbara Schreur, Texas A&I University 1986 Caroline P. Benjamin, Southwest Texas State University R. John Prevost, Southwest Research Institute Sectional Officers I — Mathematical Sciences: Barbara Schreur, Texas A&I University II — Physical Sciences: C. A. Quarles, Texas Christian University III — Earth Sciences: James L. Carter, University of Texas at Dallas IV — Biological Sciences: Robert S. Baldridge, Baylor University V — Social Sciences: David A. Edwards, San Antonio College VI — Environmental Sciences: Bennett J. Luckens, Austin VII — Chemistry: Rodney Cate, Midwestern State University VIII — Science Education: Dick E. Hammond, Southwest Texas State University IX — Computer Sciences: Ronald King, Baylor University X — Aquatic Sciences: Paula Dehn, University of Texas at San Antonio Counselors Collegiate Academy: Shirley Handler, East Texas Baptist College Helen Oujesky, University of Texas at San Antonio Junior Academy: Ruth Spear, San Marcos Peggy Carnahan, San Antonio Billy J. Franklin, Lamar University Lamar Johanson, Tarleton State University Owen T. Lind, Baylor University William J. Clark, Texas A&M University Michael J. Carlo, Angelo State University J. Knox Jones, Jr., Texas Tech University Ann Benham, University of Texas at Arlington 2nd CLASS POSTAGE PAID AT LUBBOCK TEXAS 79401 3024 I .87 Li hr ary Acqui si t i ons Smith 5 □ n i a n I n s t i t u t e uj ps p. p. n ton D -7 2 0 3 6 0 PUBLISHED QUARTERLY BY THE TEXAS ACADEMY OF SCIENCE ^tfSONId/v I'/iAY 28 1987 LiBRAR^- / dL.Tfy Ntt- Volume 39 Number 2 May 1987 GENERAL INFORMATION MEMBERSHIP. — Any person or group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. Dues for members are $20.00 annually; student members, $12.00 annually; sustaining members, at least $30.00 in addition to annual dues; life members, at least $400.00 in one payment; patrons, at least $500.00 in one payment; corporate members, $250.00 annually; corporate life members, $2000.00 in one payment. Library subscription rate is $45.00 annually. Payments should be sent to Dr. Michael J. Carlo, P.O. Box 10986, Angelo State University, San Angelo, Texas 76909. The Texas Journal of Science is a quarterly publication of the Academy and is sent to all members and subscribers. Inquiries regarding back issues should be sent to Dr. William J. Clark, Drawer H6, College Station, Texas 77844. AFFILIATED ORGANIZATIONS Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers American Association for the Advancement of Science Texas Society of Mammalogists The Texas Journal of Science (USPS 616740) is published quarterly at Lubbock, Texas U.S.A. Second class postage paid at Post Office, Lubbock, TX 79401. Please send form 3579 and returned copies to Texas Tech Press, Box 4240, Lubbock, TX 79409. ISSN 0040-4403 THE TEXAS JOURNAL OF SCIENCE Volume 39, No. 2 May 1987 CONTENTS Aggressive and defensive propensities of Solenopsis invicta (Hymenoptera: Formicidae) and three indigenous ant species in Texas. By Stanley R. Jones and Sherman A. Phillips, Jr . . . 107 Sampling effort and bias in soft-sediment benthic investigations using the Peterson dredge. By Frank S. Shipley . . . . 117 Lampropeltis similis from the Coffee Ranch local fauna (Hemphillian Land Mammal Age) of Texas. By Dennis Parmley . . . 123 An assessment of geographic and seasonal biases in systematic mammal collections from two Texas universities. By Frederick B. Stangl, Jr., and Elizabeth M. Jones . 129 A comparison of the water chemistry and benthic macroinvertebrate communities of two oxbow lakes in the Red River Basin, northwestern Louisiana. By Jack D. McCullough and Clarence W. Reed . 139 Evaluation of volcanic ash as a stratigraphic marker in playa basins, western Texas. By Kevin Cornwell and Calvin G. Barnes . 155 Karyotypes of five cricetid rodents from Honduras. By Robert D. Bradley and Jan Ensink . 171 Freshwater bivalves of the Baffin Bay drainage basin, southern Texas. By Raymond W. Neck . 177 Effect of salt on grain and forage intake in cattle. By Robert D. Brown and Alejandro Creixell . 183 General Notes Caloric content of an excavated food cache of Perognathus flavescens. By Kent M. Reed . 191 An unusual “woodrat nest.” By Walter W. Dalquest and Greg J. Coin . 192 Distributional record of Lasiurus seminolus (Chiroptera: Vespertilionidae). By Thomas E. Lee, Jr . 193 Notommata allantois in northeastern Texas lakes: a rotifer previously known only from Europe. By John Shoemaker, Sally H. Davis, and Robert K. Williams . 194 Two noteworthy populations of the fiddler crab, Uca sub cylindrical in southern Texas. By Raymond W. Neck . 196 A record of the western small-footed myotis, Myotis ciliolabrum Merriam, from the Texas Panhandle. By Robert R Hollander and J. Knox Jones, Jr . 198 Instructions to Authors 199 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Editor: J. Knox Jones, Jr., Texas Tech University Assistant to the Editor: Marijane R. Davis, Texas Tech University Associate Editor for Botany: Randy Moore, Baylor University Associate Editor for Chemistry: Marvin W. Rowe, Texas A&M University Associate Editor for Computer Science: Ronald K. Chesser, Texas Tech University Associate Editor for Mathematics and Statistics: George R. Terrell, Rice University Associate Editor for Physics: Charles W. Myles, Texas Tech University Editorial Assistants: Robert R. Hollander, Texas Tech University Richard W. Manning, Texas Tech University Scholarly papers in any field of science, technology, or science education will be considered for publication in The Texas Journal of Science. Instructions to authors are published one or more times each year in the Journal on a space-available basis, and also are available from the Editor (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409, 806 / 742-2487, Tex-an 862-2487). AGGRESSIVE AND DEFENSIVE PROPENSITIES OF SOLENOPSIS INVICTA (HYMENOPTERA: FORMICIDAE) AND THREE INDIGENOUS ANT SPECIES IN TEXAS Stanley R. Jones and Sherman A. Phillips, Jr. Department of Entomology, Texas Tech University, Lubbock, Texas 79409 Abstract. — Since its accidental introduction into Mobile, Alabama, the red imported fire ant, Solenopsis invicta Buren, has become a pest in much of the southern United States. Each year this ant continues to expand its range. The reasons »S. invicta is able to colonize successfully areas already inhabited by multifarious ant species are not understood; however, possible explanations include its aggressive behavior, reproductive capacity, large colony size, and foraging efficiency. This laboratory study was conducted to compare the aggressive and defensive behaviors of S. invicta and three native ant species of central Texas with one another. Native ants studied were Pheidole dentata Mayr, Forelius foetidus (Buckley), and Monomorium minimum (Buckley). We found interspecific differences in aggressive and defensive abilities. Solenopsis invicta and P. dentata were more aggressive than F. foetidus and M. minimum. Although the latter two species exhibited the least aggression, they displayed the greatest defensive abilities. Key words: Solenopsis invicta ; aggression; defense; native ants. The invasion and disruption of long-established populations of indigenous ant species by foreign ant species have been recorded on numerous occasions. Way (1953) reported the near total replacement of the dominant Oecophylla longinoda Latreille of eastern Africa by Anoplolepis longipes (Jerdon) introduced from India. Similarly, Pheidole megacephala (F.) has successfully displaced native ant species from numerous tropical and semitropical oceanic islands (Zimmerman, 1970). Iridomyrmex humilis Mayr also has undergone massive range expansion at the expense of other ant species (Haskins and Haskins, 1965; Fluker and Beardsley, 1970; Erickson, 1971). More recently, Clark et al. (1982) and Lubin (1984) found that Wasmannia auropunctata (Roger) has a negative impact on many of the ant species of the Galapagos Islands through interference competition. Finally, MacKay and MacKay (1982) stated that Formica haemorrhoidalis Emery can invade and completely eliminate the nests of Camponotus laevigatus (F. Smith) in southern California. However, unlike the previous examples cited, both of these sympatric species are native and could perhaps coexist indefinitely. The red imported fire ant, Solenopsis invicta Buren, has had similar success displacing ants indigenous to the southern United States (Wilson and Brown, 1958). Reagan et al. (1972) reported that S. invicta is the only ant species found by extensive pitfall trapping in Louisiana sugarcane fields following invasion by this species. Whitcomb et al. (1972) also reported that the presence of S. invicta in the fields of northern Florida reduces the densities of the native ant fauna. Buren et al. (1974) reported the The Texas Journal of Science, Vol. 39, No. 2, May, 1987 108 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 apparent elimination of Solenopsis xyloni McCook by S. invicta from several Louisiana pastures. Despite the apparent success and aggressiveness of S. invicta , several studies have revealed that this ant may not be so successful during individual encounters with other species, or when interspecific colony sizes are equal. Bhatkar (1973) demonstrated that the defensive abilities of at least six ant species indigenous to Florida exceed those of S. invicta. Also, Bhatkar et al. (1972) presented evidence that a single Lasius neoniger Emery worker is capable of killing two to three S. invicta workers. The success of S. invicta , therefore, is most likely due to high fecundity and overwhelming numbers, rather than to individual aggressive ability. An additional factor that may affect the ability of native ant species to successfully compete with, and coexist with, S. invicta is defensive ability. Several authors have alluded to this concept, but few have dealt with it specifically. Howard and Oliver (1979) stated that many ants in a Louisiana pasture tended to avoid conflict. Bhatkar (1973) stated that individuals of various ant species often demonstrate momentary defensiveness prior to escape from those of a more aggressive species. Several ant species apparently possess specialized behaviors or morphological adaptations allowing superior defensive ability. For example, Pheidole dentata Mayr has an enemy-specific alarm-recruitment system that alerts colony members to approaching intruders (Wilson, 1976). Monomorium minimum (Buckley) uses a drop of venom on its sting as an effective deterrent and irritant to would-be attackers (Baroni-Urbani and Kannowski, 1974; Adams and Traniello, 1981). A sticky fluid, discharged from the repugnatorial gland of Forelius foetidus (Buckley), acts as both repellent and irritant and, in some cases may be fatal to attacking ants (Wheeler, 1910). The three myrmicine species S. invicta , P. dentata , and M. minimum and the dolichoderine species F. foetidus consistantly occurred at high frequencies during an extensive, year-long sampling study conducted by Rogers (1984) in Kerr and Bandera Counties, Texas. That study indicated that these ant species coexist in the same habitats. Given that S. invicta has been established in Kerr and Bandera counties for approximately 10 years, we may reasonably assume that the other three species are coexisting with, and successfully competing with S. invicta. Therefore, the following laboratory study was conducted to compare the aggressive and defensive abilities of these four ant species and to understand better the mechanisms involved in their apparent coexistence. Materials and Methods Colonies were collected from Kerr, Bandera, and Lubbock counties, Texas, during their diurnal periods of lowest foraging activity, corresponding to early morning for F foetidus and M. minimum, and early afternoon for P. dentata and S. invicta (authors observation). Colonies of M. minimum and F. foetidus contain monomorphic workers, whereas workers AGGRESSION AND DEFENSE IN ANTS 109 of P. dentata are dimorphic, resulting in majors and minors. However, S. invicta workers are polymorphic, consisting of minors (head width 0.72 mm or less), medias (head width 0.73-0.92 mm), and majors (head width 0.93 mm or more) (Wilson, 1978). Aggression. — Propensity for aggression or combative ability was ascertained by placing a predetermined number of ants of different species in plastic containers (11 by 11 by 4 centimeters). The bottom surface of each container was lined with moist Castone dental plaster, providing a rough substratum as well as a moisture reservoir to prevent ant desiccation. One milliliter of distilled water was added to the containers prior to each test. The sides of each container were coated with Fluon to prevent ant escape. Each test was conducted at 30° C under fluorescent lighting and 70 to 75 percent relative humidity. Ten individuals of one species-caste were simultaneously placed in a container with 10 individuals of another species-caste. Castes of S. invicta as defined by Wilson (1978) were determined by measuring head capsule widths. Individuals were then left undisturbed for three hours, after which the number alive and dead were recorded. Individuals that retained movements but were incapacitated, were recorded as dead. Ten individuals of each species-caste were maintained in separate containers and served as controls. These procedures resulted in 17 confrontation pair groups, each of which was replicated eight times. Data were analyzed with two-tailed Student t-tests, first for each species group, then for each caste within species, thus yielding a comparative indication of combative ability for each species and each species-caste. Defensiveness. — Propensity for defensiveness or avoidance of confrontation was tested by placing a single individual of one species-caste with that of another. Containers for these trials consisted of petri dishes (60 by 20 millimeters) coated on the sides with Fluon. Temperature and humidity conditions were as above. Again, the 17 species-caste combinations were made, each with 1 5 replications. Following placement of two individuals into a container, the number of contacts made between each was recorded every 60 seconds for a maximum of 10 minutes. If a contact resulted in overt aggression leading to death, the time of this occurrence was recorded. Data were analyzed with a nested ANOVA followed by Student-Newman-Kuel’s multiple comparison test. Results Aggression Species combative performances are presented in Figure 1. No significant differences were detected between the combative abilities of P dentata and S. invicta (t = 0.35), R dentata and M. minimum (t = 0.78), or F. foetidus and M. minimum (t = 1.93). However, S. invicta was significantly greater in combative ability over F. foetidus (t — 4.55) and M. minimum (t = 7.07), whereas P. dentata was superior in ability only to F foetidus (t = 8.55). Results of combative abilities for the 17 species-caste combinations are presented in Figure 2. The general trend during three hours of confrontation between larger species-castes of similar size (pair groups 1-6) was that, regardless of species, the larger caste of the pair group was more aggressive and had greater combative ability. This trend can be noted for S. invicta majors and R dentata minors (t = 36.56), S. invicta medias and P dentata majors (t = 22.54), S. invicta medias and P. dentata minors (t = 14.76), and S. invicta minors and P. dentata majors (t = 39.00). However, the combative abilities of the latter two species were the same when equal- 110 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 CONFRONTATION SPECIES PAIRS I P dentata S. invicta o P. dentata F. foetidus 3 P. dentata M. minimum A S. invicta • F. foetidus _ S. invicta M. minimum 6 F. foetidus M minimum MEAN NUMBER ALIVE AFTER 3 HOURS (STUDENT'S PAIRED t~ TESTi P<0.05 ; df = 7) Figure 1. Results of two-species confrontation trials among four ant species. Data analyzed by Students’ t-tests. Means with the same letter are not significantly different ( P > 0.05) within species pair (n = 10 ants/ species). sized castes were compared (pair groups 1 and 6). No significant difference was detected between P. dentata and S'. invicta majors (t = 0.42) nor between R dentata and S. invicta minors (t =2.05). Comparing combative ability of the larger species-castes versus the smaller species (pair groups 7-12), significant differences were detected in every instance. Pheidole dentata majors were greater in combative ability than those of M. minimum and F. foetidus (t = 22.29; t = 99.00), as were S. invicta majors (t = 19.94; t = 3.47). In addition, S. invicta medias were significantly greater in ability than those of M. minimum (t = 7.76) and F. foetidus (t = 3.47). However, whereas P. dentata majors killed all F. foetidus (pair group 8), the impact of S. invicta majors and medias on F foetidus (pair groups 10 and 12) was not as great. Comparing smaller species-castes of similar size (pair groups 13-17), no difference in aggression was detected between S. invicta minors and M. minimum minors (t = 1.36) nor between those of M. minimum and F. foetidus (t = 1.93). Although P dentata minors and S. invicta minors were superior to F. foetidus minors (t = 3.98; t = 5.31), M. minimum minors were greater in ability than minors of P. dentata (t = 10.12). All ants maintained as controls during each of the 17 trials survived the three-hour trial durations. Defensiveness The results of trials involving defensive abilities are presented in Figure 3. Analysis of defensive abilities or avoidance of aggression indicates that AGGRESSION AND DEFENSE IN ANTS 111 CONFRONTATION SPECIES-CASTE PAIRS S. invicta (MAJOR) L P. dentata (MAJOR) 0 S. invicta (MAJOR) d' P. dentata (MINOR) ■3 S. invicta (MEDIA ) P. dentata (MAJOR) . S. invicta (MEDIA ) ^ P dentata (MINOR) c S. invicta (MINOR) ° P. dentata (MAJOR) - S. invicta (MINOR) b P. dentata (MINOR) P. dentata (MAJOR) r M. minimum „ P. dentata (MAJOR F. foetidus 9 S. invicta (MAJOR) M. minimum S. invicta (MAJOR) 0 F. foetidus 1 1 S. invicta (MEDIA ) M. minimum ip S. invicta (MEDIA ) F. foetidus |3 P. dentata (MINOR) ' M minimum R dentata (MINOR) F. foetidus |5 S. invicta (MINOR) ' M. minimum S. invicta (MINOR) lb F. foetidus |7 M minimum ' F foetidus ^V////////////////////////////7)r+ b V/////////////////////////////////////^2^ b a m b Z//////////S//////F- . 0 2 4 6 8 10 MEAN NUMBER ALIVE AFTER 3 HOURS (STUDENT'S PAIRED t- TEST : P < 0.05; df = 7 ) Figure 2. Results of confrontation trials for all 17 species-caste combinations. Data analyzed by Students’ t-tests. Means with the same letter are not significantly different (P > 0.05) within species-caste pair (n = 10 ants/ species-caste). differences do occur in species-caste pairs (F = 24). Although the mean number of contacts of pairs 1-8 was large, no significant differences were detected among them prior to overt aggression. Seven of the pairs from that group involved M. minimum and F foetidus paired with either P dentata or S. invicta , whereas one pair involved M. minimum and F. foetidus. Note that, within this group, either both species-castes within a pair were small or the size discrepencies among species-castes within a pair were large, resulting in a greater number of contacts than detected in pairs 9-17. Results from this latter group indicated that, in general, as size of the species-caste increases and the size difference of the species-caste within 112 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 AVOIDANCE SPECIES- CASTE PAIRS S, invicta (major ) 1 F. foetidus M. minimum F. foetidus 3 S. invicta (minor) ' F . foetidus 4. S. invicta (media) F . foetidus 5 S. invicta (major ) ’ M. minimum 0 S. invicta (media) M. minimum 2 P . dentata ( minor ) F . foetidus g P . dentata (minor ) M . minimum 9 S invicta (minor ) M. minimum 10 P. dentata (major ) M. minimum 11 P dentata (major ) F. foetidus 12 P dentata (minor ) S. invicta (major ) 13 P. dentata (minor ) S. invicta (minor ) I4.P. dentata (major ) S. invicta (major ) 15 P. dentata (minor ) S. invicta (media ) 10 P. dentata (major ) S. invicta (minor ) 17 P dentata (major ) S. invicta (media ) Figure 3. Comparison of defensive abilities within all 17 species-caste combinations. Data analyzed by Student-Newman-Kuels’ test. Means with the same letter are not significantly different ( P> 0.05) within species-caste pair (n = 10 ants/ species-caste). each pair decreases, the number of contacts prior to overt aggression also decreases. This trend was particularly evident between P. dentata and S. invicta (pairs 16 and 17). Discussion Laboratory observations revealed that in tests of both aggression and defense, S. invicta and R dentata were the aggressor species, whereas F. foetidus and M. minimum were more defense oriented. Though forced, prolonged interactions between these ant species-castes are not necessarily the typical interactions occurring in nature, they do provide useful insight into the methods used and success obtained during interspecific confrontations. In situations of forced, prolonged interaction, R dentata and S. invicta were superior in terms of species combative ability. In general, smaller castes were less successful in competition with larger castes. AGGRESSION AND DEFENSE IN ANTS 113 Individuals tended to be most aggressive toward ants slightly smaller than themselves. Ants of similar size were less aggressive toward each other, and were better able to survive aggressive encounters than were individuals of disparate sizes. Pheidole dentata majors were more aggressive than all other species and castes tested with the single exception of 5. invicta majors. If aggression and defense alone are considered, one would expect S. invicta to have difficulty displacing colonies of P. dentata in nature. However, S. invicta colony members outnumber those of P. dentata several-fold. In addition, the major caste of P. dentata comprises a mere 20 percent of the total worker force (Wilson, 1976), thus limiting their effectiveness during large-scale encounters. Under natural circumstances, individual foragers usually meet once, followed by immediate directional changes and separation (authors observation). Often, an ant of the more aggressive species will attempt to attack an individual of a less aggressive species, but these attempts are seldom successful. The smaller, less aggressive species such as F. foetidus and M. minimum employ a highly effective chemical interference, which deters attack and allows for escape from aggressors (Wheeler, 1910; Baroni- Urbani and Kannowski, 1974). Note that F. foetidus and M. minimum exhibited the greatest number of contacts prior to fatal combat during defensive ability trials (Fig. 3). In fact, M. minimum and F. foetidus appeared in all of the top eight trials. This finding demonstrates that during individual foraging encounters, the probability of escape or avoidance of conflict by F. foetidus and M. minimum is high, but decreases with increasing number of contacts. Because the probability of numerous encounters over short time periods is unlikely during normal foraging activity, the benefits obtained by effective chemical defense are great. The chemical repellent used by M. minimum and F. foetidus is generally referred to as defensive in nature (Baroni-Urbani and Kannowski, 1974; Adams and Traniello, 1981). However, observations of these two species in confined arenas made the distinction between defensive and offensive use of these chemicals less clear. The two species often appeared to use their chemical “defense” in an offensive manner. In communities where S. invicta population densities are low, P. dentata may compete successfully with S . invicta because of their highly specialized and aggressive soldier caste. Similarly, F. foetidus and M. minimum may compete successfully by avoiding direct aggression via effective chemical defense. Potts et al. (1984) suggested that S. invicta colony number and population densities may be severely affected by periods of water stress. Xeric conditions may limit food availability as well as reproductive output of fire ant colonies. Francke et al. (1986) demonstrated that S. invicta are not freeze tolerant and that their northern limits of distribution appear to be correlated to winter temperatures. These two factors suggest that as S. invicta expands its range northward and westward into more xeric 114 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 and gelid conditions, colony densities (mound and individual numbers within each mound) should decrease. This is, of course, barring any rapidly acquired ability of S. invicta to survive these adverse abiotic conditions. As the densities of S. invicta decrease along the leading edge of range expansion, their numbers may more closely approximate those of the indigenous ant species, specifically P. dentata , F. foetidus , and M. minimum. Interspecific combative abilities then could approximate those found during this study. Because the indigenous ant species of Texas north and west of the current range of S. invicta are well adapted to their respective climatic conditions, their ability to effect a more significant degree of resistance on the further expansion of S. invicta may become apparent. Acknowledgments We thank Drs. A. A. Sorenson, L. Chandler, O. Francke, R. Sites, and Mr. J. Cokendolpher for their reviews and helpful comments on the manuscript. We are also indebted to Dr. M. Willig for his suggestions as to the statistical analyses. This study was supported by the Texas Department of Agriculture Interagency Agreement I AC (84-85)- 1 753. Contribution No. T- 10-171, College of Agricultural Sciences, Texas Tech University. Literature Cited Adams, A. S., and J. F. A. Traniello. 1981. Chemical interference competition by Monomorium minimum (Hymenoptera: Formicidae). Oecologia, 51:265-270. Baroni-Urbani, C., and P. B. Kannowski. 1974. Patterns in the red imported fire ant settlement of a Louisiana pasture: some demographic parameters, interspecific competition, and food sharing. Environ. Entomol., 3:755-760. Bhatkar, A. P. 1973. Confrontation behavior between Solenopsis invicta Buren and certain ant species native to Florida. Unpublished Ph.D. dissertation, Univ. Florida, Gainesville, 181 pp. Bhatkar, A., W. H. Whitcomb, W. F. Buren, P. Callahan, and T. Carlysle. 1972. Confrontation behavior between Lasius neoniger (Hymenoptera: Formicidae) and the imported fire ant. Environ. Entomol., 1:274-279. Buren, W. F., G. E. Allen, W. H. Whitcomb, F. E. Lennartz, and R. N. Williams. 1974. Zoogeography of the imported fire ants. J. New York Entomol. Soc., 82:113-123. Clark, D. B., C. Guayasamin, O. Pazmino, C. Donoso, and Y. Paez de Villacis. 1982. The tramp ant Wasmannia auropunctata : autecology and effects on ant diversity and distribution on Santa Cruz Island, Galapagos. Biotropica, 14:196-207. Erickson, J. M. 1971. The displacement of native ant species by the introduced Argentine ant Iridomyrmex humilis Mayr. Psyche, 78:257-266. Fluker, S. S., and J. W. Beardsley. 1970. Sympatric associations of three ants: Iridomyrmex humilis , Pheidole megacephala and Anoplolepis longipes, in Hawaii. Ann. Entomol. Soc. Amer., 63:1290-1296. Francke, O. F., J. C. Cokendolpher, and L. R. Potts. 1986. Supercooling studies on North American fire ants (Hymenoptera: Formicidae). Southwestern Nat., 31:81-94. Haskins, C. P., and E. F. Haskins. 1965. Pheidole megacephala and Iridomyrmex humilis in Bermuda-equilibrium or slow replacement? Ecology, 46:736-740. AGGRESSION AND DEFENSE IN ANTS 115 Howard, F. W., and A. D. Oliver. 1979. Field observations of ants (Hymenoptera: Formicidae) associated with red imported fire ants, Solenopsis invicta Buren, in Louisiana pastures. J. Georgia Entomol. Soc., 14:259-263. Lubin, Y. D. 1984. Changes in the native fauna of the Galapagos Islands following invasion by the little red fire ant, Wasmannia auropunctata. Biol. J. Linnaean Soc., 21:229-242. MacKay, W., and E. MacKay. 1982. Coexistence and competitive displacement involving two native ant species (Hymenoptera: Formicidae). Southwestern Nat., 27:135-142. Potts, L. R., O. F. Francke, and J. C. Cokendolpher. 1984. Humidity preferences of four species of fire ants (Hymenoptera: Formicidae: Solenopsis ). Insectes Sociaux, 31:335-339. Reagan, T. E., G. Coburn, and S. D. Hensley. 1972. Effects of mirex on the arthropod fauna of a Louisiana sugarcane field. Environ. Entomol., 1:588-591. Rogers, W. M. 1984. Variations in seasonal abundance of the red imported fire ant, Solenopsis invicta Buren, in relation to other ant species of central Texas. Unpublished M.S. thesis, Texas Tech Univ., Lubbock, 46 pp. Way, M. J. 1953. The relationship between certain ant species with particular reference to biological control of the coreid Theraptus sp. Bull. Entomol. Res., 44:669-691. Wheeler, W. M. 1910. Ants: their structure, development and behavior. Columbia Univ. Press, New York, 663 pp. Whitcomb, W. H., H. A. Denmark, A. P. Bhatkar and G. L. Green. 1972. Preliminary studies of the ants of Florida soybean fields. Florida Entomol., 55:129-142. Wilson, E. O. 1976. The organization of colony defense in the ant Pheidole dentata Mayr (Hymenoptera: Formicidae). Behav. Ecol. Sociobiol., 1:63-81. - . 1978. Division of labor in fire ants based on physical castes (Hymenoptera: Formicidae: Solenopsis). J. Kansas Entomol. Soc., 51:615-636. Wilson, E. O., and W. L. Brown, Jr. 1958. Recent changes in the introduced population of the fire ant Solenopsis saevissima (Fr. Smith). Evolution, 12:21 1-218. Zimmerman, E. C. 1970. Adaptive radiation in Hawaii with special reference to insects. Biotropica, 2:32-38. SAMPLING EFFORT AND BIAS IN SOFT-SEDIMENT BENTHIC INVESTIGATIONS USING THE PETERSON DREDGE Frank S. Shipley Texas Water Commission, 400 Mann St., Suite 905, Corpus Christi, Texas 78401 Abstract. — Eight replicate Peterson grabs of uniform mud substrate were taken in central Matagorda Bay, Texas, to assess within-site variability in organism densities among replicates. Based upon Morisita indices calculated on organism counts in the replicates, 41 percent of species were significantly aggrigated; none was distributed uniformly in spite of the homogeneous substrate at the study site. Most large species, like mollusks, and some small but infrequent species were undersampled by single grabs. Cumulative number of species increased asymptotically with sampling effort, indicating a need for multiple replicates to adequately characterize species composition. However, 97.9 percent of all individuals taken belonged to species taken in the first two replicates, indicating that the numerical composition of the community could be characterized in relatively few grabs. Key words: sampling effort; sampling bias; Peterson dredge; benthic; Morisita index; Matagorda Bay, Texas. Representative sampling of benthic marine organisms is difficult because many factors cause their distributions to depart from random. Species distributions may be influenced by substrate preference (Aller and Dodge, 1974), response to salinity or other water quality regimes (Rosenberg and Moller, 1979), and tolerance of disturbance (Boesch et al., 1976). Besides these physical factors, additional distributional patterns are imposed by biological interactions such as competition, predation, or coloniality (Woodin, 1974; Peterson, 1977). Finally, even randomly distributed organisms are likely to be undersampled, if they occur at low density in relation to the capacity of the sampling apparatus. As a result of nonrandom distributions in benthic organisms, biological patterns of interest may be difficult to distinguish from sampling variability. Most investigators are interested in demonstrating community differences between benthic habitats, or perhaps temporal changes in a particular habitat, but to do so, sampling error must be factored out. For example, statistical tests like ANOVA require estimates of within-site variability in order to distinguish differences between sites. In this investigation, I attempted to characterize within-site variability in Peterson sample counts to increase confidence in between-site community characterizations for a study in the same area to be reported elsewhere. Materials and Methods Eight Peterson grabs, each sampling one square foot (929 square centimeters) of substrate were taken near channel marker 16, central Matagorda Bay, Texas (Fig. 1). Sediment at this open bay location is uniform soft clay, allowing complete penetration of the Peterson apparatus to procure samples of about 20 liters. The site has stable salinities of approximately The Texas Journal of Science, Vol. 39, No. 2, May, 1987 118 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 23 parts per thousand (surface) and a water depth of three to four meters. Samples were washed through a 0.59-millimeter sieve on board the sampling vessel, and material retained by the sieve was narcotized in isotonic MgCl solution for about half an hour, preserved in buffered 10 percent formalin, and returned to the laboratory for vital staining with rose bengal, sorting, and identification. Organisms were identified to species in most cases, excepting nemertines. Variability in counts for each species among repicates was calculated as the coefficient of variation (CV), which corrects for gross differences between rare and common organisms. Morisita indices (Morisita, 1959) were calculated to test for nonrandom dispersions as a source of bias in sampling. The Morisita index is a dimensionless value for which <1 indicates ordered dispersion; exactly 1, random; and >1, clumped. The index is calculated as: q 2 nj (ni — 1) i= 1 18 = - > N(N — 1) where m equals the number of individuals in the ith replicate (i = l,2,3...q), q equals the number of replicates, and N equals the number of individuals in all replicates. 15 is independent of distribution type, number of replicates, and count magnitude differences, and lends itself to variance ratio testing for departure from random dispersion. Finally, cumulative number of species was plotted against sampling effort to quantify the diminishing returns associated with increasing effort. The cumulative percent of total individuals represented by these species in successive grabs was compared on this plot to assess the numerical portion of the community that could potentially be underrepresented at each sampling effort. Results and Discussion Nonrandom dispersion of organisms could contribute to sampling variability if some replicates included aggregations of organisms and others did not. The Morisita index calculated for each species (Table 1) resulted in significantly more values >1, compared to <1 (23 as opposed to four; binomial probability P <0.01). This result indicated a trend toward aggregation (mean species Morisita value, 1.42). Total organism counts (all species) ranged from 88 to 214 per replicate, indicating variability in overall organism densities. Individual species varied in dispersion. For the 36 species, coefficients of variation on counts across replicates ranged from 0.23 ( Paraprionospio pinnata) to 2.83, in four species from just a single replicate each. Eleven of the 27 species for which sufficient data existed showed a significant departure from one (random dispersion) in Morisita value (two-tailed variance ratio P <0.05). In each of these species, the value was greater than one, indicating aggregation. Other species, for example the abundant polychaete, Paraprionospio pinnata , apparently were dispersed at random and showed good repeatability among the grab samples. There were no clear cases of ordered dispersion among the species taken, even though the substrate was apparently uniform. BIAS IN BENTHIC INVESTIGATIONS 119 Figure 1. Location where samples were taken, with inset showing the location of Matagorda Bay on the Texas Coast. For some low-density species, single grabs sampled insufficient substrate to allow reliable density estimates. Arthropods, and particularly mollusks, tended to be large-bodied and infrequent, and some polychaetes, though small, were rare enough to appear in just one or several samples. Some of these species produced high CV and Morisita values, but the small sample size preculuded statistical conclusions regarding their dispersion. A mean of 23.5 species occurred per replicate, or 65.3 percent of the 36 species occurring in all replicates, indicating substantial undersampling of species by single grabs. In this case, taking two grabs instead of one would have (on the average) increased the number of species collected by 26 percent. The slope of the cumulative species plot decreased with sampling effort (Fig. 2, lower curve) toward an asymptote somewhere at or beyond eight replicates, at which point essentially all species would be represented. 120 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Table 1. Benthic organisms obtained in replicate Peterson dredge samples of soft sediment, Matagorda Bay, Texas Organisms 1 2 Sample Replicate 3 4 5 6 7 8 Mean (CV) Morisita Phylum Nemertina Sp. A 19 7 3 5 7 3 6 1 6.38(0.87) 1.53** Sp. B 5 5 3 1 4 3 2 2 3.13(0.47) 0.91 Sp. C 2 9 2 5 2 13 6 5 8.00(0.84) 1.29* Sp. D 3 0 1 0 0 0 0 0 0.88(1.55) 4.00* Sp. E 0 0 1 0 0 0 1 1 0.38(1.38) — Phylum Mollusca Polinices duplicatus 0 0 0 0 1 0 0 1 0.25(1.85) — Nassarius vibex 0 0 0 1 1 2 2 1 0.88(0.95) 0.76 Nuculana concentrica 3 1 1 3 1 0 2 0 1.38(0.86) 1.02 Mulinia lateralis 0 0 0 1 0 0 0 1 0.25(1.85) — Tellina texana 1 0 0 0 2 3 0 1 0.88(1.28) 1.52 Abra aequalis 1 0 0 0 0 1 0 0 0.25(1.85) — Phylum Annelida Leitoscoloplos fragilis 3 0 4 0 1 9 2 0 2.38(1.29) 2.15** Aricidea sp. 4 0 2 1 8 1 3 2 2.63(0.95) 1.49* Cossura delta 12 19 16 8 20 18 20 10 15.38(0.31) 1.03 Paraprionospio pinnata 16 20 19 11 15 14 19 11 15.63(0.23) 0.99 Prionospio cirrifera 1 0 0 1 3 2 0 1 1.25(0.83) 1.14 Streblospio benedicti 0 2 0 0 0 0 0 0 0.25(2.83) — Magelona cf. phyllisae 0 0 0 0 1 0 0 1 0.25(1.85) — Mediomastus ambiseta 0 16 2 19 60 46 53 42 29.75(0.79) 1.51** Maldane sp. 1 0 0 0 0 0 0 0 0.13(2.83) — Asychis elongatus 6 1 2 2 1 1 3 4 2.63(0.67) 1.10 Pokarkeopsis levifuscina 1 1 5 0 5 4 3 4 2.88(0.68) 1.11 Sigambra sp. 2 2 4 0 2 1 4 1 2.00(0.71) 1.00 Glycinde solitaria 9 7 3 2 5 6 10 8 6.25(0.45) 1.04 Pseudeurythoe ambigua 1 7 8 4 11 11 17 5 8.00(0.62) 1.23** Euphrosine sp. 4 2 2 1 14 0 6 3 4.00(1.11) 1.89** Gattyana cirrosa 0 3 0 1 2 1 0 3 1.28(1.02) 1.24 Diopatra cuprea 0 0 0 1 3 1 0 0 0.63(1.70) 2.40 Onuphis magna 1 0 0 0 0 0 0 0 0.13(2.83) — Lumbrineris tenuis 2 3 4 5 7 4 4 3 4.00(0.38) 0.90 Phylum Sipuncula Phascolion strombi 2 1 1 0 0 3 0 0 0.88(1.29) 1.52 Phylum Arthropoda Eudorella hispida 16 5 13 7 9 9 6 21 10.75(0.51) 1.15** Leptognatha caeca 0 7 11 6 22 4 16 6 9.00(0.79) 1.45** Sp. A 0 0 0 0 0 3 0 0 0.38(2.83) — Ogyrides limicola 0 1 0 0 5 4 3 1 1.75(1.13) 1.67* Phylum Echinodermata Micropholis atra 0 1 8 3 2 3 2 5 3.00(0.84) 1.33 Total Individuals 115 120 115 88 214 170 190 144 144.50(0.30) 1.07** * Significantly different from 1.0, P<0.05. **Significantly different from 1.0, P<0.01. BIAS IN BENTHIC INVESTIGATIONS 121 Number of replicated samples Figure 2. Cumulative number of species captured is plotted as a function of sampling effort (left axis). The cumulative percent of all individuals represented by the species captured in successive grabs also is shown (right axis). The axes are calibrated so the curves begin at the same point. However, the species added with successive replications represented a strongly decreasing numerical proportion of the organisms composing the community (Fig. 2, upper curve). For example, the 29 species taken in just two replicates represented 97.9 percent of all individuals, whereas the seven additional species added in the subsequent six replicates, combined, made up the remaining 2.1 percent of all individuals. By these results, sampling efforts aimed at determining species composition of a community would require more replications than those aimed at determining patterns for numerically dominant members of the community. Benthic sampling design involves a trade-off between effort per replicate and number of replicates. For each sampling site, the greatest information per unit effort occurs with the least replication, but for the information to be meaningful in comparing locations, multiple samples become necessary to overcome system noise. Variability in organism counts among samples in this study was influenced by the propensity for some species to aggregate in spite of apparently homogeneous substrate. For such species, within-site variability in density is high, requiring numerous replications if valid comparisons are to be made with other communities. Several species, however, showed a nearly random dispersion, allowing reliable density estimation with much less work. Finally, grab volumes were insufficient in relation to the low densities of some Infrequent species, for 122 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 which density estimates could not be evaluated. For these species, alternative sampling apparatus should be considered. While there is little reason to assume these results apply exactly to other substrate types, or even to other soft bottom locations, they do emphasize the need for knowledge of species dispersion patterns prior to attempting community comparisons. Acknowledgments I thank James Bowman for help in sampling, and Bob Trebatoski and Tom Calnan for their constructive comments on the work. Literature Cited Aller, R. C, and R. E. Dodge. 1974. Animal-sediment relations in a tropical lagoon. Discovery Bay, Jamaica. J. Mar. Res., 32:209-232. Boesch, D. F., R. J. Diaz, and R. W. Virnstein. 1976. Effects of tropical storm Agnes on soft-bottom macrobenthic communities of the James and York estuaries and the lower Chesapeake Bay. Chesapeake Sci., 17:246-259. Morisita, M. 1959. Measuring of the dispersion of individuals and analysis of the distributional patterns. Mem. Fac. Sci. Kyushu Univ., ser. E (Biol.), 2:215-235. Peterson, C. H. 1977. Competitive organization of the soft bottom communities of southern California lagoons. Mar. Biol., 43:343-359. Rosenberg, R., and P. Moller. 1979. Salinity stratified benthic macrofaunal communities and long-term monitoring along the west coast of Sweden. J. Exper. Mar. Biol. Ecol., 37:175-203. Woodin, S. A. 1974. Polychaete abundance patterns in a marine soft-sediment environment: the importance of biological interactions. Ecol. Monogr., 44:171-187. LAMPROPELTIS SIMILIS FROM THE COFFEE RANCH LOCAL FAUNA (HEMPHILLIAN LAND MAMMAL AGE) OF TEXAS Dennis Parmley Department of Biology, Midwestern State University, Wichita Falls, Texas 76308 Abstract. — Matrix recently collected from the Coffee Ranch local fauna of Texas (Hemphillian Land Mammal Age) has yielded a well-preserved trunk vertebra of Lampropeltis similis. This record is important because current knowledge of the herpetofauna of the Hemphillian Land Mammal Age is deficient. Moreover, this record extends the paleogeographic occurrance of this species approximately 350 kilometers to the south and extends its geologic range approximately three million years into the mid-Hemphillian Land Mammal Age (about 6.6 million years B.P.). Key words : Lampropeltis similis', Toffee Ranch local fauna; Hemphillian Land Mammal Age; Texas. Matrix recently collected from the Coffee R.anch fossil site (Hemphillian Land Mammal Age), Hemphill County, Texas, has yielded a vertebra of the extinct snake Lampropeltis similis. Little is known about the small amphibians and reptiles of the Hemphillian of Texas (or elsewhere in North America), thus the record is of considerable interest. Traditionally, the Hemphillian Land Mammal Age was believed to belong to the middle Pliocene (Wood et al., 1941). More recently, the correlations of Berggren and Van Couvering (1974) indicated that the latest part of the Hemphillian straddles the Miocene-Pliocene boundry. Land mammal age names are used here in lieu of classical geochronologic nomenclature. The Coffee Ranch contains the type local fauna of the Hemphillian Land Mammal Age (Wood et al., 1941; Evernden et al., 1964; Dalquest, 1969), and an age of 6.6 million years B.P., based on Zircon fission tracks, is available for the hard volcanic ash that covers the quarry (Izett, 1975). Dalquest (1983) considered this to be a minimum date and believed the age of the Coffee Ranch sediments and fauna might be greater than 6.6 million years B.P. Further, based on his informal three-part division of the Hemphillian Land Mammal Age, the Coffee Ranch local fauna is considered mid-Hemphillian. The geology and mammalian fauna of the Coffee Ranch local fauna have been described in detail by Dalquest (1969, 1980, 1981, 1983) and are not discussed here. Parmley (1984) reported one salamander, turtle shell fragments ( Geochelone ), and six species of snakes (two boids and four colubrids) from the Coffee Ranch local fauna. Lampropeltis similis , here reported, is new to the paleoherpetofauna of the Coffee Ranch. Species Account and Discussion Lampropeltis similis was described by Holman (1964) on the basis of trunk vertebrae from the Valentine Formation (Barstovian) of Nebraska. The Texas Journal of Science, Vol. 39, No. 2, May, 1987 124 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 A single trunk vertebra of L. similis (Midwestern State University Collection of Fossil Vertebrates 12304; Fig. la-d) was recovered from approximately 400 kilograms of Coffee Ranch matrix. The vertebra is nearly complete and the basis of identification used here are those discussed by Holman (1964) for the holotype vertebra (University of Nebraska 61035). The combination of characters that identify the Coffee Ranch vertebra to species follows. In dorsal view, the vertebra is somewhat square in appearance (cl/naw 1.30 mm). The right accessory process is broken, but the left process is nearly complete. It is short, rounded at the end, and positioned slightly oblique to the long axis of the centrum. The dorsal surface of the neural spine is rounded. The right side of the zygosphene is broken but this structure appears to be crenate in shape. In ventral view, the hemal keel is well developed, narrow, and relatively uniform in width along its entire length. The subcentral ridges are well developed, with deep valleys between them and the hemal keel. In lateral view, the neural spine is longer (2.7 mm) than high (1.2 mm measured from the top of the zygosphene to the top of the neural spine), has a slight posterior overhang, and the anterior edge slopes gently outward. The subcentral ridges arch slightly upwards. The hemal keel expands posteriorly (see Fig. lc) indicating the vertebra was positioned near the cervicals. The condyle is slightly oblique (points dorsally) to the long axis of the centrum. In anterior view, the zygosphene is slightly, but noticeably, convex and relatively thin. The cotyle is oval in shape. The neural canal is dome-shaped and slightly larger than the cotyle. In posterior view, the neural arch is moderately depressed. The condyle is well developed and rounded. The Coffee Ranch vertebra agrees well with the holotype but represents a larger snake. The length through the zygapophyses is 4.7 mm (4.0 for holotype), the width through the zygapophyses is 5.5 mm (4.8 for holotype), and the height from the lower lip of the cotyle through the top of the zygosphene is 3.5 mm (3.1 for holotype). Lampropeltis similis previously has been reported from the Barstovian through the Clarendonian of Saskatchewan (Barstovian), South Dakota (Barstovian), Nebraska (Barstovian), and Kansas (Clarendonian) by Green and Holman (1977), and Holman (1964, 1970, 1973, 1975). The occurrence of this species in the Coffee Ranch local fauna extends its known paleogeographic range approximately 350 kilometers (217 miles) south of the Kansas locality (Fig. 2), and extends its temporal chronologic range approximately three million years into the medial Hemphillian Land Mammal Age (about 6.6 million years B.P.). The phylogenetic relationship of Lampropeltis similis to other species of fossil and Recent Lampropeltis from the United States is speculative at present. Holman (1964) postulated that L. similis may be ancestral to L. intermedius Brattstrom, from the Blancan of Arizona and Mexico, which in turn may be the stem stock for the Recent species L. triangulum, L. FOSSIL LA M PR OPELTIS FROM TEXAS 125 Figure 1. Coffee Ranch Lampropeltis similis vertebra in dorsal (a), ventral (b), lateral (c), anterior (d), and posterior (e) views. Line equals 1 mm. 126 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 2. Fossil (1-6) occurrence of Lampropeltis similis and Recent (boxed) range of L. triangulum. See text for fossil records. zonata , and L. pyromelana. Brattstrom’s (1955) description and illustrations of L. intermedius vertebrae seem somewhat subjective. He stated that L. intermedius differs from Recent L. triangulum in that the centrum ridges, when viewed from below, are narrower towards the condyle in L. triangulum. Numerous Recent L. triangulum skeletons were examined during this study and the centrum ridges appear to be individually variable. I have seen L. triangulum vertebrae that match perfectly Brattstrom’s (1955) diagramatic views and description of L. intermedius. L. similis vertebrae indicate a rather generalized Lampropeltis that appears to be linked to the Recent species L. triangulum, L. zonata, L. pyromelana , and L. mexicana, in that all have small vertebrae with low, relatively long, neural spines and depressed neural arches. FOSSIL LAMPROPELTIS FROM TEXAS 127 How L. getulus and L. calligaster fit into this phylogeny is not clear. Both species have higher neural spines and usually more vaulted neural arches than the above-listed species. L. getulus and L. calligaster , however, sometimes have neural spines, neural arches, and subcentral ridges identical to those in Elaphe obsoleta and E. guttata , making them impossible to distinguish from one another. A close relationship between Lampropeltis and Elaphe has been reported (Bury et al., 1970; Baker et al., 1972; Garstka, 1982). Thus, based on information currently at hand, I suggest the following: 1) L. intermedius probably should be considered indeterminant; 2) a common ancestor gave rise to Elaphe and Lampropeltis , with L. getulus and L. calligaster representing a separate line of Lampropeltis evolution; 3) L. similis is probably the common ancestor to the Recent species L. triangulum, L. zonata, L. pyromelana , and L. mexicana\ 4) L. similis became extinct sometime during mid-to-late-Hemphillian times. While the ecological requirements of Lampropeltis similis are not known, speculations inferred from the known requirements of Recent taxa thought to be modern relatives are suggested here. Based on its large geographic range during Barstovian through mid-Hemphillian times, it must have been a snake with broad ecological tolerances. The only Recent North American Lampropeltis with a north-to-south geographic range approaching that of L. similis is L. triangulum (Fig. 2). It may be that L. similis was similar to L. triangulum and was a nocturnal constrictor that fed chiefly on reptiles and small mammals. Its relatively low, long neural spine is indicative of fossorial or secretive forms (Holman, 1973). L. similis may have been secretive, spending a great part of its time beneath rocks or logs, or burrowed into soft soils. Acknowledgments W. W. Dalquest, R. Moline, J. Cargill, and J. Parmley helped collect matrix from the Coffee Ranch. Walter W. Dalquest provided storage and washing facilities for the matrix and reviewed an early version of this manuscript. Mr. Walter Coffee of Miami, Texas, permitted access to the fossil quarry. J. Alan Holman examined the fossil for me. Drawings of the vertebra are by Mrs. Corina Zalace. Literature Cited Baker, R. J., G. A. Mengden, and J. J. Bull. 1972. Karyotypic studies of thirty-eight species of North American snakes. Copeia, 1972:257-265. Berggren, W. C., and J. Van Couvering. 1974. The late Neogene: biostratigraphy, geochronology and paleoclimatology of the last 15 million years in marine and continental sequences. Paleogeogr. Paleoclimatol. Paleoecol., 16:1-216. Brattstrom, B. H. 1955. Pliocene and Pleistocene amphibians and reptiles from southeastern Arizona. J. Paleont., 29:50-54. Bury, R. B., F. Gress, and G. C. Gorman. 1970. Karyotypic survey of some colubrid snakes from western North America. Herpetologica, 26:461-466. Dalquest, W. W. 1969. Pliocene carnivores of the Coffee Ranch. Bull. Texas Mem. Mus., 15:1-44. 128 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 - . 1980. Camelidae from the Coffee Ranch local fauna (Hemphillian Age) of Texas. J. Paleont., 54:109-117. - . 1981. Hesperohipparion (Mammalia: Equidae), a new genus of horse from the Hemphillian of North America, with description of a new species. Southwestern Nat., 25:505-512. - . 1983. Mammals of the Coffee Ranch local fauna Hemphillian of Texas. Occas. Publ. Texas Mem. Mus., 38:1-41. Evernden, J. F., D. E. Savage, G. H. Curtis, and G. T. James. 1964. Potassium-argon dates and the Cenozoic mammalian chronology of North America. Amer. J. Sci., 262:145-198. Garstka, W. R. 1982. Systematics of the mexicana species group of the colubrid genus Lampropeltis, with an hypothesis mimicry. Mus. Comp. Zool., Breviora, 466:1-35. Green, M., and J. A. Holman. 1977. A late Tertiary stream channel fauna from South Bijou Hill, South Dakota. J. Paleont., 51:543-547. Holman, J. A. 1964. Fossil snakes from the Valentine Formation of Nebraska. Copeia, 1964:631-637. - . 1970. Herpetofauna of the Wood Mountain Formation (upper Miocene) of Saskatchewan. Canadian J. Earth Sci., 7:1317-1325. 1973. Herpetofauna of the Mission local fauna (lower Pliocene) of South Dakota. J. Paleont., 47:462-464. - . 1975. Herpetofauna of the WaKeeney local fauna (lower Pliocene:Clarendonian) of Trego County, Kansas. Univ. Michigan Papers Paleont., 12:49-66. Izett, G. A. 1975. Late Cenozoic sedimentation and deformation in northern Colorado and adjoining areas. Mem. Geol. Soc. Amer., 114:179-209. Parmley, D. 1984. Herpetofauna of the Coffee Ranch local fauna (Hemphillian Land Mammal Age) of Texas. Pp. 97-106, in Festschrift for Walter W. Dalquest, (N. V. Horner, ed), Midwestern St. Univ, xx+163 pp. Wood, H. E. (chairman), et al. 1941. Nomenclature and correlation of the North American Continental Tertiary. Bull. Geol. Soc. Amer., 52:1-48. AN ASSESSMENT OF GEOGRAPHIC AND SEASONAL BIASES IN SYSTEMATIC MAMMAL COLLECTIONS FROM TWO TEXAS UNIVERSITIES Frederick B. Stangl, Jr., and Elizabeth M. Jones Department of Biology, Midwestern State University, Wichita Falls, Texas 76308, and The Museum, Texas Tech University, Lubbock, Texas 79409 Abstract. — Geographic and seasonal biases are inherent to regional mammal collections, and are generally reflective of faculty research interests and the academic calendar of institutions housing collections. We assessed the extent of seasonal and geographic biases for 1 1 species of mammals, represented by museum study skins, in the collections of two Texas universities — Texas Tech University and Midwestern State University. Causes and possible solutions of such biases are discussed so that curators and collection managers may improve geographic and seasonal representation of their collections. Key words : systematic mammal collections; collection biases; collection management; specimen collections. Systematic collections of mammals may include comprehensive accumulations of local species. Proximity of an institution to favored habitats of locally occurring taxa affords an opportunity to acquire series of specimens that characterize seasonally variable features of a given species. Many such features also vary geographically, and a regional collection thus becomes an important component of a network of similar collections for such studies as geographic variation, reproductive activity, molt, and growth and development. Regional mammal collections affiliated with colleges or universities usually reflect past and present faculty and graduate student research activities (voucher specimens), student collections, financial support for collecting and curatorial efforts, and the academic calendar of the institutions. These factors insure the individual nature of different collections. The collections of Texas Tech University (TTU) and Midwestern State University (MWSU) are illustrative of this. A brief historical perspective of the two mammalogy programs and resulting collections is useful in understanding collection composition and collecting patterns of Texas mammals documented in this study. The TTU mammalogy program was established by the late Robert L. Packard in 1962, and currently numbers seven professional mammalogists and their more than a dozen M.S. and Ph.D. students. The collection currently contains about 45,000 specimens and is rapidly expanding. Classes involving field collecting currently are offered in the autumn, and student collections consist mostly of animals taken on class field trips away from the immediate Lubbock vicinity and often out-of-state. An increasing emphasis on extra-regional representation is mostly reflective of current faculty and graduate student research interests. The Texas Journal of Science, Vol. 39, No. 2, May, 1987 130 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Although the MWSU collection has some important holdings from outside Texas, it is primarily regional in coverage and currently contains about 15,000 specimens. The mammalogy program was initiated in 1952 by Walter W. Dalquest, who currently is joined by the senior author. The graduate program is small, and M.S. students with interest in mammalogy are occassional. Classes involving field work traditionally have been offered in the spring. Student collecting efforts frequently are directed locally. Faculty and graduate student research interests mostly are restricted to Texas mammals. Many specimens also have been collected incidental to paleontological surveys throughout the state by Dalquest and his students. Although biases in systematic mammal collections may be expected, we have found that seasonal and regional misrepresentation may be unexpectedly great in some instances. We realize that total alleviation of a documented imbalance may be difficult, but attempts to minimize such biases are necessary if the usefulness of regional collections is to be enhanced. Methods and Materials Eleven species of Texas mammals, represented minimally by museum study skins, were recorded from the collections of Texas Tech University and Midwestern State University in March 1986. Specimens were categorized by month of capture (Fig. 1), and further subdivided into groups of those taken in the immediate vicinity of the respective institutions (herein defined as resident and immediately adjacent Texas counties) and those taken elsewhere in the state (Fig. 2). The TTU vicinity included Crosby, Floyd, Garza, Hale, Hockely, Lamb, Lubbock, Lynn, and Terry counties. Because MWSU is located in a county bordering Oklahoma, only the Texas counties of Archer, Baylor, Clay, Wichita, and Wilbarger were included. Selection of Surveyed Species We initially selected 1 1 taxa that were intuitively thought to be represented in each of the two collections by sufficiently meaningful sample sizes. The diversity of taxa was intended to reveal any subsequent biases due to variance of such factors as abundance, distribution, life history strategies, ease and means of collection and preparation, and storage space requirements. Taxa also were chosen to represent mammals with known ranges that include all or part of the defined vicinity of each school ( Sylvilagus floridanus, Spermophilus tridecemlineatus, Dipodomys ordii, Peromyscus leucopus, Sigmodon hispidus), and others that occur well beyond the vicinity of each ( Spermophilus variegatus, D. spectabilis, P. pectoralis, and Sigmodon ochrognathus). Myotis velifer and Bassariscus astutus were included with the latter taxa because so few locally-taken specimens are available, although each is found in suitable habitats near each institution. Sources for distributional data were Dalquest and Horner (1984), Davis (1974), and Schmidly (1977, 1984). Selection of Characters Mammal specimens may be stored in a collection in various forms (for example, study skins, skulls, skeletons, fluid-preserved bodies, or in some combination of the above). The type of preparation may in part be biased to reflect the kinds of data desired by the collector or preparator. Therefore, even a numerically well-represented taxon may have limited use for some studies. For purposes of this study, we selected specimens prepared as museum BIASES IN SYSTEMATIC MAMMAL COLLECTIONS 131 study skins because this preservation technique is standardized among research collections and was therefore, expected to insure an adequate sample size. Results and Discussion A distinct seasonal bias was noted in both collections surveyed (Fig. 1). For each of the 11 species examined, there was a preponderance of spring-taken animals. The summer months were most poorly represented. The two collections are to some extent complimentary. Individual assessment of the TTU and MWSU collections (Fig. 2) revealed contrasts in geographic biases for the five locally occurring taxa. Geographically, the MWSU collection is more heavily biased towards regional representation. Seasonally, the spring bias is particularly pronounced for MWSU specimens (Figs. 1 and 2). Seasonal Biases Regional mammal collections often are affiliated with colleges or universities. As a result, seasonal representation of taxa in these collections usually is reflected in the research interests of past and present workers and the academic calendar. Biological or climatic phenomena often are related to unresolvable seasonal biases. Unfavorable environmental situations may elicit such responses as hibernation, aestivation, torpor, or migration, which directly affect the availability (even occurrence) of some mammals. Furthermore, seasonally inclement weather can preclude accessability to a region or simply discourage collecting efforts. Seasonal vegetation may lessen the attractiveness of trap bait or provide increased cover for some species, whereas others may be naturally cyclic in abundance. Geographic Biases Texas is a large and ecologically diverse state. It would be unreasonable to expect equal representation of each region in any collection of mammals. However, we would expect there would be a greater emphasis for locally collected animals than those taken from any other region in the state. With exceptions, we found this to be true (Fig. 2). From resident and adjacent counties of each institution, biases in representation also exist. MWSU is more heavily biased towards Wichita County, with lesser (but relatively balanced) collections from adjoining counties. Representation of local counties is more sporadic in the TTU collection. For the five locally occurring species examined, some counties are scarely represented, or not at all. Geographic biases due to small sample sizes may result for a variety of reasons other than lack of collecting efforts. Species may be scarce or rare in some areas (for example, eastern Piney Woods, or South Plains 132 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 1. Histogram reflecting seasonal representation (percentages, by month) of 1 1 species of Texas mammals from the collections of Texas Tech University and Midwestern State University. Open areas of bars indicate TTU specimens; dark areas indicate MWSU specimens. BIASES IN SYSTEMATIC MAMMAL COLLECTIONS 133 agricultural counties) or throughout their range. Others may be more common than records indicate, but are not readily collected by usual means. Because class field trips should be rewarding experiences for students, they often are planned with scenery and wildlife diversity in mind. For this reason, combined with past and present research interests, the Trans- 134 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Sylvilaqus f lor idanus Spermophi lus t r i deceml i neat us Dipodomys ordi i Figure 2. Histogram reflecting seasonal representation (percentages, by month) of five species of Texas mammals occurring in the vicinity of both Texas Tech University and Midwestern State University. Open areas of bars indicate specimens collected outside the defined vicinity of each school; dark areas indicate specimens taken in the vicinity of each school. BIASES IN SYSTEMATIC MAMMAL COLLECTIONS 135 Peromyscus leucopus Sigmodon hispidus Pecos is, aside from the local areas, the best represented region in both collections. Addressing Collection Biases For reasons discussed above, some biases are difficult and perhaps even impractical to alleviate. However, once recognized, others may be successfully addressed by the curator or collection manager. The first step toward correcting, or at least minimizing, any collection bias is the identification of the gap in representation. Our methods were tedious and time-consuming, and dealt only with representation by museum study skins of 1 1 species. There is, however, a trend toward computerization of systematic mammal collections. The ease with which information retrieval is accomplished makes possible identification of any number of existing biases, thus proving an important curatorial tool in this respect. 136 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Existing biases may be approached on a priority basis. Taxonomic gaps probably receive more attention than other biases; usually they are considered priority issues that can be remedied by specific collecting efforts or by exchanging specimens with other institutions. For seasonal and geographic biases, especially for locally occurring taxa, additional field collecting is the obvious solution. The following suggestions may be considered as possible solutions to better representation in collections. If classes involving field collecting traditionally have been offered during the same semester, rescheduling of those classes may be possible. Planned field trips also may be scheduled at different times of the semester. Summer collecting with snap traps sometimes is undesirable, due to specimen spoilage and insect damage, and may, along with the academic calendar, contribute to the scarcity of summer-taken specimens noted in this study. The use of live traps will alleviate some of the problems. It may be desirable to maintain some level of collecting locally the year around, if only for an occassional weekend. Emphasis could be placed on areas identified as geographically or seasonally underrepresented. Collectors should assess their own prejudices. These are most likely expressed when more specimens have been taken than can be prepared. Juveniles often are discarded under such circumstances. Species that are large and require considerable storage space or are difficult to prepare (for example, lagomorphs and carnivores) are more likely to be represented by “skulls only.” Others that are known to be abundant and widespread may be discarded (the “trash rat” designation). Although such taxa may be numerically well represented in a given collection, additional specimens may prove valuable if taken from certain localities or at certain times of the year. Our study was a limited foray into determining the extent of seasonal and geographic biases of museum study skins in two systematic collections. While the existence of such biases may be anticipated, we feel the magnitude of some biases will surprise investigators who similarly evaluate their own collections. Acknowledgments For critical comments and useful discussion, we thank George D. Baumgardner, Walter W. Dalquest, Norman V. Horner, J. Knox Jones, Jr., Robert D. Owen, and Stephen L. Williams. Literature Cited Dalquest, W. W., and N. V. Horner. 1984. Mammals of north-central Texas. Midwestern State Univ. Press, 261 pp. Davis, W. B. 1974. The mammals of Texas. Bull. Texas Parks and Wildlife Dept., 4:1- 294. BIASES IN SYSTEMATIC MAMMAL COLLECTIONS 137 Schmidly, D. J. 1977. Mammals of Trans-Pecos Texas. Texas A&M Univ. Press, xii + 225 pp. . 1984. The furbearers of Texas. Bull. Texas Parks and Wildlife Dept., IILvii + 55 pp. Current address of Jones: Department of Mammals, Transvaal Museum, P.O. Box 413, Pretoria, South Africa 0001. A COMPARISON OF THE WATER CHEMISTRY AND BENTHIC MACROINVERTEBRATE COMMUNITIES OF TWO OXBOW LAKES IN THE RED RIVER BASIN, NORTHWESTERN LOUISIANA Jack D. McCullough and Clarence W. Reed Department of Biology, Stephen F. Austin State University, Nacogdoches, Texas 95962 Abstract.— The effects of periodic flooding by the Red River on two oxbow lakes were studied for one year. Wilson Lake was isolated from the river by a levee and was not flooded, whereas Old River Lake was connected to the river by a small channel. The two lakes were drastically different in both physicochemical conditions and benthic community structure. Wilson Lake was highly eutrophic, characterized by high phytoplankton production, BOD, ammonium nitrogen, orthophosphate, and color. High benthic standing crop and productivity, and low species diversity further reflect eutrophic conditions in Wilson Lake. Both lakes receive runoff from agricultural land. Occasional flooding of Old River Lake by the Red River appears to play an important role in the ecology of this oxbow lake. Key words : oxbow lakes; benthic macroinvertebrates; secondary productivity; water chemistry; species diversity; eutrophication. Kalkomey (1979) pointed out that oxbow lakes are quite common in the Gulf Coast states, yet studies on those environments are relatively rare. Moore (1963) reviewed the literature on oxbow lakes in the central Gulf states and revealed that these typically are warm-water monomictic lakes (Moore, 1950; Geagan and Fuss, 1959) and typically support a high standing crop of plankton, benthos, and fish (Geagan and Allen, 1961; Lambou, 1960). Buckley and Sublette (1964) investigated the dipteran fauna of Cane River Lake, a lateral lake in Louisiana. Harrel (1973), Marsh et al. (1978), Reece (1979), and Kalkomey (1979) have conducted studies in oxbow lakes in eastern Texas. Bingham (1969) found a high build up of insecticide in a Mississippi oxbow completely separated from the river, but such high levels were not found in a lake occasionally flooded by the river. The purpose of the present study was to investigate the importance of periodic flooding by the parent river on the water quality and the benthic community in oxbow lakes. Study Area The two oxbow lakes chosen for this study are cutoff meanders of the Red River in northwestern Louisiana (Fig. 1) and are relatively shallow, with depths only occasionally exceeding three meters. One of the lakes is connected to the parent river, whereas the other is isolated from river flooding by a levee. Old River Lake is located in Natchitoches Parish, approximately 13 kilometers northwest of Natchitoches, Louisiana. The lake has a surface area of 72 hectares and is connected to the Red River by a narrow channel at the southern arm. Dominant land use around Old River Lake is agricultural, mainly cotton, although soybeans and other crops are produced The Texas Journal of Science, Vol. 39, No. 2, May, 1987 140 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 1. Old River Lake, Natchitoches Parish, and Wilson Lake, Red River Parish, in northwestern Louisiana, and location of collecting sites. in the region. A majority of the shore line has a narrow band of trees, which supplies a considerable amount of detritus. Wilson Lake is in Red River Parish, approximately 23 kilometers east of Mansfield, Louisiana. This lake has a surface area of 45 hectares and is completely separated from the river by a levee. Land surrounding the lake also is used for cotton and soybean production and for cattle pasture. LOUISIANA OXBOW LAKES 141 Each lake was sampled at four locations — one deep station and one shallow station near the end of each arm of the two lakes. Deep stations were given the letter designation A and shallow stations were designated B. Methods and Materials Physicochemical Methods Water for physicochemical analysis was collected monthly for one year from the surface at all eight sites and from one meter above the bottom at the deep stations. Samples were collected with a Kemmerer Water Sampler. Dissolved oxygen and conductivity profiles were done at each site using a Yellow Springs Oxygen Meter, model 54, and a Yellow Springs Conductivity Meter, model 33. Carbon dioxide and alkalinity were determined titrimetrically (APHA, 1980) in the field. Depth and Secchi disc transparancy also were determined at each station. Laboratory analyses were performed following procedures reported in APHA (1980). Colorimetric determinations were done using a Bausch and Lomb spectrophotometer, model 70. Phytoplankton chlorophyll a concentrations were determined on surface samples using a Turner flourometer, model 1 10. Benthic Macroinvertebrate Methods Benthic macroinvertebrates were collected monthly for one year at each collecting site using an Ekman dredge (231 square centimeters). At each site, five grabs were collected and pooled for analysis. Samples were washed in a benthic bucket having a no. 30 (0.59 mm) mesh screen bottom, then preserved in FAA (formalin-aceto-alcohol). In the laboratory, organisms were identified using keys by Edmunds et al. (1976), Edmondson (1959), Sublette (1964), Klemm (1972), Brown (1972), Mason (1973), and Pennak (1978). The drained wet weight and dry weight of each organism (except entoprocts and ectoprocts) were determined using a Mettler analytical balance, model H10. Those weights were used to estimate secondary productivity by the size frequency method (Menzie 1980): j - - 17 Pb = i X (Hj -nj+1) • (Wj • WJ+I)/2 If1 where Pb is the biomass production per square meter per unit of time; i represents the number of size groups; nj is the mean number of organisms in size group j; and Wj represents the mean weight of the size group j. Species diversity was calculated for each sample using the Shannon equation (Shannon and Weaver, 1963): jl, [ ] [3 where d is the species diversity; n* is the number of individuals of the ith taxon; and n represents the number of individuals of all taxa in the sample. Statistical methods The multivariate Hotellings T2 and univariate t statistics (Dixon and Brown, 1979) were used to compare the physicochemical and benthic data between the two oxbow lakes. Pearson’s correlation (Nie et al., 1975) was used to identify significant statistical correlations between chemical and biological parameters. Computations were done using a Honeywell CP6 mainframe computer. 142 THE TEXAS JOURNAL OF SCIENCE — VOL. 39, NO. 2, 1987 Table 1. Annual means and ranges of physiochemical data on Wilson and Old River lakes, Louisiana. Wilson Lake Old River Lake Mean Max Min Mean Max Min pH 8.7 9.6 7.0 8.2 8.9 7.2 co2 3 15 0 2 5 0 HCOs 77 115 9 108 114 63 co3 16 76 0 6 28 0 Turbidity (NTU) 9 43 1 5 17 1 Calcium 7 15 2 13 26 10 Sodium 14 17 11 16 19 14 Chloride 20 28 14 23 28 20 Iron 0.31 0.80 0.02 0.13 0.94 0.02 Sulfate 9 17 1 20 53 14 Ammonium-Nitrogen 1.07 2.12 0.58 0.59 1.54 0.16 Nitrate-Nitrogen 0.049 0.373 0.002 0.043 0.148 0.002 Total Kjeldahl nitrogen 4.86 10.81 1.4 3.21 7.24 0.74 Orthophosphate 0.311 1.02 0.07 0.194 0.58 0.04 Chlorophyll a (ug/L) 50 112 15 16 47 4 BOD 5.5 10.9 0.9 2.5 5.3 0.2 True Color (cu) 9 27 3 3 5 1 Apparent Color (cu) 23 98 8 11 24 2 Depth (meters) 2.6 4.2 2.3 2.1 3.4 2.0 Dissolved Oxygen 7.6 18.6 0.4 8.8 13.4 0.7 Conductivity (umho/cm) 237 329 190 278 338 210 All values are expressed as milligrams per liter except as indicated; pH units (1-14). Results Physicochemical Means and extremes values for physicochemical variables are given in Table 1. During the summer months, dissolved oxygen values in both lakes fell to less than one milligram per liter, but more often in Wilson Lake. Secchi disc transparancy was lower, whereas turbidity and apparent color values were higher in Wilson Lake. Ammonium nitrogen and orthophosphate were higher in Wilson Lake, but nitrate nitrogen concentrations were low in both oxbows. According to Carlson’s Trophic State Index (Carlson, 1977), Wilson Lake was eutrophic and Old River Lake was mesotrophic. Using the index based on phytoplankton chlorophyll a values, Old River Lake had an index of 57 and Wilson Lake had a value of 69. Benthic Community A checklist of benthic organisms found in the lakes is given in Table 2. Average species diversity at Old River Lake was 2.14 and ranged from an average of 2.87 at IB to 1.54 at 2A. Temporal variations in species diversity for each Old River Station are illustrated in Figure 2. The mean numerical density was 3723 organisms per square meter (org/m2) ranging LOUISIANA OXBOW LAKES 143 Table 2. List of benthic macroinvertebrates collected from Old River Lake (O) and Wilson Lake (W), Louisiana. Entoprocta Urnatella gracilis O Ecotprocta Fredericella sultana O Potsiella erecta O Plumatella repens o Oligochaeta Tub if ex o,w Limnodrilus o,w Lumbriculus w Hirudinea Placobdella o,w Helobdella stagnalis w Helobdella sp. o,w Hydracarina o,w Amphipoda Hyalella azteca 0 Pelecypoda Sphaerium o,w Gastropoda o,w Ephemeroptera Caenis o,w Hexagenia o Odonata Perithemis o,w Gomphus o,w Megaloptera Sialis o,w Trichoptera Setodes 0,W Orthotrichia O Coleoptera Haliplus O Dineutus O Berosus o Dubiraphia 0 Diptera Heleidae o,w Culicidae Chaoborus o,w Chironomidae Chironomus o,w Cryptochironomus sp. o,w C. edwardsi o,w Dicrotendipes o,w Einfeldia 0 Xenochironomus w Polypedilum o,w Lauterborniella o,w Microspectra o,w Coelotanypus o,w Tanypus o,w Procladius o,w Ablabesmyia o from 894 in March at Station 1A to 11655 in September at 1A. Average biomass was 2.92 grams wet weight per square meter (wet wt/m2). Mean annual secondary productivity was 3.86 grams dry weight per square meter per year (dry wt/m2/yr). The productivity of each station was 1A, 5.61; IB, 1.68; 2A, 3.73; 2B, 4.41. The culicid fly Chaoborus (Diptera) made up 39.7 percent of the benthic fauna at Old River Lake with an average of 1480 org/m2, and 25.3 percent of the biomass or a mean biomass of 0.74 grams wet wt/m2. Second to Chaoborus in numerical density were the chironomids, representing 33.7 percent of the numbers and 28.3 percent of the biomass. Tanypus (41.1 percent), Procladius (19.7), Coelotanypus (9.9), Chironomus (9.9) and Cryptochironomus (5.8 percent) were the most common genera collected. Ceratopogonid larvae comprised 3.9 percent of the benthic fauna and 3.2 percent of the biomass with an average of 145 org/m2 and 0.08 grams wet wt/m2. 144 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 2. Monthly species diversity values for each collecting site on Old River Lake, Natchitoches Parish, Louisiana. Members of the insect orders Ephemeroptera and Trichoptera accounted for only 2.7 and 1.2 percent of the total number of organisms collected at Old River Lake and 4.1 and 0.8 percent of the total biomass. Caenis was the most commonly collected mayfly, with an occasional specimen of Hexagenia. The caddis fly, Setodes , commonly was collected with occasional collections of Orthotricha. Hyalella azteca was the most commonly collected amphipod. Oligochaetes represented 15.1 percent of the organisms collected from Old River Lake and 31 percent of their biomass. Average numerical density was 560 org/m2 and ranged from 3216 org/m2 at station 1 A in October to 43 at 2 A in May. Limnodrilus accounted for 96.8 percent of the oligochaetes and Tubifex were collected only in small numbers. Leeches were collected only at the shallow sites and represented approximately 1.5 percent of the population, with Helobdella and Placobdella being the most common genera. The Odonata were represented in low numbers by the dragonfly naiads, Perithemis and Gomphus. The megalopteran alderfly, Sialis , was collected in low numbers on six occasions at IB, and the elmid beetle, Dubiraphia (larvae), commonly was collected at the shallow stations only. Urnatella gracilis , the fresh-water entoproct, was collected commonly at station IB in numbers ranging from 26 colonies per square meter to 447 colonies per square meter. The ectoprocts, Fredericella sultana, LOUISIANA OXBOW LAKES 145 Pottsiella erecta, and Plumatella repens , were collected only at stations 1A and IB. Mean species diversity at Wilson Lake was 1.72 ranging from an annual mean of 2.07 at station IB to 1.35 at station 2B. Figure 3 illustrates the temporal variations in benthic diversity at Wilson Lake. Numerical density of the macroinvertebrates at Wilson Lake averaged 3815 org/m2 ranging from 440 at 2B in May to 13,072 org/m2 at 2B in October. Total biomass averaged 6.76 grams wet wt/ m2 with a minimum of 0.61 at 2B in September and a maximum of 18.35 at station 1A in January. The annual mean productivity was 9.56 grams dry wt/m2/yr. The maximum rate was 14.76 at station IB and the lowest 5.04 at station 2A. Chironomid larvae were the dominant taxonomic group representing 37.1 percent of the total number and 53.7 percent of the biomass. Mean numerical density was 1416 org/m2 and mean biomass was 3.63 grams wet wt/m2. Chironomus accounted 38.6 percent of the chironomids collected at Wilson Lake, followed by Tanypus (25.1), Procladius (16.3) and Coelotanypus (14.8). Chaoborus larvae made up 27.5 percent (1050 org/m2) of the total numbers but only 6.6 percent (0.45 grams wet wt/ m2) of the total biomass. Ceratopogonid larvae comprised 6.1 percent of the total numbers and 3.3 percent of the biomass at Wilson Lake. Oligochates represented 25.9 percent of the benthic fauna of Wilson Lake with an annual mean of 987 org/m2 and 26.8 percent of the biomass (1.81 grams wet wt/m2). Limnodrilus accounted for 91.4 percent of the oligochetes collected, Tubifex accounted for 8.4 percent and Lumbriculus was present only in low numbers. Leeches were found most commonly at Wilson Lake station IB where they averaged 13.9 org/m2. Helobdella stagnalis was most common followed by Helobdella sp. and Placobdella. The pelecypod, Sphaerium (Sphaeridae), commonly was collected with an average population of 66 org/m2 representing 4.0 percent of the total biomass. Water mites were frequently collected and the average density was 79 org/m2. Caenis (Ephemeroptera), Perithemis (Odonata), Sialis (Megaloptera), and Setodes (Trichoptera) were collected on only a few occasions and only at station IB. Discussion Old River and Wilson lakes were different in both physicochemical nature and benthic community structure. Wilson Lake was a highly eutrophic lake characterized by high phytoplankton production, BOD, and color. High benthic standing crop, productivity, and low species diversity further reflect this highly eutrophic status. Runoff from agricultural land is thought to play a significant role in the eutrophication of Wilson Lake. Old River Lake was considerably less productive and classed as mesotrophic. Benthic 146 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 3. Monthly species diversity values for each collecting site on Wilson Lake, Red River Parish, Louisiana. productivity was less than half, and average phytoplankton chlorophyll a concentration less than one third that of Wilson Lake. Benthic standing crop biomass at Old River Lake was significantly lower, whereas species diversity was significantly higher. Old River Lake also receives runoff from the cotton fields along the shoreline, but the Red River appears to play an important role in the ecology of this oxbow through occasional flushing of the lake by way of a narrow channel in the southern arm. When the river does not rise enough to flush the lake, the channel provides an outlet for excess runoff thus reducing the possible build up of nutrients, as in Wilson Lake. Physicochemical Comparison Between Lakes The means of all physicochemical parameters except temperature, and depth were simultaneously tested for equality between Old River and Wilson lakes using the multivariate Hotellings T2 analysis. This statistic was found to be significant with P< 0.001. To determine which parameters contributed to this significance, two-sample t tests were applied to each parameter (Table 3). Using an alpha level of 0.05, 19 of the 22 means for chemical parameters were found significantly different between the two lakes. Several parameters found in significantly higher concentrations at Old River Lake may provide evidence of the Red River’s influence on that oxbow. Conductivity, calcium, sodium, sulfate, and chloride all were found LOUISIANA OXBOW LAKES 147 Table 3. Hotellings T2 and two-sample t tests for selected chemical parameters of Old River and Wilson lakes, Louisiana. Hotelling T2 F value 1639.6 62.9 P<.001 Parameter t P Parameter t P Secchi 9.60 <.001 PO4 -4.85 <.001 oxygen 1.80 .075* Chlorophyll a -12.04 <.001 pH -2.46 .015 Chloride 5.33 <.001 Conductivity 5.56 <.001 Turbidity -3.18 .002 Calcium 6.48 <.001 BOD -8.88 <.001 Sodium 7.26 <.001 HCO3 Aik. 3.77 <.001 Sulfate 10.42 <.001 CO2 -0.33 .740 nh4-n -9.63 <.001 CO2 Aik. -3.63 <.001 no3-n -0.71 .481 True color -9.44 <.001 N02-N -1.71 .244 App. color -6.40 <.001 TKN -5.57 <.001 Iron -2.36 .020 P values correspond to two-tailed tests of significance. * one-tailed test is significant, P =.0375 to be significantly higher at Old River Lake. These parameters were reported in high concentrations in the Red River near Powhatan, Louisiana, by USGS (1980) (Table 4). Secchi disc transparancy, color, turbidity, BOD, phytoplankton chlorophyll a , ammonium, and orthophosphate concentrations were all significantly higher at Wilson Lake, reflecting eutrophic conditions. Neither nitrate nor nitrite nitrogen was significantly different between the two oxbows. It is hypothesized, however, that the input of these nutrients into Wilson Lake was greater, but the higher input is rapidly utilized by the large algal population there. The accumulation of nutrients from agricultural runoff contributed to the higher ammonium and phosphate levels at Wilson Lake. Omernik (1976) found relatively high nutrient levels in streams draining agricultural land. Jones et al. (1976) found livestock as an identifiable source of phosphorus and ammonium-nitrogen within watersheds. Table 4. Chemical data from the Red River near Powhatan, Louisiana, (USGS, 1980). Time-weighted Parameter Maximum Minimum average Conductivity (umhos/cm) 1500 272 757 Calcium (mg/ 1) 109 27 51 Sodium (mg/ 1) 192 19 78 Sulfate (mg / 1) 214 16 92 Chloride (mg/ 1) 290 28 123 148 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 WILSON LAKE milligrams/liter 0 2 milligrams/liter 0 2 Figure 4. Histograms of monthly oxygen values at Old River Lake and Wilson Lake, Louisiana. Although the annual mean for oxygen was slightly higher at Old River Lake, that difference was not significant. However, an inspection of a histogram (Fig. 4) of the distribution of the oxygen values recorded during the year at each lake reflects the more eutrophic conditions at Wilson Lake. There were several more oxygen values both in the low range and in the high range at Wilson Lake. Larger phytoplankton populations and respiration rates at Wilson Lake accounted for those differences. LOUISIANA OXBOW LAKES 149 Table 5. Hotellings T2 and two-sample t tests for selected components of the benthic macroinvertebrate communities of Old River Lake and Wilson Lake, Louisiana. Hotelling T2 F value 128.5 7.9 P<.001 Parameter t P Parameter t P Diversity 2.81 .006 Ephemeroptera Redundancy -1.96 .053 numbers 2.38 .021 Oligochaeta biomass 2.16 .036 numbers -2.85 .006 Trichoptera biomass -3.17 .002 numbers 3.03 .004 Chaoborus biomass 3.49 .001 numbers 1.10 .275 All benthos biomass 1.48 .142 numbers -0.20 .839 Chironomidae biomass -4.73 <.001 numbers -0.52 .602 biomass -4.13 <.001 P values correspond to two-tailed tests of significance Benthic Macroinvertebrates Differences in benthic macroinvertebrate community structure between the two oxbow lakes also were tested using Hotellings T2. The statistic was significant with P < 0.001. Two sample t tests on each parameter revealed that species diversity, ephemeropteran and trichopteran numbers, and biomass were significantly higher at Old River Lake, whereas oligochaete numbers and biomass, chironomid biomass, and total benthic biomass were significantly higher at Wilson Lake (Table 5). Species diversity was higher at Old River Lake indicating generally a less stressed environment than at Wilson Lake (Fig. 5). The low oxygen values often recorded for Wilson Lake were most likely a major contributing factor. However, species diversity index ranges reported by Wilhm and Dorris (1968) indicate that both oxbow lakes are moderately polluted. The greater abundance of tubificid oligochaetes at Wilson Lake is another indication of highly stressed conditions as members of the family Tubificidae are often found in large numbers in polluted areas (Goodnight, 1973). A comparison of the chironomid populations at the two lakes reflects the differences in trophic conditions. Chironomus , a detritivore, and a genus associated with eutrophic lakes (Brinkhurst, 1974), dominated the chironomid populations at Wilson Lake. Tanypus , a hervivorus chironomid, and Coelotanypus, a predator, were the two most common chironomids in Old River Lake. Heuschele (1969) also found Tanypus stellatus to be the dominant benthic organism in a Minnesota floodplain lake. Numerically, chironomids were the dominant benthic organism in Wilson Lake, but the higher numbers were not statistically significant. 150 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 5. Average biomass and species diversity of the benthic macroinvertebrate community at collecting sites on Old River and Wilson lakes, Louisiana. Large larvae of Chironomus accounted for the statistically higher chironomid biomass at Wilson Lake. Both lakes had large populations of Chaoborus punctipennis, which was the numerically dominant benthic organism collected at Old River Lake. Again, numbers were not significantly higher. In both lakes, greater numbers of Chaoborus were found at the deeper sites at littoral stations. Chaoborus migrate vertically to deeper water during the day and return to the surface to feed (Welch, 1968). This behavioral mechanism probably allows the larva to avoid heavy grazing by planktivorous fish during the day. Lower numbers of Chaoborus in littoral areas in this study may be due to increased grazing by fish. Other reports (Paloumpis and Starrett, LOUISIANA OXBOW LAKES 151 Table 6. Comparison of annual benthic macroinvertebrate productivity (gms dry wt/m2/ yr) reported from several lakes. p Location Source 27.5 Lake Taltowisko, Poland Kajak and Rybak (1966) 21.3 Lake Mikolajskie, Poland " " 12.5 Lake Sniardwy, Poland " " 2.5 Lake Lusine, Poland " " 1.9 Lake Flosek, Poland " " 14.9 Lake Beloie, Russia Borutzky (1939) 8.4 Wyland Lake, Indiana Gerking (1962) 7.2 Linsley Pond, Connecticut Deevey (1942) 6.9 Livingston Reservoir, Texas McCullough and Jackson (1985) 5.7 Afon Hirant, N. Wales Hamilton (1969) 3.1 North Lake, Texas Durret (1973) 0.5 Fairfield Reservoir, Texas Oliphant (1977) 10.0 Wilson Lake, Louisiana This study 4.0 Old River Lake, Louisiana " I960; Harrel, 1973) also indicated that. Chaoborus was predominantly found in deeper water of oxbow lakes during daylight hours. Ephemeropteran and trichopteran numbers and biomass were significantly higher at Old River Lake, reflecting better water quality. Mayflies, caddis flies, and odonates rarely were collected in Wilson Lake and only at one collecting site (lB-a littoral area). Benthic Productivity Comparison of productivity of the benthic macroinvertebrates at Old River and Wilson lakes with several other reported values (Table 6) shows that the productivity of Wilson Lake is relatively high, whereas that of Old River is among the lower rates. Mann (1980) stated that the productivity of the benthic community was proportional to phytoplankton production rates. The phytoplankton standing crop was significantly higher in Wilson Lake, which would support that statement. However, Wilson Lake is thought to have the potential for higher benthic productivity, but is limited by extremely stressful conditions in summer. Northern lakes, without such stress during the growing season, are capable of much higher production provided other factors are not limiting. As Moore (1980) noted, algal availability is a key factor regulating invertebrate production in the temperate zone; however, benthic communities may be unable to take advantage of abundant algal growth in lakes where diversity is low owing to factors other than food supply. At Wilson Lake, where phytoplankton was always abundant, correlations between chlorophyll a and various components of the benthic community were extremely low. At Old River Lake, however, where chlorophyll a values were significantly lower, significant correlations were found between chlorophyll a and chironomid numbers (r =0.54) and biomass (r =0.37). The chlorophyll a correlation 152 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 with Tanypus , a herbaceous chironomid, was 0.52, whereas the correlation with Coelotanypus (a predator) was —0.18. In summary, periodic flooding or flushing of oxbow lakes by the parent river may reduce eutrophication of those lakes, at least along the Red River in the Gulf Coastal plains. The results of this study also suggest a mitigating factor in building impoundments on rivers. Releases from impoundments would flush oxbow lakes with relatively high productivity, reducing fertility and, in summer, those releases may reduce stress caused by low dissolved oxygen. Literature Cited American Public Health Association. 1980. Standard methods for the examination of water and waste water, APHA, New York, 15th ed., 1 193 pp. Bingham, R. 1969. Comparative study of two oxbow lakes. Completion report F19-R, Mississippi Game Fish Comm., Jackson, 216 pp. Borutsky, E. V. 1939. Dynamics of the total benthic biomass in the profundal of Lake Beloie, U.S.S.R. [Trans, by M. Ovchynnyk, Michigan State Univ.], Trudy-limnol. Sta. Kosine, 22:196-218. Brinkhurst, R. O. 1974. The benthos of lakes. St. Martin’s Press, New York, 190 pp. Brown, H. P. 1972. Biota of freshwater ecosystems identification manual no. 6. Aquatic dryopoid beetles (Coleoptera) of the United States. EPA, Proj. no. 18050 ELD, Supt. Doc., Washington D.C., 82 pp. Buckley, B. R., and J. E. Sublette. 1964. Chironomidae (Diptera) of Louisiana II. The limnology of the upper part of Cane River Lake, Natchitoches Parish, Louisiana, with particular reference to the emergence of Chironomidae. Tulane Stud. Zool., 1 1:151-166. Carlson, R. E. 1977. A trophic state index for lakes. Limnol. Oceanogr., 22:361-369. Deevey, E. S. 1942. Studies on Connecticut lake sediments III. The biostratomy of Linsley Pond. Amer. J. Sci., 240:233-238. Dixon, W. J., and M. B. Brown (eds). 1979. BMDP-79 Biomedical computer programs P-series. Univ. California Press, Berkeley, 880 pp. Durret, C. W. 1973. Density, distribution, production, and drift of benthic fauna in a reservoir receiving thermal discharges from a steam-electric generating plant. Unpub¬ lished M.S. thesis, North Texas State Univ., Denton, 74 pp. Edmonds, G. F., Jr., S. L. Jensen, and L. Berner. 1976. The mayflies of North and Central America. Univ. Minnesota Press, Minneapolis, 330 pp. Edmondson, W. T., ed. 1959. Fresh-water Biology. John Wiley & Sons, Inc., New York, 2nd ed., 1248 pp. Geagan, D. W., and T. D. Allen. 1961. An ecological survey of factors affecting fish production in Louisiana waters. Dingell-Johnson Proj. F6-R, Louisiana Wildlife and Fish Comm., 100 pp. Geagan, D. W., and C. M. Fuss. 1959. Thermal and chemical stratification of some Louisiana Lakes. Proc. Louisiana Acad. Sci., 22:32-43. Gerking, S. D. 1962. Production and food utilization in a population of bluegill sunfish. Ecol. Monogr., 32:31-78. Goodnight, C. J. 1973. The use of Aquatic macroinvertebrate as indicators of stream pollution. Trans. Amer. Micros. Soc., 92:1-13. Hamilton, A. L. 1969. On estimating annual production. Limnol. Oceanogr., 14:771-782. Harrel, R. C. 1973. Limnological studies on a southeast Texas meander scar lake. Texas J. Sci., 24:517-533. LOUISIANA OXBOW LAKES 153 Heuschele, A. S. 1969. Invertebrate life cycle patterns in the benthos of a floodplain lake in Minnesota. Ecology, 50:998-101 1. Jones, J. R., B. P. Borofka, and R. W. Bachmann. 1976. Factors affecting nutrient loads in some Iowa streams. Water Res., 10: 1 17-122. Kalkomey, K. W. 1979. The effects of flood control impoundments on sedimentation in six oxbow lakes within the Angelina and Neches River basins. Unpublished M.S. thesis, Stephen F. Austin State Univ., Nacogdoches, 39 pp. Kajak, Z., and J. I. Ryback. 1966. Production and some trophic dependences in benthos against primary production and zooplankton production of several masurian lakes. Verh. Int. Verein. Theor. Angew. Limnol., 16:441-451. Klemm, D. J. 1972. Biota of freshwater ecosystems identification manual no. 8. Freshwater leeches (Annelida: Hirundinea) of North America. EPA Project no. 18050 ELD, Supt. Doc., Washington D.C., 53 pp. Lambou, V. W. 1960. Fish populations of Mississippi River oxbow lakes in Louisiana. Proc. Louisiana. Acad. Sci., 23:52-64. Mann, K. H. 1980. Benthic secondary production. Pp. 103-118, in Fundamentals of aquatic ecosystems R.S.K. Barnes and K.H. Mann, eds.), Blackwell Sci. Publ., Oxford, 229 pp. Marsh, C. E., J. L. McGraw, Jr., and R. C. Harrel. 1978. Rotifer population in a southeast Texas oxbow lake with emphasis on cyclomorphosis of Keratella cochlearis. South¬ western Nat., 23:633-640. Mason, W. T. 1973. An introduction to the identification of chironomid larvae. EPA., Nat. Environ. Res. Center, Cincinnati, Ohio, 90 pp. McCullough, J. D., and D. W. Jackson. 1985. Composition and productivity of the macroinvertebrate community of a subtropical reservoir. Int. Revue Ges. Hydrobiol., 70:221-235. Menzie, C. A. 1980. A note on the Hynes method of estimating secondary production. Limnol. Oceanogr., 25:770-773. Moore, J. W. 1980. Composition of benthic invertebrate communities in relation to phytoplankton populations in five subarctic lakes. Int. Revue Ges. Hydrobiol., 65:657- 671. Moore, W. G. 1950. Limnological studies of Louisiana lakes I. Lake Providence. Ecol¬ ogy, 31:86-99. - . 1963. Central Gulf States and the Mississippi Embayment. Limnology in North America (D. G. Frey, ed.), Univ. Wisconsin Press, Madison, 734 pp. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner, and D. H. Bent. 1975. SPSS Statitical package for the social sciences. McGraw and Hill, New York, 2nd ed., 675 pp. Oliphant, R. P. 1977. Some effects of a thermal effluent on secondary productivity of the macroinvertebrate benthic community in Fairfield Reservoir, Texas. Unpublished M.S. thesis, Stephen F. Austin State Univ., Nacogdoches, 1 16 pp. Omernik, J. M. 1976. The influence of land use on stream nutrient levels. EPA-600/ 3-76-014, EPA, Oregon, 106 pp. Paloumpis, A. A., and W. C. Starrett. 1960. An ecological study of benthic organisms in three Illinois River floodplain lakes. Amer. Midland Nat., 64:406-435. Pennak, R. W. 1978. Fresh-water invertebrates of the United States. John Wiley and Sons, New York, 2nd ed., 803 pp. Reece, C. H. 1979. A physicochemical and phytoplankton productivity study of six oxbow lakes in the Neches River Basin in East Texas. Unpublished M.S. thesis, Stephen F. Austin State Univ., Nacogdoches, 149 pp. Shannon, C. E., and W. Weaver. 1963. The mathematical theory of communication. Univ. Illinois Press, Urbana, 117 pp. 154 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Sublette, J. E. 1964. Chironomidae (Diptera) of Louisiana I. Systematics and immature stages of some lentic chironomids of west-central Louisiana. Tulane Stud. Zool., 1 1:109- 150. United States Geological Survey. 1980. Water resources data for Louisiana, water year 1980 volume 1. Central and northern Louisiana. USGS, Water Resources Division, Baton Rouge, Louisiana, 526 pp. Welch, H. E., Jr. 1968. Energy flow through the major macroscopic components of an aquatic ecosystem. Unpublished Ph.D. dissertation, Univ. Georgia, 97 pp. Wilhm, J. L., and T. C. Dorris. 1968. Biological parameters for water quality criteria. Bioscience, 18:477-481. EVALUATION OF VOLCANIC ASH AS A STRATIGRAPHIC MARKER IN PLAYA BASINS, WESTERN TEXAS Kevin Cornwell and Calvin G. Barnes Geotechnical Services, Inc., 5730 South 86th Circle, Omaha, Nebraska 68127, and Department of Geosciences, Texas Tech University, Lubbock, Texas 79409 Abstract. — A volcanic ash deposit is exposed in Skeen Lake, a small playa basin in the larger Guthrie Lake basin on the Southern High Plains, Lynn County, Texas. The ash has been characterized to determine provenance and age. Characterization was accomplished through petrographic and major-element geochemical analysis. Bubble-wall and bubble- junction glass shard morphologies, average refractive index of 1.497, microphenocryst assemblage that included quartz, sanidine, magnetite, ilmenite, zircon, amphibole, clinopyroxene, allanite, chevkinite, biotite, and plagioclase, and distinctive concentrations of CaO (0.54 weight percent) and FeO (1.54 weight percent) helped “fingerprint” the ash. The ash is correlative with Lava Creek B ash deposits that originated in the Yellowstone area of Wyoming. This correlation suggests an approximate age of 621 thousand years (Pleistocene) for the Skeen Lake ash. Sedimentary features suggest several episodes of reworking and weathering during ash accumulation. Key words: volcanic ash; major-element concentration; shard morphology; microphenocryst assemblage. Correlation of lacustrine sediments between playa basins on the Southern High Plains of Texas is difficult because correlation usually is based on the highly variable lithologic character of the sediments (C. C. Reeves, Jr., personal communication, 1983). Late Cenozoic volcanic ash deposits are preserved locally in some of these playa basins and respresent stratigraphic markers, age and source of which can be determined by laboratory study. Ash deposits can be used to correlate lacustrine sediments, estimate sedimentation rates, and provide minimum ages for these basins. Volcanic ash derived from eruptions at Long Valley, California (Izett et al., 1970; Borchardt et al., 1972; Kortemeier, 1982), Yellowstone, Wyoming and Idaho (Swineford and Frye, 1946; Izett et al., 1971; Naeser et al., 1973; Miller, 1974; Boellstorff, 1976), and the Jemez Mountains, New Mexico (Izett et al., 1972, and Izett, personal communication, 1984) have been recognized on the Southern High Plains. Radiometric dating of ash deposits has been done by Evernden and Curtis (1951), Fleischer and Price (1964), Dalrymple et al. (1965), Naeser et al. (1971), and Boellstorff (1976). Dating techniques include fission track analysis of zircon, apatite, and glass and potassium-argon analysis of feldspar and biotite microphenocrysts. Uncertainties in such dates arise from counting statistics and also from problems related to deposition and sampling. For example, the abundance of radioactive minerals needed for radiometric dating is typically low in distal volcanic ash beds; thus, large The Texas Journal of Science, Vol. 39, No. 2, May, 1987 156 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 quantities of ash are needed to concentrate a sufficient number of grains for analysis. In addition, post-depositional alteration can leach radioactive elements from the mineral to be dated and contamination can be introduced from detrital radioactive materials during reworking of the ash. In order to overcome problems inherent in radiometric dating, it is possible to determine the geochemical characteristics of an ash unit and to match this “fingerprint” with volcanic source areas that have been dated previously. This technique has been an effective method for dating ash deposits on the Great Plains (Swineford and Frye, 1946; Czamanske and Porter, 1965; Theisen et al., 1968; Randle et al., 1970; Izett et al., 1970, 1971, 1972; Borchardt et al., 1972; Miller, 1974; Kortemeier, 1982; Izett, 1981) because the volcanic source areas generally lack detrital contamination and have fresh, abundant exposures. The purpose of this project was to determine the geochemical characteristics of a volcanic ash deposit that occurs in the Skeen Lake basin and to determine the probable volcanic source and age. The ash should provide a useful stratigraphic marker in this and other Southern High Plains basins and its age provides constraints on the timing of basin formation. Such information is particularly important in view of the interest in siting a nuclear waste repository on the Southern High Plains (see Reeves and Temple, 1986). Methods Field investigation incorporated drill hole and outcrop data in order to determine the stratigraphic relationship of the ash bed. Several pits were dug at various levels within the ash bed to reveal as much of the deposit as possible. Stratigraphic level was determined by tracing beds with oscillatory ripple marks. A detailed petrographic inspection was conducted to determine glass shard morphologies and microphenocryst assemblage. Glass shards were separated from the ash using a water tower separation technique (Harris, 1965) and by hand picking. Immersion oils were used to determine the index of refraction of glass shards and the indices of immersion oils were checked using an Abbe refractometer with a sodium source. The microphenocryst assemblage was separated from the ash using the water tower apparatus, standard heavy liquid techniques, a Franz isodynamic separator, and hand picking. The microphenocrysts were identified using optical techniques. Chemical compositions of individual glass shards were determined using a JEOL model JXA-733 electron microprobe at Southern Methodist University. The operating conditions were an accelerating voltage of 15 kilovolts, beam current of five nanoamps, and a beam diameter of 20 micrometers. The microprobe was calibrated using a suite of natural and synthetic standards. Study Area The Southern High Plains of Texas are characterized by thousands of playa lake basins which range in size from several hundred square meters to several square kilometers. The playa basin chosen for this study is located in Lynn County, Texas, approximately 45 kilometers south of Lubbock, Texas, on U.S. Highway 87. At the time of this study, a small VOLCANIC ASH IN PLAYA BASINS 157 body of water (Skeen Lake) occupied this basin. The Skeen Lake basin is one of several small basins within the larger Guthrie Lake basin (Fig. 1). The Skeen Lake playa basin was selected for three reasons: 1) the occurrence of volcanic ash in the basin; 2) the relative abundance of ash (more than five meters thick in some areas); and 3) the accessibilility of the ash in drainage cuts and barrow pits. Ash from two other Southern High Plains deposits, Buffalo Springs Lake in Yellow House Canyon, near Lubbock, and a playa basin approximately three miles south of Tulia, Texas, and pumice samples from the Jemez Mountains of New Mexico (Bandelier tuff) were collected for comparison with the ash in Skeen Lake basin. In this report, the ash bed that underlies part of the Skeen Lake basin is referred to as the “Skeen Lake ash,” the ash occurring in Yellow House Canyon is referred to as the “Buffalo Springs ash,” and the ash that occurs south of Tulia, Texas is referred to as the “Tulia ash.” These names are not formal stratigraphic terms, but serve to identify each of the deposits. Stratigraphy An ash bed is exposed on the north and west sides of a small playa lake basin just north of Skeen Lake (Lynn County, Texas) in a man-made drainage cut and several barrow pits. The drainage cut is approximately six meters wide and 34 meters long and exposes the top 1.8 meters of the 5.2-meter-thick ash bed. Layering within the ash bed strikes N25W and dips one to four degrees to the northeast. A zone of white, calcareous sediment, 3.1 meters thick, occurs on the east side of the basin at approximately the same elevation as the ash deposit but does not contain ash. A small power auger was used to drill two holes to determine the thickness of the ash deposit. One hole was drilled just north of the drainage cut and revealed an ash thickness of approximately 5.2 meters. The second hole was drilled in the floor of one of the barrow pits and approximately 4.6 meters of ash was encountered there. In both localities, the ash lies unconformably on the Duck Creek Formation of Cretaceous age. Underlying Sedimentary Rocks The Cretaceous Duck Creek Formation is a gray-green to yellow-brown sandy shale. Regionally, Cretaceous sediments are unconformably overlain by the Ogallala Formation. The conformity is typically exposed in larger pluvial basins such as the Gutherie Lake basin. However, erosion removed the Ogallala Formation from most of the Skeen Lake basin. Subsequent deposition and reworking of ash have complicated the unconformity by depositing ash directly on the Duck Creek Formation in peripheral parts of the basin but over lacustrine sediments in more basinward areas (C. C. Reeves, Jr., personal communication). Skeen Lake Ash Drill cuttings show that the lower 4.0 meters of the Skeen Lake ash is predominantly gray with cavities and solution channels filled with brown sediment (Fig. 2). The ash is fine grained and ranges in color from pale 158 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 TAHOKA Figure 1. Location of the Guthrie Lake basin, Texas. Outline of the Guthrie basin shows the location of Skeen Lake and Guthrie Lake. gray to gray. No phenocrysts are visible to the naked eye. Small (0.16 centimeter), black inclusions occur in isolated pockets and apparently represent organic material that was trapped in the ash during reworking of the initial airfall deposit. Five fresh-water limestone lenses occur from 15 to 107 centimeters from the top of the unit. These limestone lenses are dark gray, indurated, discontinuous, and range in thickness from 0.3 to 3.8 centimeters. At a height of 4.2 meters, oscillatory ripple marks occur and can be traced at least 24.4 meters. The ripple marks trend N89W, have an amplitude of about 0.65 centimeter, and a wavelength of 2.55 centimeters. In localities exposed to weathering, a layer of calcareous cement, 0.30 centimeter thick, caps the ripple marks. Alternating layers of pale gray and gray ash occur above the ripple marks. Swirled flow structures, horizontal laminations, small (0.15 centimeter) black inclusions, and a small (2.5 centimeters thick) limestone lens all occur in this 20.3-centimeter section. VOLCANIC ASH IN PLAYA BASINS 159 0 F - G H Figure 2. Stratigraphy of the Skeen Lake ash deposit. Ash is predominantly gray except as marked with (L), where it is pale gray. Units are: A, massive ash; B, calcareous ash; C, laminated ash; D, flow structure; E, Cretaceous Duck Creek Formation; F, pale gray ash zones; G, clay- and organic-rich pockets; H, oscillatory ripple marks; I, green ash. 160 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 The upper 23 centimeters of the ash bed represents the contact between ash and overlying sediment and are predominantly platy, calcareous, interbedded layers of pale gray and gray ash that range in thickness from 5.1 to 7.6 centimeters. The most widespread bedding structures in the ash deposit are the swirled flow structures. These consist of thin parallel laminations contorted into swirls and circular shapes. Deposition by debris flow or sheet flow is the most likely explanation of these structures. The presence of swirled flow structure in all sampled localities suggests that the process responsible operated on a basin-wide scale and was not restricted to one or two sites. In the drainage cut, joints strike between N6E and N45E. These joints could be desiccation cracks or could be the result of loading and unloading forces caused by fluctuating levels of water in the basin. Thin (0.64 centimeter thick), white, indurated calcareous cement fills these fractures. Discontinuous zones of weathered green ash occur at various locations within the deposit and are thought to represent hiatuses within the depositional cycles of the ash. Weathered ash shows several diagenetic effects. In some localities a thin, white, indurated, calcareous layer plates the ash. In other localities, the upper eight to 15 centimeters of ash have become indurated with platy, calcareous cement. In barrow pits where the ash has been exposed to weathering for some time, it is yellowish green in color and is locally clay- rich. Overlying Sediments The ash is unconformably overlain by small (15 centimeters in diameter) discontinuous pockets of red-brown fine sand to sandy clay loam. Our work and more recent drilling by C. C. Reeves, Jr. (personal communication) show that these sediments are part of the eolian Blackwater Draw Formation (Reeves, 1976). Regionally, the Blackwater Draw Formation is permeated with white, semi-indurated caliche, which ranges in thickness from 15 to 76 centimeters and is overlain by soils from the Mansker and Portales Group (Mowery and McKee, 1958). These soils vary in thickness from 15 to 91 centimeters. Petrography Shards comprise more than 99 percent of the ash and are composed of semihydrated optically isotropic glass. Shard morphologies are primarily bubble wall and bubble junction (approximately 96 percent) with the remainder being pumiceous. Heiken (1972) considered shard morphology to be governed by the viscosity of the original magma, which in turn is dependent on temperature, chemical composition, and volatile content (principally water and the halogens). Bubble-wall and bubble-junction VOLCANIC ASH IN PLAYA BASINS 161 shard morphologies are generally associated with magmas having relatively low viscosities, whereas pumiceous shards are characteristic of more viscous magmas (Izett, 1981). Some of the glass shards appear slightly cloudy due to numerous gas bubbles. These gas bubbles are spheroidal in shape and are relatively abundant in the bubble-wall and bubble-junction shards. Rare shards are deep gray to black in color. Steen-Mclntyre (1977) suggested that this phenomenon could result from exsolution of finely divided mafic minerals (for example, magnetite — Schlinger et al., 1986), or from contamination by magma mixing. The refractive index of the glass shards ranges from 1.495 to 1.498 and averages 1.497. Refractive indices of glass shards from several ash beds on the Southern High Plains are shown in Table 1. Note that the refractive index of a glass sample depends, to a significant extent, on whether or not the glass is hydrated (Ross and Smith, 1955) because hydration tends to cause a change in refractive index. The microphenocryst assemblage within the Skeen Lake ash bed is similar to assemblages in many other deposits on the Southern High Plains. Table 1 also shows the microphenocrysts that occur in ash deposits from several source areas. Note that the absence of plagioclase in the Bandelier tephra is considered to be characteristic of the unit (Izett et al., 1972) as is the occurrence of mica in the “Pearlette” ash. (“Pearlette ash” collectively refers to several ash beds that crop out on the Great Plains and were erupted from vents in the Yellowstone Park area.) The microphenocryst assemblage of the Skeen Lake ash is quartz, sanidine, magnetite, ilmenite, zircon, amphibole, clinopyroxene, allanite, chevkinite, biotite, and plagioclase. Crystals that did not have glass adhering to their surfaces were classified as detrital grains although some may be microphenocrysts. Detrital grains were probably introduced during deposition and reworking of the ash. In plane light, microphenocrystic clinopyroxene is dark green and slightly pleochroic. The Skeen Lake ash also contains detrital clinopyroxene that lacks adhering glass shards. The detrital clinopyroxene is pale green in color, diopsidic in composition, and has a sharp, angular morphology that suggests a minimum amount of abrasion during transport. The detrital clinopyroxene could have been entrained during eruption by erosion of the vent walls. Entrained xenocrysts would be transported and deposited with the ash but would not be cognate to the ash-forming magma and would lack attached glass. The pale clinopyroxene could also be tephra from a different (mafic) source area, such as New Mexico, Mexico, or Central America. If so, glass from such an eruption was not observed or has devitrified, leaving only its microphenocryst assemblage. No other angular microphenocryst phase lacking adhering glass was observed. 162 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Table 1. Refractive indices of glass shards and microphenocryst assemblages from several Great Plains ash deposits. Range Average Source “Pearlette ” ash 1.498-1.502 1.501 Swineford and Frye, 1946 1.499-1.501 N/A Frye et al., 1948 1.498-1.500 1.499 Izett et al., 1970 quartz, oligoclase, sanidine, ferroaugite, hornblende, allanite, chevkinite, zircon, apatite, magnetite, ilmenite, sphene, mica, and clinopyroxene Bishop ash 1.492-1.499 1.495 Izett et al., 1970 biotite, hornblende, quartz, plagioclase, sanidine, zircon, apatite, allanite, sphene, ilmenite, and magnetite Bandelier tephra (Guaje ash) 1.497-1.499 1.497 Izett et al., 1972 quartz, sanidine, clinopyroxene, zircon, chevkinite, allanite, magnetite, and ilmenite 1.496-1.498 (Guaje pumice) 1.497 Izett et al., 1972 1.496-1.498 (Tsankawi pumice) 1.497 Izett et al., 1972 1.501-1.518 Mount Mazama 1.508 Randle et al., 1970 plagioclase, hypersthene, magnetite, hornblende, clinopyroxene, trace apatite Allanite is yellow-brown to pale green in plane light. Allanite grains exhibit subhedral to anhedral morphologies and are the least abundant mineral present. Chevkinite occurs as long, pencil-shaped euhedral crystals and is red-brown in plane and polarized light. Amphibole is dark brown to dark green in color and is anhedral. The anhedral morphology suggests resorption of amphibole by the host magma. Anhedral amphibole grains from other ash beds on the Great Plains have not been reported. Garnet also occurs in the ash bed. The grains show little evidence of reworking and fluvial transport; however, the garnet grains are not attached to glass and are likely detrital. Izett (1981) noted that the occurrence of garnet in a magmatic suite probably is related to vapor-phase crystallization or incorporation of xenocrysts. Conversely, the garnet and diopsidic clinopyroxene could reflect detrital contamination from a high-grade metamorphic terrain (for example, the Sangre de Cristo Range, New Mexico). Mineral assemblages separated from the Buffalo Springs and Tulia ashes and the Guaje and Tsankawi pumice beds of the Bandelier Tuff all contained quartz, sanidine, and clinopyroxene. Clinopyroxene similar to VOLCANIC ASH IN PLAYA BASINS 163 the detrital clinopyroxene in the Skeen Lake ash was not present in these ash and pumice beds. The Tulia ash contains rare anhedral amphibole but lacks magmatic biotite. Geochemistry Most siliceous ash beds in the western and central United States can be classified as either rhyolitic or dacitic. Rhyolitic ashes can be further subdivided into G- and W-types. Izett (1981) showed that ash types also could be distinguished on the basis of shard types, mineralogy, glass composition, and refractive index. These features are summarized in Table 2. Izett (1981) showed that G-type and W-type rhyolitic centers are built on areas of continental crust that have undergone extensional deformation. These include Long Valley, California, Yellowstone, Wyoming, and the Valles Caldera, New Mexico. Dacitic centers typically are located near convergent plate margins such as the Cascade Range of the Pacific Northwest. Major element compositions of glass from the three ash deposits under study are presented in Table 3. For purposes of correlation, elements that exhibit a noticeable degree of variation within and among tephra layers of different ages and origins are considered most useful. These elements are Ca, Fe, Mn, Mg, and Ti. Na and K are strongly affected by post- depositional diagenetic processes and yield results that are not characteristic of the original ash chemistry. The Skeen Lake ash is clearly rhyolitic but does not exactly fit into either G- or W-type. However, the Skeen Lake ash shows greater similarity to G-type rhyolites than to W-type rhyolites, as is evident from its FeO content (Tables 2 and 3). Average major-element concentrations of Skeen Lake, Tulia, and Buffalo Springs glasses and of tephra from other late Tertiary and early Quaternary volcanic systems are presented in Table 4. Comparison of FeO contents shows that the Skeen Lake and Tulia glasses are more FeO-rich than glasses from the Long Valley (Bishop), Los Chocoyos, or Crater Lake (Mazama) eruptive centers. CaO contents of the Skeen Lake and Tulia ashes are greater than those in Bandelier (Tsankawi and Guaje) eruptive products and less than those in Mazama and Los Chocoyos glass (Table 4). Both Skeen Lake and Tulia glasses are quite similar to glass from Yellowstone eruptive centers (Lava Creek B, Mesa Falls, and Huckleberry Ridge). The Lava Creek B glass most closely resembles the Skeen Lake and Tulia glass, suggesting that both of these playa lake deposits were erupted during the Lava Creek B event. Buffalo Springs ash is nearly identical to Guaje glass, supporting the correlation of the deposit with the Guaje eruption made by Izett (1981). 164 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 of £ % o ^ cl -3 o -<-* .S "O u. .22 03 Q "S e c cd cd <*5 1/5 s -s a v • & g * 2 £ 6 — s- cd O >. •s cd r- o o ON CM CO d- in r- no ON On On ■d- -d- -d- s S © ^ V ‘C o o V V 3 cd O £ « ^ S £ I cd 13 3 2 u | % ob a a >N ^ 6 VOLCANIC ASH IN PLAYA BASINS 165 Table 3. Major-element compositions of glass shards from Skeen Lake, Tulia, and Buffalo Springs ashes. Si02 A 1203 FeO* MgO CaO Na20 K20 Ti02 MnO P205 Total SKEEN LAKE 72.05 11.36 1.63 0.03 0.45 2.96 5.24 0.16 0.12 0.14 94.14 72.25 11.71 1.66 0.05 0.41 2.69 5.20 0.21 0.09 0.26 94.53 72.33 11.42 1.25 0.03 0.57 2.71 5.06 0.14 0.00 0.06 93.57 71.97 11.42 1.30 0.05 0.61 3.21 4.79 0.31 0.13 0.43 94.22 71.11 11.39 1.33 0.04 0.55 2.88 5.38 0.26 0.11 0.14 93.19 70.96 11.21 1.67 0.04 0.52 3.21 5.22 0.12 0.06 0.00 93.01 72.98 11.60 1.14 0.04 0.47 3.26 5.17 0.17 0.00 0.00 94.83 70.41 11.18 1.52 0.06 0.53 3.28 4.79 0.04 0.00 0.28 92.09 Ave. 71.76 11.41 1.44 0.04 0.51 3.02 5.11 0.18 0.06 0.16 S.D. 0.73 0.15 0.15 0.01 0.04 0.29 0.18 0.05 0.05 0.05 TULIA 72.74 11.30 1.20 0.01 0.42 3.14 5.13 0.12 0.09 0.06 94.21 70.83 11.14 1.61 0.06 0.47 3.02 4.75 0.14 0.01 0.33 92.36 72.68 11.65 1.57 0.02 0.55 3.13 5.44 0.07 0.01 0.26 95.38 71.81 11.27 1.52 0.02 0.57 3.14 4.81 0.21 0.10 0.13 93.58 Ave. 72.02 11.34 1.47 0.03 0.50 3.11 5.03 0.14 0.05 0.20 S.D. 0.90 0.22 0.19 0.02 0.07 0.06 0.32 0.06 0.05 0.12 BUFFALO SPRINGS 72.21 11.69 1.04 0.02 0.17 4.13 4.59 0.17 0.07 0.12 94.26 71.79 11.50 1.35 0.02 0.18 3.81 4.77 0.00 0.06 0.10 93.58 70.96 11.16 1.29 0.01 0.23 3.53 4.86 0.00 0.12 0.00 93.68 70.73 11.53 1.11 0.00 0.23 3.66 4.00 0.31 0.27 0.21 92.05 70.51 11.16 1.25 0.02 0.11 3.54 5.02 0.04 0.00 0.31 91.96 Ave. 71.23 11.40 1.20 0.02 0.18 3.73 4.64 0.07 0.07 0.11 S.D. 0.33 0.13 0.10 0.02 0.04 0.18 0.32 0.10 0.10 0.06 *Total Fe as FeO Environment of Deposition The playa lake basin in which the ash occurs is located at the southeastern end of the much larger (approximately 16 kilometers long and eight kilometers wide) Guthrie Lake basin. Reeves (1966a, 1966b) suggested that many of the large pluvial lakes on the Southern High Plains were once part of an open lake system interconnected by early Pleistocene streams. These early Pleistocene drainage channels trend east to southeast (Gazdar, 1981). The lake basins formed over Cretaceous topographic highs where the Ogallala Formation was unusually thin. Once the drainage penetrated the Ogallala Formation and encountered the more competent Cretaceous shales and limestones, downward erosion slowed and lateral erosion predominated, carving out the lake basins. The Guthrie Lake basin overlies Cretaceous sediment (predominantly Kiamichi Shale and Edwards Limestone) which, in areas surrounding the 166 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Table 4. Major-element compositions of glass from selected pyroclastic rocks, normallized to 100 percent, anhydrous. Ash Age* Si02 A 1203 FeO MgO CaO Na20 K20 Ti02 MnO 1 Bishop Tuff 0.74 76.9 13.1 0.71 0.05 0.53 3.72 4.80 0.06 — 2 Bishop ash 0.74 77.3 13.0 0.72 0.07 0.56 3.43 4.80 0.07 3 Lava Creek B 0.61 76.8 12.2 1.57 0.02 0.54 3.57 4.98 0.10 0 0 i 4 Mesa Falls 1.27 76.8 12.2 1.49 0.04 0.58 3.21 5.41 0.11 0.03 5 Huckleberry Ridge 2.01 76.5 12.2 1.78 0.02 0.58 3.34 5.21 0.12 0.05 6 Skeen Lake see text 76.7 12.2 1.54 0.04 0.54 3.23 5.46 0.19 0.06 7 Tulia see text 76.8 12.1 1.64 0.02 0.56 3.25 5.47 0.14 0.04 8 Buffalo Springs see text 76.9 12.3 1.30 0.02 0.19 4.03 5.01 0.08 0.11 9 Guaje ash 1.47 76.4 12.4 1.38 0.05 0.28 3.73 5.61 0.05 0.08 10 Guaje pumice 1.47 76.2 12.9 1.32 0.03 0.21 4.06 5.14 0.04 0.08 1 1 Tsankawi pumice 1.15 76.5 12.9 1.45 0.04 0.26 4.00 4.73 0.04 0.08 12 Los Chocoyos 85,000 77.0 12.9 0.90 0.31 1.18 3.98 3.60 0.14 — 13 Mazama 6,500 75.1 14.6 1.79 0.48 1.79 4.09 1.56 0.46 0.06 Sources for compositional data: 1 and 2 (Izett et al., 1970; 3, 4 and 5 (Christiansen and Blank, 1972); 6, 7, and 8 (this paper); 10 and 11 (Izett et al., 1972); 12 (Drexler et al., 1980); 13 Ritchey, 1980). Sources for age data: see Table 1 in Izett (1981). *Age in millions of years, except for Chocoyos and Mazama (in years) basin, is unconformably overlain by sand and gravel of the Ogallala Formation. The age of the Guthrie Lake basin has not been clearly determined. Winkler (1985) considered the Ogallala Formation to be late Miocene-Pliocene in age. Thus the Guthrie Lake basin has a maximum age of mid- to late Miocene (assuming that the basin developed after the Ogallala Formation was deposited and did not exist during Ogallala deposition). In addition, the absence of sediments of Nebraskan and Kansan age in the basin suggests a late Pleistocene age (Fig. 2). Because Quaternary drainage has been to the southeast, one would expect the deepest part of the Guthrie basin to develop in the southeast (that is, Skeen Lake basin). Numerous playa basins similar to the Skeen Lake basin occur within the Guthrie Lake basin, some of which are deeper than the Skeen Lake basin. However, the absence of ash in these playas suggests that they formed after the ash fall as the result of deflationary processes (Reeves, 1965) and are younger than the Skeen Lake basin. The occurrence of ash in the Skeen Lake basin and its absence from basins elsewhere within the Guthrie Lake basin may indicate that Skeen Lake was the only significant body of water present (in the Guthrie basin) during deposition and reworking of the ash, or that the major part of the Guthrie basin did not exist during early Pleistocene time. Outcrops of ash 4.6 to 6.1 meters above the present water level demonstrate that Skeen Lake was larger at the time of deposition than at present. VOLCANIC ASH IN PLAYA BASINS 167 Discussion and Conclusions Chemical analysis shows that glass from the Skeen Lake and Tulia ash deposits are compositionally similar to tephra erupted during the Lava Creek B event in the Yellowstone volcanic province (Table 4). This correlation is supported by the absence of Mesa Falls tephra elsewhere on the Southern High Plains (Izett, 1981). The Lava Creek B eruption occurred 610,000 years before present (see Izett, 1981). Thus the presence of this ash in the Skeen Lake and Tulia basins restricts basin formation to ages greater than 610,000 years. In spite of the fact that older Pleistocene ash deposits occur elsewhere on the Southern High plains (Izett et ah, 1972; Izett, personal communication, 1984), drilling shows that older ash does not occur in the Skeen Lake basin (Reeves, personal communication, 1986). With the exception of the Bishop ash, all of these eruptions were older than 1.15 million years (Table 4). Thus it is evident that the Skeen Lake basin formed in the interval between the last voluminous Bandelier eruption (1.15 million years) and the Lava Creek B eruption (610 thousand years). This interpretation is supported by the apparent absence of sediments of Nebraskan age in the basin. Further drilling within the Guthrie basin is necessary to confirm its chronology and the exact relationship to the Skeen basin. Acknowledgments We wish to thank C. C. Reeves, Jr., and J. R. Giardino for their assistance in the field and for reviews of an early version of this paper. Dr. Reeves generously shared data from recent drilling in the Skeen Lake basin that have helped to clarify our original data set. Mr. Leighton Knox and Mr. Cleve Littlepage graciously provided access to the Skeen Lake basin. Dwight Deuring assisted in microprobe analyses. We are grateful for financial support from the Summer Research Program of the Graduate School, Texas Tech University, from the Department of Geosciences, Texas Tech University, and from Sigma Xi. Literature Cited Boellstorff, J. D. 1976. The succession of Late Cenozoic volcanic ashes in the Great Plains: a progress report. Pp. 35-71, in Guidebook, 24th Annual Meeting, Midwestern Friends of the Pleistocene, Stratigraphy and Faunal Sequence— Meade County, Kansas, Guidebook Ser. 1, Kansas Geol. Surv., 85 pp. Borchardt, G. A., P. J. Aruscavage, and H. T. Millard, Jr. 1972. Correlation of the Bishop Ash, a Pleistocene marker bed, using INAA. J. Sed. Petrol., 42: 301-306. Christiansen, R. L., and H. R. Blank, Jr. 1972. Volcanic stratigraphy of the Quaternary rhyolite plateau in Yellowstone National Park. U.S. Geol. Survey Prof. Paper, 729-B: 1-18. Czamanske, G. K., and S. C. Porter. 1965. Titanium dioxide in pyroclastic layers from volcanoes in the Cascade Range. Science, 150: 1022-1025. Dalrymple, G. B., A. Cox, and R. R. Doell. 1965. Potassium-argon age and paleomagnetism of the Bishop Tuff, California. Geol. Soc. Amer. Bull., 76:665-673. 168 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Drexler, J. W., W. I. Rose, Jr., R. S. J. Sparks, and M. T. Ledbetter. 1980. The Los Chocoyos Ash, Guatemala: a major stratigraphic marker in middle America and in three ocean basins. Quat. Res., 13: 327-345. Evernden, J. F., and K. Curtis. 1951. Potassium-argon dating of Pleistocene vol- canics. Quaternia, 4: 13-17. Fleischer, R. L., and P. B. Price. 1964. Glass dating by fission fragment tracks. J. Geophys. Res., 69: 331-339. Frye, J. W., A. Swineford, and A. B. Leonard. 1948. Correlation of Pleistocene deposits of the central Great Plains with the glacial section. J. Geol., 56: 501-525. Gazdar, M. N. 1981. Tertiary and Quaternary drainage of the southern High Plains. Ph.D. dissertation. Dept. Geosciences, Texas Tech Univ., Lubbock, 110 pp. Harris, R. L., Jr. 1965. A water tower apparatus to improve zircon separation technique. Geol. Soc. Amer. Bull., 76: 971-974. Heiken, G. 1972. Morphology and petrology of volcanic ashes. Geol. Soc. Amer. Bull., 83: 1961-1988. Izett, G. A. 1981. Volcanic ash beds: Recorders of upper Cenozoic silicic pyroclastic volcanism in the western United States. J. Geophys. Res., 86: 10200-10222. Izett, G. A., R. E. Wilcox, and G. A. Borchardt. 1972. Correlation of volcanic ash in Pleistocene deposits near Mount Blanco, Texas, with the Guaje pumice beds: near the Jemez Mountains, New Mexico. Quat. Res., 2: 554-578. Izett, G. A., R. E. Wilcox, J. D. Obradovich, and R. L. Reynolds. 1971. Evidence for two Pearlette-like beds in Nebraska and adjoining areas. Geol. Soc. Amer. Abs. Prog., 3: 265-266. Izett, G. A., R. E. Wilcox, H. A. Power, and G. A. Desborough. 1970. The Bishop ash bed, a Pleistocene marker bed in the western United States. Quat. Res., 1: 121-132. Kortemeier, C. P. 1982. Occurrence of Bishop Ash near Grama, New Mexico. New Mexico Geol., 4: 22-24. Miller, P. A. 1974. An occurrence of Type B Pearlette Ash in Scurry County, Texas. Geol. Soc. Amer. Abs. Prog., 6: 459. Mowery, I. C., and G. S. McKee. 1958. Soil survey of Lynn County, Texas. USDA Soil Survey Series 1953, 71 pp. Naeser, C. W., G. A. Izett, and R. E. Wilcox. 1971. Zircon fission track ages of Pearlette- like volcanic ashes in the Great Plains. Geol. Soc. Amer. Abs. Prog., 3: 657. - . 1973. Zircon fission track ages of Pearlette family ash beds in Meade County, Kansas. Geology, 1: 93-95. Randle, K., G. G. Goles, and L. R. Kittleman. 1970. Geochemical and petrographical characterization of ash samples from Cascade Range volcanoes. Quat. Res., 1: 261-282. Reeves, C. C., Jr. 1965. Chronology of west Texas pluvial lake dunes. J. Geol., 73: 502- 508. - 1966a. Pluvial basins of west Texas. J. Geol., 74: 269-291. - . 1966b. Pleistocene climate of the Llano Estacado II. J. Geol., 74: 642-647. - . 1976. Quaternary stratigraphy and geologic history of the southern High Plains, Texas and New Mexico. Quat. Strat. North America, Dowden, Hutchison, and Ross Inc., pp. 213-234. Reeves, C. C., Jr., and J. M. Temple. 1986. Permian salt dissolution, alkaline lake basins, and nuclear-waste storage, southern High Plains, Texas and New Mexico. Geology, 14: 939-942. Ritchey, J. L. 1980. Divergent magmas at Crater Lake, Oregon: products of fractional crystallization and vertical zoning in a shallow, water-undersaturated chamber. J. Volcan. Geotherm. Res., 7: 373-386. Ross, C. S., and R. L. Smith. 1955. Water and other volatiles in volcanic glasses. Amer. Mineral., 40: 1071-1089. VOLCANIC ASH IN PLAYA BASINS 169 Schlinger, C. M., R. M. Smith, and D. R. Veblen. 1986. Geologic origin of magnetic volcanic glasses in the KBS tuff. Geology, 14: 959-962. Steen-Mclntyre, V. 1977. A manual for tephrochronology. Unpublished Ph.D. disser¬ tation, Colorado State Univ., 167 pp. Swineford, A., and J. C. Frye. 1946. Petrographic comparison of Pliocene and Pleistocene volcanic ash from western Kansas. Bull. State Geol. Surv. Kansas, 64: 1-32. Theisen, A. A., G. A. Borchardt, M. E. Harwards, and R. A. Schmitt. 1968. Neutron activation for distinguishing Cascade Range pryoclastics. Science, 161: 1009-1011. Winkler, D. A. 1985. Stratigraphy, vertebrate paleontology and depositional history of the Ogallala Group in Blanco and Yellowhouse canyons, northwestern Texas. Unpublished Ph.D. dissertation, Univ. Texas, Austin, 243 pp. KARYOTYPES OF FIVE CRICETID RODENTS FROM HONDURAS Robert D. Bradley and Jan Ensink Department of Biological Sciences and The Museum, Texas Tech University, Lubbock, Texas 79409, and Department of Biology, Texas A&M University, College Station, Texas 77843 Abstract. — Karyotypes of five species of cricetid rodents ( Nyctomys sumichrasti florencei, Oryzomys cousei cousei, Peromyscus boylii sacarensis, Peromyscus mexicanus saxatilis, and Reithrodontomys sumichrasti modestus) are reported for the first time from Honduras. The standard karyotypes of Oryzomys cousei cousei and Peromyscus mexicanus saxatilis revealed that these two subspecies possess the same diploid numbers as those previously reported for their respective species but possess different numbers of autosomal arms. The standard karyotypes of the remaining three taxa showed no variation from those initially described for their species. Key words : standard karyotypes; Nyctomys sumichrasti; Oryzomys cousei; Peromyscus boylii; Peromyscus mexicanus; Reithrodontomys sumichrasti ; Honduras. During a collecting trip to the Republic of Honduras, cricetid rodents were obtained from several localities and karyotyped. Among those collected were five subspecies for which the karyotypes, to our knowledge, have not been reported previously. Two of these subspecies had karyotypes that differed from those initially described for other races of the same species. The description of standard karyotypes of the five subspecies is the basis of this paper. Materials and Methods Somatic cell suspensions were prepared from bone marrow samples of wild-caught individuals as described by Baker et al. (1982). The frozen suspensions were thawed and standard metaphase spreads were prepared upon return to the laboratory. The descriptive morphology of chromosomes follows that of Patton (1967). The diploid number (2N) and total number of autosomal arms, designated as the fundamental number (FN), were determined for each taxon. Voucher specimens were deposited in the Texas Cooperative Wildlife Collections, Texas A&M University. Results and Discussion Oryzomys cousei cousei (Alston, 1877) The karyotype (2N =56, FN =60 — Fig. 1) of this taxon possesses three small biarmed pairs and 24 pairs of large to small acrocentric chromosomes. The X is a large subtelocentric chromosome; the morphology of the Y is unknown. This karyotype differs from that of O. c. aquaticus as reported by Benson and Gehlbach (1979) in that it has two additional biarmed pairs of autosomes. The X and Y were described as a large submetacentric chromosome and a medium-sized acrocentric or subtelocentric chromosome, respectively. The differences in FN and chromosome The Texas Journal of Science, Vol. 39, No. 2, May, 1987 172 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 O. c. cousei M r H" IIP II# ii ■ifi ft ft il A*. I** A| Aft H A H m a a* a • m A % h% Aft - 11 * * * n 8 XX Figure 1. Standard karyotype of Oryzomys cousei cousei. Asterisks indicate biarmed autosomal pairs. morphology of the biarmed pairs suggest that this karyotype is substantially different from that of O. c aquaticus ; however, further investigations (differential staining techniques) are necessary to determine the significance of this variation. Specimen examined. — Dept. Olancho: 2.4 mi. SW Dulce Nombre de Culmi, 1 2. Nyctomys sumichrasti florencei Goldman, 1937 This karyotype (2N =50, FN =52) consists of one large and one small pair of biarmed chromosomes and 22 pairs of acrocentric chromosomes. The X is a large subtelocentric chromosome; the description of the Y is unknown as no males were examined. The karyotype of this subspecies is identical to that reported by Lee and Elder (1977) from Jalisco, Mexico, for Nyctomys sumichrasti sumichrasti. Our purpose for reporting it stems from the fact that Honduras is a considerable distance geographically from Jalisco. This also represents the first reported karyotype for the subspecies, N. s. florencei. Specimen examined. — Dept. Francisco Morazan: El Hatillo, 1 $. Reithrodontomys sumichrasti modestus Thomas, 1907 This karyotype (2N=40, FN=76) possesses 38 metacentric to subtelocentric autosomes. The X is a large metacentric chromosome and KARYOTYPES OF CRICETID RODENTS 173 P. b. sacarensis (40fl )>« 1» * P Hi fin m ii it «• ab ii at r* m si ii ii XX Y Figure 2. Standard karyotypes of Peromyscus boylii sacarensis (FN =52). Asterisks indicate biarmed autosomal pairs. The Y chromosome (from another individual) also is illustrated. the Y is a medium-sized metacentric chromosome. The 2N —40 and FN =76 in this karyotype is similar to that reported by Carleton and Meyers (1979) and Hood et al. (1984) for Reithrodontomys sumichrasti australis from Costa Rica, but differs from the 2N =42 and FN =80 reported by Engstrom et al. (1981) for R. s. nerterus from Mexico. Specimen examined. — Dept. Francisco Morazan: 2 mi. NE El Hatillo, 1(5- Peromyscus boylii sacarensis Dickey, 1928 The karyotype of this taxon (2N =48, FN =52 and 54 — Fig. 2) exhibits a polymorphic condition in that three individuals examined possessed one large biarmed pair, two small biarmed pairs, and 19 acrocentric pairs of chromosomes (FN =52), whereas two other individuals possessed two large biarmed pairs, two small biarmed pairs, and 18 acrocentric pairs of chromosomes (FN =54). The difference between these two karyotypes involves a polymorphic condition (either an acrocentric or biarmed condition) for the second largest chromosomal pair. In both cases, the X is a large submetacentric chromosome, and the Y is a small metacentric chromosome. The polymorphism described herein is similar to that reported by Davis et al. (1986) and Houseal et al. (1987) for P. b. beatae (FN =48- 54) from southern Mexico. Although only the FN =52 and 54 conditions were apparent in this study of five individuals, a larger sample size should reveal mice with FN =53. Specimens examined. — Dept. Francisco Morazan: 2 mi. NE El Hatillo, 1 (5, 3 9; 1.3 mi. NE, 0.8 mi. E El Hatillo, 1 9. 174 ★ M * Aft Figure 3. Standard karyotype of Peromyscus mexicanus saxatilis. Asterisks indicate biarmed autosomal pairs. Peromyscus mexicanus saxatilis Merriam, 1898 The karyotype (2N =48, FN =56 — Fig. 3) of this subspecies is comprised of two large pairs, one medium pair, and two small pairs of biarmed chromosomes in addition to 18 pairs of acrocentric chromosomes. The X is a large submetacentric chromosome and the Y is a small metacentric chromosome. This karyotype differs from that originally described for P. mexicanus (Rogers et al. 1984) by possessing an acrocentric condition for the third largest pair of chromosomes instead of a submetacentric condition. This acrocentric condition results in a FN =56 rather than the FN =58 as described by Rogers et al. (1984). The FN =56 condition of P. m. saxatilis is not only unique to P. mexicanus , but is unique to the entire P mexicanus group, because the nine species of this group previously studied all have a karyotype of FN =58 (Rogers et al., 1984; Smith et al., 1986). Specimens examined . — Dept. Francisco Morazan: 6 mi. NE El Hatillo, 2 3, 1 5, 1.3 mi. NE, 0.8 mi. E El Hatillo, 1 <$, 2 2. Acknowledgments We thank Thomas E. Lee for aid in specimen collection. We also thank Meredith J. Hamilton, Karen McBee, Calvin A. Porter, and Ronald A. Van Den Bussche for reviewing a previous draft of this manuscript. We especially acknowledge the Departmento de Recursors Naturalles for providing us with collecting permits. This study was funded by a National Science Foundation Grant (NSF-PCM-8202794) to Drs. John W. Bickham and Priscilla K. Tucker. THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 P. m. saxatilis w * fl< to 10 n tt fl# fik AT QA Aft I# ft A *n> * * I. XY KARYOTYPES OF CRICETID RODENTS 175 Literature Cited Baker, R. J., M. W. Haiduk, L. W. Robbins, A. Cadena, and B. F. Koop. 1982. Chro¬ mosomal studies of South American bats and their systematic implications. Pp. 303- 327, in Mammalian biology in South America (M.A. Mares and H.H. Genoways, eds). Spec. Publ. Ser., Pymatuning Lab. Ecol., Univ. Pittsburgh, 6:1-539. Benson, D. L., and F. R. Gehlbach. 1979. Ecological and taxonomic notes on the rice rat ( Oryzomys cousei ) in Texas. J. Mamm., 60:225-228. Carleton, M. W., and P. Meyers. 1979. Karyotypes of some harvest mice, genus Reithrodontomys. J. Mamm., 60:307-313. Davis, K. M., S. A. Smith, and I. F. Greenbaum. 1986. Evolutionary implications of chromosomal polymorphisms in Peromyscus boylii from southwestern Mexico. Evo¬ lution, 40:645-649. Engstrom, M. D., R. C. Dowler, D. S. Rogers, D. J. Schmidly, and J. W. Bickham. 1981. Chromosome variation within four species of harvest mice ( Reithrodontomys ). J. Mamm., 62:159-164. Hood, C. S., L. W. Robbins, R. J. Baker, and H. S. Shellhammer. 1984. Chromosomal studies and evolutionary relationships of an endangered species, ( Reithrodontomys raviventris). J. Mamm., 65:655-667. Houseal, T. W., I. F. Greenbaum, D. J. Schmidly, S. A. Smith, and K. M. Davis. 1987. Karyotypic variation in Peromyscus boylii from Mexico. J. Mamm., 68: in press. Lee, M. R., and F. F. B. Elder. 1977. Karyotypes of eight species of Mexican rodents (Muridae). J. Mamm., 48:479-487. Patton, J. L. 1967. Chromosome studies of certain pocket mice, genus Perognathus (Rodentia: Heteromyidae). J. Mamm., 48:27-37. Rogers, D. S., I. F. Greenbaum, S. C. Gunn, and M. D. Engstrom. 1984. Cytosystematic value of chromosomal inversion data in the genus Peromyscus (Rodentia: Cricetidae). J. Mamm., 65:457-465. Smith, S. A., R. D. Bradley, and I. F. Greenbaum. 1986. Karyotypic conservatism in the Peroymscus mexicanus group. J. Mamm., 67:584-586. FRESHWATER BIVALVES OF THE BAFFIN BAY DRAINAGE BASIN, SOUTHERN TEXAS Raymond W. Neck Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, Texas 78744 Abstract. — The freshwater bivalve fauna of the streams draining into Baffin Bay in southern Texas consists of six species of unionids and the introduced Asiatic clam. Although all of the unionids are native to southern Texas, only one species is believed to be native to the drainage basin of Baffin Bay. Key words : freshwater bivalves; Baffin Bay, Texas. The coastal plain of Texas is drained by a series of rivers ranging from the Sabine River on the eastern boundary to the Rio Grande on the southwestern boundary. A relatively large area north of the Rio Grande drainage and south of the Nueces River contains no sizable drainages. The southern part of this area is known as the Llano Mesteno. No cohesive drainages occur on the Llano, the northern area of which contains a series of intermittent streams that flow into a hypersaline estuary known as Baffin Bay (Fig. 1). Limited biological surveys have been carried out in this area of southern Texas, and no freshwater mussels have been reported from the study area (Strecker, 1931). Herein are reported the results of a survey to determine the distribution of freshwater mussels of this region. Regional studies of mussel faunas have covered the Nueces system (D. W. Taylor, unpublished data) and the lower Rio Grande (R. W. Neck and A. L. Metcalf, unpublished data). The environment is moderately xeric. Rainfall is moderate (average annual total for Kingsville is 628 mm), but temperatures are also high (average temperatures for January and July are 13.3° C and 28.9° C, respectively). Most of the streams are intermittent in flow; low runoff periods wash out salt accumulations (Simmons, 1957), whereas torrential runoffs following tropical depressions may scour out normally stable creek bottoms (Suhm, 1974; Russell and Wood, 1976). Some evidence exists that the climatic regime has become more xeric during the past 150 years, but the natural tendency toward low, irregular flows in area creeks may have been exacerbated by land use policies — for example, land clearance for farming, damming of certain streams as stock tanks for ranching, and withdrawal of ground water for municipal and industrial use (Bollaert, 1850; Price and Gunter, 1942). Geological evidence indicates that an extensive river system with moderate water flow occurred in this area during the late Pleistocene (Suhm, 1978). Water pollution is slight to moderate and involves sewage, pesticides, herbicides, and industrial sources (Breuer, 1957). The Texas Journal of Science, Vol. 39, No. 2, May, 1987 178 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Figure 1. Map of drainages of Baffin Bay, Texas, with sampling localities for freshwater bivalves. Bivalve Fauna Comprehensive field sampling over the drainage area of Baffin Bay has revealed a highly localized, low-diversity freshwater mussel fauna. Of 31 sites surveyed, only five supported populations of freshwater bivalves (Table 1). Voucher specimens have been deposited in the Dallas Museum of Natural History. Anodonta imbecilis Say, 1829. — This small, thin-shelled mussel is known from a single pond (locality 6) in the study area. Shells have a brown periostracum with few or no green rays or suffusion. This population is probably referable to the central Texas subspecies A. i. horda (Gould, 1855). Quadrula apiculata speciosa (Lea, 1862). — For many years, Texas populations were referred to under the names Quadrula quadrula (Rafinesque, 1820) or Q. q. apiculata (Say, 1829). A single valve of this species was found in Dairy Lake (locality 18). Pustules are reduced in height above valve surface and are less common on peripheral areas of the shell. BIVALVES OF BAFFIN BAY, TEXAS 179 Table 1. Distribution of freshwater bivalves of Baffin Bay drainage area by habitat, southern coastal plain of Texas. Species Fresh creek Salt creek Fresh pond Estuarine Total Anodonta imbecilis - - 1 - 1 Quadrula apiculata speciosa - - 1 - 1 Uniomerus tetralasmus 2 - 2 - 4 Cyrtonaias tampicoensis - - 1 - 1 Lampsilis teres - - 1 - 1 Toxolasma texasensis - - 3 - 3 Corbicula fluminea - - 2 - 2 Sites with bivalves 2 0 3 0 5 Total sites 11 10 6 4 31 Total species 1 0 7 0 7 The shell resembles those of Q. a. speciosa (Lea, 1862) from the Nueces River and the Rio Grande. Uniomerus tetralasmus (Say, 1831). — This species is the most widely occurring and abundant clam in the Baffin Bay drainage area (localities 6, 8, 11, 18). Creek specimens are only moderate-sized, reaching a shell length of 78.9 mm. Lake specimens reach a greater length (up to 121.2 mm) and develop a rounded ventro-posterior tip as the posterior ridge becomes noticeable. Nacre color is white. The above characters indicate tetralasmus as the proper assignment according to analyses by Frierson (1903) and Morrison (1977). Both of these authors concluded that Uniomerus declivus (Say, 1831) and U. tetralasmus are valid species whereas Johnson (1970) synonymized the two taxa. Cyrtonaias tampicoensis (Lea, 1838). — Shells of this species were found only in Lake Alice (locality 12), an artificial impoundment in the San Fernando Creek system. Several individual valves and one articulated pair were found in one area three meters in diameter. The intense localization of shells and absence of small individuals suggest strongly that these represent animals imported for use as fish bait. A likely source is the Nueces River (especially Lake Corpus Christi) to the north. No evidence of an established breeding population was uncovered. The recovered shells formerly were referable to C. berlandieri (Lea, 1857), which is either synonymous with, or a subspecies of, C. tampicoensis. Lampsilis teres (Rafinesque, 1820). — Individuals of this species from a single site (locality 6) are rather thin-shelled and have green rays. No infraspecific epithet is deemed appropriate at this time because the green- rayed character may be a unifactorial genetic polymorph. For many decades 180 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 this species was referred to as L. anodontoides (Lea, 1831), but Johnson (1972) pointed out that teres has priority. The rejection of teres by Pilsbry (in Ortmann and Walker, 1922) as “not identifiable” was due to “specious reasons” according to Johnson (1972). Toxolasma texasensis (Lea, 1857). — For many years this taxon was referred to the genus Carunculina Simpson, 1898. Recently, Valentine and Stansbery (1971) utilized Toxolasma Rafinesque, 1831, because it is an older name (Morrison, 1969). Shell characters discussed by Valentine and Stansbery (1971) indicate that the species present in the Baffin Bay area is texasensis. Shells are moderate in size (length to 49.6 mm) with a peach- colored nacre. Assignment of a subspecific name would be premature at this time. Names available from southern Texas (all type localities in the Rio Grande drainage) include: Unio bair dianus Lea, 1857; Lampsilis texasensis compressus Simpson, 1900; and Lampsilis mearnsi Simpson, 1900. Corbicula fluminea (Muller, 1774). — The Asiatic clam was found only in two lake environments (Lake Alice and Dairy Pond). Individuals of this species were not common at either of these localities; most shells were from 35-40 mm in length. While the restricted size also could indicate a nonbreeding introduction, a more likely explanation involves the rarity of successful spawning. Shells found in the study area have a pale purple nacre and pale honey-brown periostracum. All Corbicula from the Baffin Bay drainages are referable to the “white form” of Hillis and Patton (1982). Discussion Field surveys have revealed an isolated, low-density bivalve fauna of low species diversity inhabiting the creeks and ponds (stock tanks) of the Baffin Bay system. Low precipitation and a high evaporation rate produce intermittent streams in areas of soft sediments. Analysis of the collecting sites as to habitat type reveal that all species were found in freshwater ponds; only one, U. tetralasmus , was found in freshwater creek habitats (Table 1). This distributional pattern indicates the likelihood that all mussel species recorded from this area are introduced except for U. tetralasmus , a species that is rare in both the Nueces and Rio Grande systems. Introduction of the other species probably occurred via introduced fish stock. Although U. tetralasmus is well adapted to periodic desiccation, murky water, and soft substrates, this species was conspicuously absent from Lake Alice (locality 12). In fact, Lake Alice was the only site with mussels that did not support a population of U. tetralasmus. This species, however, does not appear to adapt well to large bodies of permanent water (Murray, 1979). BIVALVES OF BAFFIN BAY, TEXAS 181 The portion of southern Texas that is drained by streams flowing into Baffin Bay can be divided into three broad zones in reference to general suitability of the environment for freshwater bivalves. A coastal zone (inland to an approximate elevation of 28 meters above mean sea level) is characterized by salinity levels sufficient to cause soil precipitate. Freshwater bivalves might be able to live there for short periods of time in temporary freshwater ponds following heavy rainfall. A middle zone (between 16 and 28 meters above mean sea level) exhibits salinity levels that do not result insoil precipitate, but streams in this zone have frequent periods of no flow. An inland zone (generally above about 28 meters mean sea level) is characterized by tolerable salinity levels, although water is still rather scarce. This inland zone also contains large stock tanks that support populations of the locally introduced unionid species and the Asiatic clam. The inland zone also may experience lower rates of siltation due to slightly higher elevation (yielding more rapid water movement) and greater percentage of pasture land as opposed to cultivated land (yielding less silt because of vegetation cover). Acknowledgments I thank Harold D. Murray for first introducing me to the fauna of this area and T. B. Samsel, III, for drafting the map. Literature Cited Bollaert, W. 1850. Observations on the geography of Texas. J. Royal Geograph. Soc., 20:113-135. Breuer, J. P. 1957. An ecological survey of Baffin and Alaz.an bays, Texas. Publ. Inst. Mar. Sci., Univ. Texas, 4:134-155. Frierson, L. S. 1903. The specific valve of Unio declivus. Say. The Nautilus, 17:49-51. Hillis, D. M., and J. C. Patton. 1982. Morphological and electrophoretic evidence for two species of Corbicula (Bivalvia: Corbicultidae) in North America. Amer. Midland Nat., 108:74-80. Johnson, R. I. 1970. The systematics and zoogeography of the Unionidae (Mollusca: Bivalvia) of the southern Atlantic Slope region. Bull. Mus. Comp. Zool., 140:263-449. - . 1972. The Unionidae (Mollusca: Bivalvia) of peninsular Florida. Bull. Florida State Mus., Biol. Sci., 16:181-249. Morrison, J. P. E. 1969. The earliest names for North American naiads. Ann. Rpt. Amer. Malacol. Union, 1969:22-24. - . 1977. Species of the genus Uniomerus. Bull. Amer. Malacol. Union, 42:10-11. Murray, H. D. 1979. Freshwater mussels of Lake Corpus Christi, Texas. Bull. Amer. Malacol. Union, 44:5-6. Ortmann, A. E., and B. Walker. 1922. On the nomenclature of certain North American naiades. Occas. Papers Mus. Zool., Univ. Michigan, 112:1-75. Price, W. A., and G. Gunter. 1942. Certain recent geological and biological changes in South Texas, with consideration of probable causes. Proc. Trans. Texas Acad. Sci., 26:138-156. Russell, J. L., and C. E. Wood. 1976. The effects of Tropical Storm Fern (September, 1971) on Baffin Bay, Texas. Texas A&I Univ. Studies, 9:133-145. 182 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Simmons, E. G. 1957. An ecological survey of the Upper Laguna Madre of Texas. Publ. Inst. Mar. Sci., Univ. Texas, 4:156-200. Strecker, J. K. 1931. The distribution of the naiades or pearly fresh-water mussels of Texas. Baylor Univ. Spec. Bull., 2:1-71. Suhm, R. W. 1974. Shoreline and beach sediment characteristics of western Baffin Bay, Texas. Texas A&I Univ. Studies, 7:3-27. - . 1978. Preliminary investigation of the La Paloma Mammoth Site (late Pleistocene), Kenedy County, Texas. Texas A&I Univer. Studies, 11:13-36. Valentine, B. D., and D. H. Stansbery. 1971. An introduction to the naiads of the Lake Texoma region, Oklahoma, with notes on the Red River fauna (Mollusca: Unionidae). Sterkiana, 42:1-40. EFFECT OF SALT ON GRAIN AND FORAGE INTAKE IN CATTLE Robert D. Brown and Alejandro Creixell Caesar Kleberg Wildlife Research Institute and College of Agriculture and Home Economics, Texas A&I University, Kingsville, Texas 78363 Abstract. — Grain rations containing 0, 20, 30, 40, or 50 percent salt were offered to 15 steers receiving chopped coastal Bermuda grass ( Cynodon dactylori) or Kleberg bluestem ( Dichanthium annulatum ) hay ad libitum. Grain intake (DMI/BW0 75kg) decreased (/><.01) with the addition of salt, but there was no difference ( P >.05) in grain intake between salt levels. The overall salt intake was 0.189 percent (SE = 0.016) of body weights. The effect of salt on grain intake for cattle receiving coastal Bermuda grass hay was predicted by the equation DMIg/ BW0 75kg = 0. 148 - 0.060 (% NaCl) + 0.007 (% NaCl)2, R2 = 94.98%; whereas that of cattle receiving Kleberg bluestem hay was predicted by the equation DMIg/BW0 75kg = 0.151 — 0.053 (% NaCl) + 0.005 (% NaCl)2, R2 = 92.4%. Intake of coastal Bermuda grass was higher (/><.01) than that of Kleberg bluestem when hays were fed alone. Cattle consuming grain ate less ( P <.05) hay than those not receiving grain. There was no difference ( P >.05) in hay intake (DMI/BW0 75kg) between cattle receiving any grain ration. Cattle overconsumed grain when grain was not salted. At salt levels above 20 percent, intake of the hay-grain mixes was inadequate to meet requirements. It is thus recommended that salt should not be added to grain rations above 20 percent to limit grain intake in beef cattle. Key words : salt; livestock; supplement; feed; digestibility. In areas where forages tend to be of low quality, the use of supplemental feeds for beef cattle is a necessary practice. However, supplements commonly are expensive, and the daily feeding of range cubes or grain mixes is labor intensive. Sodium chloride (NaCl) frequently has been used to regulate the intake of supplements provided free-choice to cattle on rangelands (Weir and Torrel, 1953; Riggs et al., 1953; Cardon, 1953). The advantages of feeding salted grains to cattle are that overconsumption can be controlled, supplement for several days can be placed in self-feeders, and labor can be reduced (Cunha, 1980). Unfortunately, salt also can be toxic if consumed in excessive amounts. Surprisingly little research has been done on the specific percentages of salt required to reduce grain intake in beef cattle (Morris, 1980). The objectives of ths investigation were: 1) to determine the consumption of concentrates and forages by beef cattle at various levels of NaCl for two forages; and 2) to generate a regression equation for each forage to predict grain intake at different salt levels. Materials and Methods The experimental facility consisted of an open pen, 30 meters long by 12 meters wide, with a three-sided shed at the north end housing 16 electronic feeding gates (Calan Broadbent Inc., Northwood, New Hampshire 03261). The gates allowed each steer access to its own The Texas Journal of Science, Vol. 39, No. 2, May, 1987 184 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 feed bunk and thus the collection of individual data on intake. At the south end of the pen, there was a self-filling water tank to which the animals were allowed free access. Fifteen Santa Gertrudis steers with an average beginning weight of 285 kilograms were used. The animals were dewormed, weighed, and ear-tagged before the start of the experiments. The cattle were fed either baled coastal Bermuda grass ( Cynodon dactylon ) or baled Kleberg bluestem ( Dichanthium annulatum). Both forages were chopped to a length of three centimeters before feeding to reduce animal waste. Before the start of the intake trials, in vivo digestion trials were conducted on both forages. Six steers were used in each trial. Seven-day adjustment periods were followed by seven- day periods of data collection. Feces were collected by means of fecal bags, and aliquots of feces and feed were analyzed to determine digestibility (Schneider and Flatt, 1975: table 1). For the intake trials, five grain rations containing 0, 20, 30, 40, or 50 percent NaCl were prepared for each forage (Table 2). The rations, balanced by the Least Cost Ration Balancing Program at the Texas A&I University Computer Center, were designed to meet the requirements of 273-kilograms, growing-finishing steers gaining 0.45 kilograms per day (NRC 1976). Therefore, the steers required 16.2 Mcal/day digestible energy (DE), 0.34 kg/day digestible protein (DP), and a calcium (Ca) to phosphorus (P) ratio of 1.1:1. Initial analysis of duplicate samples of the coastal Bermuda grass and Kleberg bluestem hays showed them to contain 8.5 and 4.0 percent crude protein (CP) and 49.3 and 46.0 percent total digestible nutrients (TDN), respectively. The latter values were calculated after chemical determination of neutral detergent fiber (NDF). The computer program balanced the grain rations by estimating the forage intake, subtracting from the steers’ requirements the nutrients supplied by forage, and determining the grain mix required to supply the deficient nutrients. To calculate the salt-containing grain mixes, the program substituted salt for a portion of the corn and cottonseed meal (CSM). Protein and energy contents of the grain mixes declined proportionately as the salt content in the mixes increased. Each of the four intake trials consisted of a 10-day adjustment period followed by a seven- day period of data collection. The 15 steers were randomly divided into five groups of three steers each. During the first two trials, chopped coastal Bermuda grass was offered to all animals ad libitum at 0800 daily. At 1700, feed bunks were cleaned, and hay orts were collected and weighed. The five grain mixes of Ration 1 were offered to the five groups of cattle at 1700 daily, and the grain orts were collected at 0800. The steers were re-randomized at the end of each trial. During the second two trials, chopped Kleberg bluestem was substituted as the forage, as well as the five grain mixes of Ration 2. The hay, grain, and fecal samples were analyzed for dry matter (DM) according to AOAC (1970) methods. Crude protein was determined by the micro Kjeldahl procedure (AOAC 1970; Hertel, 1975). Acid detergent fiber (ADF), cellulose (CELL), and lignin (LIG) were determined by the procedure outlined by Goering and Van Soest (1970). Sodium (Na) determinations were performed by flame photometry at the Texas Agricultural Extension Service Soil Testing Laboratory, Texas A&M University, College Station (AOAC, 1970). Chloride (Cl) determinations were performed as outlined by Harris (1970). Gross energy was determined using the Parr adiabatic bomb calorimeter (Parr Instrument Company, 1975). The water offered to the cattle was analyzed by flame photometry and found to contain 315 parts per million of Na and 253 of Cl. A completely randomized two-factor factorial design was employed to investigate differences in the effects of salted grain mixes and forages. The model used was: Yyk = U + Si + Fj + (SF)ij + Eijk, where Yijk — dependent variable, U = true overall mean, i = 1, 2, 3, 4, 5 salt levels, j = 1 , 2 forages, and k = 1, 2, 3 animals. EFFECT OF SALT ON CATTLE FEED 185 Table 1. Chemical composition and digestibility of forages. Feed analysis' Coastal Bermuda grass2 DC3 Kelberg bluestem2 DC3 Dry matter, % 93.9 0.52 94.2 0.56 Crude protein, % 5.7 0.51 3.3 0.33 Acid detergent fiber, % 44.8 0.50 57.1 0.55 Cellulose, % 31.7 0.63 37.0 0.67 Lignin, % 7.1 0.36 7.7 0.30 Gross energy, Meal/ kg 4.2 0.53 4.1 0.60 Crude fiber, % 33.0 37.0 Sodium, % 0.05 0.02 Chlorine, % 0.53 0.49 'Dry matter basis. 2Values are means of six observations of a seven-day collection period. 3 Digestion coefficient. The above model was used to analyze the following dependent variables: HDMI . . . Hay intake (DMIg/BW075kg); GDMI . . . grain intake (DMIg/ BW0 75kg). In order to predict GDMI at different levels of salt, two regression equations were generated using GDMI as the dependent variable. In both equations a second term (quadratic) was added on the basis of its ability to provide significantly more predictive power than the single term. To obtain a quantitative measure of how well the second term predicted the dependent variable, the square multiple correlation coefficient (R2) was obtained. A highly significant improvement (R<.01) in the percentage change of R2 was obtained by the quadratic term. Addition of a third term (cubic) to the equation was not significant ( P >.05). Duncan’s Multiple Range Test was employed to compare means. Results and Discussion Digestibilities of coastal Bermuda grass and Kleberg bluestem when fed alone are presented in Table 1. The digestibility of DM, ADF, CELL, and energy was higher for Kleberg bluestem. The CP digestibility, on the other hand, was higher for coastal Bermuda grass. Thus the quality of the two hays, as measured by their digestible nutrients, was much closer than originally estimated. The digestion coefficient for lignin was found to be 0.36 in coastal Bermuda grass and 0.30 in Kleberg bluestem. This may have been attributed to the procedure used (Potassium permanganate), which may not have accounted for all of the fecal lignin (Wallace and Van Dyne, 1970). Grain DMI/BW0 75 of the salt-free rations was higher (P<.01) than that of the salted grain mixes (Table 3). While there was a tendency for grain intake to decline with increasing salt concentrations, the differences in GDMI were not significant between the salted grain mixes. When coastal Bermuda grass was offered as the forage, GDMI was lower at each level of salt than it was for Kleberg bluestem, although not significantly (Table 3). Previous research suggests that the percentage of salt necessary to regulate intake of supplements depends on such factors as age, weight, 186 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Table 2. Composition of salted grain rations.1 Ingredients Salt, % Ration l3 0.0 20.0 30.0 40.0 50.0 Corn, % 99.0 79.2 69.3 59.4 49.5 Limestone, % Nutrients2 1.0 0.8 0.7 0.6 0.5 DP, % 7.4 5.9 5.2 4.5 3.7 DE, Meal/ kg 3.9 3.1 2.7 2.3 1.9 Ingredients Salt, % Ration 24 0.0 20.0 30.0 40.0 50.0 Corn, % 82.7 66.2 57.9 49.6 41.4 Cottonseed meal, % 16.0 12.8 11.2 9.6 8.0 Limestone, % Nutrients 2 1.3 1.0 0.9 0.8 0.6 DP, % 12.0 9.6 8.4 7.2 6.0 DE, Meal/ kg 3.8 3.0 2.6 2.3 1.9 'Rations balanced to meet NRC (1976) requirements for 273-kg steers gaining 0.46 kg/ day. 2CP and DE were calculated according to NRC (1976). 3Fed to steers receiving coastal Bermuda grass. 4Fed to steers receiving Kleberg bluestem. and type of animal (Riggs et al., 1953). Rich et al. (1976) reported the requirement of salt in cattle to be roughly 28 grams per head per day or 0.1 pound salt per 100 pounds body weight. Fowler (1969), Ensminger (1976), and Newmann (1977) all reported a standard of 25 percent salt in protein rations to restrict intake to two pounds a day. Ensminger (1976) also stated that 33 1/3 percent would result in an intake of 1.5 pounds per day, or “a reduction in salt level of 33 1/3 percent to 24 percent will increase consumption 50 percent.” In order to predict GDMI at predetermined levels of salt, two equations were generated. A regression analysis using GDMI as the dependent variable generated the following equations for coastal Bermuda grass and Kleberg bluestem respectively: 1) GDMIg/BW0 75kg = 0.148 — 0.060 (% NaCl) + 0.007 (% NaCl)2, R2 = 94.9%; 2) GDMIg/BW075kg = 0.151 - 0.053 (% NaCL) + 0.005 (% NaCl)2, R2 = 92.4%. Average dry matter consumption of coastal Bermuda grass per (BW°'75kg) was significantly higher than the consumption of Kleberg bluestem (Table 3). Coastal Bermuda grass and Kleberg bluestem dry matter consumption (5.95 and 5.20 kilograms per head per day, respectively) was significantly higher when fed alone in the preliminary digestion trials than when grain (salt excluded) was fed in the feeding trials (4.92 and 3.73 kilograms per head per day) (Creixell, 1982). Coastal Bermuda grass consumption (kg/BW°'75kg) increased slightly but not significantly until the grain mixture contained 30 percent salt, when hay intake leveled off, then declined with the addition of 50 percent salt EFFECT OF SALT ON CATTLE FEED 187 Table 3. Consumption of coastal Bermuda grass (CB), Kleberg bluestem (KBS) and salted grain mixes fed ad libitum1. Grain intake Hay intake Salt intake Salt intake % NaCl Forage DMI/BW0 75kg SE DMI/BW0 75kg SE g/ head/ day % BW SE 0 CB 0.68a 0.019 0.73a 0.011 0 0 0 KBS 0.70a 0.001 0.53b 0.004 0 0 0 20 CB 0.2 1 b 0.007 0.75a 0.003 330 0.230a .018 KBS 0.26b 0.020 0.61b 0.009 400 0.21T .054 30 CB 0.1 4b 0.003 0.78a 0.001 340 0.2283 .012 KBS 0.21 b 0.005 0.58b 0.007 490 0.343a .028 40 CB 0.08b 0.002 0.78a 0.005 250 0.183ab .026 KBS 0.1 4b 0.002 0.59b 0.013 400 0.288ab .027 50 CB 0.06b 0.002 0.66a 0.011 230 0. 1 5 lb .016 KBS 0.07b 0.004 0.61b 0.006 260 0.192b .023 'Values are means of seven-day observations on six steers. Numbers in columns with different superscripts are significantly different (.P<0.05). (Table 3). Kleberg bluestem consumption (kg/BW°'75kg) was non significantly higher for all the salt levels than for the zero percent salt- grain mix. Analysis of variance showed the percentage salt did not affect HDMI significantly ( P >.05) compared with the controls, and that there was no significant interaction (P >.05) between the salt levels and the hay quality. Thus, as salt was added to the grain mix, the hay digestible energy (DE) intake (Meal/ day) changed little, whereas the grain DE intake decreased. Analysis of variance showed no significant interaction between the two forages and five grain mixes in salt intake per unit body weight. (Table 3). Salt intake as a percent of the body weight was greatest at 20 percent (CB) or 30 percent (KBS) salt and least at 50 percent salt in the grain mix for both forages. This decline in salt intake per unit body weight at the 50 percent salt level could signify a palatability problem. The overall salt intake per unit body weight for both forages over all grain mixes was 0.189 percent (±0.016). This finding is in general agreement with Corah (1980), who suggested that cattle have a salt requirement of 0.1 percent of their body weight. Although the highest salt intake was recorded at the 30 percent level (490 grams), no serious effects were noted in any of the groups of steers fed the grain-salt ration. Scouring, although never serious, occurred in two steers of the salt-fed groups. There was some decline in the condition of the hair and fleshing at the close of the experiments because the nutrient content of the forages and grain was enough only to maintain the steers at a constant weight. 188 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Summary and Conclusions Average daily consumption of coastal Bermuda grass (5.96 kg) and Kleberg bluestem (5.20 kg) when fed alone did not provide adequate levels of DP (0.17 and 0.05 kg) or DE (13.01 and 12.81 Meal/ day) for the steers to meet their requirements. Thus, the addition of supplemental feed was necessary. When supplemental feed was included in the diet, cattle in the control group consumed enough of the hay-grain mix to obtain adequate DE (estimated 28.57 and 27.67 Meal/ day) for the two forages, respectively. These figures, when compared with the requirements of DE (16.20 Meal/ day) indicate, however, that cattle consumed more grain than necessary in the control group. On the other hand, when salt was added to the grain mix, grain intake was restricted such that cattle failed to consume adequate total DE. Because 20 percent salt was the lowest salt level considered in this study, it appears that the level of grain intake needed for a balanced diet when either coastal Bermuda grass or Kleberg bluestem is offered as the forage, would be between 0 and 20 percent salt. Acknowledgments The authors would like to thank Mauricio Pacheco and Andres Garza, Jr., for assistance with animal handling and Dr. Ralph Bingham for advise on statistical analyses. We are indebted to the Houston Livestock Show and Rodeo for their financial support of this project and to King Ranch, Inc., for the loan of experimental animals. Literature Cited A.O.A.C. 1970. Official methods of analysis. Assoc. Off. Agric. Chemists, Washington, D.C., 10th ed., 1015 pp. Cardon, B.P. 1953. Influence of a high salt intake on cellulose digestion. J. Anim. Sci., 12:536-540. Creixeli, A. 1982. Effect of sodium chloride in the regulation of grain intake in beef cattle. M.S. thesis, Texas A&I Univ., Kingsville, 55 pp. Cunha, T. J. 1980. Salt. Animal Nutrition and Health, 14 pp. Ensminger, M. E. 1976. Beef cattle science. The Interstate Publ. Co., Danville, Illinois, 1556 pp. Fowler, S. H. 1969. Beef production in the south. The Interstate Publ. Co., Danville, Illinois, 858 pp. Goering, H. K., and P. J. Van Soest. 1970. Forage fiber analysis (apparatus, reagents, procedures, and some applications). Agric. Handbook, USDS-ARS, Beltsville, Maryland, 379:1-20. Harris, L. E. 1970. Nutrition research techniques for domestic and wild animals. Utah State Univ., Logan, Vol. 1, 163 pp. Hertel, J. M. 1975. Laboratory determination of protein. Texas A&I Univ., College of Agriculture, Kingsville (mimeo), 8 pp. Morris, J. G. 1980. Assessment of sodium requirements of grazing beef cattle. J. Anim. Sci., 50:145-152. National Research Council. 1976. Nutrient requirements of domestic animals, No. 4. Nutrient requirements of beef cattle. Fifth revised ed. National Academy Sciences, National Resarch Council, Washington, D.C., 56 pp. EFFECT OF SALT ON CATTLE FEED 189 Neumann, A. L. 1977. Beef cattle. John Wiley & Sons, New York, 7th ed., 883 pp. Parr Instrument Company. 1975. Instructions for the 1241 and 1242 adiabatic calorimeters. Parr Instrument Co., Moline, Illinois, Manual no. 153, 23 pp. Rich, T. D., S. Armbruster, and D. R. Gill. 1976. Great Plains beef cattle handbook factsheet no. 1950, Oklahoma State Univ., Stillwater. Quoted in: L.H. Corah, 1980. Feeding and nutrition of the beef breeding herd. Pp. 91-105, in D.C. Church, (ed)., Digestive physiology and nutrition of ruminants, vol. 3-Practical nutrition, O&B Books, Corvallis, Oregon, 416 pp. Riggs, J. K., R. W. Colby, and K. V. Sells. 1953. The effect of self-feeding salt cottonseed meal mixtures to beef cows. J. Anim. Sci., 12:379-393. Schneider, B. J., and W. P. Flatt. 1975. The evaluation of feeds through digestibility experiments. Univ. Georgia Press, Athens, 423 pp. Wallace, J. D., and G. M. Van Dyne. 1970. Precision of indirect methods for estimating digestibility of forage consumed by grazing cattle. J. Range Manag., 23:424-430. Weir, W. C., and D. T. Torrell. 1953. Salt-cottonseed meal mixture as a supplement for breeding ewes on the range. J. Anim. Sci., 12:353-358. GENERAL NOTES CALORIC CONTENT OF AN EXCAVATED FOOD CACHE OF PEROGNATHUS FLAVESCENS Kent M. Reed Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas, 77843 The climate of the Great Plains region is unpredictable and, in winter, may include prolonged subzero temperatures. Accordingly, small mammals that occur there, with their high ratio of surface area to body volume, must have mechanisms by which to cope with periods when foraging above ground is disadventageous energetically. One of the smaller mammals on the central and northern Great Plains (Jones et al., 1983), the plains pocket mouse ( Perognathus flavescens) is not known to hibernate and does not accumulate body fat to use as a source of energy in winter (Hibbard and Beer, 1960). Instead, it has a propensity to cache seeds within its underground burrows for use in winter. Several food caches of P flavescens have been examined (Bailey, 1929; Hibbard and Beer, 1960; Reed and Choate, 1986), but none has been subjected to caloric analysis. A food cache of a plains pocket mouse was uncovered on 20 July 1985 (Reed and Choate, 1986), affording the opportunity to obtain preliminary data on energy storage in this species. Seeds in the food cache were sorted by phena. Samples were homogenized using a CRC micro-mill and weighed to the nearest 0.0001 of a gram using a Mettler analytical balance. Three of the seed types collected ( Croton texensis, Lithosperma incisum , and unidentified Compositae) were sufficiently numerous to be analyzed separately. The remaining seeds were combined into one sample for analysis. Because of the high oil content of certain seeds, samples (excepting that consisting of the combined seeds) were not pressed into pellets but were packed loosely in the crucibles for burning. Gross caloric content then was determined using a Parr adiabatic calorimeter standardized with benzoic acid. The results of the caloric analysis are sumarized in Table 1. All samples, excepting those of L. incisum , had energy values greater than 4000 calories per gram. Two seed types, C. texensis and unidentified Compositae, had energy values greater than 5800 calories per gram. Caloric content for the samples averaged 5068.3 calories per gram. Gross caloric content for the cache totaled 356,335 calories. Although no values of basal metabolic rate for P. flavescens are available, estimates for a similar-sized, related species (P. flavus) were reported by Wolff and Bateman (1978). They found that P. flavus (mean body weight, 8.8 grams) consumed from 0.193 to 0.320 grams Table 1. Weight, caloric content, total calories, and percent biomass of seeds removed from an excavated cache. Grams Calories per gram Calories Percent biomass Croton texensis 47.30 5829.08 275,715.48 67.3 Lithosperma incisum 18.44 3141.90 57,936.64 26.2 Compositeae 1.78 5879.86 10,466.15 2.5 Other 2.79 4378.87 12,217.05 4.0 Total 70.31 356,335.35 100.0 The Texas Journal of Science, Vol. 39, No. 2, May, 1987 192 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 of food per day per gram body weight (mean caloric value, 4207 calories per gram) at temperatures ranging from 15° to 1°C. This pocket mouse thus consumes 7145 to 11,847 calories per day without entering torpor. Using these values, P. flavescens (weight, 7-12 grams according to Jones et al., 1986) would require from 7713 to 12,789 calories per day under similar conditions. At this rate of energy consumption, assuming no period of torpor, the cache would have lasted approximately 35 days. Presumably, the mouse that accumulated the food cache would have added to it throughout summer and into autumn. Perognathus flavescens is not known to be active above ground in winter (Hibbard and Beer, 1960; Jones et al., 1986), and above-ground activity is thought to be temperature regulated; in Kansas, for example, P. flavescens was not found above ground for four months from mid-November to mid-March (Reed and Choate, 1986). Apparently, frequency and duration of torpor in P. flavus, are related to temperature and availability of food (Wolff and Bateman, 1978). By reducing its energy requirement by entering torpor periodically, P. flavescens could ration its stores and thereby avoid the need for foraging above ground during periods of extremely cold temperatures. I thank D. A. Knabe, Department of Animal Science, Texas A&M University, for use of laboratory facilities, Ed Gregg for technical assistance in the analyses, and J. R. Choate and D. J. Schmidly for critically reviewing a preliminary draft of the manuscript. Literature Cited Bailey, V. 1929. Mammals of Sherburne County, Minnesota. J. Mamm., 10:153-164. Hibbard, E. A., and J. R. Beer. 1960. The plains pocket mouse in Minnesota. Flicker, 32:89-94. Jones, J. K., Jr., D. M. Armstrong, R. S. Hoffmann, and C. Jones. 1983. Mammals of the northern Great Plains. Univ. Nebraska Press, Lincoln, xii + 379 pp. Jones, J. K., Jr., D. M. Armstrong, and J. R. Choate. 1986. Guide to mammals of the plains states. Univ. Nebraska Press, Lincoln, xvii + 371 pp. Reed, K. M., and J. R. Choate. 1986. Natural history of the plains pocket mouse in agriculturally disturbed sandsage prairie. Prairie Nat., 18:79-90. Wolff, J. O., and G. C. Bateman. 1978. Effects of food availability and ambient temperature on torpor cycles of Perognathus flavus (Heteromyidae). J. Mamm., 59:707-716. AN UNUSUAL “WOODRAT NEST” Walter W. Dalquest and Greg J. Coln Midwestern State University, Wichita Falls, Texas 76308 and Box 11, Eagle Nest, Creede, Colorado 81130 On 9 February 1986, the junior author discovered a bird’s nest composed mostly of the mummified body of an adult woodrat, Neotoma mexicana. The site was on a ridge or shoulder a mile south of the summit of Bristol Head Mountain, approximately 9600 feet elevation, 11.5 miles south of Creede, Mineral County, Colorado. The nest was in a tall wax current ( Ribes ccreum Dougl.) bush, 1.5 meters from the ground, and 6 meters out from the base of a tall (50 meter) cliff. The ground beyond the bush sloped swiftly to a rocky slide area. An active woodrat nest was located less than 100 meters away. The mummified carcass was well preserved, except that the fur of the head and part of the chest was missing. The body was curled gently. The bird, approximately the size of a junco to judge from the size of the nest cup, had added bits of grass and vegetation above the mummy, and tucked other bits beneath it (Fig. 1). There was no evidence to indicate The Texas Journal of Science, Vol. 39, No. 2, May, 1987 GENERAL NOTES 193 Figure 1. Mummified carcass of woodrat utilized as nest by small bird. that eggs actually were incubated, but vegetation was packed and the nest obviously had been used. We speculate that in the previous winter an owl had captured and killed the woodrat, but dropped it in flight. The dense bush prevented retrieval and the cold weather at that high elevation prevented attack of the carcass by insects. The woodrat body was freeze-dried by spring, forming a soft platform for the small bird to use for its nest. We are grateful to Gregory Pogue for identifying the bush and to Dr. Norman Horner for the photograph of the nest. DISTRIBUTIONAL RECORD OF LASIURUS SEMINOLUS (CHIROPTERA: VESPERTILIONIDAE) Thomas E. Lee, Jr. Department of Biology, Angelo State University, San Angelo, Texas 76909 In April 1985, Mr. C. Harrison found a Seminole bat, Lasiurus seminolus, caught by a wing on a barbed-wire fence, 6 mi. N Clay, Burleson Co., Texas. The area where the bat was found consists mainly of a large cotton field with oak-hickory forest (sometimes favored by L. seminolus ) located nearby. This specimen (ASNHC 1984) represents the first record of this species taken west of the Brazos River (Schmidly, 1983; although see Hall, 1981, for discussion of an extralimital specimen taken near Brownsville, Cameron Co., Texas) and extends the known distribution of the Seminole bat approximately 32 kilometers to the west. Literature Cited Hall, E. R. 1981. The mammals of North America. John Wiley & Sons, New York, l:xv + 1-600 + 90. Schmidly, D. J. 1983. Texas mammals east of the Balcones Fault Zone. Texas A&M Univ. Press, College Station, xviii + 400 pp. The Texas Journal of Science, Vol. 39, No. 2, May, 1987 194 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 NOTOMMATA ALLANTOIS IN NORTHEASTERN TEXAS LAKES: A ROTIFER PREVIOUSLY KNOWN ONLY FROM EUROPE John Shoemaker, Sally H. Davis, and Robert K. Williams Department of Biological Sciences, East Texas State University, Commerce, Texas 75428 A comprehensive study of rotifers in Lake Fork, Wood Co., Texas, has been conducted since the autumn of 1982. Lake Fork is a water-supply and recreational lake administered by the Sabine River Authority and covers nearly 28,000 acres (about 11,330 hectares). The reservoir drains a watershed of sandy pastures, dairies, farms, and oak forests. During the study period from autumn 1982 to summer 1985, the pH of the water averaged 6.9 from measurements made by the authors. Occasional algae blooms resulted in slightly alkaline measurements. Specimens of the rare rotifer, Ptygura elsteri Koste, 1972, were collected from this lake in October and November, 1984, and in January and March, 1985 (Shoemaker and Williams, 1986). The authors have identified 157 species of rotifers from Lake Fork collections during these studies. Specimens, which were identified to the genus Notommata Ehrenberg, 1830, were found in samples collected on 26 November 1984, 5 March 1985, and 12 March 1985. The species could not be determined by use of keys, illustrations, or descriptions of several authors (Harring and Myers, 1922; Taft, 1932; Edmondson, 1959; Stemberger, 1979; and others). The organism varied greatly from N. silpha (Gosse, 1887) and N. aurita (Muller, 1786), which had been identified previously from Lake Fork collections. A striking feature of the organism was a broad posterior projection of the body wall forming a dorsal covering or hood over the proximal segments of the foot, with only the terminal segment and toes projecting beyond the hood (Fig. 1). The organism resembled N. collaris Ehrenberg, 1832, and N. pachyura (Gosse, 1886) in this regard. These two species (Fig. 1) have the posterior projection of the body wall but neither is as wide as the Lake Fork specimen and N. collaris has a shorter covering. In general appearance, the Lake Fork specimens were more rounded and broad of body than either N. pachyura or N. collaris. The formula key to the Notommata in Voigt and Koste (1978) was used to identify the Lake Fork specimens as members of the collaris-pachyura-allantois association (=formenkreis). Use of the key to this association led to the conclusion that the Lake Fork specimens were members of N. allantois Wulfert, 1935. According to the formula key and description of Voigt and Koste (1978) the anal appendage is wide, ligulate, and without a Figure 1. The collaris-pachyura-allantois association of Notommata. Dorsal: a, N. allantois (from Lake Fork, Wood Co.; Texas); b, N. collaris (redrawn from Voigt and Koste, 1978); c, N. pachyura (from Lake Wright Patman, Bowie Co., Texas). The Texas Journal of Science, Vol. 39, No. 2, May, 1987 GENERAL NOTES 195 Figure 2. Notommata allantois from Lake Fork, Wood Co., Texas, profiles and structures: a, lateral, foot extended; b, foot withdrawn; c, dorsal, corona partially withdrawn; d, posterior structures showing bilobed vesicle; e, asymmetrical trophi. posterior projecting appendage. There is no caudal process on the foot. The toes are greater than 40 micrometers in length and are pointed. The corona auricles are large and ciliated. There is an asymmetrical trophi, the retrocerebral sac extends above the posterior third of the body, and the bladder is two-parted (Fig. 2). Specific differences among N. allantois, N. collaris, and N. pachyura are found in the foot and toes (Fig. 1). In N. allantois , the toes are pointed, greater than 40 micrometers in length, project outward, and are not segmented. There is no projection on the terminal foot segment. In N. collaris , the toes are less than 35 micrometers, conical, and straight. In N. pachyura, the toes ae greater than 40 micrometers, pointed, and appear to be two- segmented. There are spurs on the last joint of the foot. Since the collection of N. allantois in Lake Fork in 1984 and 1985, specimens also have been found in abundance in collections from Lake O’ the Pines in Harrison Co., Texas, made in November 1985. Voigt and Koste (1978) stated that N. allantois was known only from Europe. With the discovery of this organism in Lake Fork and Lake O’ the Pines, we believe the distribution of this rotifer is poorly known only because of the lack of extensive studies in North America, other than in the Great Lakes region, and elsewhere. Edmondson (personal communication from W. T. Edmondson, Department of Zoology, University of Washington, Seattle, March, 1985) stated that due to the cosmopolitan nature of rotiferans, it would be expected that such intercontinental range extensions as this would occur as more investigations were conducted. This research was partially supported by an East Texas State University Faculty Research Grant for the illustrations by Sally H. Davis. 196 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Literature Cited Edmondson, W. T. 1959. Rotifera. Pp. 420-494, in Freshwater biology (W. T. Edmondson, ed.), Wiley, New York, 2nd ed., 1248 pp. Harring, H. K., and F. J. Myers. 1922. The rotifer fauna of Wisconsin. Trans. Wisconsin Acad. Sci., Arts, Litt., 20:553-662. Shoemaker, J. H., and R. K. Williams. 1986. Occurrence of the rare rotifer, Ptygura elsteri Koste, 1972, in Lake Fork, Wood County, Texas. Texas J. Sci., 38:55-58. Stemberger, R. S. 1979. A guide to the rotifers of the Laurentian Great Lakes. EPA- 600/4-79-021, U.S. Environmental Protection Agency, Cincinnati, Ohio, 186 pp. Taft, C. E. 1932. Oklahoma rotifers. Ohio J. Sci., 32:492-504. Voigt, M., and W. Koste. 1978. Rotatoria. Die Radertiere Mitteleuropas. Ein Bestimmungswerk, begrundet Max Voigt. Uberordnung Monogononta 2. Auflage neubearbeitet von Walter Koste. Gebriider Borntraeger, Berlin. I. Textband, 673 pp. II. Tafelband mit 234 Tafeln. TWO NOTEWORTHY POPULATIONS OF THE FIDDLER CRAB, UCA SUBCYLINDRICA , IN SOUTHERN TEXAS Raymond W. Neck Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, Texas 78744 At least seven species of fiddler crabs (Ocypodidae: Uca) are known from Texas; most populations are found in restricted habitats of the interidal and supratidal portions of coastal areas (Powers, 1977). One species of fiddler crab, Uca subcylindrica (Stimpson, 1859), is endemic to hypersaline, brackish, and even quasi-freshwater habitats of semiarid portions of southern Texas and northeastern Mexico (Hildebrand, 1958; Barnwell and Thurman, 1984; Thurman, 1984). This species is known as the puffed fiddler crab (Bright and Hogue, 1972). Uca subcylindrica is able to exist in a wide variety of habitats over a large part of southern Texas because of highly adaptable physiological capabilities. U. subcylindrica is able to withstand greater loss of body water than other species of Uca (Rabalais and Cameron, 1985a). Habitats with salinities above 90 parts per thousand can be successfully occupied by this species (Thurman, 1984; Rabalais and Cameron, 1985a). Larval development in U. subcylindrica is quite abbreviated and may require only eight days (Rabalais and Cameron, 1983); a typical nursery area is a temporary rain puddle (Rabalais and Cameron, 1983; 1985b). Despite the existence of a highly adaptable physiology, U. subcylindrica appears to be an inferior competitor against other Uca because multi-species fiddler crab communities generally do not contain U. subcylindrica (Thurman, 1984). Field work in southern Texas has revealed an inland locality for U. subcylindrica that is farther from tidal habitats than any other previously reported. Additionally, a human- made pond created in the early twentieth century has become sufficiently saline to support a population of the species. On 30 March 1981, I collected several adult U. subcylindrica in Brooks County, nine kilometers east-southeast of Falfurrias. Collection site was the FM 2191 crossing of Los Olmos Creek, which is a saline marsh dominated by sacahuiste ( Spartina spartinae). This general area is characterized by high levels of gypsum associated with the Gyp Hill Salt Dome; water was clear brown. This locality is approximately 25 kilometers west of the previously reported westernmost locality (site V on Los Olmos Creek in Thurman, 1984). Wetland areas are not continuous along the bed of intermittent Los Olmos Creek, but distances between neighboring wetland areas evidently can be traversed by the crab during moist periods. The Texas Journal of Science, Vol. 39, No. 2, May, 1987 GENERAL NOTES 197 On 25 September 1980, I collected a series of U. subcylindrica from a saline pond in Cameron County, 14 kilometers northwest of Brownsville. This pool, which is located just east of the eastern boundary of Resaca de la Palma State Park, was created in the early twentieth century. A portion of a bend in a resaca (abandoned channel of the Rio Grande) was separated from the rest of the resaca by an eathern dike. Aerial photographs taken on November 1938 revealed an established brush line on this dike that was probably constructed during the 1920’s during peak agricultural development of the Lower Rio Grande Valley (Foscue, 1934). Following the creation of this two-hectare pond, runoff accumulation and subsequent evaporation have slowly produced a saline environment. Salinity levels obviously vary with water volume but are sufficiently low to allow growth of water milfoil ( Myriophyllum sp.). The dominant terrestrial plant is sea ox-eye daisy ( Borrichia frutescens). Dispersal from naturally occurring saline areas (within 10 kilometers of Resaca de la Palma site) probably occurs during periods of high water level and high ambient relative humidity. The ability of U. subcylindrica to maintain populations in non tidal habitats far from coastal areas and to colonize newly-created saline habitats (also nontidal) results from an adaptable complement of physiological processes. This adaptability has allowed U. subcylindrica to invade more terrestrial habitats than other speices of the generally semiterrestrial genus Uca. I thank Nancy N. Rabalais for verification of identify of specimens. Specimens have been deposited in the Texas Memorial Museum. Literature Cited Barnwell, F. H., and C. L. Thurman, II. 1984. Taxonomy and biogeography of the fiddler crabs (Ocypodidae: genus Uca ) of the Atlantic and Gulf coasts of eastern North America. Zool. J. Linn. Soc., 81:23-87. Bright, D. B., and C. L. Hogue. 1972. A synopsis of the burrowing land crabs of the world and list of their arthropod symbionts and burrow associates. Los Angeles Co. Nat. Hist. Mus. Contrib. Sci., 220:1-58. Foscue, E. J. 1934. Agricultural history of the lower Rio Grande Valley region. Agric. Hist., 9:124-137. Hildebrand, H. H. 1958. Estudios biologicos preliminares sobre La Laguna Madre de Tamdaulipas. Ciencia (Mex.), 17:151-173. Powers, L. W. 1977. Crabs (Brachyura) of the Gulf of Mexico. Contrib. Marine Sci., 20 (suppl.): 1-190. Rabalais, N. N., and J. N. Cameron. 1983. Abbreviated development of Uca subcylindrica (Stimpson, 1859) (Crustacea, Decapoda, Ocypodidae) reared in the laboratory. J. Crustacean Biol., 3:519-541. - . 1985a. Physiological and morphological adaptations of adult Uca subcylindrica to semi-arid environments. Biol. Bull., 168:135-146. - . 1985b. The effects of factors important in semi-arid environments on the early development of Uca subcylindrica. Biol. Bull., 168:147-160. Thurman, C. L., II. 1984. Ecological notes on fiddler crabs of south Texas, with special reference to Uca subcylindrica. J. Crustacean Biol., 4:665-681. 198 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 A RECORD OF THE WESTERN SMALL-FOOTED MYOTIS, MYOTIS CILIOLABRUM MERRIAM, FROM THE TEXAS PANHANDLE Robert R. Hollander and J. Knox Jones, Jr. The Museum and Department of Biological Sciences, Texas Tech University, Lubbock, Texas 79409 Although reported from the Wichita Mountains and Black Mesa of Oklahoma (Hall, 1981), the western small-footed myotis, Myotis ciliolabrum ciliolabrum Merriam, 1886, has been known previously in Texas only from the western counties of the Trans-Pecos region (Davis, 1974; Schmidly, 1977). A record from the Panhandle of that state is, therefore, noteworthy. On 30 January 1960, the late Donald R. Tinkle and a field party collected hibernating bats in a sinkhole on the Hedgecoke Ranch, 29 mi. SSW Claude, Armstrong Co., Texas. Among the specimens taken was a single M. c. ciliolabrum , a skin alone of unrecorded sex (TTU 120), that was orignially misidentified as a pipistrelle (see also Mollhagen, 1973). The golden brown dorsal color, narrow and pointed tragus (5 mm long), small feet, and blackish ears, however, leave no doubt as to the identity of the bat. External measurements (mm) recorded by the preparator are: total length, 77; length of tail, 33; length of hind foot, 6; length of ear, 13. The forearm measures 32.5. We follow Jones et al. (1986) in use of the specific name ciliolabrum for this bat in Texas. Myotis ciliolabrum is a saxicolous species. It is likely, therefore, that individuals hibernate from time to time in the many caves and sinkholes in the Panhandle region and occur there in summer in suitable rocky habitats. Literature Cited Davis, W. B. 1974. The mammals of Texas. Bull. Texas Parks and Wildlife Dept., 41:1- 294. Hall, E. R. 1981. The mammals of North America. John Wiley & Sons, New York, l:xv+ 1-600+90. Jones, J. K., Jr., D. C. Carter, H. H. Genoways, R. S. Hoffmann, D. W. Rice, and C. Jones. 1986. Revised checklist of North American mammals north of Mexico, 1986. Occas. Papers Mus., Texas Tech Univ., 107:1-22. Mollhagen, T. 1973. Distributional and taxonomic notes on some West Texas bats. Southwestern Nat., 17:427-430. Schmidly, D. J. 1977. The mammals of Trans-Pecos Texas. Texas A&M Univ. Press, College Station, xiii+225 pp. The Texas Journal of Science, Vol. 39, No. 2, May, 1987 INSTRUCTIONS TO AUTHORS Scholarly manuscripts in any field of science or technology, including science education, will be considered for publication in The Texas Journal of Science. Prior to acceptance, each manuscript will be reviewed by at least two knowledgeable critics, the appropriate Associate Editor, and the Editor. Manuscripts intended for publication in the Journal should be submitted to the Editor, J. Knox Jones, Jr. (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409), in accordance with the following instructions. No manuscript submitted to the Journal is to have been published or submitted elsewhere. Manuscripts must be double-spaced throughout (including tables, legends, and cited literature) and submitted in triplicate on typed or clear machine-produced copies on 8.5 x 1 1-inch paper, with margins of approximately 1.5 inches. The centered title of the article (usually 10 words or less) should be followed by the name(s) of the author(s) and institutional or business address(es), including zip-code, both also centered on the title page. Each manuscript should have a brief, concise Abstract, terminating with up to five key words. The following text can be subdivided into sections as appropriate (examples follow): introductory information is self evident and thus usually needs no heading; materials and methods (acknowledgments frequently can be placed here as well); results; discussion; summary or conclusions; literature cited. Major sections are centered and capitalized; secondary divisions are italicized (underlined) left flush; tertiary headings are italicized at the beginning of paragraphs. Cite all references in text by author and date in chronological order — Jones (1971); Jones (1971, 1975); (Jones, 1971); (Jones, 1971, 1975); (Jones, 1971; Smith, 1973; Davis, 1975); Jones (1971), Smith (1973), Davis (1975); Smith and Davis (1985); (Smith and Davis, 1985). If more than two authors, use Jones et al. (1976) or (Jones et al., 1976). Citations to publications by the same author(s) in the same year should be designated alphabetically (1979a, 1979b). Be sure all citations in text are included in the Literature Cited section and vice versa. Hypothetical examples of proper citations are given below. Davis, G. L. 1975. The mammals of the Mexican state of Yucatan. Unpublished Ph.D. dissertation, Texas Tech Univ., Lubbock, 396 pp. Jones, T. L. 1971. Vegetational patterns in the Guadelupe Mountains, Texas. Amer. J. Bot., 76:266-278. - . 1975. An introduction to the study of plants. John Wiley and Sons, New York, xx+386 pp. Jones, T. L., A. L. Bain, and E. C. Burns. 1976. Grasses of Texas. Pp. 205-265, in Native grasses of North America (R. R. Dunn, ed.), Univ. Texas Studies, 205:xx+l- 630. Smith, J. D. 1973. Geographic variation in the Seminole bat, Lasiurus seminolus . J. Mamm., 54:25-38. Smith, J. D., and G. L. Davis. 1985. Bats of the Yucatan Peninsula. Occas. Papers Mus., Texas Tech Univ., 97:1-36. Consecutively-paged journal volumes and other serials should be cited only by volume number and pagination. Serials with more than one number and that are not consecutively paged should be cited by number as well (Smiths. Misc. Coll., 37(3): 1-30). Illustrations are acceptable only as original inked line drawings or photographic prints. They normally should be no larger than 4.5 x 6.5 inches and mounted on 8.5 x 11 paper or backing. Each figure should be marked on the back with the name of the author(s) and figure number. If confusion might result as to arrangement of a figure, label “top.” All legends for figures must be typed (double-spaced) on a separate sheet(s) of paper. All figures must be referred to in text — as “Figure 3” or “(Fig. 3).” All tables are to be typed, double-spaced, and headed by the legend, on a single page(s) for each table. All should be cited at the appropriate place in text as “Table 1” or “(Table 1).” 200 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 2, 1987 Some important specific points for authors: (1) do not break words at the right-hand margin of text; (2) footnotes are to be avoided except as absolutely needed in tables; (3) scales for illustrations should be on the figure, not in the legend, to avoid errors when illustrations are reduced for publication; (4) be sure all lettering or other symbols on illustrations will be clearly evident after reduction of them to Journal page size; (5) the editor should be notified immediately of any change in address of the responsible author, whose telephone number also should appear on correspondence; (6) in order to make papers more readable for the general scientific public, abbreviations are to be avoided in text except for standard mathematical or chemical formulae (where an abbreviation might be used many times to save space, write out the full term the first time used and give the abbreviation, which can be used thereafter, in parentheses); (7) except where fractions are used, write out numbers through nine in text and use numerals for 10 and beyond; (8) consult recent issues of the Journal , beginning with 38(1), for all matters of style. The principal author will receive galley proofs along with edited typescript and a reprint order form. Proofs must be corrected and returned to the editor within five days; failure to return proof promptly will result in delay of publication. Reprint order forms should be returned directly to Texas Tech Press (P.O. Box 4240, Lubbock 79409), not to the editor. Charges of $35 per printed page (or part thereof), or partial payment, strongly are encouraged by members of the Texas Academy of Sciences when grant or institutional funds are available for that purpose. Some contribution, even if modest, is expected for any paper that exceeds 10 printed pages. Nonmembers of the Academy are required to cover all page costs except as rarely excepted by the Treasurer. Authors are provided with page-charge information when their manuscript is accepted for publication. General Notes. — Beginning with volume 39 of the Journal , a section for noteworthy but short contributions may appear at the end of each number. Manuscripts published as General Notes normally will not exceed four or five typed pages. The format is the same as for feature articles except no abstract is included and the only subheading in text is a centered Literature Cited (if needed) unless italicized paragraph subheadings are absolutely essential, as in the case, for example, of more than one account for individual species of plants or aimals. While the decision as to whether a manuscript is best suited for a feature article or a note will be made by the editoral staff, authors are encouraged to indicate their preference at the time a manuscript is submitted to the Editor. THE TEXAS ACADEMY OF SCIENCE, 1986-87 Officers President: President-Elect: Vice-President: Immediate Past President: Executive Secretary: Treasurer: Editor: A AS Council Representative: Directors 1985 George B. McClung, San Angelo Barbara Schreur, Texas A&I University 1986 Caroline P. Benjamin, Southwest Texas State University R. John Prevost, Southwest Research Institute 1987 John P Fackler, Jr., Texas A&M University David R. Gattis, Freese and Nichols, Inc. Sectional Chairpersons I — Mathematical Sciences : Barbara Schreur, Texas A&I University II — Physical Sciences: C. A. Quarles, Texas Christian University III — Earth Sciences : James L. Carter, University of Texas at Dallas IV — Biological Sciences: Robert S. Baldridge, Baylor University V — Social Sciences: David A. Edwards, San Antonio College VI — Environmental Sciences: Bennett J. Luckens, Austin VII — Chemistry: Rodney Cate, Midwestern State University VIII — Science Education: Dick E. Hammond, Southwest Texas State University IX — Computer Sciences: Ronald King, Baylor University X — Aquatic Sciences: Paula Dehn, University of Texas at San Antonio Counselors Collegiate Academy: Shirley Handler, East Texas Baptist College Helen Oujesky, University of Texas, San Antonio Junior Academy: Ruth Spear, San Marcos Peggy Carnahan, San Antonio Lamar Johanson, Tarleton State University Owen T. Lind, Baylor University Glenn Longley, Southwest Texas State University Billy J. Franklin, Lamar University William J. Clark, Texas A&M University Michael J. Carlo, Angelo State University J. Knox Jones, Jr., Texas Tech University Ann Benham, University of Texas at Arlington THE TEXAS JOURNAL OF SCIENCE USPS 616740 The Texas Academy of Science Box 4240, Texas Tech University Lubbock, Texas 79409-3151, U.S.A. BOUND PRINTED MATERIAL RETURN POSTAGE GUARANTEED 2nd CLASS POSTAGE PAID AT LUBBOCK TEXAS 79401 THE TEXAS JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP. — Any person or group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. Dues for members are $20.00 annually; student members, $12.00 annually; sustaining members, at least $30.00 in addition to annual dues; life members, at least $400.00 in one payment; patrons, at least $500.00 in one payment; corporate members, $250.00 annually; corporate life members, $2000.00 in one payment. Library subscription rate is $45.00 annually. Payments should be sent to Dr. Michael J. Carlo, P.O. Box 10986, Angelo State University, San Angelo, Texas 76909. The Texas Journal of Science is a quarterly publication of the Academy and is sent to all members and subscribers. Changes of address and inquiries regarding back issues should be sent to Dr. William J. Clark, Drawer H6, College Station, Texas 77844. AFFILIATED ORGANIZATIONS Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers American Association for the Advancement of Science Texas Society of Mammalogists The Texas Journal of Science (US PS 616740) is published quarterly at Lubbock, Texas U.S.A. Second class postage paid at Post Office, Lubbock, TX 79401. Please send form 3579 and returned copies to Texas Tech Press, Box 4240, Lubbock, TX 79409. ISSN 0040-4403 THE TEXAS JOURNAL OF SCIENCE Volume 39, No. 3 August 1987 CONTENTS A framework for plant community classification and conservation in Texas. By David D. Diamond, David H. Riskind, and Steve L. Orzell . 203 Intrapopulational variation in two samples of arid-land foxes. By Jerry W. Dragoo, Jerry R. Choate, and Thomas P. O’ Farr ell . 223 Comparison of canopy position and other factors on seedling growth in Acacia smallii. By R. J. Lohstrom and O. W. Van Auken. . . . . 233 Natural history sketches, densities, and biomass of breeding birds in evergreen forests of the Rio Grande, Texas, and Rio Corona, Tamaulipas, Mexico. By Frederick R. Gehlbach . 241 Clear Fork vertebrates and environments from the Lower Permian of north-central Texas. By Phillip A. Murry and Gary D. Johnson . . . 253 A predatory terrestrial flatworm, Bipalium kewense, in Texas: feral populations and laboratory observations. By Raymond W. Neck . 267 Validation of daily ring deposition in otoliths of wild young-of-the-year largemouth bass. By J. Jeffery Isely and Richard L. Noble . 273 Notes on distribution and natural history of some bats on the Edwards Plateau and in adjacent areas of Texas. By Richard W. Manning, J. Knox Jones, Jr., Robert R. Hollander, and Clyde Jones . . . 279 General Notes Additional records of tick (Acari: Ixodidae, Argasidae) ingestion by whiptail lizards, genus Cnemidophorus. By Chris T. McAllister and James E. Keirans . 287 A specimen of white-winged dove, Zenaida asiatica, from Archer County, Texas. By Frederick B. Stangl, Jr., and Warren Pulich . . 288 Anolis sagrei (Sauria: Iguanidae) established in southern Texas. By Ken King, David Cavazos, and Frank W. Judd . . . 289 First record of Centropages typicus Kroyer (Copepoda: Centropagidae) in the Gulf of Mexico. By David C. Me Aden, George N. Greene, and William B. Baker, Jr. . 290 Evidence of communal nesting and winter-kill in a population of Baiomys taylori from north-central Texas. By Frederick B. Stangl, Jr., and Stephen Kasper . 292 New western distributional record of Terrapene Carolina triunguis. By Michael R. Gannon . . 293 First records of the flammulated owl ( Otus flammeolus) in the central Trans-Pecos of Texas. By Donna Burt, D. Brent Burt, Terry C. Maxwell, and Ross C. Dawkins . 293 Instructions to Authors . 295 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Editor: J. Knox Jones, Jr., Texas Tech University Assistant to the Editor: Marijane R. Davis, Texas Tech University Associate Editor for Botany: Chester M. Rowell, Marfa, Texas Associate Editor for Chemistry: Marvin W. Rowe, Texas A&M University Associate Editor for Computer Science: Ronald K. Chesser, Texas Tech University Associate Editor for Mathematics and Statistics: Patrick L. Odell, University of Texas at Dallas Associate Editor for Physics: Charles W. Myles, Texas Tech University Editorial Assistants: Robert R. Hollander, Texas Tech University Richard W. Manning, Texas Tech University Scholarly papers in any field of science, technology, or science education will be considered for publication in The Texas Journal of Science. Instructions to authors are published one or more times each year in the Journal on a space-available basis, and also are available from the Editor (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409, 806/742-2487, Tex-an 862-2487). A FRAMEWORK FOR PLANT COMMUNITY CLASSIFICATION AND CONSERVATION IN TEXAS David D. Diamond, David H. Riskind, and Steve L. Orzell Texas Natural Heritage Program, General Land Office, Stephen F. Austin Building, Austin, Texas 78701, and Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, Texas 78744 (DHR) Abstract. — Seventy-eight late serai stage plant community types are described and classified at the series level (characterized by dominant species or genera). The classification framework used can be expanded to include finer subdivisions at the plant association (defined as a plant community of definite floristic composition within a uniform habitat) level. Community types are ranked according to conservation needs from endangered (1) to secure (4), and 10 (13 percent) are threatened or endangered. Ranking remains partially subjective, but depends on the estimated number of late serai stage relicts, the estimated number of relicts protected, the estimated area occupied by the community type, and the relative threat of severe disturbance. The classification and conservation rankings presented should stimulate debate among users and result in successive refinements. Quantitative data on vegetation are inadequate to provide a finer scale classification of communities for most regions of Texas, and quantitative inventories of resources already in public or private managed natural areas also are lacking. The need for these data, the need for enlightened stewardship of existing natural areas, and the need to consider landscape ecology in preserve design and selection mark the greatest challenges in the future for Texas conservationists. Key words: classification; conservation; landscape ecology; plant commun¬ ity. The land area of Texas is more than 370,000 square kilometers. Climatic regimes range from humid subtropical to arid subtropical to continental steppe (Larkin and Bomar, 1983; Owen and Schmidly, 1986), while geologic substrate ranges from Recent sands and silts to Cretaceous limestones to Precambrian granite (Sellards et al., 1966). This wide variation in climate and geology results in a wide variation in landform, soils, and vegetation. Soils have been classified and mapped to a relatively fine degree across most of Texas. However, there has been no attempt at a fine-scale, state-wide classification of vegetation, although several authors have classified the vegetation at a physiognomic or natural region level (see Tharp, 1939; Gould, 1975; Kuchler, 1964; LBJ School of Public Affairs, 1978; Brown et al., 1979; Fig. 1). An informal approach to classification has generally been adopted by North American ecologists, who have used different approaches and frameworks for different vegetation types and purposes. However, there is merit in acceptance of a standardized classification, because this favors more efficient organization and communication of information among users (see Driscoll et al., 1984). Because conservation efforts generally concentrate on late serai stage, rather than early or mid-successional The Texas Journal of Science, Vol. 39, No. 3, August, 1987 204 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 ROLLING PLAINS 9a. Mesquite Plains 9b. Escarpment Breaks 9c. Canadian Breaks HIGH PLAINS TRANS PECOS I la. Mountain Ranges lib Desert Grassland lie Desert Scrub lid. Salt Basin lie. Sand Ilf. Stock! I. PINEY WOODS la Longleaf Pine Forest lb. Mixed Pine-Hardwood Forest OAK WOODS a PRAIRIES 2a. Oak Woodlands 2b Eastern Cross Timbers 2c. Western Cross Timbers BLACKLAND PRAIRIES 3a. Blackland Prairie 3b. Grand Prairie SOUTH TEXAS BRUSH COUNTRY 6a Brush Country 6b. Bordas Escarpment 6a Subtropical Zone EDWARDS PLATEAU 7a Live Oak -Mesquite Savanna 7h Balcones Conyonlands 7c Lampasas Cut Plain LLANO UPLIFT 8a Mesquite Savanna 8b Oak 8 Oak-Hickory Woodlands GULF COAST PRAIRIES & MARSHES 4a Dunes /Barrier 4b. Estuarine Zone 4c Upland Prairies a Woods COASTAL SAND PLAINS Figure 1. The natural regions of Texas (after LBJ School of Public Affairs, 1978). communities, a standardized classification of such communities is appropriate for present purposes. History suggests that conservation efforts often overlook the most rare and endangered community types (Gehlbach, 1975). The result of well- meaning but poorly informed and directed protection efforts is often the preservation of numerous examples of relatively common and secure types, usually because they are the most easily acquired or the most aesthetically appealing. A well-founded classification of community types, together with ranks that are based on the rarity and need for protection of each type, should help direct conservation efforts. Awareness of the need for conservation recently has increased in Texas. The Texas Natural Heritage Program was established within the General Land Office in 1983 to provide a biological inventory of rare species and habitats. A Resource Protection Division was legislatively mandated and established within the Texas Parks and Wildlife Department in 1985. The Texas Organization for Endangered Species established a community PLANT COMMUNITIES OF TEXAS 205 protection committee in 1986. Despite growing public awareness, however, surprisingly little is known about the status of plant community protection in Texas (Gehlbach, 1975; Carls and Ludeke, 1984). The purpose of this review is to provide a classification of communities, plus propose conservation ranks for community types and forward suggestions on the direction of future conservation efforts in Texas. Classification Framework The classification framework used is modified from Driscoll et al. (1984), who, in turn, used a framework provided by UNESCO (1973). A similar framework has been applied in other regions (see Brown et al., 1979; Paysen et al., 1980; Buck and Paysen, 1984). The top level of this hierarchial framework is the class, which is based on dominant growth form. These are: 1. Forest — communities formed by trees at least three meters tall, with a canopy of 61 percent or more. 2. Woodland — communities dominated by trees with 26 to 60 percent canopy (this includes “open forest” and “savannah” types of some authors). 3. Shrubland — communities composed of shrubs from half a meter to three meters tall with 26 percent or more canopy cover. 4. Dwarf Shrubland — communities with 26 percent or more canopy cover of shrubs less than half a meter tall (this category includes heathlands and shrub-dominated wetlands — none are described for Texas). 5. Herbaceous Vegetation — communities dominated by grasses, graminoids, or forbs with less than 25 percent canopy cover of woody plants. Two additional classes associated with aquatic regimes are recognized here, but were not described by Driscoll et al. (1984): 6. Swamps — forested or shrub-dominated wetlands, with standing water at the surface at least 50 percent of the year and hydric soils. 7. Marshes — herbaceous-dominated wetlands, with standing water at least 50 percent of the year and hydric soils. The next level in the hierarchy is the subclass. Nineteen subclasses are recognized based on morphological characteristics of dominant species, such as evergreen or deciduous habit or adaptation to temperature and water. Two levels of hierarchy below the subclass, the group and formation, will not be adopted here. They are based on different criteria for different vegetation types, and there is often difficulty is assigning lower level taxa (series or associations, described below) to formations or groups. Assignment of lower level taxa to groups or formations would provide little additional understanding of their interrelationships, and might even cause confusion. The lowest levels in the hierarchy are the series and the association. Series are characterized by specificity of structure and physiognomy of the vegetation (Mueggler and Stewart, 1980; Pfister and Arno, 1980). They are named for typical dominant or co¬ dominant species or, in a few cases, genera. Series are defined to allow for some differences in species composition and dominance relationships among representitive stands, and thus local variants are circumscribed. Associations, subdivisions of series, are plant community types of definite floristic composition within a uniform habitat. They are named and recognized by dominant or co-dominant species. Associations have been quantatively defined for parts of Texas, but are unknown across most of the state. Hence for uniformity's sake, vegetation is classified only to the series level in the following sections. The term “community type” is applied to both series and associations. 206 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Classification of Late Seral Vegetation Forest and Woodland Fourteen forest and 19 woodland community types were defined (Table 1). Forests of both uplands and floodplains of the East Texas Piney Woods are relatively well known (Tharp, 1926; Nixon et ah, 1973; Chambless and Nixon, 1975; Nixon and Raines, 1976; Nixon et ah, 1977; Watson; 1979; Nixon et ah, 1980; Hinton, 1981; Marks and Harcombe, 1981; Marietta and Nixon, 1983; Nixon et ah, 1983), and 10 community types are defined. The sweetbay ( Magnolia virginiana) series is among the most broadly defined of these, and may range from mainly deciduous forest in the north to mainly evergreen shrubland (“baygalls”) in the south. Vegetation of the Post Oak Savannah and Cross Timbers to the west is not as well known, and one forest and one woodland series is indicated based on data from McBryde (1933), Dyksterhuis (1948), McCaleb (1954), Rice and Penfound (1959), Risser and Rice (1971), Marcy (1982), Marietta and Nixon (1983), Ward (1983), and others. Two coastal live oak ( Quercus virginiana) co-dominated woodlands are indicated within the Coastal Prairie. These are for the most part restricted to the Upper Coastal Prairie, with the coastal live oak-post oak (Quercus virginiana- Quercus stellata) usually over clay pan soils with a loamy surface layer and the coastal live oak-pecan ( Quercus virginiana- Carya illinoensis) usually over clayey soils. The coastal live oak-seacoast bluestem ( Quercus virginiana-Schizachyrium scoparium var. littoralis) series inhabits sandy soils of the Coastal Sand Plains. Subtropical vegetation of the lower Rio Grande Valley is likewise not well known, but one forest and one woodland series is defined based on field surveys and data from Clover (1933), Davis (1942), Everitt and Gonzales (unpublished data) and Neck et ah (unpublished data). Four woodland community types of the Edwards Plateau, Balcones Canyonlands, and Llano Uplift are defined by data from Beuchner (1944), Tolstead and Cory (1946), Van Auken et ah (1978, 1979, 1981), Ford and Van Auken (1982), Van Auken and Bush (1985), Gehlbach (1988), Riskind and Diamond (1988), and others. Riparian forests and woodlands west of the Piney Woods are poorly known, but often contain some combination of sugarberry (Celtis laevigata ), hackberry ( Celtis reticulata ), elm (Ulmus sp.), ash ( Fraxinus sp.), oaks (Quercus sp.), and pecan among the dominants. Hence, a sugarberry-elm series is indicated. Likewise, water and coastal live oaks are often among the dominants of Upper Coastal Prairie floodplains, and a water oak-coastal live oak series is designated. Additional data on these community types are in Beuchner (1944), Nixon (1975), Allen (1974), Ford and Van Auken (1982), Mohler (1979), Van Auken and Bush (1985), Gehlbach (1988), and Riskind and Diamond (1988). PLANT COMMUNITIES OF TEXAS 207 Deciduous riparian woodlands of the Trans-Pecos generally contain some combination of ash, willow ( Salix sp.), cottonwood ( Populus sp.), hackberry, and walnut ( Juglans sp.) as dominants. The three series defined are based primarily on data from Webster (1950), Gehlbach (1967), Burgess and Northington (1974), Johnston (1974), Riskind (1976), and Brown et al. (1979). Four series dominated by pine ( Pinus sp.), juniper ( Juniperus sp.) or oak are described for the Trans-Pecos mountains, and are based on data from Gehlbach (1967), Brown et al. (1979), and Johnston and Henrickson (1987). Shrubland Shrublands dominate the contemporary landscape of the South Texas Plains and Trans-Pecos (Table 2). Desert shrublands and deciduous shrublands of the Trans-Pecos are relatively well known, and 10 series are indicated based on data from Webster (1950), Burgess and Northington (1974), Johnston (1974), Brown (1982), Burgess and Klein (unpublished data), Johnston and Henrickson (1987), and others. In contrast, composition of late serai stage shrublands of the South Texas Plains is poorly known. The four series listed are based primarily on field surveys and data from Clover (1933), Johnston (1952), Davis and Speicer (1965), Fanning et al. (1965), Drawe et al. (1979b), Everitt and Gonzales (unpublished data), Lonard et al. (1987), and Neck and Riskind (unpublished data). Communities dominated by Havard shin oak ( Quercus havardii) and tall grasses inhabit stabilized dunes in the High Plains, northern Rolling Plains, and northeastern Trans-Pecos. Slightly finer textured sandy soils, or disturbed sites, have sandsage ( Artemisia filifolia) and mid-grasses. These are described as belonging to a Havard shin oak-tallgrass or sand sage-midgrass series. Over thin limestone soils in the High Plains and Rolling Plains, especially slopes of the Red River Valley, Mohr’s shin oak ( Quercus mohriana) is often dominant, and hence a Mohr’s shin oak series is defined. Additional information is found in Johnston (1974), Warnock (1974), Eyre (1980), McMahan et al. (1984), and unpublished data from the U. S. Soil Conservation Service. The redberry juniper-midgrass {Juniperus pinchotii) and one seed juniper-midgrass (/. monospermd) series inhabit the Rolling Plains and High Plains, with juniper-dominated shrubland especially common over caliche or shallow soils. The former also occurs on the Stockton Plateau and Trans-Pecos Mountains, along with the scrub oak-mountain mahogany {Quercus pungens — Cer cocarpus montanus) series. There is variation with exposure and substrate, and a number of different shrub and grass species are associated with different species of juniper (Webster, 1950; Warnock, 1970; Adams, 1979; unpublished data, U. S. Soil Conservation Service). 208 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 1. — Late serai stage forest (dominants are trees more than three meters tall, forming 61 percent or more canopy) and woodland (dominants are trees forming 26-60 percent canopy) community types of Texas, with conservation ranks (1, endangered; 2, threatened; 3, apparently secure; 4, secure). Evergreen and deciduous forests are followed by evergreen and deciduous woodlands. Primary region of occurrence after LBJ School of Public Affairs (1978). Community type Primary region of occurrence Conservation rank Mainly Evergreen Forest Douglas Fir-Pine Series 11a 4 ( Pseudotsuga menziesii-Pinus sp.) Ponderosa Pine Series 11a 4 ( Pinus ponderosa ) Texas Palmetto Series 6c 1 (Sabal texana) Mainly Deciduous Forest American Beech-Southern Magnolia Series la 2 ( Fagus grandifolia- Magnolia grandiflora) Baldycypress-Sycamore Series 7b 3 ( Taxodium distichum- Plat anus occidentalis) Loblolly Pine-Oak Series lb, la 4 ( Pinus taeda-Quercus sp.) Overcup Oak Series la, lb, 2a, 3a 3 ( Quercus lyrata) Post Oak-Black Hickory Series 2a, 2b, 3a 3 ( Quercus stellata-Carya texana) Shortleaf Pine-Oak Series lb 4 ( Pinus echinat a- Quercus sp.) Sugarberry-Elm Series 2, 3, 4, 6, 7 3 (Celt is laevigata/ C. reticulata- Ulmus sp.) Swamp Chestnut Oak-Willow Oak Series 8, 9 la, lb 3 ( Quercus michauxii-Q. p hellos ) Sweetbay Series la, lb 3 (Magnolia virginiana ) Water Oak-Sweetgum Series la, lb 3 (Quercus nigra- Liquidambar styraciflua) Water Oak-Coastal Live Oak Series 4c 3 (Quercus nigra- Q. virginiana) Mainly Evergreen Woodland Ashe Juniper-Oak Series 7a, 7b 4 (Juniperus ashei-Quercus sp.) Coastal Live Oak-Seacoast Bluestem Series 5 3 ( Quercus virginiana- Schizachyrium soparium var. littoralis) Longleaf Pine-Rhynchospora Series la 3 (Pinus palustris-Rhynchospora sp.) Longleaf Pine-Tallgrass Series la 2 (Pinus palustris) Pinyon Pine-Oak-Juniper Series 11a 4 (Pinus cembroides/ P. edulis- Quercus sip -Juniperus sp.) PLANT COMMUNITIES OF TEXAS 209 Table L — Continued. Rocky Mountain Jumper Series (Juniperus scopulorum) 9b, 10 4 Texas Ebony-Anacua Series ( Pithecellobium flexicaule-Ehretia anacua ) Mainly Deciduous Woodland 6c 1 Bluejack Oak-Pine Series ( Quercus incana-Pinus sp.) la, lb, 2a 3 Coastal Live Oak-Pecan Series ( Quercus virginiana-Carya illinoensis) 4c 3 Coastal Live Oak-Post Oak Series ( Quercus virginiana-Q. stellata) 4c 3 Freemont Cottonwood-Willow Series ( Populus freemontii-Salix sp.) 11a, lib, 11c, Ilf 3 Gray Oak-Oak Series (' Quercus grisea- Quercus. sp.) 11a, 11b, Ilf 4 Lacey Oak Series {Quercus glaucoides ) 7a, 7b 3 Mesquite-Huisache Series ( Prosopis glandulosa- Acacia farnesiana) 4c, 6a, 11c, Ilf 4 Netleaf Hackberry-Little Walnut Series (Celt is reticulata- Juglans microcarpa) lib, 11c, Ilf 3 Pine-Bluejack Oak Series {Pinus sp .-Quercus incana) la, lb 3 Plateau Live Oak Series ( Quercus fusiformis ) 7a, 8 3 Post Oak-Blackjack Oak Series ( Quercus stellata- Q. marilandica) 2a, 2b, 2c 4 Texas Oak Series ( Quercus texana) 7b, 7c 3 Velvet Ash-Willow Series ( Fraxinus velutina-Salix sp.) 11a, lib, 11c 3 Herb- Dominated Communities Seventeen herb-dominated community types were identified (Table 3). Six are tallgrass communities, with four inhabiting the Blackland, Fayette, and Upper Coastal prairies. They are described based on data from Dyksterhuis (1946), Launchbaugh (1955), Diamond (1980, 1983), Dunlap (1983), and Diamond and Smeins (1984, 1985). The seacoast bluestem series often contains gulfdune paspalum ( Paspalum monosta- chyum) and sedges, and inhabits stabilized dunes and flats of the barrier islands, plus the Coastal Sand Plains. It is defined on data from Johnston (1955, 1963) and Judd et al. (1977). The cenicella-beach morning glory ( Sesuvium portulacastrum-Ipomoea stolonifora) series occurs on partially stabilized coastward barrier island dunes, and is defined primarily on data from Chabreck (1972), Judd et al. (1977), Drawe et al. (1979a), and Lonard and Judd (1980). 210 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 2. — Late serai stage shrublands (dominants are shrubs or small trees half a meter to three meters tall, forming 26 percent or more canopy) of Texas, with conservation ranks (1, endangered; 2, threatened; 3, apparently secure; 4, secure). Evergreen shrublands are followed by deciduous shrublands and xeromorphic shrublands. Primary region of occurrence after LBJ School of Public Affairs (1978). Community type Primary region of occurrence Conservation rank Mainly Evergreen Shrubland Ceniza series 6a 3 ( Leucophyllum frutescens ) Havard Shin Oak-Tallgrass Series 9, 10, lie 3 ( Quercus havardii) Oneseed Juniper-Midgrass Series 9a, 9b, 10 4 (Juniperus monosperma) Redberry Juniper-Midgrass Series 9a, 9b, 10, 11a, Ilf 4 {Juniperus pinchotii) Sandsage-Midgrass Series 9a, 10 4 ( Artemisia filifolia) Scrub Oak-Mountain Mahogany Series 11a, lib 4 {Quercus pungens-Cercocarpus montanus) Mainly Deciduous Shrubland Apache-plume Series lib, 11c, Ilf 4 {Fallugia par ado xa) Blackbrush Series 6a 4 {Acacia rigidula) Fern Acacia Series 6a, 6b, Ilf 4 {Acacia berlandieri ) Mesquite-Sandsage Series lib, 11c, lie, Ilf 4 {Prosopis glandulosa- Artemisia filifolia ) Mohr’s Shin Oak Series 9a, 9b, 7a, Ilf 4 {Quercus mohriana) Rough Tequilia Series 11c, lid 3 ( Tequilia hispidissima ) Texas Ebony-Snake-eyes Series 6c 2 {Pithecellobium flexicaule- Phaulothmnus spinescens ) Xeromorphic (Desert) Shrubland Creosotebush-Mariola Series lib, 11c 4 {Larrea tridentata- Parthenium incanum) Creosotebush-Tarbush Series 11c 4 {Larrea tridentata- Florensia cernua) Giant Dagger Series 11a, 11c 4 ( Yucca faxoniana) Lechuguilla-Sotol Series lib, 11c 4 {Agave lecheguilla- Dasylirion leiophyllum) Mesquite-Saltbush Series 11c 4 {Prosopis glandulosa- Atriplex sp.) Whitethorn Acacia Series 11a, 1 1c 4 {Acacia neovernicosa ) PLANT COMMUNITIES OF TEXAS 211 Table 3. — Herb-dominated (dominants are herbaceous, with less than 26 percent canopy of woody species) community types of Texas, with conservation ranks (1, endangered; 2, threatened; 3, apparently secure; 4, secure). Tall grasslands are followed by medium tall and short grassland and herbaceous community types. Primary region of occurrence after LBJ School of Public Affairs (1978). Community type Primary region of occurrence Conservation rank Tall Grassland (dominants more than 1 meter) Gammagrass-Switchgrass Series 3 a, 4c 1 ( Tripsacum dactyloides-Panicum virgatum) Little Bluestem-Brownseed Paspalum Series 4c 2 (Schizachyrium scop arium- Paspalum plicatulum ) Little Bluestem-Indiangrass Series 3a, 4c 1 (, Schizachyrium scoparium-Sorghastrum nutans) Cottonwood-Tallgrass Series 9a, 10 2 ( Populus delt aides) Silveanus Dropseed Series 3a 2 ( Sporobolus silveanus) Seacoast Bluestem Series 4a, 5 3 {Schizachyrium scoparium var. littoralis) Medium Tall Grassland (dominants 0.5- 1.0 meter) Alkali Sacaton-Fourwing Saltbush Series 11c, lid, lie, Ilf 3 {Sporobolus air aides- A triplex canescens) Black Grama-Sideoats Grama Series 11a, lib 3 {Bouteloua eriopoda-B. curtipendula) Cane Bluestem-Mesquite Series 6a 3 {Bothriochloa barbinodis- Prosopis glandulosa) Curlymesquite-Sideoats Grama Series 7a 3 {Hilar ia be longer i- Bouteloua curtipendula) Little Bluestem-Sideoats Grama Series 9a 3 {Schizachyrium scoparium- Bouteloua curtipendula) New Mexico Little Bluestem-Wolftail Series 11a 3 {Schizachyrium scoparium var. neomexicana- Lycurus p hie aides) Sideoats Grama-Mesquite Series 9a, 10 3 {Bouteloua curtipendula- Prosopis glandulosa) Short Grassland (dominants less than 0.5 meter) Blue Grama-Buffalograss Series 7a, 10 4 {Bouteloa gracilis- Buchloe dactyloides) Tobosa Series 7a, 10, 11b, 11c 3 {Hilar ia mutica) Forb-Dominated Vegetation Cenicilla-Beach Morning Glory Series 4a 4 {Sesuvium portulacastrum-Ipomoea stolonifera) Bogs Sphagnum-Rhynchospora Series la, 2a 3 {Sphagnum sp .-Rhynchospora sp.) 212 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 The cottonwood-tallgrass ( Populus deltoides) series of the Rolling Plains and High Plains has been little studied, but because of unique species composition probably warrants recognition as a separate community type (Bruner, 1931; Tolstead and Cory, 1946; Allred, 1956; unpublished data, U. S. Soil Conservation Service). Likewise, the sphagnum-rhynchospora ( Sphagnum sp -Rhynchospora sp.) series, found primarily in eastern Texas, denotes bogs that have not been intensively studied, but general descriptions can be found in Rowell (1949), Penfound (1952), Ajilvsgi (1979), Watson (1979), and Nixon and Ward (1986). The seven midgrass and two shortgrass community types are mainly confined to central and western Texas. They often are intermixed with shrublands, with shrubs being more prevalent in the contemporary landscape. Descriptions are based on Beuchner (1944), Dyksterhuis (1948), Webster (1950), Allred (1956), Brown and Schuster (1969), Smeins et al. (1976), Brown (1982), Dunlap (1983), Fowler and Dunlap (1986), Riskind and Diamond (1988), and unpublished data from the U. S. Soil Conservation Service. Swamps and Marshes Within eastern Texas, one forested and two shrub-dominated swamps are defined (Table 4) based on data from Ajilvsgi (1979), Watson (1979), Marks and Harcombe (1981), and Nixon and Ward (1986). A black mangrove ( Avicennia germinans) series is indicated for the South Texas coast based on descriptions from Clover (1933), McMillan (1971), and Lonard et al. (1987). Black mangroves are well established only along the far South Texas coast, and even there they are subject to periodic die- backs due to freezing temperatures (McMillan, 1971). The five marsh series defined generally correspond to differences in salinity, with the saltgrass ( Distichlis spicata ), gulf cordgrass ( Spartina spartinae ), and rush-sedge ( Juncus sp.) series occurring inland as well as along the coast, and the rest confined to coastal areas. Descriptions can be found in Chabreck (1972), Judd et al. (1977), Drawe et al. (1979a), and Lonard and Judd (1980). Conservation Ranks Community types were ranked according to conservation priorities by considering the following criteria, range-wide: (1) estimated number of late serai stage relicts — A=0-5, B=6-20, C=21-100, D=more than 100; (2) relative threat of severe disturbance — A=extreme, B —moderate, C=probably none, D=none; (3) estimated number of protected relicts — A=0-1, B=2-5, C=6-I0, D=more than 10; and (4) estimated area occupied by the community type, or serai stages of the community type PLANT COMMUNITIES OF TEXAS 213 Table 4. — Swamp (wetlands with woody dominants) and marsh (wetlands with herbaceous dominants) community types of Texas, with conservation ranks (1, endangered; 2, threatened; 3, apparently secure; 4, secure). Primary region of occurrence after LBJ School of Public Affairs (1978). Community type Primary region of occurrence Conservation rank Swamps Black Mangrove Series 6c 4 {Avicennia germinans) Baldcypress Series la, lb 3 ( Taxodium distichum) Buttonbush Series la, lb, 4c 4 ( Cephalanthus occidentalism Water Elm-Swamp Privet Series la, lb 4 {Planer a aquatica-Forestiera acuminata ) Marshes Rush-Sedge Series All regions 4 {Juncus sp.) Gulf Cordgrass Series 2a, 4a, 4b, 4c, 5, 6 3 ( Spartina spartinae) Marshhay Cordgrass Series 4a, 4b, 4c 3 {Spartina patens) Saltgrass Series 5, 6, 7, 9, 10, 11 4 {Distichlis spicata) Smooth Cordgrass Series 4a, 4b 3 {Spartina alterniflora) that are recoverable to near-climax — A=less than 1000 hectates, B=1000- 5000 hectares, C=5000-25,000 hectares, D=more than 25,000 hectares. Unpublished information from the Texas Natural Heritage Program, which represents a synthesis of data from federal, state, and major private conservation-oriented land owners, was the primary source used to determine ranks. Community types were ranked as endangered (1) if they received an A in two categories, threatened (2) if they received at least one A no D ranks, apparently secure (3) if they received no A and one or no D ranks, and secure (4) if they received two or more D ranks. These catagories were mutually exclusive for the community types listed in Tables 1-4. Three community types are considered endangered (Table 5). These include one grassland within the Blackland and Fayette prairies, and two subtropical types of the Rio Grande delta. Most of the area formerly occupied by these community types has been converted to cropland (McMahan et al., 1984). Seven additional community types are considered threatened. All five of the tallgrass grasslands of the Blackland, Fayette, and Costal prairies are endangered or threatened, primarily due to conversion of these types to row crops. 214 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 5. — Conservation ranking for endangered (1) and threatened (2) community types of Texas. Community type Total area1 Estimated number of relicts2 Number of relicts protected3 Relative threat of disturbance4 Conservation rank American Beech-Southern Magnolia Series (Fagus grandifolia- Magnolia grandiflora) C C B A 2 Cottonwood-Tallgrass Series ( Populus deltoides) B B A B 2 Gammagrass-Switchgrass Series ( Tripsacum dactyloides-Panicum virgatum) A B A A 1 Little Bluestem-Brownseed Paspalum Series ( Schizachyrium scoparium- Paspalum plicatulum) B B B A 2 Little Bluestem-lndiangrass Series (Schizachyrium scoparium-Sorghastrum nutans) B C B A 2 Longleaf Pine-Tallgrass Series (Pinus palustris) C C B A 2 Silveanus Dropseed Series (Sporobolus silveanus) B B A B 2 Texas Ebony-Anacua Series (Pithecellobium flexicaule-Ehretia anacua) B A B A 1 Texas Ebony-Snake-eyes Series ( Pithecellobium flexicaule- Phaulothamnus spinescens) B B B A 2 Texas Palmetto Series (Sabal texana) A A A A 1 'A = less than 1000 hectares, B = 1000-5000 hectares, C = 5000-25,000 hectares, D = more than 25,000 hectares. 2 A = 0-5, B = 6-20, C = 21-100, D = more than 100. 3A = 0-1, B = 2-5, C = 6-10, D = more than 100. 4 A - extreme, B = moderate, C = probably none, D = none. Discussion Several limitations of the plant community classification are apparent: (1) determination of the composition of late serai stage plant communities is difficult due to lack of relicts and lack of quantitative data; (2) many serai community types, which have come to dominate the contemporary landscape due to various cultural impacts, are not included in the classification; and (3) many plant associations are threatened or endangered, but are not included here because plant communities are only classified at the series level. Hence, the classification presented is dynamic and can be altered, pending improved knowledge of the vegetation of Texas. Development of an inclusive classification is an iterative process that will probably not be accomplished in the near future. Conservation ranks are subject to change with increased knowledge and changing land use patterns. The ranking system is not meant to be dogmatically applied. Numerous indices have been proposed for setting priorities, and the conservation rank of a community is one of a number of criteria to consider (Tubbs and Blackwood, 1971; Trans, 1974; Gehlbach, 1975; Klopatek et al., 1981; Game and Peterken, 1984). Examples of community types that may become endangered or PLANT COMMUNITIES OF TEXAS 215 threatened because of changing land use patterns include East Texas bottomland forests, due to building of dams or building of levees and clearing for agriculture; East Texas upland forests and bogs, due to timber production activities; woodlands along the Balcones escarpment, due to urbanization caused by the growth of San Antonio, New Braunfels, San Marcos, Austin, Temple, and Waco; and barrier island grasslands due to expanding recreational and urban development. Research Needs Few comprehensive qualitative or quantitative surveys of natural resources within managed areas exist. Carls and Ludeke (1984) surveyed all state parks, but this analysis was based on qualitative secondary data provided by the Texas Parks and Wildlife Department. Smeins and Diamond (1986) surveyed all federal, state, and private-managed natural areas in central Texas, but this, again, was based on secondary data and the geographic scope of the study was limited. One of the purposes of the Texas Natural Heritage Program, now housed within the General Land Office, is to gather information on all managed natural areas in Texas. This data base is extremely useful as a synthesis of existing information, but it is not yet complete and partly based on secondary information, though qualitative surveys will be completed over time. The secondary qualitative data used in all of these cases is of undetermined reliability. Published quantitative data on plant communities are incomplete. Regions that are especially lacking include the South Texas Plains, lower Rio Grande Valley, Edwards Plateau, and Rolling Plains. Data on riparian forests outside of the Piney Woods are generally lacking as are data on the ecophysiology and life history of important species. Also, the taxonomy of important species is in some cases controversial. A long and continued history of overgrazing, cultivation, timber production, and other changes in land use makes reconstruction and recognition of late serai plant communities tentative, and the opportunity to study and conserve relatively natural vegetation is constantly diminishing. There is an immediate need for more quantitative studies on the composition, distribution, and ecology of the vegetation of Texas. These data then could be synthesized and incorporated into a revised classification and, together with more complete information about natural areas already protected, would serve to direct future conservation efforts. Stewardship Conservation of significant natural areas already under federal or state ownership or in private nature preserves is not guaranteed. Sometimes 216 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 relict plant communities are not recognized, or are recognized but given a low priority in the overall management scheme. Even where nature preserves are established to maintain the integrity of relict plant communities, neglect due to lack of funds or improper management due to lack of expertise has in many cases caused degradation of the resource. Hence, there is a need to identify and manage significant natural areas on public lands, and to provide funds and expertise for management of private nature preserves. Private owners who are good land stewards have been, and continue to be, identified and honored by organizations such as The Nature Conservancy, but the fate of the native vegetation within private ownership is at best tentative. Landscape Ecology Considerable attention has been given to the problem of insularity of nature preserves and concomitant problems of long-term genetic isolation and the attenuation of recruitment and extinction relationships of species (Pickett and Thompson, 1978; Harris, 1984). The size of natural areas and their proximity to each other (degree of fragmentation) are important landscape considerations that influence these processes. Concepts borrowed from the theory of island biogeography indicate that, all else being equal, larger areas, or small areas with close neighbors, provide for increased diversity and dispersal potential and lower extinction rates (Diamond, 1975; Wilson and Willis 1975; Harris, 1984). An interesting approach toward solving the problem of habitat fragmentation and concomitant species extinction on small preserves would be management of corridors, such as highway right-of-ways and stream floodplains, for native vegetation in order to provide migration routes connecting natural areas (Ode, 1972). Thus, landscape considera¬ tions regarding the size and placement of nature preserves are important considerations. Unfortunately, we are far from achieving an integrated, biogeographic approach to natural area protection and management, and decisions usually must be made on a local, case-by-case basis (Godron and Forman, 1983). Summary The plant community classification presented in Tables 1-4 is the only state-wide classification at so fine a scale of resolution. Preliminary conservation ranks, based on rarity of relicts and current protection status, have been assigned to each community type. These should be subject to modification, pending acquisition of additional quantitative data, which are lacking for much of Texas. Significant natural communities on state and federal lands and private preserves are not always recognized or managed with conservation of the resource as the PLANT COMMUNITIES OF TEXAS 217 primary goal. An analysis of the late serai stage communities already in public ownership or in private preserves would ensure that significant communities are recognized and would help in setting conservation priorities. Funds and expertise are needed for the proper management of existing preserves and the establishment of new preserves. A biogeogra¬ phic approach toward natural area conservation would improve the probability for effective long-term protection, and conservationists should make landscape ecology an important consideration. Acknowledgments My thanks go to the numerous reviewers of the initial community classification, and to Mr. Edwin L. Bridges, Dr. Paul A. Harcombe, Dr. Fred R. Gehlbach, and Dr. Fred E. Smeins for review of the completed manuscript. Literature Cited Adams, D. B. 1979. Vegetation-environment relationship in Palo Duro Canyon, West Texas. Unpublished Ph.D. dissertation, Univ. Oklahoma, Norman, 134 pp. Ajilvsgi, G. 1979. Wild flowers of the Big Thicket. Texas A&M Univ. Press, College Station, 361 pp. Allen, H. G. 1974. Woody vegetation of the lower Navasota River waterhsed. Unpublished M.S. thesis, Texas A&M Univ., College Station, 80 pp. Allred, B. E. 1956. Mixed Prairie in Texas. Pp. 209-254, in Grasslands of the Great Plains (J. E. Weaver and T. J. Fitzpatrick, eds.), Johnsen Publ. Co., Lincoln, Nebraska, 395 pp. Beuchner, H. K. 1944. The range vegetation of Kerr County, Texas, in relation to livestock and white-tailed deer. Amer. Midland Nat., 31:697-713. Brown, D. E. 1982. Biotic communities of the American southwest — United States and Mexico. Desert Plants, 4:1-342. Brown D. E., C. H. Lowe, and C. P. Pace. 1979. A digitized classification system for the biotic communities of North America, with community (series) examples for the southwest. Appendix I, Arizona-Nevada Acad. Sci., 14:1-16. Brown, J. W., and J. L. Schuster. 1969. Effects of grazing on a hardland site in the southern High Plains. J. Range Manag., 22:418-423. Bruner, W. E. 1931. The vegetation of Oklahoma. Ecol. Monogr., 1:99-188. Buck, M. C., and T. E. Paysen. 1984. A system of vegetation classification applied to Hawaii. U.S.D.A., F.S. Tech. Rept., PSW-76: 1-11. Burgess, T. L., and D. K. Northington. 1974. Desert vegetation of the Guadalupe Mountains region. Pp. 229-242, in Biological resources of the Chihuahuan Desert region (R. H. Wauer and D. H. Riskind, eds.), Nat. Park Serv. Trans. Proc. Ser., 6:xxii + 1-658. Carls, E. G., and A. K. Ludeke. 1984. State park contributions to natural area protection in Texas, USA. Biol. Conserv., 28:95-110. Chabreck, R. H. 1972. Vegetation, water and soil characteristics of the Louisiana coastal region. Louisiana St. Univ. Agric. Exp. Sta. Bull., 664:1-72. Chambless, L. F., and E. S. Nixon. 1975. Woody vegetation — soil relations in a bottomland forest of East Texas. Texas J. Sci., 26:407-416. Clover, E. U. 1933. Vegetational survey of the lower Rio Grande Valley, Texas. Madrono, 4:41-66,77-100. Davis, A. M. 1942. A study of Boscaje de la Palma in Cameron County, Texas, and Sabal texana. Unpublished M.S. thesis, Univ. Texas, Austin, v + 6-1 1 1 pp. 218 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Davis, R. B., and R. L. Speicer. 1965. Status of the practice of brush control on the Rio Grande Plains. Texas Parks and Wildlife Bull., 46:1-40. Diamond, D. D. 1980. Remnant plant communities of the Fayette Prairie. Unpublished M.S. thesis, Texas A&M Univ., College Station, viii + 80 pp. - . 1983. Composition, diversity, and interspecific relationships of grasslands within the True and Upper Coastal Prairies of North America. Unpublished Ph.D. dissertation, Texas A&M Univ., College Station, xii + 155 pp. Diamond, D. D., and F. E. Smeins. 1984. Remnant grasslands and ecological affinities of the Upper Coastal Prairie. Southwestern Nat., 29:321-334. - . 1985. Composition, classification and species response patterns of remnant tallgrass prairies in Texas. Amer. Midland Nat., 113:94-308. Diamond, J. M. 1975. The island dilema: lessons of modern biogeographic studies for the design of nature reserves. Biol. Conserv., 8:73-88. Drawe, D. L., K. R. Krather, W. H. McFarland, and D. D. Netter. 1979a. Vegetation and soil properties of five habitat types of North Padre Island. Texas J. Sci., 33:145-157. Drawe, D. L., R.D. Chamrad, and T. W. Box. 1979b. Plant communities of the Welder Wildlife Refuge. Welder Wildlife Found. Contrib., 5-1:1-38. Driscoll, R. S., E. L. Merkel, D. L. Radloff, D.E. Snyder, and J. S. Hagihara. 1984. An ecological land classification framework for the Unived States. U.S.D.A., F.S. Misc. Publ., 1439:1-56. Dunlap, O. W. 1983. A quantitative descriptive study of the grassland vegetation and soils of the eastern Edwards Plateau. Unpublished M. S. thesis, Univ. Texas, Austin, 133 pp. Dyksterhuis, E. J. 1946. The vegetation of the Fort Worth Prairie. Ecol. Monogr., 16:1- 29. - . 1948. The vegetation of the western Cross Timbers. Ecol. Monogr., 18:325-375. Eyre, F. H. 1980. Forest cover type of the United States and Canada. Soc. Amer. Foresters, Washington, D.C., vi + 148 pp. Fanning, C. D., C. M. Thomas, and D. Issacs. 1965. Properties of saline range soils of the Rio Grande Plain. J. Range Manag., 18:190-193. Ford, A. L., and O. W. Van Auken. 1982. The distribution of woody species in the Guadalupe River floodplain forest on the Edwards Plateau of Texas. Southwestern Nat., 27:383-392. Fowler, N. L., and D. W. Dunlap. 1986. Grassland vegetation of the eastern Edwards Plateau. Amer. Midland Nat., 115:146-155. Game, M., and G. F. Peterken. 1984. Nature reserve selection strategies in the woodlands of central Lincolnshire, England. Biol. Conserv., 28:157-181. Gehlbach, F. R. 1967. The vegetation of the Guadalupe Escarpment. Ecology, 48:404-419. - . 1975. Investigation, evaluation, and priority ranking of natural areas. Biol. Conserv., 8:79-88. - . 1988. Woodlands and forests of the Balcones Scarp Zone in central Texas: structure, succession and evaluation for landscape planning. In Vegetational landscapes of the Balcones Escarpment and Edwards Plateau (B. B. Amos and F. R. Gehlbach, eds.), Baylor Univ. Press, Waco, in press. Godron, M., and R. T. Forman. 1983. Landscape modification and changing ecological characteristics. Pp. 12-28, in Disturbance and ecosystems (H. A. Mooney and M. Godron, eds.). Springer- Verlag, New York, ciii + 291 pp. Gould, F. W. 1975. Texas plants — a checklist and ecological summary. Texas Agric. Exp. Sta., MP-585(revised):l-121. Harris, L. D. 1984. The fragmented forest: island biogeography theory and the preservation of diversity. Univ. Chicago Press, Chicago, xviii + 211 pp. Hinton, J. Z. 1981. Vegetation, edaphic, and topographic relationships in an East Texas forest. Unpublished M. S. thesis, Texas A&M Univ., College Station, 78 pp. PLANT COMMUNITIES OF TEXAS 219 Johnston, M. C. 1952. Vegetation of eastern Cameron County, Texas. Unpublished M. S. thesis, Univ. Texas, Austin, 127 pp. - . 1955. Vegetation of the Eolian Plain and associated coastal features of southern Texas. Unpublished Ph.D. dissertation, Univ. Texas, Austin, 167 pp. - . 1963. Past and present grasslands of southern Texas and northern Mexico. Ecology, 44:456-466. - . 1974. Brief resume of botanical, including vegetational, features of the Chihuahuan Desert region with special reference to uniqueness. Pp. 351-359, in Biological resources of the Chihauhuan Desert region (R. H. Wauer and D. H. Riskind, eds.), Nat. Park Serv. Trans. Proc. Ser., 3:xii + 1-658. Johnston, M. C., and J. Henrickson. 1987. Chihauhuan Desert flora. Chihuahuan Desert Res. Inst., Alpine, in press. Judd, F. W., R. I. Lonard, and S. L. Sider. 1977. The vegetation of South Padre Island in relation to topography. Southwestern Nat., 22:31-48. Klopatek, J. M., J.T. Kitchings, R. J. Olson, K. D. Kumar, and L. K. Mann. 1981. A hierarchial system for evaluating regional ecological resources. Biol. Conserv., 20:271- 286. Kiichler, A. W. 1964. Potential natural vegetation of the conterminous United States. Amer. Geogr. Soc. Spec. Publ., 36:1-116. Larkin, T. J., and G. W. Bomar. 1983. Climatic atlas of Texas. Texas Dept. Water Resources, Austin, ix + 151 pp. Launchbaugh, J.R. 1955. Vegetational changes in the San Antonio Prairie associated with grazing, retirement from grazing, and abandonment from cultivation. Ecol. Monogr., 25:39-57. LBJ School of Public Affairs. 1978. Preserving Texas’ natural heritage. LBJ School Publ. Aff., Pol. Res. Proj. Rept., 3 1 :ix + 1-34. Lonard, R. I., and F. W. Judd. 1980. Phytogeography of South Padre Island, Texas. Southwestern Nat., 25:313-322. Lonard, R. I., J. H. Everitt, and F. W. Judd. 1987. Woody plants of the lower Rio Grande Valley, Texas. Texas Memorial Museum, Univ. Texas, Austin, in press. Marcy, L. E. 1982. Habitat-types of the Eastern Cross Timbers of Texas. Unpublished M.S. thesis, Texas A&M Univ., College Station, 175 pp. Marietta, K. L., and E. S. Nixon. 1983. Vegetational analysis of a post oak-black hickory community in eastern Texas. Texas J. Sci., 35:197-204. Marks, P. L., and P. A. Harcombe. 1981. Forest vegetation of the Big Thicket, Southwest Texas. Ecol. Monogr., 51:287-305. McBryde, J. B. 1933. The vegetation and habitat factors of the Carizo sands. Ecol. Monogr., 3:247-297. McCaleb, J.E. 1954. An ecological and range vegetation analysis of the upland sites of the southern extension of the Oak-Hickory Forest Region in Texas. Unpublished Ph.D. dissertation, Texas A&M Univ., College Station, 99 pp. McMahan, C. A., R. G. Frye, and K. L. Brown. 1984. The vegetation types of Texas, including cropland. Texas Parks and Wildlife Dept., Austin, ii + 40 pp. McMillan, C. 1971. Environmental factors affecting seedling establishment of black mangrove on the central Texas coast. Ecology, 52:927-930. Mohler, C. L. 1979. An analysis of floodplain vegetation of the lower Neches River drainage, Southeast Texas, with some consideration of the use of regression and correlation in plant synecology. Unpublished Ph.D. dissertation, Cornell Univ., Ithaca, 681 pp. Mueggler, W. F., and W. L. Stewart. 1980. Grassland and shrubland habitat types of western Montana. U.S.D.A., F.S. Gen. Tech. Rept., INT-66T-154. Nixon, E. S. 1975. Successional stages in a hardwood bottomland forest near Dallas, Texas. Southwestern Nat., 20:232-235. 220 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Nixon, E. S., L. F. Chambless, and J. L. Malloy. 1973. Woody vegetation of a Palmetto [Sabal minor (Jacq.) Pers.] area in East Texas. Texas J. Sci., 24:535-541. Nixon, E. S., B. L. Ehrahart, J. S. Jasper, and R. L. Ward. 1983. Woody streamside vegetation of Prairie Creek in East Texas. Texas J. Sci., 35:205-213. Nixon, E. S., K. L. Marietta, R. O. Littlejohn, and H.B. Weyland. 1980. Woody vegetation of an American beech ( Fagus grandifolia) community in eastern Texas. Castaenea, 45- 171-180. Nixon, E. S., and J.A. Rains. 1976. Woody creekside vegetation of Nacogdoches County, Texas. Texas J. Sci., 27:443-452. Nixon, E. S., and J. R. Ward. 1986. Floristic composition and management of East Texas pitcher plant bogs. Pp. 283-293, in Wilderness and natural areas in the eastern United States: a management challenge (D. L. Kulhavy and R. N. Conner, eds.), Center Appl. Studies, S. F. Austin St. Univ., Nacagoches, 416 pp. Nixon, E. S., R. L. Whillett, and P. W. Cox. 1977. Woody vegetation of a virgin forest in an eastern Texas river bottom. Castaenea, 42:227-239. Ode, A. H. 1972. A rationale for the use of prairie species in roadside vegetation management. Midwest Prairie Conf. Proc., 2:174-179. Owen, J. G., and D. J. Schmidly. 1986. Environmental variables of ecological importance in Texas. Texas J. Sci., 38:99-1 19. Paysen, T. E., J. A. Derby, H. Black, Jr., V. C. Bleich, and J. W. Mincks. 1980. A vegetation classification system applied to southern California. U.S.D.A., F.S. Tech Rept., PSW-45:1-31. Penfound, W. T. 1952. Southern swamps and marshes. Bot. Rev., 7:413-446. Pfister, R. D., and S.F. Arno. 1980. Classifying forest habitat types based on potential climax vegetation. For. Sci., 26:52-70. Pickett, S. T. A., and J. N. Thompson. 1979. Patch dynamics and the design of nature preserves. Biol. Conserv., 13:27-37. Rice, E. L., and W. T. Penfound. 1959. The upland forests of Oklahoma. Ecology, 40:593- 608. Riskind, D. H. 1976. Big Bend ranch, Cienega Creek and environs, Presidio County, Texas: an appraisal of the botanical resources. Unpublished report for Texas Parks and Wildlife Dept., Austin, 8 pp. Riskind, D. H., and D. D. Diamond. 1988. Introduction to the natural environment. In Vegetational landscapes of the Balcones Escarpment and Edwards Plateau (B. B. Amos and F. R. Gehlbach, eds.), Baylor Univ., Press, Waco, in press. Risser, P. G., and E. L. Rice. 1971. Phytosociological analysis of an Oklahoma upland forest. Ecology, 52:940-945. Rowell, C. M. 1949. A preliminary report on the floral composition of a Sphagnum bog in Robertson County. Texas J. Sci., 1:50-53. Sellards, E. H., W. S. Adkins, and F. B. Plummer. 1966. The geology of Texas, vol. 1. Stratigraphy. Univ. Texas, Austin, 146 pp. Smeins, F. E., and D. D. Diamond. 1986. Grasslands and savannahs in central Texas: protection, status and management. Pp. 381-394. in Wilderness and natural areas in the eastern United State: a management challenge (D. A. Kulhavy and R. N. Conner, eds.). Center Appl. Studies, S. F. Austin St. Univ., Nacogdoches, 416 pp. Smeins, F. E., T. W. Taylor, and L. B. Merrill. 1976. Vegetation of a 25 year exclosure on the Edwards Plateau, Texas. J. Range Manag., 29:24-29. Tharp, B. C. 1926. Structure of the Texas vegetation east of the 98th meridian. Univ. Texas Bull., 2606:1-97. - . 1939. The vegetation of Texas. Anson Jones Press, Houston, xvi + 74 pp. Tolstead, W. Z., and V. L. Cory. 1946. Notes on the flora of Taylor County, Texas. Field and Lab., 14:32-48. PLANT COMMUNITIES OF TEXAS 221 Trans, W. 1974. Priority ranking of biotic natural areas. Michigan Bot., 13:31-39. Tubbs, C. R., and J. W. Blackwood. 1971. Ecological evaluation of land for planning purposes. Biol. Conserv., 3:169-171. UNESCO. 1973. International classification and mapping of vegetation. UNESCO, Paris, 37 pp. Van Auken, O. W., and J. K. Bush. 1985. Secondary succession on terraces of the San Antonio River. Torry Bot. Club, 112:158-166. Van Auken, O. W., A. L. Ford, and J. L. Allen. 1981. An ecological comparison of upland deciduous and evergreen forests of central Texas. Amer. J. Bot., 68:1249-1256. Van Auken, O. W., A. L. Ford, and A. Stein. 1979. A comparison of some woody upland and riparian plant communities of the southern Edwards Plateau. Southwestern Nat., 24:115-180. Van Auken, O. W., A. L. Ford, A. Stein, and A. G. Stein. 1978. Woody vegetation of upland plant communities in the southern Edwards Plateau. Texas J. Sci., 32:24-35. Ward, J.R. 1983. Woody vegetation of the dry uplands in east Texas. Unpublished M. S. thesis, Stephen F. Austin St. Univ., Nacogdoches, 145 pp. Warnock, B. H. 1970. Wildflowers of the Big Bend county, Texas. Sul Ross St. Univ., Alpine, vix + 151 pp. - . 1974. Wildflowers of the Guadalupe Mountains and the Sand Dunes country, Texas. Sul Ross St. Univ., Alpine, xxviii + 176 pp. Watson, G. 1979. Big Thicket plant ecology, an introduction. Big Thicket Museum, Saratoga, 2nd ed., x + 65 pp. Webster, G. L. 1950. Observations on the vegetation and summer flora of the Stockton Plateau in northeast Terrell County, Texas. Texas J. Sci., 2:234-242. Wilson, E. O., and E. O. Willis. 1975. Applied biogeography. Pp. 522-534, in Ecology and evolution of communities (M. L. Cody and J. M. Diamond, eds.), Belknap Press, Cambridge, Massachusetts, 545 pp. I ' 1 INTRAPOPULATIONAL VARIATION IN TWO SAMPLES OF ARID-LAND FOXES Jerry W. Dragoo, Jerry R. Choate, and Thomas R O’Farrell Museum of the High Plains, Fort Hays State University, Hays, Kansas 67601 (JWD, JRC); EG&G Energy Measurements, Inc., 130 Robin Hill Road, Goleta, California 93117 (TPO) Abstract. — Intrapopulational variation was assessed for 14 cranial measurements in 205 specimens of the San Joaquin kit fox ( Vulpes macrotis mutica) and 157 specimens of the swift fox ( Vulpes velox velox). Variance components analysis revealed that less of the variability within both samples was the result of secondary sexual variation and age variation than of individual variation. Multivariate analysis of variance demonstrated significant differences among age categories but nonsignificant differences among sexes in both species. Tukey’s studentized range tests showed that subadult males and females can be pooled with adult males and females for taxonomic study of geographic variation of these foxes. Old adults are significantly larger than subadults and adults, and should be analyzed separately. The two nominal species exhibit similar patterns of intrapopulational variation. The taxonomic relationship between the swift fox ( Vulpes velox) and the kit fox ( Vulpes macrotis) is uncertain. Most authorities (Creel and Thornton, 1971; Egoscue, 1979; Packard and Bowers, 1970; Thornton and Creel, 1975) regarded them as separate species, but Hall (1981) classified them as subspecies of the same species (the name V velox having priority) based on what was interpreted as gene flow between them in eastern New Mexico and western Texas (Rohwer and Kilgore, 1973). Previous taxonomic studies on these foxes (cited above) dealt with relatively small samples from large geographic areas. The extent of variation within populations was poorly known, and this hindered attempts to evaluate the differences found in specimens from widely separated localities. To resolve this problem, large samples from restricted areas were needed. Two such samples now are available. One was collected from a population of the endangered San Joaquin kit fox ( Vulpes macrotis mutica) on the U.S. Department of Energy’s (DOE) Naval Petroleum Reserve no. 1 (Elk Hills), Kern Co., California, and the other was obtained in western Kansas ( Vulpes velox velox) with cooperation of the Kansas Fish and Game Commission. Dragoo et al. (1986) characterized mensural variation of cranial measurements in the San Joaquin kit fox. The purpose of this study is to compare intrapopulational variation found in that taxon with that found in the sample of the swift fox from western Kansas. Data from these two large samples will be included in another manuscript detailing The Texas Journal of Science, Vol. 39, No. 3, August, 1987 224 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 the systematic and evolutionary relationships of swift and kit foxes based on both morphometric and genetic analyses. Materials and Methods We examined skulls of 362 arid-land foxes (205 of V. m. mutica and 157 of V v. velox). Most (181) of the 205 kit fox skulls were from animals that had been live-trapped and ear- tagged between 1980 and 1986 and subsequently had been found dead of various causes. Twenty-four untagged animals were found dead incidental to other investigations. Researchers of the EG&G Energy Measurements survey team at Elk Hills characterized individuals as puppies, subadults, or adults when they were necropsied based on patterns of tooth eruption and wear, weight, and characteristics of the pelage. Live-trap, mark, and release data indicated that puppies were less than five months old, subadults were between five and 10 months old, and adults were more than 10 months old. The utility of nine sutures as ageing criteria for the few specimens from California not aged at necropsy, and for all specimens from western Kansas, was tested based on the known-age specimens from California. The sutures tested were the occipito-parietal, squamoso-parietal, interparietal, interfrontal, coronal, basioccipito-basisphenoid, maxillary (palate), basisphenoid-presphenoid, and premaxillary-maxillary. These sutures have been used to age other carnivores (Orr et al., 1970); however, we found that all but two (basioccipito-basisphenoid and basisphenoid-presphenoid) are completely fused before the eruption of the permanent dentition in swift and kit foxes. Also, the interfrontal, coronal, and maxillary sutures sometimes separate when skulls are cleaned and dried. Therefore, only the basioccipito-basisphenoid and basisphenoid-presphenoid sutures were used to age specimens. Waithman and Roest (1977) considered specimens adults if the basioccipito-basisphenoid suture was fused. We found that, in subadults, the basioccipito-basisphenoid suture is fused and the basisphenoid-presphenoid suture still is open. In adults, both sutures are fused. As adult foxes mature, the suture lines increasingly become indistinguishable. Therefore, we subdivided the adult age class into adults and old adults based on the distinguishability of these sutures. Dental attrition and date of collection also were evaluated in borderline instances. Attrition of occlusal surfaces of the teeth is pronounced in certain old adults. Date of capture was considered because the San Joaquin kit fox and swift fox produce only one litter per year, the former in February (Morrell, 1972) and the latter in early April (Kilgore, 1969). Specimens from Kansas were acquired from trappers and fur dealers during two harvest seasons (15 November through 15 January of 1981-82 and 1982-83). Therefore, no puppies were included in that sample and, for purposes of comparison, it was necessary to exclude the puppy age class from the California sample. Dragoo et al. (1986) previously ascertained that this age class was highly variable and not useful for taxonomic studies. Fourteen cranial measurements (illustrated by Dragoo et al., 1986) were recorded to the nearest 0.1 mm using dial calipers: greatest length of skull (GLS) — from anterior tip of premaxillary to posterior point of inion (back of skull); nasal length (NL) — from anterior notch to tip of posterior extension of nasal; zygomatic breadth (ZB) — greatest distance across zygomata; breadth of braincase (BB) — maximal breadth of braincase across level of parietal-squamosal suture; postorbital constriction (PC) — least width across frontals at constriction behind postorbial processes; lyre breadth (LB) — distance between temporal ridges at frontal-parietal suture; palatal length (PL) — from anterior surface of premaxillary to posterior border of palate; alveolar length of maxillary toothrow (AL) — from anterior border of alveolus of PI to posterior border of alveolus of M2; rostral breadth at first molar (RBM1) — breadth of rostrum across alveoli of Ml; rostral breadth at canines VARIATION IN ARID-LAND FOXES 225 (RBC) — breadth of rostrum across alveoli of canines; greatest bullar width (GBW) — greatest width of bulla perpendicular to long axis of skull; greatest bullar length (GBL) — from sharp anterior end of tympanic bulla near pterygoid process diagonally to posterior end of bulla adjacent to paraoccipital process; bullar depth (BD) — from dorsal surface of external auditory meatus to bottom of bulla; external auditory meatus height (EAMH) — greatest height of external auditory meatus perpendicular to long axis of skull. All measurements were taken by one person (Dragoo) to minimize sampling error. These measurements have been used in morphometric studies of other canids (Nowak, 1979) and in previous investigations of swift and kit foxes (Waithman and Roest, 1977; Rowher and Kilgore, 1973). We used the VARCOMP procedure of Statistical Analysis System (SAS) 84.2 (SAS Institute Inc., 1982) to determine the proportion of variation attributable to sex, age, sex by age interaction, and “residual” (variation resulting from factors other than sex or age). Descriptive statistics (X, range, SE, and CV) were determined for each variable in each age class for both sexes using the UNIVARIATE procedure of SAS. Animals of unknown sex were not included because one purpose of this analysis was to determine whether sexes could be pooled for taxonomic study of swift and kit foxes. When a particular measurement could not be taken, that specimen was excluded from the sample for that measurement but was included for other measurements. For reasons enunciated by Willig et al. (1986), we used multivariate analysis of variance (procedure GLM — SAS Institute Inc., 1982) to determine whether overall group differences between sexes, among age categories, or among a combination of sexes and age categories (sex by age interaction) were significant. Tukey’s studentized range test (HSD) of the TUKEY option of the GLM procedure then was used to group nonsignificant subsets of sexes within each age class and age classes within each sex while protecting the experimentwise error rate for multiple comparisons. Only 264 specimens were used for this analysis because specimens with missing measurements were not included. Results The two ways of partitioning variance (variance components and sums of squares) of the VARCOMP procedure yielded comparable results. Therefore, only variance components data are listed (Table 1). Variance components estimates generally were larger than sums of squares except for “residual,” which was reversed. “Residual” (which we interpreted as individual variation and sampling error) accounted for well over one-half of the mensural variation in both samples. Less than 16 percent of the mensural variation in the samples was the result of sexual dimorphism. Means for mensural variation resulting from age were higher when partitioned by variance components (18 percent in V. velox and 27 percent in V. macrotis ) than when partitioned by sums of squares (6 percent and 19 percent, respectively). Descriptive statistics for each sex and age class in the samples are given in Table 2. When these samples are compared side by side, V macrotis appears to have a slightly longer skull and broader braincase than V. velox. V. macrotis also has deeper bullae and a larger external auditory meatus. The bullae are approximately the same length and width in the two nominal species. V velox has slightly longer nasals and a wider rostrum at the canines and molars. In old adults of both taxa, lyre 226 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 1. — Percent contribution to the total variance of two populations of arid-land foxes ( Vulpes velox and V. macrotis) by sex (S), age (A), sex by age interaction ( S X A), and residual (R) for 14 morphometric characters. Variance Components S A SX A R velox macrotis velox macrotis velox macrotis velox macrotis GLS 17.52 30.61 61.29 45.59 0.00 3.46 21.19 20.34 NL. 24.92 10.88 32.24 23.23 0.00 12.02 42.84 53.87 ZB 34.51 19.75 5.49 40.83 0.00 7.63 60.00 31.79 BB 18.59 13.50 4.94 19.92 0.00 1.06 76.47 65.52 PC 7.07 1.75 10.37 22.20 0.00 2.43 82.56 73.62 LB 11.18 10.35 27.33 52.84 0.00 6.27 61.49 30.54 PL 24.69 21.57 39.92 46.97 0.00 1.08 35.39 30.38 AL 28.19 23.10 28.72 35.60 0.00 4.93 43.09 36.37 RBM1 22.05 23.99 17.00 25.84 0.00 4.41 60.95 45.76 RBC 22.48 27.61 22.83 20.86 0.00 4.29 54.69 47.24 GBW 0.00 1.72 2.64 12.72 0.00 0.00 97.36 85.56 GBL 1.10 0.94 1.75 20.09 0.00 13.04 97.15 65.93 BD 0.00 5.76 0.00 11.75 6.88 0.00 93.12 82.49 EAMH 10.00 3.33 2.22 3.33 0.00 0.00 87.78 93.34 MEAN 15.88 13.92 18.34 27.27 0.49 4.33 65.29 54.48 breadth usually was less than 10 mm, greatest length of skull generally was greater than 120 mm, and postorbital constriction became narrower, proportionally, as length of skull increased. These characters also may be used to age specimens. Multivariate analysis of variance revealed that differences among age categories were significant (P > F = 0.0001 in V macrotis, P> F = 0.01 in V. velox , both based on Wilks’ criterion for calculation of exact F) but that differences between sexes or resulting from sex by age interaction were nonsignificant (P > 0.05). Tukey’s studentized range tests (Table 3) grouped adult males and females in the same nonsignificant subset for all but one multiple comparison (GLS in V. macrotis ), and grouped subadult males and females in the same nonsignificant subset for all multiple comparisons. The four sex and age classes (subadult and adult males and females) were grouped in the same nonsignificant subset for 12 multiple comparisons for V. velox and for four multiple comparisons for V macrotis. The F values for three measurements (GBW, GBL, and BD) for V. velox were less than 1.0 and were regarded as nonsignificant. Discussion Grinnell et al. (1937) presented a table of measurements comparing males with females and demonstrated, using descriptive statistics, that 2, subadults; 3, adults; and 4, old adults. VARIATION IN ARID-LAND FOXES O'- ON >n « ON o vo >n on vo on — ' ON on -d OO in (N vq o r- o o d - n Tf on n Tf n OV oo - oo vq vq n ^ ov — ' — od oi n on on on on on Tfr vo Tf Tf on c- r-- — n o d o n © o o % o o r- ^ Ov ON O — Tf n- Ti¬ ed o x> o 5 VO O — ON g o on m in o — ' — ; o d 00 >v N Tt Ov on r~- Tf Tf Tf Tf Tt on on >n © cd on" >n in TfTfTtTt o cfv cd on m in vo on «n o r-" oo’ on" Tf © o o — — ON ON Ov ON Tf on ON oo Tt ON ov tt on r- — on tt in vd od on on on on on on © m n- on on — > oo ti- ON -OOOOOOvl oo oo Tf on m oo oo vo m ov — r- Tf — ov oo on vo m o-oqovn ON ON ON ON ON ■rr Tt Tf Tf Tt Ov Ov O Tf — CO tT cd --3 cd — — C cd ^ cd cd g £ § £ £ on on on on < Postorbital constriction female 10 22 23.7 22.5 0.60 0.50 22.1- 24.8 20.6- 25.0 4.01 5.23 male 3 10 23.6 22.7 0.87 0.57 22.7- 24.1 21.1- 24.0 3.21 3.95 female 72 83 23.1 21.6 0.32 0.21 19.6- 26.1 19.8- 23.8 5.88 4.47 228 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 m o Os Os l"* sO sc> — ; © ^ Os ON o o on n© r~~’ (N (N (N - o rt < — < ON — on o on © ON Os no on t~~ oo on © m rf nO OO Os O ON ^ iq vo « on « O 6 « - ni on on on on on m no O oo cd cd 6 £ on on on on -rt S c £ c c aj c c c ON ON on no tJ- «J -© cd — « — g g g cd cd , p « vq fn oo vo so wd so rd in on_ in n on in w m rn m ^ OO rt VO VO CN wd yd vo in vd VO p p so — ; cn cd cn cn vO cn © oo rd o© on on vo oo nr oo p vd r~~ vo‘ vd on © p p wn d in vq p rd rd cn p p rd rd nf m cn cn cn cn cn m m cn oo cn vo r» © — < © cn on cn cn — i vo CN >n CN »n — i oo vo m oo q oc « nr r - nr nr nr" © cn in © •n h q q in -d nf nf p p vq CN rd rd vq on ed rd oo — cn m O -d © -d — * — ; oq oq r - CN -d CN CN o ^ © -d — h m © -d CN CN m — i © © o oo £ o* © © o* © m © © © © O J3 o el 73 o o o o o o o o x> OO nf vo VO co CN OO — — i •H O O O O c3 o o o o -£ oo m nr nr 3 n « OQ O O O O o — ; p DO ON nf vq vq O nr" nt A «n p —< p p p rd ©o rd o© on CN — I OO CN cn >n vq «n p nf Nr nr* nf nf ~ vo cn in cn cn vo vo vo *n CN CN CN 04 CN p —i p CN VO CN CN CN CN — i vo cn nr >n o o 0) £ On Fx o •3 8 § ° 13 W ^ oj> cl ^ £2 Cd g £ g £ £ cn cn m cn nr 73 ^ 73 73 cj gs cs cd cd g £ g £ £ cn cn m to ^ £2 cl ^ ^ g £ g £ £ cn cn to cn tT --j £ ^ 0.05). Mean dry weight for Acacia smallii plants grown in the open (full light) was 1.79 ± 0.99 FACTORS AFFECTING SEEDLING GROWTH IN ACACIA 235 Figure 1. Bar graph showing mean dry weight (+ 1 SD) of Acacia smallii grown for 12 weeks under various field conditions. Treatments were as follows: 1, open; 2, under canopy; 3, herbivory; 4, no herbivory; 5, competition; 6, no competition (trenched plots); 7, nutrient supplement; 8, no nutrients added. Plants in treatment 1 were significantly different from those in treatment 2 (ANOVA, LSD, P < 0.05). There were no significant differences between the other treatments. grams, whereas the mean dry weight of those under the canopy (shade plants) was 0.96 ± 0.44 grams. Mean dry weight of the zero-time plants was 1.07 ± 0.26 grams. Canopy plants were slightly lower in mean dry weight than the zero-time plants (P > 0.05). Means of all treatments were separated using the least significant difference test (Table 1). The main differences are in canopy position. Other differences should not be considered because the major effects were not significant when tested with ANOVA. Rainfall during the experiment was 62 percent greater than normal and did not seem to be a factor. There were only three weeks (5, 10, and 11) during the experiment that did not have any rainfall and plant mortality was only two percent during the experiment (Fig. 2). In addition, the mortalities occurred in the canopy treatments and they did not occur during brief periods of drought. Discussion During most successional events, biotic and abiotic conditions change (Stevens and Walker, 1970; Brady, 1974; Armson, 1977; Gorham et al., 1979). Surface light-intensity is high early in succession and decreases as 236 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 1. Mean aboveground dry weight (± 1 SD) for Acacia smallii grown in 16 treatments and zero-time measurements. Mean values followed by the same letter are not significantly different (ANOVA and Least Significant Difference Test P> 0.05). Treatment Dry-weight (grams) Zero time Under canopy, herbivory, competition, nutrients Under canopy, herbivory, competition, no nutrients Under canopy, herbivory, no competition, nutrients Under canopy, herbivory, no competition, no nutrients Under canopy, no herbivory, competition, nutrients Under canopy, no herbivory, competition, no nutrients Under canopy, no herbivory, no competition, nutrients Under canopy, no herbivory, no competition no nutrients Open, herbivory, competition, nutrients Open, herbivory, competition, no nutrients Open, herbivory, no competition, nutrients Open, herbivory, no competition, no nutrients Open, no herbivory, competition, nutrients Open, no herbivory, competition, no nutrients Open, no herbivory, no competition, nutrients Open, no herbivory, no competition, no nutrients 1.07 ± 0.27AB 0.81 +0.43A 0.97 ± 0.30AB 0.75 ± 0.36A 1.05 ± 0.60AB 1.15 ± 0.61 AB 0.99 ± 0.32AB 1.04 + 0.46 AB 0.98 ± 0.44AB 1.16 ± 0.27AB 1.72 + 0.74BCD 1.74 + 1.22BCD 1.75 + 0.61BCD 2.15 ± 0.94CD 2.04 ± 0.85CD 1.32 ± 1.37ABC 2.47 + 1.46D forest community development proceeds (Bazzaz, 1979). Usually soil nitrogen is low during the early stages of succession and increases in time (Gorham et al., 1979). Consequently, colonizers are usually sun plants (heliophytes) tolerant of low soil nitrogen, whereas species of mature communities are usually shade plants (sciophytes) that require higher levels of soil nitrogen (Bormann, 1953; Grime, 1965; Loach, 1967; Ormsbee et al., 1976). Van Auken and Bush (1985) showed that Acacia smallii occured early during secondary succession on river terraces in southern Texas. In addition, they showed that A. smallii was a sun plant (Bush and Van Auken, 1986a). Acacia smallii is also tolerant of low soil nitrogen (Van Auken et al., 1985), and soil nitrogen increased during this successional sequence (Bush and Van Auken, 1986b). However, the above studies were either field observations or greenhouse experiments. The present experimental field study showed that canopy position was the most important of the factors evaluated concerning establishment and growth of Acacia smallii. Seedlings planted in the open grew 1.86 times faster than those under the A. smallii canopy. This confirms the previous light- limited growth experiments carried out in the greenhouse and the proposed successional position of A. smallii. Other factors examined, including competition, nutrient supplements, and herbivory, did not cause significant changes in A. smallii growth. Competition for soil-borne resources, mainly water and nutrients, can FACTORS AFFECTING SEEDLING GROWTH IN ACACIA 237 O 2 Z < Figure 2. Rainfall and mortality bar graph of Acacia smallii seedlings planted in the field. play a significant role in establishment and subsequent growth of species (Harper, 1977). Trenching can quite effectively eliminate root competition from study plots (Ehrenfeld, 1980; Horn, 1985). We did not see a difference in A. smallii growth between trenched and nontrenched plots and concluded that competition between adult and seedling plants was unimportant. Competition with woody plant roots may have been masked by competition with herbaceous plants that were not removed from the study plots (both canopy and open plots). Competition between woody plant seedlings and herbaceous plants can cause significant growth reduction in one or the other species (Harper, 1977). Cohn and Van Auken (unpublished results) showed significant growth reduction of A. smallii by Cynodon dactylon (bermuda grass) in a greenhouse study. Similar studies with Prosopis glandulosa (honey mesquite) and various grasses had the same results (Glendening and Paulsen, 1955; Van Auken and Bush, 1987). Nutrient additions were not expected to have a great effect on the growth of A. smallii because of previous studies (Van Auken et al., 1985). However, because of the higher levels of nitrogen and carbon found below the A. smallii canopy, a nutrient treatment was included. Additional nitrogen in the canopy soil could have ameliorated the growth of A. smallii in spite of the shading effects caused by the canopy. The 20 percent decrease in growth of A. smallii in the open when nutrients were added was probably the result of stimulation of associated herbaceous species by the removal of a nutrient limitation. With the growth limitation removed, the herbaceous species may have used up a nutrient required for A. smallii growth. Herbivores can have drastic effects on plant populations and change successional sequences (Harper, 1977; Gilbert, 1985), although not shown 238 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 in the present study. This study was completed in the autumn and was of limited duration. A study completed over a longer time and including both the spring and fall growing season might show significant growth reduction due to herbivory. Although many factors may effect seedling establishment and growth of woody plants, the present field study identified seedling position relative to the canopy as the most important. Canopy shade inhibits A. smallii growth whereas the full sun of open areas promotes it. Canopy shade may be the major environmental variable required to explain the lack of A. smallii seedlings under the A. smallii canopy. Acacia smallii appears to be a heliophyte and early colonizer of old-fields and overgrazed grasslands of southern Texas and associated areas. Acknowledgments We would like to thank P. Fonteyn for the generous use of his light meter. Special thanks go to J. K. Bush for help with the planting and statistical analysis. Literature Cited Armson, K. A. 1977. Forest soil-properties and processes. Univ. Toronto Press, Toronto, 390 pp. Bazzaz, F. A. 1979. The physiological ecology of plant succession. Ann. Rev. Ecol. Syst., 10:351-371. Boardman, N. K. 1977. Comparative photosynthesis of sun and shade plants. Ann. Rev. Plant Physiol., 28:335-377. Bormann, F. H. 1953. Factors determining the role of loblolly pine and sweetgum in early oldfield succession in the Piedmont of North Carolina. Ecol. Monogr., 23:339-358. Brady, N. C. 1974. The nature and properties of soils. MacMillan, New York, 653 pp. Bush, J. K., and O. W. Van Auken. 1986a. Light requirements of Acacia smallii and Celtis laevigata in relation to secondary succession on floodplains of South Texas. Amer. Midland Nat., 115:118-122. - . 1986b. Changes in nitrogen, carbon, and other surface soil properties during secondary succession. Soil Sci. Soc. Amer. J., 50:1597-1601. Ehrenfeld, J. G. 1980. Understory response to canopy gaps of varying size in a mature oak forest. Bull. Torrey Bot. Club, 107:29-41. Gilbert, L. E. 1985. Ecological factors which influence migratory behavior in two butterflies of the semi-arid shrublands of south Texas. Pp. 724-747, in Migration: mechanisms and adaptive significance (M. A. Rankin, ed.), Contrib. Marine Sci., 27 (suppl.): 1-868. Glendening, G. E., and H. A. Paulsen, Jr. 1955. Reproduction and establishment of velvet mesquite as related to invasion of semidesert grasslands. USDA Tech. Bull., 1127:1-50. Gorham, E., P. M. Vitousek, and W. A. Reiners. 1979. The regulation of chemical budgets over the course of terrestrial ecosystem succession. Ann. Rev. Ecol. Syst., 10:53-84. Grime, J. P. 1965. Shade tolerance in flowering plants. Nature, 208:161-163. Harper, J. L. 1977. The population biology of plants. Academic Press, London, 892 pp. Horn, J. C. 1985. Responses of understory tree seedlings to trenching. Amer. Midland Nat., 114:252-258. Johnson, N. D., C. C. Chu, P. R. Ehrlich, and H. A. Mooney. 1984. The seasonal dynamics of leaf resin, nitrogen, and herbivore damage in Eriodictyon californicum and their parallels in Diplacus aurantiacus. Oecologia, 61:398-402. FACTORS AFFECTING SEEDLING GROWTH IN ACACIA 239 Loach, K. 1967. Shade tolerance in tree seedlings. I. Leaf photosynthesis and respiration in plants raised under artificial shade. New Phytol., 66: 607-621. NOAA. 1984. Local climatological data, annual summary with comparative data, San Antonio, Texas. National Climatic Data Center, Asheville, North Carolina, 8 pp. Ormsbee, P., F. A. Bazzaz, and W. R. Boggess. 1976. Physiological ecology of Juniperus virginiana in old fields. Oecologia, 23:75-82. SAS Institute. 1982. SAS user’s guide. SAS Institute Inc., Cary, North Carolina, 584 pp. Smith, R. L. 1980. Ecology and field biology. Harper and Row, New York, 835 pp. Steel, R. G. D., and J. H. Torrie. 1980. Principles and procedures of statistics: a biometric approach. McGraw-Hill, New York, 633 pp. Stevens, P. R., and T. W. Walker. 1970. The chronosequence concept and soil formation. Quart. Rev. Biol., 45:333-350. Taylor, F. B., R. B. Hailey, and D. L. Richmond. 1966. Soil Survey of Bexar County, Texas. USDA Soil Conserv. Serv., Washington, D.C., 126 pp. Van Auken, O. W., and R. Kapley. 1979. Principles of ecology: laboratory manual. Burgess Publ. Co., Minneapolis, 175 pp. Van Auken, O. W., and J. K. Bush. 1985. Secondary succession on terraces of the San Antonio River in South Texas. Bull. Torrey Bot. Club., 1 12:158-166. - . 1987. Inhibition of honey mesquite growth by bermuda grass. J. Range Manage., in press. Van Auken, O. W., E. M. Gese, and K. Connors. 1985. Fertilization response of early and late successional species: Acacia smallii and Celtis laevigata. Bot. Gaz., 146:564-569. Weaver, J. E., and F. E. Clements. 1966. Plant ecology. McGraw-Hill, New York, 601 pp. I NATURAL HISTORY SKETCHES, DENSITIES, AND BIOMASS OF BREEDING BIRDS IN EVERGREEN FORESTS OF THE RIO GRANDE, TEXAS, AND RIO CORONA, TAMAULIPAS, MEXICO Frederick R. Gehlbach Department of Biology, Baylor University, Waco, Texas 76798 Abstract. Comparative aspects of life history, including weights and densities, are presented for 24 breeding species at the Rio Grande in Texas and 40 at the Rio Corona, in Tamaulipas, 18 of which are the same. The information suggests greater stability with a greater density of birds, especially passerines, and greater ecological influence of all breeding species at the Rio Corona. Avian biomass is 2.7 times larger at the Rio Corona, where wet-season rainfall is 2.5 times greater and may account for the inter-site difference through increased primary productivity. Key words', avifaunas; life histories; densities; biomass. Over two decades I studied breeding birds at the Santa Ana National Wildlife Refuge on the Rio Grande, Texas, and near Guemez, Tamaulipas, on the Rio Corona (Gehlbach et ah, 1976; Gehlbach, 1981). Mature evergreen forest at the two locales, 330 kilometers apart, was essentially the same; for example, the tree stratum averaged 16.1 and 17.8 meters (m) tall, 29.0 and 33.1 centimeters (cm) in diameter, and contained 161 and 140 trees per hectare with species diversities (H'log2) of 2.43 and 2.31, respectively. However, avifaunas were temperate (Rio Grande) and tropical (Rio Corona) in nature. Unfortunately, study plots at the Rio Corona were destroyed by bulldozing in 1979-1980, so my general investigation of climate and vegetation in relation to avifaunal dynamics ceased in 1978. I here present life history sketches of breeding birds as background to brief, inter-site comparisons of avifaunal biomass and density. Methods Investigations at each locale centered on two, lineal, eight-hectare plots in riparian evergreen forest, bordered by thorn woodland and mesquite grassland on the nonriver side. Breeding (April, June) and winter (December-January, March) community structures were examined, and the habits of breeding birds observed. The latter are categorized as making local movements (LM, absent for a few days or weeks but not the entire season), being migratory (M, absent in the non-breeding season), or resident (R, present in slightly varying numbers year-round). Nomenclature follows the 1983 edition of the “Check-list of North American Birds” except for Vireo flavoviridis ( sensu Johnson and Zink, 1985). Mean weights are based on personal captures and personally examined museum specimens from the region unless credited otherwise. The mean population densities plus or minus one standard deviation concern adults per eight hectares, calculated from four seasonal or eight total spot-map censuses per plot, three to five consecutive days each, in The Texas Journal of Science, Vol. 39, No. 3, August, 1987 242 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 the period 1973-1978. Biomass is mean weight times mean density per species, summed for all species per season per site, except that ducks and kingfishers are excluded because they were not seen to feed in the forest. Breeding Species Black-bellied whistling-duck ( Dendrocygna autumnalis ), 838 grams (E. Bolen, personal communication). — Corona: LM; 2.1 ± 1.3 in April, June; four fledglings on three days in June 1976; nesting likely just outside plots. Grande: LM; June 1974 nest with eggs 3.5 m high in broken limb of tepeguaje ( Leucaena pulverulenta); otherwise absent. Some winter near both study locales. Muscovy duck ( Cairina moschata ), ca. 2000 grams (literature). — Corona:LM; absent December-January; 8-10 sporadically in March; 2.3 ± 1.6 in April, June. Nest not located, but birds persistent in forest interior, April and June. Gray hawk ( Buteo nitidus ), 528 grams N = 13. — Corona: R (1.7 ± 0.7). Defended nest, contents uncertain, in June 1973; another with two feathered nestlings, June 1976; single dependent fledgling, June 1976. Nests 15 m high in coma ( Bumelia lanuginosa) crown, 175 m from river, and 18 m high in upper half of Montezuma cypress ( Taxodium mucronatum) at river bank; both nests about 35 cm diameter, similar in aspect to those of B. platypterus in central Texas (Bush and Gehlbach, 1978). Grande: visitor; immature April 1977, June 1975; adult June 1977, which suggests nesting in the region. Plain chachalaca ( Ortalis vetula ), 563 grams (W. Marion, personal communication). — Corona: R(8.0 ± 4.2). Notably warier than at Rio Grande; mostly in tree crowns, rarely on ground; nests undiscovered. At 0800 in April, six flocked with eight great kiskadees, green jays, and altamira orioles and ate fruit 15 m high in coma and brasil ( Condalia hookeri). Grande: R; 10-16 before 1977 when fed grain, 5-9 thereafter (10.5 ± 3.3). Commonly on ground, less so in shrubs and trees where certain leaves as well as fruits are eaten. Nests, two m (eggs) and 10 m (contents uncertain) high in granjeno ( Celtis pallida) and brasil, respectively, June 1974, 1976; fledglings 18 June 1976 (see Marion and Fleetwood, 1978). Red-billed pidgeon ( Columba flavirostris ), 244 grams, N = 18. — Corona: LM; December-March birds flock and sun in tree crowns (11.1 ± 3.6); nesting population is but 5.3 ± 0.9 in April, June. Persistent singing from ebony (Pithecellobium flexicaule) crowns in April and June, copulation 8 April 1974; nests not located. Grande: LM; absent December-January; five appeared 25 March 1978 but disappeared the next day. Nest 12 m high in ebony crown, contents uncertain, June 1974; calling adult June 1976, but no mate or nest located (0.5 ± 0.7). BREEDING BIRDS IN EVERGREEN FORESTS 243 White-winged dove ( Zenaida asiatica ), 159 grams N = 16. — Corona: M (6.0 ± 2.3). Earliest return 1 1 April 1974. Three nests with two feathered nestlings each in huisache ( Acacia famesiana), 3-5 m high, June 1973 on forest-woodland border. These trees were bulldozed in 1975; doves visited plots but did not nest thereafter. Grande: M; earliest return 4 April 1977, but a few sometimes winter in region. Nests in small trees of the forest-woodland border; not a primary inhabitant of evergreen forest at either study site. White-tipped dove ( Leptotila verreauxi ), 206 grams (W. Shifflet, personal communication). — Corona: R (5.8 + 1.2). Foraged on ground, usually in open areas at dawn and dusk; always flew below the canopy and perched below 4 m. Singing increased March through June; nest under construction 2 m high in dense vine tangle 4 m from river, 10 April 1974. Grande: LM; 1-2 more adults in June than December-April (3.0 ± 1.5). April (two eggs) and June (two nestlings) nests in dense subcanopy, 1-5 m high. General behavior as at Rio Corona. Green parakeet ( Aratinga holochlora ), 145 grams, N = 5. — Corona: M; absent December-early March, although present throughout winter 125 kilometers south. Flocks in March and pairs in golden-fronted woodpecker and other cavities, above 7 m, by 8 April 1974, (8.0 + 2.8). April 1974 nest (eggs?) in river bank cypress; June 1976 nest in ebony 75 m from river had single, feathered juvenile; simultaneously, another pair and single helper tended two feathered nestlings in a natural cypress hole in which all five roosted together (also see elegant trogon). Red-crowned parrot ( Amazona viridigenalis) 345 grams, N = 2. — Corona: LM, especially in December-january (9.3 + 0.8). Eight in January 1974, 3-10° C, left for two days when air temperatures dropped below freezing. In April, June, most left plots by 1000 hours, returning to roost at 1700-1800 hours, although one or two pairs often stayed. Copulation on 10 April; pairs in natural cypress cavities and those of lineated woodpeckers, 8 April-21 June 1973 and 1976. Yellow-headed parrot ( Amazona oratrix), 575 grams, N = 2. — Corona: LM; like the red-crowned parrot in daily movements and temporary emigration coincident with freezing weather (4.7 ± 0.8). Pair nested in a cypress cavity, June 1976, but by 1978 most suitable cavities had been opened by Mexicans to obtain nestlings for the pet trade and all parrots were scarce. While wary of man and intolerant of red-crowned parrots normally, in April 1974 both species fed in the same ebonies together, dropping opened seed pods on me 3-4 m below. Yellow-billed cuckoo ( Coccyzus americanus ), 56 grams, N = 13. Grande: M. (2.8 ± 0.8); returned in early April. Nests 2 and 3 m high in brasil and anacua ( Ehretia anacua ), respectively, June 1976. Seemingly the subcanopy, insect-foraging equivalent of the squirrel cuckoo. The yellow-billed cuckoo is a passage migrant at the Rio Corona. 244 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Squirrel cuckoo ( Piaya cayana ), 100 grams, N = 24. — Corona: R (2.7 ± 0.9). Possibly absent until 1975 (tentative sighting April 1974). Courtship feeding in June 1976, but nest not discovered. Foraged and drank at the riverbank in a nine-species, passerine and golden-fronted woodpecker flock, January 1975. Groove-billed ani (Crotophaga sulcirostris ), 77 grams, N = 20. — Corona: LM (3.5 + 1.1); usually absent, December-March but one in January 1975. Five-bird nesting colony in huisache grove at forest- woodland border, June 1973; three in June 1976. One-two active nesets at 2-3 m, not observed closely but seemingly communal (See, Vehrencamp, 1978). Grande: LM (1.2 + 0.5); earliest bird on plot 11 April 1978, but a few usually winter in the region. Unwatched nests below 3 m in huisache and granjeno each June; fledglings in June. Like the white-winged dove in its forest-margin breeding site. Eastern screech-owl ( Otus asio ), 172 grams, N = 52. — Grande: R (1.8 + 0.5). Called December-June unless repressed by weather or visiting great-horned owls ( Bubo virginianus). On 19-20 and 23-25 March 1978, at 1800-1930 hours, two territorial pairs contended for a cavity 4 m high in a sugarberry ( Celtis laevigata ), chiefly with descending trills. Elf owls called simultaneously 100-150 m away (screech-elf interactions were rare because elf owls selected thorn woodland or mesquite grassland, whereas screech owls preferred evergreen forest). Pair with two fledglings in June 1976. Curiously absent from Rio Corona plots but not found nesting within eight hectares of any Glaucidium in my experience. Ferruginous pygmy-owl ( Glaucidium brasilianum), 64 grams, N = 23. — Corona: R (2.5 ± 0.7). Occupied woodpecker holes by late March; on 22 March 1978, two territorial pairs began antiphonal and synchronous calling at 1900 hours with much flying and some screeching within about 900 m2. At dusk 3-7 June 1973, a pair fed cicada nymphs to at least three feathered nestlings 3 m above ground in an old golden- fronted woodpecker hole in a coma. All observed nesting activity was crepuscular or nocturnal. Elf owl ( Micrathene whitneyi ), 35 grams, N = 3. — Grande: M (0.8 ± 0.7); earliest return 19 March 1978. Primarily outside evergreen forest but nest in one plot near its border, June 1976 (see eastern screech owl). Single bird called there but screech owl drove it away at 1900 hours, 18 April 1975. Nest site a ladder-backed woodpecker hole, 3 m high, in dead anachua limb, 5 m from a noisy water pump and 400 m from the closest screech owl nest. Mottled owl ( Ciccaba virgata), 338 grams, N = 14.— Corona: R (1.3 + 0.9); one pair in area of both plots. Rarely heard December-March; more often vocal in April and June except during great-horned owl visits. Much shyer than small owls; daytime vocalizations rare. Nest not located. BREEDING BIRDS IN EVERGREEN FORESTS 245 Common pauraque ( Nictidromus albicollis ), 63 grams, N = 25. — Corona: R (1.3 ± 0.8). Normally silent December-January, sporadic calling in March, frequent April, June. Grande: R (2.2 + 0.7). One flushed at 1500 hours, 27° C, January; but another January bird appeared torpid with only its back exposed in deep leaf litter at 0900 hours and 6°C. One nesting pair per plot but 1-6 calling individuals per 400 m along a gravel road in late March to June. Pauraques aggregated for lek- type (communal) calling in open areas beginning about 1900 hours in March and 2045 hours June; dispersed in 20-40 minutes and, less frequently, reassembled at dawn. Buff-bellied hummingbird ( Amazilia yucatanensis ), 4 grams, N = 8. — Corona: LM (1.8 ± 1.3); like the parrots in emigration during severe December-January weather. Present daily March-June; fed in shrub and herb strata, particularly upon Turk’s cap ( Malvaviscus drummondii). Nest not found. Grande: LM (1.2 ± 0.8); as at Rio Corona but greater fidelity in winter, possibly because of hummingbird feeders and planted nectar sources nearby. Three unwatched nests, June 1974 and 1976, 2.0- 2.5 m high in vine tangles on small trees. Elegant trogon ( Trogon elegans), 70 grams, N = 7. — Corona: R (2.5 ± 0.7). Usually silent and solitary December-January; paired and calling March-June. Only twice in 11 mixed, winter flocks of large passerines were trogons associated. Pair excavated old, natural, ebony cavity 8-10 April; pairs in June used lineated woodpecker holes in black willows ( Salix nigra); 1973 nest two m above river and one m from bank, occupied by nesting green parakeets in June 1976. Two fledglings with male, 23 June 1976. Blue-crowned motmot ( Momotus momota ), 104 grams, N = 22. — Corona: R (3.3 ± 1.5). Hooting infrequent December-January, common March-June. Not seen in mixed species flocks. Sometimes hover-fed on foliage like elegant trogon. Pair investigated bare riverbank 21-23 March 1978, started to dig burrow 2 m above river on 23rd; another pair dug in man-made dirt bank 2 m above river level and 10 m from it, 9-11 April 1974; a third pair repeatedly investigated the soil pit used by ringed kingfishers but was driven off by them (see below). One-two fledglings with adults in June. Ringed kingfisher ( Ceryle torquata ), 330 grams, N = 8. — Corona: R (2.8 ± 0.9). Nestlings in burrow 2.5 m from bottom of 6 by 6 by 4 m (deep) soil pit in the forest, 15 m from river, March 1978. Pair dug burrow 3 m above river in natural cutbank, January 1974, but did not nest in January 1975, a more severe winter. Fledglings in April and June. Adults occasionally perched in forest. Golden-fronted woodpecker ( Melanerpes aurifrons ), 80 grams, N = 28. — Corona: R (4.2 ± 1.2). Adults fed nestlings in ebony snag 3 m high, June 1973, 5 m from old cavity housing ferruginous pygmy owl nestlings. 246 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Fledglings with adults June 1976. Grande: LM; like white-tipped dove with breeding seasonal increase (2.3 ± 0.5 in December-March to 5.0 ± 0.8 in June). June pair with nest 5 m high in tepeguaje stub assisted by adult helper; other June nests 1.5 and 3.5 m high in dead tepeguaje and anachua limbs. Fledglings 21 June 1976. As at Rio Corona, commonly foraged with green jays, altamira orioles, great kiskadees, and other resident passerines in winter. Ladder-backed woodpecker ( Pico ides scalaris ), 33 grams N = 18. — Grande: R (1.8 ± 1.1); mostly peripheral to forest. Nest 2.5 m high in dead mesquite limb at forest-woodland border, June 1976. Solitary winter forager in forest; similar habits but only occasional at Rio Corona. Lineated woodpecker ( Dryocopus lineatus ), 135 grams, N = 13. — Corona: R (2.0 ± 0); single pair all year in the area of both plots like the mottled owl. Pair fed large nestlings in black willow cavity, 3 m high and 1.5 m from river bank, June 1973; nest only 2 m from active elegant trogon nest in older lineated woodpecker cavity in same clump of willows. Brown-crested flycather ( Myiarchus tyr annulus), 38 grams, N = 15. — Grande: M (2.3 ± 0.5); first noted 19 March 1978 but none first week of April 1977. Nest in natural and golden-fronted woodpecker cavities, June 1974-1976; pair with single fledgling 18 June 1975. Species appears to be ecological equivalent of sulphur-bellied flycather at Rio Corona, where the brown-crested is a passage migrant, at least on my plots. Great Kiskadee ( Pitangus sulphur atus), 74 grams, N = 17. — Corona: R (3.8 ± 0.4). Nest building 21 March-11 April, above 3 m, in Montezuma cypress and ebony; nestlings in June 1973, 2.5 m directly below rose-throated becard nestlings. December-March fruit-eating in forest; 70 percent of time spent along river, although fish-catching infrequent by contrast with the Rio Grande. Foraged with other, large passerines but not inclined to move concurrently with jays, orioles, Couch’s kingbirds. Grande: LM? (0.8 ± 0.7). A nest, 5 m high in ebony, built first week of April 1977. No observations of fruit-eating and no known interactions with other flycatchers as at Rio Corona. Sulphur-bellied flycatcher ( Myiodynastes luteiventris ), 48 grams, N = 6. — Corona: M (2.0 ± 0). Absent late December to at least 11 April. Nests above 2 m in Montezuma cypress, black willow and ebony cavities, both woodpecker-drilled and natural, in June; three fledglings with adult pair 21 June 1976. Couch’s kingbird ( Tyrannus couchii ), 39 grams, N = 32. — Corona: LM; influx in December-April (4.5 ± 3.3) versus June (1.7 + 0.5). Ten foraged with four other passerines; ate mostly insects, rarely fruit like associated great kiskadees, 23 March 1978. Nests 2-5 m high in eastern cottonwoods ( Populus deltoides ), April 1974 and June each year. Grande: LM, outflux December-April (1.3 + 0.5) versus June (3.3 + BREEDING BIRDS IN EVERGREEN FORESTS 247 0.4). Do some individuals of this and other tropical species at their northern limits, like the red-billed pigeon and clay-colored robin, move from the Rio Grande area southward in the winter? More solitary in winter than at Rio Corona; used tree canopy and forest interior more frequently than great kiskadee. Nests in a variety of trees, 2-4 m high, late April, June. Rose-throated becard (Pachyramphus aglaiae ), 33 grams, N = 23. — Corona: R (4.2 ± 1.2). Winter mixed foraging with jays, orioles, tanagers. Nesting territories about 150 m apart along river; nests in lower limbs of tepeguaje and cypress, often within a few meters of active altamira oriole and great kiskadee nests, April and June; nestlings in June 1973. Grande: LM?; a female made three unsuccessful nesting attempts in dead lower tepeguaje branches, 2.5 m high, April and June 1977. No male seen. By March 1978, house sparrows ( Passer domesticus) took over the last nest, which then blew down. Green jay ( Cyanocorax yncas), 77 grams (D. Gayou, personal communication). — Corona: R (4.5 ± 3.3). Usually foraged with orioles, tanagers, grosbeaks, kingbirds in December-March; noisiest, most quarrelsome member of such flocks. Nests with eggs (April) and nestlings (June), 3-10 m high, in dense subcanopy trees with vine tangles, attended secretively by more than one pair (see Gayou, 1986). Nested and foraged further from river and foraged less commonly on ground than brown jay. Grande: R (4.1 ± 1.4); typically one flock of four or five per plot. Foraged with altamira orioles, Couch’s kingbirds, and golden-fronted woodpeckers, December-March, as at Rio Corona in canopy and subcanopy, but more frequently on ground at Rio Grande. Courtship in March-April; nesting unobserved. Brown jay ( Cyanocorax mono), 206 grams, N = 12. — Corona: R; usually in monospecific flocks of 4-15, December-March (9.5 ± 3.2). Nest (eggs) in April 1974, attended by five birds; four nestlings fledged 22 June 1976; both nests 11-12 m high in ebony crowns, about 30 centimeters in diameter, built of large twigs and small sticks much like nests of the gray hawk. Courtship feeding in March. On flocking and foraging behavior at the site see Morrison and Slack (1977). Mexican crow ( Corvus imparatus ), 235 grams, N = 8. — Corona: R (4.5 ± 2.3). Flocks of three to nine in December-March; paired in April and carrying sticks. Nesting off plots but fledglings present 21-23 June 1976. Foraged almost exclusively on riparian items but in March 1978 ate fruit in forest. Tufted titmouse (Parus bicolor ), 22 grams, N = 38. — Corona: R (3.7 + 0.8); foraged with migrant warblers, gnatcatchers, kinglets in canopy through winter-early spring; for example, a pair accompanied 36 small passerines of seven species at 0830 hours, 25 March 1978. Courtship in March-April but nesting unobserved; fledglings in June. Grande: R (2.2 248 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 ± 0.9). Two fed with 11 wintering passerines of six species in canopy (70 percent) and subcanopy (30 percent), 21 March; then with a northern house wren ( Troglodytes aedon ), hermit thrush ( Catharus guttatus ), and pair of long-billed thrashers ( Toxostoma longirostre ), they mobbed a calling eastern screen owl at 1700 hours. Nesting 1.5 m high in dead anacua, old woodpecker hole, April 1977. Spot-breasted wren ( Tryothorus maculipectus ), 16 grams N = 22. — Corona: R (5.3 ± 0.9). Singing birds year-round, at least 100 m apart in about 50 m-diameter territories centered on dense shrubs in clearings and plot edges. Only once seen foraging in a mixed passerine flock of warblers, wrens, and sparrows in these same thickets in winter. Apparent niche equivalent of Carolina wren at Rio Grande but much shyer; nests unobserved. Carolina wren ( Thryothorus ludovicianus), 19 grams, N = 16. — Grande: R (1.2 ± 0.7). Solitary or paired with year-long territory in subcanopy positions like spot-breasted wren. Nest (eggs?) in rotten stump of Berlandier ash ( Fraxinus berlandierana ), April 1977. Clay-colored robin ( Turdus grayi ), 76' grams, N = 23. — Corona: LM; December-March influx (4.3 ± 1.5) versus single pair in June (2.0 + 0). Nestlings June 1973, 3.5 m high in subcanopy brasil; fledglings June 1976. Winter flocks with large resident passerines plus occasional squirrel cuckoos and golden-fronted woodpeckers; three amongst 26 birds of seven species in canopy (61 percent) and subcanopy (39 percent) strata, 1 January 1975. Occasional singing March, frequent in April and June. Grande: one on both plots December-June 1978. Yellow-green vireo ( Vireo flavoviridis) 19 grams, N = 11. — Corona: M; one pair with nestlings, 3 m high in a 4 m pata de vaca ( Bauhinia divaricata ) 10 m from river, 3-7 June 1973, but apparently not in the plots thereafter. Tropical parula ( Parula pitiayumi ), 8 grams, N = 7. — Corona: LM; December-January outflux (2.2 ± 0.2) compared to March-June (5.3 ± 1.6) when nested in ballmoss ( Tillandsia recurvata) above 10 m in riverside trees; two nests (eggs?) 350 m apart, April 1974. Flocked with migrant canopy-users, chiefly warblers, gnatcatchers, and kinglets, December-March, when more numerous than the wintering northern parula (P. americana). Grimson-collared grosbeak (Rhodothraupis celaeno ), 50 grams, N = 2. — Corona: LM; December-April (1.5 ± 0.7). One accompanied a 26 bird flock, 1 January 1975 (see clay-colored robin); regular in winter censuses. A pair with two fledglings nested in forest immediately outside one plot, June 1973, but not noted otherwise in June. Increasing disturbance by campesinos, hunting, and wood-cutting, may have been a factor with this and other species like the yellow-green vireo. BREEDING BIRDS IN EVERGREEN FORESTS 249 Blue bunting ( Cyanocompsa parellina ), 16 grams, N = 20. Corona: R (2.0 ± 1.2). Occupied shrub thickets of forest clearings; rarely at river or roadside hence somewhat segregated from olive sparrow and spot¬ breasted wren. Territorial males no closer than about 30 m from territorial olive sparrows. Mottled and blue males sang in March-June; nests unobserved. Occasionally entered winter, ground-shrub foraging flocks; for example a male accompanied nine, small, migrant passerines, 18 March 1978 (see below). Olive sparrow ( Arremonops rufivirgatus ), 24 grams, N = 19. — Corona: R (6.8 + 1.2). Foraged with Lincoln’s sparrow ( Melospiza lincolnii ) in winter and white-tipped dove year-round (70 percent time on ground versus 50 percent on ground in blue bunting and northern cardinal, Car¬ dinal is cardinalis); song perches to two m in dense shrubs. Grande: LM: December-January outflux (1.7 ± 0.9) compared to April, June (4.7 ± 0.9). Habits as at Rio Corona; nests unobserved at both sites. Bronzed cowbird ( Molothrus aeneus ), 54 grams, N = 31. — Corona: LM; usually absent December-March but common April, June (6.5 ± 2.5), when “helicopter hovering” courtship of males frequent in man¬ made clearings. Grande: LM; like Rio Corona but notably less abundant (4.0 ± 0.8). A female repeatedly tried to enter an altamira oriole nest under construction, 1 1 April 1974, but was repulsed by the pair. Hooded oriole ( Icterus cucullatus ), 25 grams, N = 8. — Corona: R (2.7 ± 0.9). Usually flocked with altamira orioles and similar large passerines in tree canopy December-March, although a pair foraged with Lincoln’s sparrows, northern house wrens, and orange-crowned warblers ( Vermi - rora celata) in 1-2 m shrubs, March 1978. Nestlings 12 m high in ebony, 300 m from the river, June 1973. Grande: R; common in late 1950’s, scarcer in 1960’s, and absent during this study (Gehlbach, 1981:25). Altamira oriole (Icterus gularis ), 67 grams, N = 25. — Corona: R (6.0 ± 2.2). Core member of larger passerine flocks in winter, but monospecific flock of 21 on 22 March 1978. Persistent singing in March, nest-building 9-11 April and 3-7 June at about 100 m intervals along river in drooping cypress and tepeguaje branches at least 5 m high. Grande: LM; December-March outflux (0.8 ± 0.8) versus April, June (2.2 ± 0.9). Nestlings 6.5 m high in tepeguaje, June 1976; nest building in April (see bronzed cowbird). Winter flocks with other large passerines cover 1.5-3. 5 times more lineal area/ time than at Rio Corona, suggesting scarcer food at this marginal northern locale. Audubon’s oriole (Icterus graduacauda ), 46 grams, N = 20. — Corona: R (2.5 ± 0.8). In winter flocks less frequently than other orioles (27 percent of flocks versus 73 percent average for hooded and altamira, N = 11). April and June birds usually stayed 50 m or more from river as did hooded orioles, compared to river-nesting altamira orioles; sang year- round like congeners, suggesting permanent territory; nest unobserved. 250 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 1. Biomass and density of breeding birds (excluding ducks and kingfishers) in riparian evergreen forests of the Rio Corona and Rio Grande, 1973-1978, and concurrent, monthly climatic means plus or minus one standard deviation from Victoria, Tamaulipas (Rio Corona), and McAllen, Texas (Rio Grande). Seasons Parameters Rio Corona Rio Grande Winter (December- Biomass (kg/ ha) 2.8 0.9 March) % nonpasserine 76.4 92.4 x ± SD/ species (g/ha)1 94.2 ± 140.4 75.2 + 200.6 Density (n/ha) 16.1 4.2 % nonpasserine 47.0 64.1 x ± SD/ species2 0.5 ± 0.3 0.3 ±0.3 Breeding (April- Biomass (kg/ ha) 2.8 1.2 June) % nonpasserine 80.6 87.5 x ± SD/ species (g/ha) 80.9 + 125.9 52.9 + 155.3 Density (n/ha) 17.8 7.2 % nonpasserine 51.8 55.9 x ± SD/ species (g/ha) 0.5 + 0.3 0.3 + 0.2 Dry (November-April) Temperature (°C) 20.0+ 1.1 19.1 + 1.2 Precipitation (cm) 3.1 +0.3 2.9 + 0.4 Wet (May-October) Temperature(°C) 26.2 ± 0.8 28.5 + 0.7 Precipitation (cm)3 12.3+ 1.3 5.0 + 0.6 'inter-site biomass < comparisons insignificant in winter and breeding seasons (P>0.05); species = 30 (Corona) versus 13 (Grande) and 35 versus 22, respectively. 2Inter-site density comparisons signficant in winter ( F = 3.9, P = 0.05) and breeding ( F = 5.2, P = 0.02) seasons. 3Only significantly different climatic feature {F — 15.3, P = 0.01). Grande: R; one pair on territory in cedar elms ( Ulmus crassifolia), hung with Spanish moss ( Tillandsia usneoides ), January-June 1976; nest unobserved. No other records from study plots. Much rarer in 1970’s than earlier (Gehlbach, 1981:25). Biomass, Density, and Climate The Rio Corona forest supported 2.3 (breeding season) to 3.1 (winter) times more avian biomass per unit area with 2.5 (breeding) to 3.8 (winter) times higher density of breeding birds than the Rio Grande forest (Table 1). Interestingly, nonpasserine species are relatively more important at the Rio Grande, despite that fact that they are usually considered to be so in the tropics. Nevertheless, they dominate both avifaunas. Birds at the Rio Grande are no larger on average, but they are less dense and more variable in both density and biomass per species than at the Rio Corona. Of particular interest is the 2.7 times overall increase in biomass at the Rio Corona; because it corresponds closely to 2.5 times greater wet- season rainfall, the only major climatic feature that differs between the two locales (Table 1). BREEDING BIRDS IN EVERGREEN FORESTS 251 Earlier, I postulated that more precipitation is the key to understand¬ ing increased food production and hence avifaunal diversity at the Rio Corona (Gehlbach et al., 1976). Subsequently, I also suggested that added rainfall, especially in the nesting season, explains the general regional increase in nonpasserines; because it enhances food-type diversity as well as productivity, which permits more nonpasserines to coexist. Conversely, passerines are more strongly separated by topographic diversity, which is minimal in the Rio-Grande-Rio Corona lowlands (Gehlbach, 1987). Thus, despite their considerable thermal, vegetative, and bird-species similarities, the Rio Corona avifaunal is richer and more productive than its Rio Grande counterpart. Acknowledgments Allan Phillips, Brian Chapman, and John Rappole made constructive comments on this manuscript. My helpers in the field were Jon Peterson in 1956, the co-authors of Gehlbach et al. (1976), and Nancy Gehlbach since 1959. Wayne and Libby Shifflett were especially generous hosts at the Santa Ana National Wildlife Refuge, the location of my Rio Grande study plots. The late George Sutton stimulated my interest in the Rio Corona (see Sutton, 1951). Literature Cited Bush, M. E., and F. R. Gehlbach. 1978. Broad-winged hawk nest in central Texas: geographic record and novel aspects of reproduction. Bull. Texas Ornithol. Soc., 11:41- 43. Gayou, D. C. 1986. The social system of the Texas green jay. Auk, 103:540-547. Gehlbach, F. R. 1981. Mountain islands and desert seas: a natural history of the U.S.- Mexican borderlands. Texas A&M Univ. Press, College Station, xvi + 298 pp. - . 1987. Avian biotic provinces of the Texas borderlands: new techniques for synthetic resource assessment and mapping. Southwestern Nat., in press. Gehlbach, F. R., D. O. Dillon, H. L. Harrell, S. E. Kennedy, and K. R. Wilson. 1976. Avifauna of the Rio Corona, Tamaulipas, Mexico: northeastern limit of the tropics. Auk, 93:53-65. Johnson, N. K. and R. M. Zink. 1985. Genetic evidence for relationships among the red¬ eyed, yellow-green, and Chivi vireos. Wilson Bull., 97:421-435. Marion;, W. R., and R. J. Fleetwood. 1978. Nesting ecology of the plain chachalaca in south Texas. Wilson Bull., 90:386-395. Morrison, M. L., and R. D. Slack. 1977. Flocking and foraging behavior of brown jays in northeastern Mexico. Wilson Bull., 89:171-173. Sutton, G. M. 1951. Mexican birds, first impressions. Univ. Oklahoma Press, Norman, xv + 282 pp. Vehrencamp, S. L. 1978. The adaptive significance of communal nesting in groove-billed anis ( Crotophaga sulcirostris). Behav. Ecol. Sociobiol., 4:1-33. , CLEAR FORK VERTEBRATES AND ENVIRONMENTS FROM THE LOWER PERMIAN OF NORTH-CENTRAL TEXAS Phillip A. Murry and Gary D. Johnson Tarleton State University, Stephenville, Texas 76402, and University of South Dakota, Vermillion, South Dakota, 57069 Abstract. — Vertebrate fossils were collected from the Arroyo, Vale, and Choza formations of the Clear Fork Group (early Permian) in north-central Texas. The Arroyo, lowest of the three Clear Fork units, yielded the most diverse vertebrate faunas. In the lowermost Arroyo, rare occurrences of fusulinids and certain sharks indicate a marine influence. In the youngest beds of the lower Arroyo, the shark Xenacanthus disappears, but other taxa such as the lungfish Gnathorhiza, the amphibian Lysorophus, and terrestrial amphibians and reptiles increase in abundance and diversity. By middle Arroyo deposits, this condition is reversed and the assemblages become dominated by Orthacanthus or Diplocaulus. This suggests that environmental conditions, or at least preservational factors, had deteriorated. Except for one locality in the lower Vale, faunal abundance continues to decrease and both faunal diversity and abundance further decline above middle Vale deposits. Although environmental factors were probably in part responsible for the decline with initiation of sabkha conditions, preservational factors also played a role in the decrease of relative faunal abundance in the Vale and Choza units. Influences on the environment would have included eustatic changes in sea level related to basin subsidence and Gondwanaland glaciation, possible fluctuations in the intertropical convergence zone, and orographic effects. Key words : Texas geology; Lower Permian, Clear Fork Group; vertebrate paleontology; paleoecology. The Clear Fork Group (Permian, Leonardian) is traditionally divided into three formations. In ascending order these are the Arroyo, Vale, and Choza. To the south of the study area these formations are separated by the Standpipe Limestone at the top of the Arroyo Formation and the Bullwagon Dolomite at the top of the Vale Formation (Olson, 1958). However, in north-central Texas there are no marine marker beds by which the terrestrial Clear Fork facies may be separated. The objective of this study was to document changes in vertebrate abundance and diversity throughout the Clear Fork Group in the vicinity of the Wichita River in north-central Texas (Fig. 1). Because there are no marker beds to differentiate the Clear Fork Group in north-central Texas and the lithologies are not distinct enough to distinguish the three formations as separate mappable units, the beds above the Lueders Limestone and below the unconformity with the San Angelo Formation are regarded as undifferentiated Clear Fork units. The fossils recovered reveal significant changes through this portion of the section, but no biostratigraphic zones could be delineated. Using E. C. Olson’s (1951a, 1951b, 1952a, 1952b, 1954, 1955a, 1955b, 1956, 1958) interpretations of this part of the Clear Fork as a guide, we grouped our localities into corresponding Arroyo, Vale, and Choza categories for comparative purposes. The Texas Journal of Science, Vol. 39, No. 3, August, 1987 254 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Surface collections of vertebrate fossils were made at 94 sites. About 20 percent of the localities yielded microvertebrates, obtained by bulk¬ sampling methods such as wet-sieving. Many taxa of fishes and tetrapods were recovered (Table 1), which constitute a larger and more diverse sample size than could possibly have been obtained by standard collecting procedures. In cases where articulated specimens were discovered, excavation by standard quarrying techniques was carried out. CLEAR FORK PERMIAN VERTEBRATES 255 Thousands of macrovertebrate and microvertebrate fossils were collected, and serve as the basis of this study (Appendix). Lower Arroyo Faunas and Environments The Arroyo unit (in the lower Clear Fork beds) contains the most prolific fossil localities in both faunal diversity and abundance. Our observations of the middle and upper lithologies of the Arroyo in the 256 THE TEXAS JOURNAL OF SCIENCE -VOL. 39, NO. 3, 1987 Table 1. Vertebrates identified from the Clear Fork Group. Class Chondrichthyes Subclass Leposondyli Subclass Elasmobranchii Order Nectridea Order Xenacanthodii Diplocaulus sp. Xenacanthus luedersensis Order Microsauria Orthacanthus platypternus Lysorophus sp. Order Chimaerida Euryodus sp. Helodusl Ostodolepis sp. Class Osteichthyes indeterminate microsaurs Subclass Actinopterygii Class Reptilia Order Palaeoniscida Subclass Anapsida indeterminate paleoniscoids Order Cotylosauria Subclass Dipnoi Captorhinus sp. Gnathorhiza sp. Labidosaurus sp. G. dikeloda indeterminate large captorhinomorph Class Amphibia Subclass Euryapsida Subclass Labyrinthodontia Order Araeoscelida Order Temnospondyli araeoscilid? Trimerorhachis sp. Subclass Synapsida Eryops sp. Order Pelycosauria Zatrachysl Varanopsl Order Anthracosauria Dimetrodon sp. Cricotusl Edaphosaurus cf. E. pogonias Seymour ia sp. indeterminate caseid Waggonerial Diadectes sp. indeterminate seymouriamorphs Indian Creek, West Coffee Creek, and Lost Lake areas agree with Olson’s (1958) description as “even red shales” (silty mudstones). Lowermost Arroyo The lower Arroyo unit in the East Coffee Creek area is lithologically more variable than noted in previous studies. Some beds in this unit consist of rather uniform, dark, reddish-brown mudstone devoid of recognizable sedimentary structures. However, the most prolific localities occur in reddish-brown, clay-pebble conglomerates or silty mudstones, which underlie light green siltstone. This lithologic diversity within a restricted stratigraphic horizon indicates significant variations in deposi- tional environment. The area may have been a coastal plain during local Arroyo deposition as there appears to be some marine influence. Only three fusulinid foraminifers were recovered from the lower part of the Arroyo unit; however, they may not have been reworked from earlier deposits. No fusulinids were recovered from the Wichita-Albany Group in north- central Texas using similar collection techniques (Johnson, 1979: appendix). CLEAR FORK PERMIAN VERTEBRATES 257 A major difference between the Clear Fork Group and the underlying marine/ nonmarine Wichita-Albany Group (Johnson, 1981) is the sparseness of shark remains in the Clear Fork. Only three helodont shark teeth were recovered from the Clear Fork; all came from the lower portion of the Arroyo. These teeth may have been reworked from the Wichita-Albany Group, although the presence of a euryhaline helodont shark cannot be discounted. In the Wichita-Albany, helodont sharks represent approximately one percent of the total chondrichthyan assemblage (Johnson, 1981: fig. 3). The most common sharks in the Wichita-Albany include Orthacanthus texensis and Xenacanthus lueder- sensis. With the exception of 40 abnormally large X. luedersensis teeth collected from the lower Arroyo, neither of these taxa appear in the Clear Fork Group. Some sharks may have been freshwater or euryhaline during Wichita-Albany time, but their near absence in the Clear Fork suggests they were marine. The abnormal X. luedersensis teeth, sparse fusulinids and helodonts, and the absence of other chondrichthyans indicate that the typical Wichita-Albany marine environment is not represented in the lower Clear Fork. The xenacanthodiids, traditionally considered as freshwater sharks, further complicate the pattern. The most common, presumably freshwater shark in the Wichita-Albany, Orthacanthus texensis , is entirely absent from the Clear Fork. Xenacanthus luedersensis , the second most common shark in the Wichita-Albany, is fairly common in the lowest part of the Arroyo; it is absent throughout the remaining Clear Fork. Only Orthacanthus platypternus , which appears throughout the Wichita- Albany, persists throughout the Clear Fork, often as the most common fossil. Our studies show that the lowest part of the Arroyo in the East Coffee Creek area exhibits more diversity than previously reported. The last (youngest) positively identified remains recovered in this study of Edaphosaurus , a single specimen of an embolomere, and Zatrachys-Mke elements may represent relict survivors of the older Wichita-Albany fauna. However, well-documented remains of Edaphosaurus pogonias have been recovered from areas west of Middle Coffee Creek and some fragments have been found in West Coffee Creek (E. C. Olson, personal communication), although the absence of this taxon in our collections from this area probably demonstrate it’s rarity in the Clear Fork Group. Middle Lower Arroyo The middle portion of the lower Arroyo contains numerous localities in the East Coffee Creek area and these are the most prolific in the Clear Fork Group. Taxa include xenacanth sharks and lungfish; amphibians, including seymouriamorphs, diadectids, Diplocaulus , Lysorophus , and microsaurs; and reptiles, such as captorhinids and pelycosaurs. Although 258 THE TEXAS JOURNAL OF SCIENCE^VOL. 39, NO. 3, 1987 similar in general faunal representation, these assemblages exhibit more diversity than should be expected. The stratigraphically lowest of these assemblages (locality 37; see Appendix) consists mostly of Orthacanthus ( O . platypternus) teeth. This assemblage occurs in a cross-laminated, clay-pebble conglomerate, representing a channel deposit. Within localities adjacent to locality 37 but that are probably slightly higher stratigraphically, diversity decreases, and there is an extreme decrease in numbers of Orthacanthus teeth. These assemblages are from reddish- brown mudstones, which represent overbank and pond deposits in which large xenacanth sharks ( Orthacanthus ) would not be present, but a wide variety of other smaller taxa could be found. The number of Diplocaulus specimens increases compared to the xenacanths in these deposits. Diversity decreases in most of these overbank deposits to the point where only isolated remains of Diplocaulus and Dimetrodon are found. Upper Lower Arroyo There is a major change in the general faunal composition in the youngest beds of the East Coffee Creek area. This is illustrated at locality 47. The last Xenacanthus teeth occur in these beds; although there is a major decrease in the number of Orthacanthus teeth, they still constitute about 20 percent of the identified material. The number of Gnathorhiza elements, the only lungfish genus recovered from all Clear Fork units, increases dramatically. No locality lower in the section contains more than eight percent lungfish elements, whereas Gnathorhiza tooth plates in these beds constitute about one-third of the identified elements. Locality 47 is one of only two assemblages in the Clear Fork Group where palaeoniscoid elements are numerous. There is also a large increase in the relative abundance of Diplocaulus at this horizon, but presumed terrestrial components such as microsaurs, seymouriamorphs, captorhino- morphs, and pelycosaurs maintain about the same relative abundance as in lower deposits. This assemblage is unlike the older Arroyo assemblages, with a much lower percentage of Orthacanthus relative to Gnathorhiza and a higher percentage of Diplocaulus. Seasonality may have increased during this period inasmuch as the first large concentra¬ tions of Lysorophus , an aestivating amphibian, occur at locality 48. This complements the large increase in Gnathorhiza , a known aestivator since Wichita-Albany time (Berman, 1976). Middle and Upper Arroyo Faunas and Environments By middle Arroyo time, there is a large decrease in vertebrate diversity and abundance. Locality 52 in the lower portion of the middle Arroyo unit contains a moderately diverse assemblage, including Orthacanthus , Diplocaulus , microsaurs, captorhinids, and Dimetrodon , but numbers of CLEAR FORK PERMIAN VERTEBRATES 259 recovered elements are much lower than from the lower Arroyo. The lithology of most middle Arroyo localities consists of reddish-brown mudstone. The only other locality with moderate diversity (locality 56) is covered by a calcareous lag gravel, which differentiates it from other middle Arroyo sites. A large number of Orthacanthus teeth were recovered from this locality, as well as palaeoniscoid fish scales, Diplocaulus and Dimetro don. It seems apparent that by middle Arroyo time, environmental conditions, or at least preservational conditions, had deteriorated. However, if seasonality differences were accentuated at this time, we found no evidence of it in the assemblages of the middle and upper Arroyo. Despite intensive searching, no Lysorophus localities were found in this portion of the section, and only a few Gnathorhiza tooth plate fragments were found in the upper Arroyo. Instead, the faunas are dominated by Orthacanthus and Diplocaulus , neither of which is believed to have been able to withstand extreme seasonality. Above late Arroyo deposits, Orthacanthus and Diplocaulus remain the dominant compo¬ nents in the Clear Fork, although their abundance tends to be inversely proportional to each other. This may indicate a preference for deeper streams for Orthacanthus and shallower streams or ponds for Diplocau¬ lus , but supporting lithologic evidence is inconclusive. Vale and Choza Faunas and Environments The lithology of our lower Vale localities consists primarily of structureless mudstone, thereby differing from the conglomerates des¬ cribed by Olson (1971). However, these apparent lithologic differences may be due to the placement of the tenuous Arroyo-Vale boundary. In lower Vale deposits, there may be some indication of wet-dry seasonality, evidenced by a large concentration of Lysorophus nodules in the Rose Hollow Creek area at locality 67. Some lower Vale localities produced a few lungfish toothplates but most yielded either Orthacanthus or Diplocaulus. A substantial fauna, consisting of numerous Diplocaulus skeletons, along with Orthacanthus, Labidosaurus , Dimetrodon , and caseid pelycosaur remains, occurs in laminated silts and shales at locality 72, which do not indicate wet-dry seasonality. Other than the relatively scarce Lysorophus localities, there is little indication of wet-dry seasonality in the lower to middle Vale. Lysorophus localities are typically isolated occurrences; although rich in numbers of nodules, the localities reported in this study contain few or no associated taxa. Therefore, the presence of Lysorophus suggests adaptation to, or preservation in, unique environments on the floodplain, in which few other species lived or could be preserved, in addition to wet-dry seasonality. However, E. C. Olson (personal communication) reports that Lysorophus , Gnathorhiza , and Diplocaulus may be associated in nodules 260 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 and burrows. It is his interpretation that they do occur together where there is strong seasonality or aestivation, at least during their early ontogeny. Therefore, the ecological interrelationships of these taxa are not apparent. The last relatively numerous fossils of any taxon in the Clear Fork Group are preserved in the middle Vale unit at localities 80 and 81. These remains consist almost entirely of Orthacanthus teeth, with a few associated elements of palaeoniscoid fish, lungfish, Diplocaulus , micro- saurs, Captorhinus , and Dimetrodon. The deposits containing these taxa occur at localities that are different from other, less prolific, middle Vale localities. The lithology consists of clay-pebble conglomerate rather than mudstones, probably representing fluvial deposits. Olson (1971) believed the presence of clay-pebble conglomerates in the middle Vale indicated the initiation of “monsoonal-type rainfall.” However, we found the presence of this facies in the middle Vale to be a relatively uncommon occurrence, although our best fossil localities in this unit are associated with it. The lower Arroyo unit contains more widely distributed clay- pebble conglomerates. However, Olson (1971:649) inferred moderate, evenly distributed rainfall throughout the year during lower Arroyo time. Abundant evaporite minerals occur in the Clear Fork Group for the first time in the middle Vale unit. From upper Vale through Choza deposits, there is an increase in the amount of evaporites, and at several localities there is textural evidence of ground-water influence on evaporite formation and dissolution. The upper Vale and Choza units contain a paucity of vertebrate remains, consisting primarily of Orthacanthus and Diplocaulus. These taxa often occur in inversely different proportions for any given locality, but the numbers of each may be too meager to be conclusive. We recovered a partial skull of a large captorhinomorph from the upper Vale at locality 89, and a relatively intact skeleton of probably the same taxon was found at locality 96. These large captorhinomorphs are uncommon, and appear to be restricted to the Vale and lower Choza. As the amount of the evaporites continued to increase, probably indicating more influence from marginal coastal sabkha environments (Smith, 1976), the deposits markedly decrease in both numbers of specimens and in diversity. Despite intensive searching, we did not find any localities above the lower Choza. Presumably, deteriorating environmental factors limited faunal distribution and affected preserva¬ tion during much of later Clear Fork time. Discussion The environmental changes that occurred during Clear Fork time may be related to one or more contemporary events. The general changes in facies from lower to upper Clear Fork are related to basin subsidence in which sabkha conditions were initiated during upper Clear Fork times. CLEAR FORK PERMIAN VERTEBRATES 261 Variance in seasonality (cyclicity), which accompanied this basin subsidence, may have been due to glacially caused eustatic changes in sea level (Presley and McGillis, 1982). Because this region was less than 20 degrees from the equator (Ziegler et al., 1977), local aridity would not have been related to the subtropical high pressure belt associated with latitudes of 20 to 30 degrees. Therefore, the climatic changes observed might be related to variations in the intertropical convergence zone, which probably controlled wet-dry seasonality. Continental effects, interrelationships of atmospheric circulation, and physical barriers, as well as eustatic sea level changes affecting oceanic circulation during glacial episodes, would have influenced both the depositional environ¬ ments and the distribution of fossil vertebrates in the Clear Fork. Acknowledgments We want to recognize the cooperation and hospitality of Gene Willingham and Glen Collier of the Waggoner Ranch. E. C. Olson has provided encouragement and information throughout this study and, along with Robert Hook, provided many useful comments in review of the manuscript. We were ably assisted in field and laboratory by Roy Frosch, Henry Huggins, Kirk Kennedy, Kent Newman, Layne Schulz, Barry Severson, Mark Vomacka, Kyle Williams, and Richard Wolfe. Colleen Odenbrett and Brian Barker typed the manuscript. Funding for this project was provided by National Science Foundation Grant EAR-8218472. Literature Cited Berman, D. S. 1976. Cranial morphology of the lower Permian lungfish Gnathorhiza (Osteichthyes: Dipnoi). J. Paleont., 50:1020-1033. Johnson, G. D. 1979. Early Permian vertebrates from Texas: Actinopterygii ( Schaefferich - thys), Chondrichthyes (including Pennsylvanian and Triassic Xenacanthodii), and Acanthodii. Unpublished Ph.D. dissertation, Southern Methodist University, Dallas, 653 pp. - . 1981. Hybodontoidei (Chondrichthyes) from the Wichita- Albany Group (early Permian) of Texas. J. Vert. Paleont., 1:1-41. Olson, E. C. 1951a. Fauna of upper Vale and Choza: 1-5. 1. A new family of the Parareptilia. 2. A new captorhinomorph reptile. 3. Lungfish of the Vale. 4. The skull of Gnathorhiza dikeloda Olson. 5. An eryopid amphibian. Fieldiana: Geol., 10:89-128. - . 1951b. Vertebrates from the Choza Formation, Permian of Texas. J. Geol., 59:178- 181. - . 1952a. Fauna of the upper Vale and Choza: 6. Diplocaulus. Fieldiana: Geol., 10:147-166. - . 1952b. The evolution of a Permian vertebrate chronofauna. Evolution, 6:181-196. - . 1954. Fauna of the Vale and Choza: 7-9. 7. Pelycosauria: family Caseidae. 8. Pely- cosauria: Dimetrodon. 9. Captorhinomorpha. Fieldiana: Geol., 10:192-218. - . 1955a. Fauna of the Vale and Choza: 10. Trimerorhachis : including a revision of pre-Vale species. Fieldiana: Geol., 10:225-274. - . 1955b. Parallelism in the evolution of the Permian reptilian faunas of the Old and New worlds. Fieldiana: Zool., 37:385-401. - . 1956. Fauna of the Vale and Choza: 11-13: 11. Lysorophus: Vale and Choza; Diplocaulus, Cacops and Eryopidae: Choza. 12. A new trematopsid amphibian from the 262 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Vale Formation. 13. Diadectes, Xenacanthus, and specimens of uncertain affinities. Fieldiana: Geol., 10:313-334. - . 1958. Fauna of the Vale and Choza. 14. Summary, review and integration of the geology and faunas. Fieldiana: Geol., 10:397-448. - . 1971. Vertebrate paleozoology. John Wiley and Sons, New York, 839 pp. Presley, M.W., and K. A. McGillis. 1982. Coastal evaporite and tidal-flat sediments of the upper Clear Fork and Glorieta formations, Texas Panhandle. Univ. Texas Bureau Econ. Geol., Rept. Inv., 115:1-50. Smith, G. E. 1976. Sabkha and tidal-flat facies control of stratiform copper deposits in north Texas. Pp. 25-39, in Stratiform copper deposits of the midcontinent region, a symposium (K. S. Johnson and R. L. Croy, eds.), Oklahoma Geol. Surv. Circ., 77. Ziegler, A. M., C. R. Scotese, M.E. Johnson, W. S. McKerrow, and R. K. Bambach. 1977. Paleozoic biogeography of continents bordering the Iapetus (Pre-Caledonian) and Rheic (Pre-Hercynian) oceans. Pp. 1-22, in Paleontology and plate tectonics with special reference to the history of the Atlantic Ocean (R. M. West, ed.), Milwaukee Public Mus. Spec. Publ. Biol. Geol., 2:1-109. Appendix Clear Fork Group Localities in Approximate Stratigraphic Order Each locality is listed by Southern Methodist University vertebrate locality number, followed by faunal description. Most taxa were identified on the basis of isolated elements. Major faunal descriptions are preceded by a description of the associated unweathered sedimentary facies (color notations are from the Geological Society of America Rock-Color Chart). Sample size for bulk processing is stated in pounds. Formation names are used in an informal sense (based on Olson, 1958) and are not meant to imply differentiation of the Clear Fork Group. All specimens, copies of field sketches, and precise locality data are reposited in the Shuler Museum of Paleontology at Southern Methodist University, Dallas, Texas. Several localities containing as yet unidentified fossils have not been assigned locality numbers, but are on file in the Shuler Museum. Lower Arroyo Formation East Coffee Creek area SMU 33 Edaphosaurus cf. E. pogonias. SMU 34 pelycosaur. SMU 35 small Diplocaulus skull and jaws plus associated fragments. SMU 36 80 lbs.; fusulinids; Spirorbis; O. platypternus; X. lueder sensis; Gnathorhiza; Diplocaulus ; amphibian scutes; Dimetrodon; coprolites. SMU 37 Moderate reddish-brown (10R 4/6), calcareous, cross-laminated siltstone with silt pebbles (“clay-pebble conglomerate”). Occurs immediately below light greenish-gray (5G 8/1), noncalcareous, cross-laminated flaggy siltstone. 2200 lbs.; Spirorbis; O. platypternus; X. luedersensis ; xenacanth spine fragments and denticles, Helodusl; ; Gnathorhiza; palaeoniscoid scales; Diadectes; Diplocaulus; Lysorophus; Euryodus (partial skull); unidentified microsaurs; Captorhinus; Labidosaurus; Dimetrodon. SMU 38 80 lbs.; fusulinid; O. platypternus; palaeoniscoids; Tr inter orhachis (small partial skull and jaws). SMU 39 Dark reddish-brown (10R 3/4), noncalcareous, slightly silty mudstone with less than one percent light green noncalcareous mudstone. 2000 lbs.; O. platypternus , CLEAR FORK PERMIAN VERTEBRATES 263 SMU 40 SMU41 SMU 42 SMU 43 SMU 44 SMU 45 SMU 46 SMU 47 SMU 48 SMU 49 xenacanth? spine fragments; Gnathorhiza ; unidentified fish-like centrum; Trimerorhachis ; Eryops; Zatrachysl; Diadectes; Seymouria; Diplocaulus; Lysoro- phus ; microsaur?; Labidosaurusl; Dimetrodon; Varanopsl xenacanth? spine fragment; large fish? vertebra; Trimerorhachis-, Eryops-, Diadectesl ; Diplocaulus; microsaur?; Labidosaurus; Dimetrodon. Diadectes; small pelycosaur. Diplocaulus. Diplocaulus ; Dimetrodon. Diplocaulus; Dimetrodon. O. platypternus; xenacanth calcified cartilage; Dimetrodon. Eryops; Dimetrodon. Pale reddish-brown (10R 5/4) to dark reddish-brown (10R 3/4), calcareous, muddy siltstone. Occurs immediately below a light greenish-gray (5G 8/1), noncalcareous, cross-laminated, flaggy siltstone. 2750 lbs.; X. leudersensis; O. platypternus; xenacanth calcified cartilage; Gnathorhiza (more than one species?); unidentified lungfish?; palaeoniscoid scales; Trimerorhachis, Cricotusl; Waggonerial; small seymouriamorph; Diplocaulus; Lysorophus; Ostodolepis; Euryodus; unidentified microsaur; Captorhinus; araeoscelid?; Dimetrodon; Varanopsl; small captorhinomorph trackway. Lysorophus nodules. Diplocaulus. SE of mouth of West Coffee Creek SMU 50 O. platypternus; Trimerorhachis; small seymouriamorph; Diplocaulus; Dime¬ trodon. Middle Arroyo Formation Indian Creek area SMU 51 60 lbs. from three horizons; seed?; Trimerorhachisl; Diplocaulus; Dimetrodon. SMU 52 1200 lbs. from six horizons; O. platypternus; xenacanth? spine fragment; palaeoniscoid scale; Trimerorhachisl; labyrinthodont; Diplocaulus; microsaur?; captorhinid; Dimetrodon. SMU 53 Diplocaulus. SMU 54 Dimetrodon. SMU 55 Diplocaulus; Dimetrodon. West Coffee Creek area SMU 56 Moderate reddish-brown (10R 4/6) to dark reddish-brown (10R 3/4), noncalcareous, slightly silty mudstone; covered by a calcareous lag gravel. 1400 lbs.; O. platypternus; xenacanth calcified cartilage; palaeoniscoid scales; Diplocaulus; Dimetrodon; coprolites. Upper Arroyo Formation Area northwest of Lost Lake SMU 57 Dark reddish-brown (10R 3/4), noncalcareous, silty mudstone that contains light greenish-gray (5G 8/1) irregular spots. Covered (upslope) by fissile to flaggy red siltstone with greenish-gray irregular spots. 500 lbs.; O. platypternus; xenacanth calcified cartilage and denticles; broken palaeoniscoid scales (common); Gnathorhiza; Diplocaulus; Dimetrodon; coprolite. Large pelycosaur bones. SMU 58 264 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Lowermost Vale Formation (Olson, 1958) Fish Creek area SMU 59 Grayish-red (10R 4/2) silty mudstone to siltstone with irregular light greenish- gray (5G 8/1) spots; overlies a light greenish-gray clay-pebble conglomerate. All occur in large-scale cross beds. 500 lbs.; O. platypternus', indeterminate bones and teeth. SMU 60 Grayish-red to reddish-brown siltstone with coarse sand-size clay clasts (“clay- pebble conglomerate”); overlain by pale green muddy siltstone. All occur in large-scale cross-beds; 40 meters across; same as SMU 59, but stratigraphically higher; 250 lbs.; O. platypternus', Gnathorhiza. SMU 61 500 lbs.; O. platypternus', Gnathorhiza', lungfish? scale; Eryops\ Diplocaulus', Dimetrodon, coprolites. Lower Vale Formation Fish Creek area SMU 62 Calcified xenacanth cartilage; pelycosaur? SMU 63 O. platypternus ; xenacanth calcified cartilage; Gnathorhiza dikeloda', Diplocau- lus; small reptile? Rose Hollow Creek area SMU 64 SMU 65 SMU 66 SMU 67 SMU 68 Indeterminate reptile. O. platypternus', Diplocaulus', indeterminate bones of large reptile. Captorhinus', large indeterminate fragments. Lag deposit (15 by 35 meters) of noncalcareous nodules weathering from dark reddish-brown (10R 3/4), calcareous siltstone, which is about 40 centimeters thick. Underlain by dark reddish-brown, blocky, silty mudstone mottled with irregular greenish-gray (5GY 6/1) spots. 1200 lbs. (surface only); O. platypternus (three teeth); Gnathorhiza (one pterygoid tooth plate, fragments); Lysorophus (about 1 1,300 nodules, 1484 vertebrae). O. platypternus. Crooked Creek area (southwest) SMU 69 SMU 70 SMU 71 SMU 72 SMU 73 SMU 74 SMU 75 SMU 76 SMU 77 SMU 78 SMU 79 Partial pelycosaur limb bone. Moderate reddish-brown (10R 4/6) slightly silty mudstone with irregular light greenish-gray (5GY 8/1) spots. 2000 lbs.; O. platypternus', xenacanth calcified cartilage, xenacanth denticle; Gnathorhiza, fish scale?; Trimerorhachis', Seymou- ria\ Diplocaulus', pelycosaur; unidentified jaw in plaster jacket; coprolites. Diplocaulus! Pale reddish-brown (10R 5/4), with some moderate reddish-brown (10R 4/6) and less moderate reddish-orange (10R 6/6), noncalcareous, cross-laminated, muddy siltstone, mottled with very light gray (N8). Contains some light greenish- gray (5G 8/1) laminae. 900 lbs.; O. platypternus', seymouriamorph?; Diplocaulus', Dimetrodon ; caseid; also a Diplocaulus bone bed currently being excavated. Diplocaulus. Diplocaulus', pelycosaur. Diplocaulus. Diplocaulus. Diplocaulus. Diplocaulus. O. platypternus', labyrinthodont; Diplocaulus. CLEAR FORK PERMIAN VERTEBRATES 265 Middle Vale Formation Crooked Creek area (northwest) SMU 80 SMU 81 SMU 82 SMU 83 SMU 84 SMU 85 SMU 86 SMU 87 SMU 88 Very pale green (10G 8/2) to pale green (5G 7/2) to light greenish-gray (5G 8/ 1), slightly calcareous mudstone dominated by clay clasts (“clay-pebble conglomerate”). Most of the clasts, together with enclosed layers of mudstone, are pale red (5R 6/2) with some pale reddish-brown (10R 5/4). Capped by light greenish-gray siltstone. 2300 lbs. (50 percent of —20 mesh concentrate not sorted); seed; Spirorbis; O. platypternus; xenacanth denticles and calcified cartilage; Gnathorhiza (more than one species?); Eryopsl; Diplocaulus; micro- saur; Captorhinus; Dimetrodon ; 38 small indeterminate tetrapod jaws; coprolites. Pale green (5G 7/2) to light greenish-gray (5G 8/1), slightly calcareous mudstone dominated by clay clasts and small (1 mm) flakes and vugs of gypsum? (“clay- pebble conglomerate”). Clasts are same color as mudstone. Layers of conglomerate alternate with red mudstone. Capped by light greenish-gray siltstone. 1200 lbs.; Spirorbis ; O. platypternus ; xenacanth spine fragments and denticles; Gnathorhiza ; palaeoniscoid teeth and fish scale; Diplocaulus; Dimetrodon', coprolites. Dimetrodon ? Pelycosaur. Large indeterminate jaw fragment. Indeterminate skull and jaw fragments. Lysorophus. Diplocaulus; captorhinid? Pelycosaur. Upper Vale Formation East end of Ignorant Ridge SMU 89 Large captorhinomorph partial skull. Ignorant Ridge area east of Mustang Creek SMU 90 Gnathorhiza', Diplocaulus', indeterminate bones in carbonate/ hematite matrix. SMU 91 Diplocaulus', pelycosaur. SMU 92 O. platypternus', Dimetrodon. SMU 93 O. platypternus', Diplocaulus. SMU 94 O. platypternus', amphibian scutes; coprolites. 3 miles north of Vera SMU 95 Diplocaulus, Lysorophus SMU 96 skeleton of a large captorhinomorph SMU 97 O. platypternus Lower Choza Formation North and south rim of Ignorant Ridge SMU 98 Diplocaulus SMU 99 Xenacanth calcified cartilage; 300 lbs. of sandstone/siltstone blocks: Gnathor¬ hiza’, indeterminate amphibian clavicles and pelycosaur jaws appear on surfaces of blocks. SMU 100 Pale reddish-brown (10R 5/4) to dark reddish-brown (10R 3/4), with some pale red (5R 6/2) and minor light bluish-gray (5B 7/1), calcareous mudstone 266 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 dominated by small clay? pebbles (“clay-pebble conglomerate”). Overlain by greenish-gray (5G 8/ 1-6/1) siltstone; overlies reddish-brown (10R 5/ 4-3/ 4), noncalcareous, slightly silty mudstone with conchoidal fractures. 1200 lbs.; O. platypternus\ palaeoniscoid scale; Gnathorhiza\ Diplocaulus\ amphibian scutes; isolated reptile teeth; coprolites. West of Mustang Creek SMU 101 Diplocaulus. SMU 102 Dimetro don. SMU 103 Small amphibian? skull fragments; small coprolites. 2 miles north of Vera SMU 104 O. platypternus\ xenacanth calcified cartilage; amphibian jaw. North side of North Wichita River SMU 105 Diplocaulus\ Labidosaurus. A PREDATORY TERRESTRIAL FLATWORM, BIPALIUM KEWENSE , IN TEXAS: FERAL POPULATIONS AND LABORATORY OBSERVATIONS Raymond W. Neck Texas Parks and Wildlife Department, 4200 Smith School Road, Austin, Texas 78744 Abstract. — Laboratory observations on behavior and records of feral populations of the terrestrial flatworm, Bipalium kewense Moseley, 1878, in Texas are reported. Key words: Bipalium kewense ; terrestrial flatworm; predatory behavior. While most biologists are aware of the existence of aquatic free-living flatworms (planarians), relatively few are familiar with terrestrial flatworms. The suborder Terricola (class Turbellaria, order Tricladida) comprises several hundred species (von Graff, 1899), several of which have become rather widespread as a result inadvertant dispersal with human commerce (Hyman, 1940, 1943, 1951, 1954). The most common of these introduced flatworms in undoubtedly Bipalium kewense Moseley, 1878, of the family Bipaliidae. Moseley (1878) described this species from a moribund greenhouse specimen collected at the Royal Botanic Gardens at Kew, England. Subsequently, B. kewense has become so widespread that it has been described as the cosmopolitan land planarian (Hyman, 1951; Winsor, 1983a). Long assumed to be a native of some part of the Oriental region (Hyman, 1940), Winsor (1983a) restricted the native range of B. kewense from northern Vietnam to Kampuchea and possibly south through Malaysia. B. kewense exhibits a honey brown background color with several dark brown to purplish brown longitudinal stripes (three narrow lines and two broad, diffuse bands). The anterior end is a characteristic semicircular shape and possesses a more uniform dark brown coloration. One of the larger specimens seen personally (from Austin) was 18 centimeters in length when moving but easily stretched to 30 centimeters when hanging down from the top of the terranium; width of body portion was approximately four millimeters. Hyman (1951) reported that B. kewense may reach a length of 35 centimeters. A summary of literature reports of B. kewense for the United States reveals localities over most of the area east of the Mississippi River, except for some of the northernmost states; populations also are known from California (Winsor, 1983a). Records from the northern and western edge of its range in the United States are generally from greenhouses rather than outdoor sites. Such a restriction indicates a susceptibility to cold and dry conditions (to be expected in a species from the warm, moist regions of southeastern Asia). The published records from Texas The Texas Journal of Science, Vol. 39, No. 3, August, 1987 268 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 are from “greenhouses” in Houston, which were the source of experimental animals (Campbell, 1965; Trammel and Campbell, 1971), although specimens have been purchased from a biological supply house in Houston (Phillips and Dresden, 1973). No published records of feral populations of B. kewense in Texas are known to me, although the presence of this species is well known to invertebrate zoologists in the state. Relatively little information is available concerning the bionomics of B. kewense. Reproduction is normally asexual by fragmentation in subtropical and temperate areas (Hyman, 1940), although Connella and Stern (1969) observed cocoon formation and emergence of young. The taxa observed in this last study could have been Bipalium adventitium , which was described from introduced material in California (Hyman, 1943). B. adventitium has since established numerous populations in the northeastern United State (Klots, 1960). Laboratory populations of B. kewense have been utilized in various physiological investigations (Campbell, 1965; Trammell and Campbell, 1971; Phillips and Dresden, 1973). Kawaguti (1932) reported that B. kewense could survive water loss up to 45 percent of body weight. The purpose of this communication is to present observations on predatory behavior and collection records of B. kewense in Texas. Laboratory Observations B. kewense feeds primarily (if not exclusively) on earthworms, on which it utilizes a collaginase to destroy the cuticle (Phillips and Dresden, 1973). It is known to be a minor pest on earthworm farms in the southern United States (Winsor, 1983a). Klots (1960) reported that B. adventitium fed on “small annelids,. . . slugs, insect larvae and the like.” However, personal observations of B. kewense crawling over living slugs, Sarasinula plebeius , revealed no reaction from the flatworm. Detailed observations on the predatory behavior of B. kewense have not been published. Wallen (1954) merely reported that “a full grown specimen could devour half of a five-inch earthworm within 30 minutes leaving a foamy mass.” No attack description was provided although Wallen (1954) reported that B. kewense “usually moved from the anterior to the posterior end of the earthworm.” Laboratory observations revealed an interesting attack method. First contact occurred when an earthworm touched the body of a B. kewense. Although the earthworm initially appeared to be stuck to the flatworm, it later withdrew and continued onward with no apparent increased rate of locomotion. At the same time, however, the B. kewense began to move its anterior end toward the earthworm until its head came into contact with the worm. Then, B. kewense wrapped its body rapidly around the earthworm, which attempted eascape by increased rate of locomotion. B. TERRESTRIAL FLATWORM IN TEXAS 269 kewense wrapped itself around the prey, flattening out its body in an attempt to envelop the body of the earthworm. The B. kewense eventually was able to get all of its body criss-crossed to an extent that the earthworm was completely encased. Volume covered by the flatworm decreased as the earthworm was apprently consumed. Approximately 55 minutes following the attack, the B. kewense was again crawling around, whereas the earthworm was reduced to a layer of bubbles. Predatory behavior of the related Bipalium adventitium was reported by Dindal (1970). Random contact followed by initial attack was similar to that observed for B. kewense , but B. adventitium did not appear to constrict earthworms. Average time of consumption was about 45 minutes. Greatest mortality factor in feral populations of B. kewense is likely to be dessication. However, predation may have been sufficiently significant in its native habitat to evolve distastefulness. Winsor (1983b) reported vomiting in response to ingestion of B. kewense by a domestic cat. Personal observations revealed that both a Texas slider ( Chrysemys concinna texana) and a tiger salamander ( Ambystoma tigrinum tigrinum) refused to eat B. kewense. After snapping at B. kewense , the tiger salamander moved its head from side to side in an apparent rejection response. Texas Feral Populations Cameron County. — B. kewense is known from two residential yards (separated by three kilometers) in Brownsville. Populations are much less dense than those found in Austin. Distribution is also much more spotty in the Brownsville area. Despite many earlier field surveys, B. kewense was not found in the Brownsville area until 24 December 1976; a subsequent population was found on 22 December 1977. Comal County. — B. kewense was abundant under old railroad crossties in an open, disturbed area of level ground above Comal Springs, Landa Park, New Braunfels, on 12 June 1976. Travis County. — B. kewense has a general distribution in the city of Austin. Whereas denser populations are characteristic of residential yards, flatworms also have been found in rock rip-rap areas on slopes below a city street (where no supplemental water is provided) and on a wooded slope along the shoreline of Town Lake. I have observed Austin populations since 1968. Kendall County. — Several large specimens of B. kewense were found underneath logs lying on the second terrace of the Guadalupe River within Guadalupe River State Park on 24 April 1975. The significance of this record is its occurrence in a totally nonurban habitat. Origin of this population was likely from flood-borne specimens from an upstream urban area. Probabilities of persistence of this and similar populations 270 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 are likely to be low, although population maintenance in protected mesic canyons is possible. San Saba County. — Several small B. kewense were collected under doomed limbs at the edge of a pool of Gorman Creek, approximately 65 meters upstream of the falls in Gorman Falls State Park on 18 July 1985. This population probably originated with soil attached to exotic plants, for example, Canna sp. and Colocasia sp., which are well established at the site. Few population sources, if any, exist upstream on Gorman Creek; downstream dispersal along the Colorado River is unlikely because of the existence of a sheer 20-meter cliff at Gorman Falls. Walker County. — A population of B. kewense was discovered in a woodland in a residential area of Huntsville near the junction of U.S. 190 and Texas 90 on 14 December 1977. Summary Populations of B. kewense can be expected in residential areas of cities in the eastern two-thirds of Texas. The northern and western boundaries will be determined by degree of winter cold and periodic xeric conditions, respectively. It is of note that the only Oklahoma record is from a greenhouse (Wallen, 1954) and no records are known from New Mexico (Winsor, 1983a). Rarity of rural populations of B. kewense in Louisiana has been reported (Dundee and Dundee, 1963). Literature Cited Campbell, J. W. 1965. Arginine and urea systhesis in the land planarian: its significance in biochemical evolution. Nature, 208:1299-1301. Connella, J. V., and D„ H. Stern. 1969. Land planarians: sexuality and occurrence. Trans. Amer. Micros. Soc., 88:309-311. Dindal, D. L. 1970. Feeding behavior of a terrestrial turbellarian Bipalium adventitium. Amer. Midland Nat., 83:635-637. Dundee, D. S., and H. A. Dundee. 1963. Observations on the and planarian Bipalium kewense Moseley in the Gulf Coast. Syst. Zool., 12:36-37. Hyman, L. H. 1940. Native and introduced land planarians in the United States. Science, 92:105-106. - . 1943. Endemic and exotic land planarians in the United States with a discussion of necessary changes of names in the Rhynchodemidae. Amer. Mus. Novit., 1241:1-12. - . 1951. The invertebrates. II. Platyhelminthes and Rhynchocoela. The acoelomate Bilateria. McGraw-Hill, New York, 550 pp. - . 1954. Some land planarians of the United States and Europe, with remarks on nomenclature. Amer. Mus. Novit., 1667:1-21. Kawaguti, S. 1932. Physiology of land planarians. Mem. Fac. Sci. Agric. Taihoku Univ., 7:28-37. Klots, A. B. 1960. A terrestrial flatworm well established outdoors in the northeastern United States. Syst. Zool., 9:33-34. Moseley, N. H. 1878. Description of a new species of land-planarian from the hothouse at Kew Gardens. Ann. Mag. Nat. Hist., ser. 5, 1:237-239. TERRESTRIAL FLATWORM IN TEXAS 271 Phillips, J., and M. H. Dresden. 1973. A collagenase in extracts of the invertebrate Bipalium kewense. Biochem. J., 133:329-334. Tramell, P. R., and J. W. Campbell. 1971. Carbamyl phosphate synthesis in invertebrates. Comp. Biochem. Physiol., 406:395-406. von Graff, L. 1899. Monographic der Turbellarien II (Tricladida Terricola). Engelmann, leipzig, 893 pp. Wallen, I. E. 1954. A land planarian in Oklahoma. Trans. Amer. Microsp. Soc., 73:193. Winsor, L. 1983a. A revision of the cosmopolitan land planarian Bipalium kewense Moseley, 1879 (Turbellaria:Tricladida:Terricola). Zool. J. Linnaean Soc., 79:61-100. - . 1983b. Vomiting of land planarians (Turbellaria:Tricladida:Terricola) ingested by cats. Australian Vet. J., 60:282. ' VALIDATION OF DAILY RING DEPOSITION IN OTOLITHS OF WILD YOUNG-OF-THE-YEAR LARGEMOUTH BASS J. Jeffery Isely and Richard L. Noble Department of Wildlife and Fisheries Sciences, Texas A&M University, College Station, Texas 77843 Abstract.— The formation of daily rings in otoliths of young-of-the-year largemouth bass ( Micropterus salmoides) was observed in fish collected from three Texas ponds. Daily ring deposition began at hatching. Rings were deposited daily over a 152-day period in one pond and over 47- and 52-day periods in two other ponds. The 95 percent confidence interval for estimating the ring count from age for an individual fish was less than seven (plus or minus) rings. Distinguishable rings were deposited with daily regularity as long as growth was observed. Errors in ageing likely resulted from an inability to distinguish sub¬ daily from daily rings or from loss of resolution of rings that were compressed due to slow growth. Key words : otoliths; daily ring deposition; largemouth bass. Until recently, ageing of fish has been restricted to the use of annular marks in bony (calcified) structures. Pannella (1971, 1974) suggested that, in addition to annual rings, the otoliths of some fish also contain daily rings, and that these daily rings were a universal property of teleost otoliths. Schmidt and Fabrizio (1980) reported a positive correlation between otolith ring count and time in samples from wild populations of largemouth bass ( Micropterus salmoides) up to 80 days of age in New York. Miller and Storck (1982) verified daily ring deposition up to 100 days from otoliths of laboratory-raised largemouth bass held at ambient temperatures in Illinois. They found that ring formation began at hatching; however, the total complement of rings formed during the prolarval stage was visible for only 10 to 15 days after swim-up. Ages determined from otoliths tended to underestimate true age of older fish because of increased opacity of the otolith, which obscured some rings, and poor resolution of closely-spaced rings caused by slow growth at low temperatures. Although daily growth rings appear to be deposited throughout the growing season and have been used to verify annulus formation in largemouth bass (Taubert and Tranquilli, 1982), previous efforts to verify the daily periodicity of ring deposition primarily have involved laboratory-raised fish. The purpose of this study was to investigate the formation of daily growth rings in otoliths of largemouth bass raised under field conditions rather than under the influence of cyclical laboratory activity. The Texas Journal of Science, Vol. 39, No. 3, August, 1987 274 THE TEXAS JOURNAL OF SCIENCE -VOL. 39, NO. 3, 1987 Materials and Methods To determine the age at first otolith ring deposition, otoliths were examined from largemouth bass larvae with a known hatching date. Largemouth bass embryos were collected on 19 April 1984 from a nest in a 0.5-hectare pond on the Southwest Texas State University campus. The embryos were incubated in two four-liter glass jars at ambient light and temperature. Upon hatching, fry were removed at approximatley two-day intervals for 1 1 days and stored in 70 percent ethanol for later analysis. Otoliths (sagittae) were later excised, placed in a small drop of immersion oil, and observed under a microscope (400X) with transmitted light. Daily deposition of rings after 10 days was investigated by comparing the increments in mean ring counts with the time in days between samples collected from three 0.1 -hectare ponds in central Texas. Pond K-2 was stocked with one male and two female mature largemouth bass in April of 1983 to produce a spawn. Brood fish were removed after fry were observed. Up to 30 fish were collected at about 25-day intervals from 26 May through 25 October 1983. Ponds SC-4 and SC-5 also were stocked in April 1983 with one male and four mature female largemouth bass to insure a spawn. As in pond K-2, brood fish were removed when fry were observed and up to 30 fry were collected at about 25-day intervals from 1 July through 22 August 1983. Fish were collected by seining and were returned to the laboratory for processing. Otoliths were removed, stored in plastic vials until dry, then mounted on glass microscope slides with a small drop of thermoplastic cement. The otoliths were then prepared according to the methods of Taubert and Coble (1977) with modifications as described by Miller and Storck (1982), except that no. 600 carborundum paper was substituted for no. 600 grinding powder on a glass plate, and sections were not polished with a felt wheel. Rings were observed through a small amount of immersion oil. Otoliths that were unreadable due to thickness and opacity were repositioned so that the ground surface was attached to the slide, and the opposite surface was ground until a thin, readable section was produced. Ring counts were made in the posterior and ventral fields of the otolith; rings in the anterior field were poorly defined. The number of growth rings in each otolith was counted five consecutive times by the senior author and the median was determined. Because several factors negatively bias ring counts whereas few factors positively bias counts, the median was considered to produce a more accurate age estimate than the mean. Results and Discussion Largemouth bass embryos began to hatch within 24 hours after collection and all had either hatched or died within 48 hours. Observations confirmed the report by Miller and Storck (1982) that otolith ring deposition begins at hatching. Ring deposition continued with daily regularity through the first 10 days of life (Table 1). Seven samples of largemouth bass were collected from pond K-2 during a 152-day period (Table 2). The difference in the mean number of otolith rings between consecutive sampling dates corresponded closely to the time interval in days. The difference in mean otolith ring counts between consecutive samples and the time in days between samples was <6 in all cases. The deviation from expected count between the first and last sample, which represents a cumulative error, was —5 (Table 2). Although rings formed prior to swim-up were less distinct than other rings, no loss of resolution of nuclear rings in older fish (as seen by RING DEPOSITION IN OTOLITHS OF BASS 275 Table 1. Number of rings visible in otoliths of largemouth bass three to 10 days old. Number of rings Age (days) N Mean Range 3 I 3.0 - 5 3 6.0 - 7 3 8.3 8-9 10 4 11.3 11-12 Miller and Storck, 1982) was observed in this study and these rings were counted. The calculated ages, therefore, are age from hatching, not age from swim-up, as reported for largemouth bass by Miller and Storck (1982). Poor resolution of rings near the margin of the otolith often accounted for much of the variability in ring count. As these rings became imbedded in the otolith, resolution increased. Errors in ageing between consecutive samples, therefore, tended to be compensatory. Variability within the first four samples was low as indicated by the range and standard error (Table 2). Variability increased in the 29 August sample and remained high throughout the remainder of the study. Otolith rings deposited during August were compressed and difficult to distinguish, even at levels of magnification higher than normally used in this study (1000 X). Poor resolution of these compressed rings may have caused an increase in the variance and an underestimation of the age of fish collected after 29 July. The narrowing of growth rings was likely a result of a decline in growth during this period due to high water temperature and the density of fish in the pond. In order to reduce density, 500 fish were removed from pond K-2 on 30 September . Following the removal of these fish, growth of fish in the pond increased and well-defined daily rings were deposited thereafter. Variability continued to remain high because the region contributing the variation in ring count already had been imbedded in the otolith. Otolith rings again became compressed near the margin in the 25 October sample, indicating that first-year growth was nearly complete by that time. Relationship between otolith ring spacing and growth may be useful in back-calculation of growth regimes. Miller and Storck (1982) found that total length of largemouth bass could be predicted from otolith radius. A relationship between the width of adjacent daily increments on otoliths and fish growth also has been observed for other species (Strusaker and Uchiyama, 1976; Taubert and Coble, 1977). Marshall and Parker (1982) found a high correlation between otolith diameter and fork length in sockeye salmon ( Oncorhynchus nerka), and suggested that otolith ring spacing might be used to reconstruct detailed growth histories of individual fish. 276 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Table 2. Number of rings observed in otoliths of young-of-the-year largemouth bass and the associated deviations from expected counts. Pond Number of rings Change Change in age (days) Error Sum of error Date N Mean Range SE in count K-2 May 26 30 38.8 35-49 2.16 June 21 27 65.9 61-68 1.73 27 26 1 1 July 6 15 81.7 80-83 0.80 16 15 1 2 July 29 19 104.2 100-105 1.34 22 23 -1 1 Aug. 29 25 132.1 122-137 4.38 28 31 -3 -2 Sept. 27 25 164.0 153-170 3.78 32 29 3 1 Oct. 25 10 186.2 176-195 4.96 22 28 -6 -5 SC-4 July 1 25 84.7 81-87 1.60 July 25 24 109.5 98-114 3.97 25 24 1 1 Aug. 22 23 139.1 131-142 2.70 29 28 1 2 SC-5 July 1 33 63.3 55-74 3.83 July 25 25 83.5 77-88 2.83 21 24 -3 -3 Aug. 17 26 106.0 98-109 2.82 22 23 -1 -4 Young largemouth bass were collected on three occasions from ponds SC-4 and SC-5 during a 52-day period (Table 2). The differences between mean otolith ring count and the time interval in days between consecutive samples was <1 from SC-4 and <3 from SC-5. The deviation from the expected difference between the first and last sample was two days in SC- 4, and —4 days in SC-5. Resolution of growth rings was poor throughout the sampling period and variability was high in all samples. Linear regression used to describe the relationship between the number of rings counted and time in days for data from all three ponds combined had a slope of 0.995 (R2 = 0.991). The regressions used to describe the number of rings at a given time in days for each pond had slopes of 0.99, 1.05, and 0.90 for ponds K-2, SC-4, and SC-5, respectively. The 95 percent confidence interval for estimating the ring count from age for an individual fish was less than ±7 rings for the combined data and less than ±6 rings for the regression from an individual pond. Whereas the slopes of the individual regressions were different from each other, and the slopes from ponds SC-4 and SC-5 were different from 1.0 (a <0.05), the power to detect a deviation in slope of 0.025 was >0.98. Although Schmidt and Fabrizio (1980) did not detect a significant deviation from daily deposition of rings, their power to detect a larger difference of 10 percent in the mean age of wild largemouth bass (estimated from their published statistics) was only 0.72. The deviation from a slope of 1.0 observed in this study was due to the power of the test and may not be biologically significant. Variation from the expected ring count of one per day likely resulted from a consistent inability to distinguish sub-daily from daily rings, or loss of resolution of rings that RING DEPOSITION IN OTOLITHS OF BASS 277 were closely spaced due to slow growth of fish in some samples, rather than from nondaily ring deposition. Conclusions Otolith rings appear to be deposited with daily regularity as long as the fish are growing. When growth slows, otolith increments become compressed and difficult to distinguish. Our results indicate that rings were generally deposited with daily regularity on otoliths in largemouth bass for up to 190 days, at which point rings became too closely spaced to accurately count due to slow growth of the fish. Cessation of growth also was used to explain why Miller and Storck (1982) could only verify the age of largemouth bass to 100 days, whereas Taubert and Tranquilli (1982) were able to do so for the length of the growing season (180 days). The use of daily growth rings to age fish becomes invalid when otolith rings become too compressed to be distinguished reliably, and this apparently occurs at different times for different populations. Acknowledgments This study was conducted as part of Texas Agricultural Experiment Station project S- 6206, with additional support provided by the Tom Slick Graduate Research Fellowship program. Appreciation is extended to Drs. William H. Neill, William E. Grant, James A. Matis, and James A. Rice for their contributions to the development of this manuscript and to Alan E. Rudd for his frequent field assistance. Literature Cited Marshall, S. L., and S. S. Parker. 1982. Pattern identification in the microstructure of sockeye salmon ( Oncorhynchus nerka) otoliths. Canadian J. Fisheries and Aquatic Sci., 39:542-547. Miller, S. J., and T. Storck. 1982. Daily growth rings in otoliths and young-of-the-year largemouth bass. Trans. Amer. Fisheries Soc., 111:527-530. Pannella, G. 1971. Fish otoliths: daily growth layers and periodical patterns. Science, 173:1124-1126. - . 1974. Otolith growth patterns: an aid in age determination in temperate and tropical fishes. Pp. 28-39, in Ageing of fish. (T. B. Bagenal, ed.), Gresham Press, Old Woking, England. Schmidt, R. E., and M. C. Fabrizio. 1980. Daily growth rings on otoliths for aging young- of-the-year largemouth bass from wild populations. Prog. Fish Culturist, 42:78-80. Strusaker, P., and J. H. Uchiyama. 1976. Age and growth of the nehu, Stolephorus purpureus (Pisces:Engraulidae), from the Hawaiian Islands as indicated by daily growth increments of sagittae. U.S. Nat. Marine Fisheries Serv., Fisheries Bull., 74:9-17. Taubert, B. C., and D. W. Coble. 1977. Daily rings in otoliths of three species of Lepomis and Tilapia mossambica. J. Fisheries Res. Board Canada, 34:332-340. Taubert, B. C., and J. A. Tranquilli. 1982. Verification of formation of annuli in otoliths of largemouth bass. Trans. Amer. Fisheries Soc., 111:531-534. Present address of authors: Department of Zoology, North Carolina State University, Raleigh, North Carolina 27695. . 1 ' NOTES ON DISTRIBUTION AND NATURAL HISTORY OF SOME BATS ON THE EDWARDS PLATEAU AND IN ADJACENT AREAS OF TEXAS Richard W. Manning, J. Knox Jones, Jr., Robert R. Hollander, and Clyde Jones The Museum and Departments of Biological Sciences and Museum Science, Texas Tech University, Lubbock, Texas, 79409 Abstract — Distributional and ecological observations are recorded for 10 species of bats from the Edwards Plateau and adjacent areas of south-central Texas. Key words : Chiroptera; distribution; natural history; Texas. Except for studies of maternity colonies in large caves, little has been published concerning bats on the Edwards Plateau of south-central Texas since Blair’s (1952) preliminary survey. For the past two years, most importantly in the summer of 1986, we have collected mammals on the Edwards Plateau and in immediately adjacent areas in order to gain a better understanding of distributional patterns in this interesting ecological region. Among the materials thus far accumulated through our field efforts are specimens of 10 species of bats, a number of which represent noteworthy distributional records. These are reported here, along with additional specimens housed in The Museum of Texas Tech University that were collected earlier by other investigators. Pertinent comments on distribution and natural history are recorded. Unless noted otherwise, all dates in text relate to the year 1986. We are grateful to the Graduate School of Texas Tech University for providing a summer research assistantship to Manning in support of this project. We also thank the office of the Vice President for Academic Affairs and Research at Texas Tech, the National Institutes of Health, and the Theodore Roosevelt Memorial Fund of the American Museum of Natural History (to Hollander) for financial support of field work. My otis velifer incautus (J. A. Allen, 1896) The cave myotis is abundant and widely distributed on the Edwards Plateau (Blair, 1952), where breeding colonies of up to several thousand bats commonly are found. This species is a year-round resident of the region. We took lactating females as early as 19 May; a pregnant female netted on 28 May carried a single fetus (30 mm in crown-rump length). Both volant young and lactating females were captured on 19 and 21 June in Kimble County as they foraged for insects beneath a canopy of The Texas Journal of Science, Vol. 39, No. 3, August, 1987 280 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 pecan trees. We observed these bats using cliff swallow nests as day roosts in August in Kinney County, a behavior previously reported by Pitts and Scharninghausen (1986). Specimens examined, including bats from several new localities of record, are as follows: Crockett Co.: 19 mi. S, 16 mi. W Ozona, 1. Kimble Co.: Texas Tech Univ. Center at Junction, 35; 1 mi. S, 0.3 mi. W Junction, 2; 2.2 mi. S, 0.8 mi. E Junction, 13; 5 mi. S Texas Tech Univ. Center at Junction, 24; 8 mi. S Texas Tech Univ. Center at Junction, 8; Flemming Cave, 9.5 mi. S, 7 mi. W Junction, 3. Kinney Co.: 19 mi. SE Del Rio, 4. Llano Co.: Enchanted Rock, 1. Mason Co.: James River Bat Cave, 11 mi. S, 6 mi. W Mason, 10. Real Co.: Leakey, 19. Sterling Co.: 3.5 mi S, 4 mi. E Sterling City, 2. Sutton Co.: 3 mi. NE Sonora, 1. Val Verde Co.: 44 mi. N, 6 mi. W Del Rio, 1; Fawcett Cave, 36 mi. N Del Rio, 10; Dolan Springs, 36 mi. N, 6 mi. W Del Rio, 2; Comstrock R. R. Tunnel, 12 mi. W, 3 mi. S Comstock, 20; Fisher’s Fissure, 2 mi. W Langtry, 9; 0.5 mi. E Langtry, 1; Mile [Eagle Nest] Canyon, 0.75 mi. E Langtry, 3; Comstock, 2. My otis yumanensis yumanensis (H. Allen, 1864) The Yuma myotis is known from the Trans-Pecos region of Texas from Brewster, El Paso, Jeff Davis, Presidio, and Val Verde counties (Schmidly, 1977). This species occurs also on the southwestern edge of the Edwards Plateau to the east of the Pecos River. We captured three females, all lactating, in Val Verde County on 30 May. Specimens examined, all from Val Verde County, are as follows: 0.5 mi. E Langtry, 3; Mile [Eagle Nest] Canyon, 0.75 mi. E Langtry, 1; R. R. Tunnel, 1.2 mi. W, 2.7 mi. N mouth of Pecos River, 2; junction of Rio Grande and Pecos rivers, 3; Comstock R. R. Tunnel, 12 mi. W, 3 mi. S Comstock, 21. Lasionycteris noctivagans (Le Conte, 1831) When this manuscript was in galley proof, Robert J. Baker and a group of students netted four silver-haired bats, all males, under a canopy of pecan trees in the floodplain of the South Llano River at the Texas Tech University Center at Junction, Kimble County. All were taken on the night of 17 May 1987. Presumably late northward migrants, these specimens provide the second locality of record for L. noctivagans from the Edwards Plateau, the other being an individual obtained on 20 March 1948 along the Medina River, about 18 mi. W Medina, Bandera County (Blair, 1952). BATS OF EDWARDS PLATEAU 281 Pipistrellus hesperus maximus Hatfield, 1936 Western pipistrelles are widely distributed in Trans-Pecos Texas, having been recorded from all counties except Reeves and Pecos (Schmidly, 1977). Davis (1974) reported specimens from Val Verde, Uvalde, and Edwards counties on the southern edge of the Edwards Plateau. While netting in a steep-walled canyon, 0.5 mi. E Langtry, Val Verde County, we captured a lactating female on 30 May. Other bats taken at this location included Myotis velifer , Tadarida brasiliensis , and Myotis yumanensis. We also netted a lactating female on 5 July 1985 at a place 3 mi. S and 5 mi. E McCamey, Upton County. This locality lies to the east of Reeves and Pecos counties and helps to delineate the distribution of this species in the state. Pipistrellus subflavus subflavus (F. Cuvier, 1832) Blair (1952) reported this subspecies from only four counties on the Edwards Plateau — Bandera, Edwards, Kendall, and Kerr. Specimens available to us that help to clarify the distribution of this pipistrelle are as follows: Edwards Co.: Devil’s Sinkhole, 3. Kimble Co.: Texas Tech Univ. Center at Junction, 6; 5 mi. S Texas Tech Univ. Center at Junction, 1. Real Co.: Leakey, 6. At Junction, pipistrelles were netted under a canopy of large deciduous trees in the floodplain of the South Llano River. Those from Leakey were taken as described in the account of Nycticeius. The specimens listed above average somewhat paler, taking age and season into consideration, than do typical individuals of P. s. subflavus (see account below). Pipistrellus subflavus clarus Baker, 1954. According to Hall (1981), this race of the eastern pipistrelle has been recorded from Comstock, Devil’s River, and Del Rio, all in Val Verde County and all east of the Pecos River, and 12 mi. W Comstock near the mouth of the Pecos. At hand are three specimens from Trans-Pecos, Texas, all males, that were taken in Fisher’s Fissure, 2 mi. W Langtry, Val Verde County, on 12 April 1968. The original description of this subspecies (Baker, 1954) was based on specimens from Coahuila, Mexico, and immediately adjacent southern Texas. Compared with P. s. subflavus , it was noted to be of approximately the same size, but paler in color and with zygomata slightly more expanded laterally. Our three bats, all adults, are paler dorsally than spring-taken specimens from the Texas Panhandle but differ less from specimens from the Edwards Plateau to the east, suggesting the distinct possibility of an east-west cline in color across 282 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 central Texas. We note little difference in the breadth of the zygomata in Texas material we have examined. Although we have not studied Mexican specimens of clarus , recognition of this race, based on information now available, should be carefully scrutinized. Lasiurus borealis borealis (Muller, 1776) Red bats are common in eastern Texas and are thought to be permanent residents of that area (Schmidly, 1983). Blair (1952) considered this species to be fairly common on the eastern part of the Edwards Plateau but reported specimens only from Kerr and Val Verde counties. We collected a lactating female on 11 June in Sterling County as it flew under a canopy of pecan trees. We also netted a series of 15 red bats in Real County (see account of Nycticeius) in July and August. This sample included adult males and females plus young-of-the-year of both sexes. A female taken in Kimble County on 21 May carried four embryos (crown-rump length 20 mm). Specimens examined are as follows: Kimble Co.: Texas Tech Univ. Center at Junction, 14; 5 mi. S Texas Tech Univ. Center at Junction, 7; 17 mi. SE Junction, 2. Real Co.: Leakey, 15. Sterling Co.: 4 mi. S, 4 mi. E Sterling City, 1. Sutton Co.: Sonora, 1. Lasiurus cinereus cinereus (Palisot de Beauvois, 1796) Davis (1974) reported this species as a “relatively rare migrant through Texas.” Schmidly (1977), however, reported the hoary bat as resident in the warm months in “wooded, montane areas” of the Trans-Pecos region. On 19 April, we netted a barren female over a small cattle pond, which was surrounded by mesquite and in desert scrub habitat, in Crockett County. A male taken on 19 September in Sutton County had numerous small cactus spines in its wing membranes and several imbedded in the abdominal fur. That same evening another hoary bat was observed to hit the mist net we had stretched over a concrete tank, fall into the water, and take to the air from the surface, a behavior previously reported by Schmidly (1983). The only information on reproduction in the state by this species outside the Chisos and Guadalupe mountains of far western Texas (pregnant females captured in April and June, respectively — Schmidly, 1977) is of a gravid female obtained on 19 April 1978 in Big Thicket National Preserve (Schmidly, 1983). It is noteworthy, therefore, that two females taken on 20 and 21 May in Kimble County were pregnant, carrying two fetuses each that measured 9 and 10 mm, respectively, in crown-rump length. While these and other females mentioned above may have been migrants, the presence of pregnant individuals in Texas as late as May and June strongly underscores the possibility that young are born in the state. BATS OF EDWARDS PLATEAU 283 Specimens examined are as follows: Crockett Co.: 5 mi. S, 5 mi. E McCamey, 1. Kimble Co.: Texas Tech Univ. Center at Junction, 3; 5 mi. S Texas Tech Univ. Center at Junction, 1. Sutton Co.: 13 mi. W Sonora, 1. Nycticeius humeralis humeralis Rafinesque, 1818 The evening bat is known to occur primarily to the east of the Balcones Escarpment in Texas (Schmidly, 1983). It has been reported only from Kerr and Bandera counties on the Edwards Plateau (Blair, 1952; Davis, 1974). We netted five evening bats under a cypress canopy along Leakey Creek at Leakey, Real County (see Fig. 1). One young female was taken there on 10 July, and a male and three females were captured on 20 August. An additional five evening bats (three females, two males) were netted over a tree-lined stretch of Pinto Creek, 19 mi. SE Del Rio, in Kinney County, on 22 August. The specimens here reported represent the westernmost records of this species in Texas. Antrozous pallidus pallidus (Le Conte, 1856) The pallid bat is known to occur eastward on the Edwards Plateau at least as far as Kerr and Kimble counties (Martin and Schmidly, 1982). It is widely distributed in the western part of the region. We took lactating females in Kimble County on 19 May, 19 June, and 21 June. A female taken there on 21 May carried two fetuses (29 mm in crown-rump length). On the night of 30 June, while netting for bats underneath a bridge over the Pecos River in Crockett County, we noted that pallid bats were using the underside of the bridge as a night roost. After collecting several, we observed bat activity for one-half hour or so. Near the end of that time, approximately two hours after sunset, a large female (forearm 51 mm, weight 21 grams) Antrozous alighted under the bridge closely followed by two smaller individuals (forearms 48 mm, weights 14 and 12 grams), which immediately began to nurse from her. We collected these three bats. Examination of the stomachs of the young animals (both males) revealed that the lower portion was filled with insects but that the upper part contained a layer of milk. This observation indicates that young Antrozous both forage for insects and continue to nurse for a time after they become volant. It also implies a strong mode of communication, even while on the wing, between members of a family group, which evidently forage together in mid-summer (see O’Shea and Vaughan, 1977). Specimens examined from the Edwards Plateau and adjacent areas, in addition to those reported by Martin and Schmidly (1982), are as follows: Crockett Co.: 5 mi. N, 4 mi. W Iraan, 8. Kimble Co.: Texas Tech Univ. Center at Junction, 64; 1 mi. S Texas Tech Univ. Center at 284 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Figure 1. Photograph of the place where bats were netted along Leakey Creek, Real County. Myotis velifer, Pipistrellus subflavus , Lasiurus borealis , Nycticeius humeralis, and Tadarida brasiliensis were collected at this site, which may be the southwesternmost locality of occurrence in Texas of bald cypress and Spanish moss. Junction, 1; 5 mi. S Texas Tech Univ. Center at Junction, 7; 17 mi. SE Junction, 1. Upton Co.: 3 mi. S, 5 mi. E McCamey, 8. Tadarida brasiliensis mexicana (Saussure, 1860) The Brazilian free-tailed bat is the most common chiropteran on the Edwards Plateau. Blair (1952) stated that T. brasiliensis is a year-round resident of the region. We, however, doubt that this is true, especially in the western areas with which we are most familiar. Short et al. (1960) suggested that southward migration of this species from the plateau may follow a westerly path. A nonpregnant female and a male captured on 18 March in Upton County probably represent early migrants into (or through) that area. Lactating females were taken as late as 10 July (Real County). Bracken Cave, Comal County, and Ney Cave, Medina County, each are known to house maternity colonies that have been estimated to contain 20 to 30 million free-tailed bats (Raun and Baker, 1959). Blair (1952), Short et al. (1960), and McCracken (1984) listed other caves on the Edwards Plateau that serve as maternity sites. Specimens examined, many of which represent new localities of record, are as follows: Blanco Co.: Davis Cave, vie. 11 mi. N, 10 mi. W Johnson BATS OF EDWARDS PLATEAU 285 City, 65. Comal Co.: Bracken Cave, vie. 2 mi. S, 11 mi. W New Braunfels, 70. Crockett Co.: 5 mi. S, 5 mi. E McCamey, 12. Kimble Co.: Texas Tech Univ. Center at Junction, 11; 1 mi. S, 0.3 mi. W Junction, 1; 5 mi. S Texas Tech Univ. Center at Junction, 6. Kinney Co.: 19 mi. SE Del Rio, 1. Mason Co.: James River Bat Cave, 11 mi. S, 6 mi. W Mason, 13. Medina Co.: Ney Cave, 61. Midland Co.: Midland, 1. Real Co.: Leakey, 1. Sterling Co.: 3.5 mi. S, 4 mi. E Sterling City, 22. Sutton Co.: 3 mi. NE Sonora, 25; 13 mi. W Sonora, 2; 3 mi. E Sonora, 1. Upton Co.: 3 mi. S, 5 mi. E McCamey, 2. Uvalde Co.: Frio Cave, 19. Val Verde Co.: 5 mi. S, 3 mi. E Pandale, 1; Pecos River, near Pandale, 2; 44 mi. N, 6 mi. W Del Rio, 2; Fern Cave, 40 mi. NW Del Rio, 1; Comstock R. R. Tunnel, 12 mi. W, 3 mi. S Comstock, 28; Fisher’s Fissure, 2 mi. W Langtry, 2; 0.5 mi. E Langtry, 2; Mile [Eagle Nest] Canyon, 0.75 mi. E Langtry, 8. Literature Cited Baker, R. H. 1954. A new bat (genus Pipistrellus) from northeastern Mexico. Univ. Kansas Publ., Mus. Nat. Hist., 7:583-586. Blair, F. W. 1952. Bats of the Edwards Plateau in central Texas. Texas J.Sci., 4:95-98. Davis, W. B. 1974. The mammals of Texas. Bull. Texas Parks and Wildlife Dept., 41:1-267. Hall, E. R. 1981. The mammals of North America. John Wiley & Sons, New York, 2nd ed., l:xv+ 1-600+90. McCracken, G. F. 1984. Communal nursing in Mexican free-tailed bat maternity colonies. Science, 223:1090-1091. Martin, C. O., and D. J. Schmidly. 1982. Taxonomic review of the pallid bat, Antrozous pallidus (Le Conte). Spec. Publ. Mus., Texas Tech Univ., 18:1-48. O’Shea, T. J., and T. A. Vaughan. 1977. Nocturnal and seasonal activities of the pallid bat, Antrozous pallidus. J. Mamm., 58:269-284. Pitts, R. M., and J. J. Scharninghausen. 1986. Use of cliff swallow and barn swallow nests by the cave bat, Myotis velifer, and the free-tailed bat, Tadarida brasiliensis. Texas J. Sci., 38:265-266. Raun, G. G., and J. K. Baker. 1959. Some observations of Texas cave bats. Southwestern Nat., 3:102-106. Schmidly, D. J. 1977. The mammals of Trans-Pecos Texas. Texas A&M Univ. Press, College Station, xiii+255 pp. - . 1983. Texas mammals east of the Balcones Fault Zone. Texas A&M Univ. Press, College Station, xviii+400 pp. Short, H. L., R. B. Davis, and C. F. Herreid, II. 1960. Movements of the Mexican free¬ tailed bat in Texas. Southwestern Nat., 5:208-216. GENERAL NOTES ADDITIONAL RECORDS OF TICK (ACARI: IXODIDAE, ARGASIDAE) INGESTION BY WHIPTAIL LIZARDS, GENUS CNEMIDOPHORUS Chris T. McAllister and James E. Keirans Renal- Metabolic Laboratory (151-G), Veterans Administration Medical Center, 4500 S. Lancaster Road, Dallas, Texas 75216, and Department of Biological Sciences, North Texas State University, Denton, Texas 76203, and Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Allergy and Infectious Diseases, Department of Entomology, Museum Support Center, Smithsonian Institution, Washington, D. C. 20560 There are few reports of reptiles ingesting ticks. In Africa, Norval (1976) reported that the yellow-throated plated lizard, Gerrhosaurus flavigularis, would feed on the ixodid or hard tick species Haemaphysalis muhsami, Amblyomma nuttalli, and A. hebraeum and Norval and McCosker (1983) observed the rainbow skink, Mabuya quinquetaeniata margaritifer, feeding on the African cattle tick, A. hebraeum. In the United States, Degenhardt and Jones (1972) recorded “ticks” in the stomach contents of the dunes sagebrush lizard, Sceloporus graciosus arenicolous, and McAllister (1987) reported the first instance of the argasid or soft tick Otobius megnini being eaten by the Texas spotted whiptail lizard, Cnemidophorus gularis gularis. Herein we report two additional cases of predation on ticks by two species of whiptail lizards. An adult male C. g. gularis (snout-vent length, 85 mm) was obtained by F. S. Hendricks on 19 May 1974, 20.8 kilometers east of Soledad at an elevation of 665 meters in the state of Coahuila, Mexico. The stomach contained a single female Dermacentor parumapertus. The presence of hair in the stomach of the lizard along with this tick suggests that the tick may have been taken from a dead host. The other situation involved an adult female Colorado checkered whiptail, C. tesselatus (pattern class “C”; snout-vent length, 98 mm), taken by J. F. Scudday on 26 June 1974 at the Mansfield Ranch, 32.5 kilometers northwest of Vega, Oldham Co., Texas. Five (two male, two female, one nymph) spinose ear ticks, Otobius megnini, were found in the stomach. Fecal material from the rectum appeared also to contain digested tick remains, which may indicate feeding on ticks on more than one occasion. The finding of two males, two females, and one nymph of O. megnini in the stomach of this lizard suggests that it probably scavenged these ticks from the ground. The senior author found no additional ticks in more than 300 whiptail lizards of six species. The frequency of ticks in the diet of Cnemidophorus sp. is estimated to be less than one percent. Further, the comprehensive dietary studies of Milstead (1957, 1965), Medica (1967), Milstead and Tinkle (1969), and Scudday and Dixon (1973) on various species of Cnemidophorus did not list ticks as a prey item of whiptail lizards. These lizards utilize the “widely-foraging” strategy of obtaining prey (Mac Arthur and Pianka, 1966) and have a seasonal variation in prey consumed, indicative of temporally abundant Arthropoda. Reference specimens are deposited in the Smithsonian Institution Acarological Collections as follows: O. megnini (RML 118495); D. parumapertus (RML 118399). The lizards are deposited in the Texas Cooperative Wildlife Collection and the Sul Ross State University Collection as follows: C. g. gularis (TCWC 46905); C tesselatus (SRSU 3392). The Texas Journal of Science, Vol. 39, No. 3, August, 1987 288 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 We thank J. R. Dixon, M. E. Retzer, and J. F. Scudday for loaning the lizards specimens used in this report and D. B. Pence for critically reviewing the manuscript. Literature Cited Degenhardt, W. G., and K. L. Jones. 1972. A new sagebrush lizard, Sceloporus graciosus, from New Mexico and Texas. Herpetologica, 28:212-217. MacArthur, R. H., and E. R. Pianka. 1966. On optimal use of patchy environment. Amer. Nat., 100:603-609. McAllister, C. T. 1987. Ingestion of spinose ear ticks, Otobius megnini (Acari: Argasidae) by a Texas spotted whiptail, Cnemidophorus gularis gularis (Sauria: Teiidae). Southwestern Nat., 32: in press. Medica, P. A. 1967. Food habits, habitat preference, reproduction, and diurnal activity in four sympatric species of whiptail lizards ( Cnemid phorus ) in south central New Mexico. Bull. So. California Acad. Sci., 66:251-276. Milstead, W. W. 1957. Some aspects of competition in natural populations of whiptail lizards (Genus Cnemidophorus). Texas J. Sci., 9:410-447. - . 1965. Changes in competing populations of whiptail lizards ( Cnemidophorus ) in southwestern Texas. Amer. Midland Nat., 73:75-80. Milstead, W. W., and D. W. Tinkle. 1969. Interrelationships of feeding habits in a population of lizards in southwestern Texas. Amer. Midland Nat., 81:491-499. Norval, R. A. I. 1976. Lizards as opportunist tick predators. Rhodesian Vet. J., 7:63. Norval, R. A. I., and P. J. McCosker. 1983. Tick predation by a rainbow skink. Zimbabwe Vet. J., 13:53. Scudday, J. F., and J. R. Dixon, 1973. Diet and feeding behavior of teiid lizards from Trans-Pecos Texas. Southwestern Nat., 18:279-289. A SPECIMEN OF WHITE-WINGED DOVE, ZENAIDA ASIATICA, FROM ARCHER COUNTY, TEXAS Frederick B. Stangl, Jr., and Warren Pulich Department of Biology, Midwestern State University, Wichita Falls, Texas 76308, and Department of Biology, University of Dallas, Irving, Texas 75062 The white-winged dove ( Zenaida asiatica) occurs in Texas primarily in the Rio Grande Valley and South Texas Plains, and locally north to Bexar County, with scattered sightings from the southern Panhandle (Arnold, 1984). In Oklahoma, there are sight records from Greer and Jackson counties (Wood and Schnell, 1984). A white-winged dove was shot by a hunter 1 mi. S Holliday, Archer Co., Texas, on 21 September 1986. Apparently, it was mistaken for a mourning dove ( Zenaida macroura), but the left wing was saved, and was presented to the Midwestern State University Collection of Birds. The specimen (MWSU 1073) represents the only voucher record known to us of a white-singed dove from as far north in Texas as Archer County. With the exception of a specimen collected from the Texas High Plains on 17 May 1890 in Armstrong County (Oberholser, 1974), all other extra-limital records are sight records. Oberholser’s original manuscript (filed on microfilm at the University of Dallas, Irving, Texas) indicated that the J. K. Strecker collection at Baylor University, Waco, Texas, may have been the repository for the specimen in question. A check at Baylor and six other The Texas Journal of Science, Vol. 39, No. 3, August, 1987 GENERAL NOTES 289 major museums with avian material from Texas did not uncover the alleged specimen. We believe the example from Armstrong County may no longer exist. We thank curators at the following museums for their assistance: American Museum of Natural History, Cincinnati Museum of Natural History, Strecker Museum of Natural History, Texas A&M University, University of Texas, U.S. National Museum, and Yale University. Literature Cited Arnold, K. A. 1984. Texas Ornithological Society Checklist of the birds of Texas. Texas Ornith. Soc., 2nd ed., 147 pp. Oberholser, H. C. 1974. The bird life of Texas. Univ. Texas Press, Austin, 530 pp. Wood, D. S., and G. D. Schnell. 1984. Distributions of Oklahoma birds. Univ. Oklahoma Press, Norman, xxi+209 pp. ANOLIS SA GREI (SAURIATGU ANIDAE) ESTABLISHED IN SOUTHERN TEXAS Ken King, David Cavazos, and Frank W. Judd Department of Biology and Coastal Studies Laboratory Pan American University, Edinburg, Texas 78539 Anolis sagrei, the brown anole, is native to Cuba, Jamaica, and the Bahamas. The species has been introduced and established in Florida and in several localities from southern Mexico to Honduras (Conant, 1975). The brown anole was first reported in the United States by Garman (1887), who listed the locality as “Florida Keys.” Godley et al. (1981) reviewed the distributional status of A. sagrei in Florida and reported that it now inhabits much of the peninsular part of that state as disjunct colonies that generally are confined to urban settings. Wilson and Porras (1983) stated that A. sagrei is now the most common reptile in urban areas of southeastern Florida. Surprisingly, there were no records of the brown anole in other Gulf Coast states until Dixon (1987) reported the species in Houston, Texas, in 1985. His specimens were found in a plant nursery by J. W. Werler. At the urging of Dr. Dixon we initiated a survey of plant nurseries in southern Texas to ascertain if A. sagrei was present. Plant nurseries were deemed likely places for residency of the brown anole because Godley et al. (1981) suggested that one of the primary means of dispersal was through the transport of lizards or their eggs in ornamental plants via the commercial nursery trade. During the fall of 1986, we searched nurseries in Brownsville and Harlingen (Cameron County), Edinburg, McAllen, and Mission (Hidalgo County) and San Antonio (Bexar County) Texas, for A. sagrei. We found brown anoles at one nursery in each of three cities — Brownsville, Harlingen, and San Antonio. We collected three males and two juveniles from the Harlingen nursery on 19 October 1986. Two of the males were sent to J. R. Dixon, Texas A&M University, who confirmed that the lizards were indeed A. sagrei , but he noted that they were not of the same genetic stock as specimens from Tampa, Florida. Both adults and juveniles were observed, captured, and released at the Brownsville nursery on 18 October 1986. Several juvenile A. sagrei were observed at the San Antonio nursery on 10 October 1986. Two were captured, but the owner denied our request to keep the specimens. On 1 November 1986, we collected 18 A. sagrei from the Harlingen nursery — nine adult males, six adult females, The Texas Journal of Science, Vol. 39, No. 3, August, 1987 290 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 and three juveniles. The population at this nursery is large. Brown anoles were seen throughout the 2.4 hectares of grounds. We could easily have captured additional specimens. Furthermore, the owner and an employee reported that they had populations established at their homes. There is no question that breeding populations of A. sagrei are now established in Cameron County, Texas, and it is likely that juveniles we saw in San Antonio reflect a breeding population there. The Brownsville and Harlingen localities are only about 48 kilometers apart and both are near ports on the coast. Conversely, the San Antonio locality is 386 kilometers north (inland) from Harlingen. Interviews with the owners and employees of the nurseries convinced us that brown anoles have been present at each nursery for about three years. In December 1983, southern Texas experienced a devastating freeze that killed most of the citrus trees and many other introduced tropical plants, including palms. Anolis sagrei may have been introduced to southern Texas when nurserymen replenished their stocks of ornamental tropical plants from Florida following the 1983 freeze. Alternatively, the lizards or their eggs may have been introduced via ships visiting the ports of Brownsville or Harlingen. Multiple introductions by both methods is a distinct possibility. Furthermore, A. sagrei may have become established at other localities in southern Texas and along the Texas Gulf Coast. Hopefully, this note will help stimulate interest in surveys for the brown anole so that its range in Texas can be ascertained and its spread or decline documented. We thank James R. Dixon for confirming our identification of Anolis sagrei , suggesting this study, and reviewing an earlier draft of the manuscript. Literature Cited Conant, R. 1975. A field guide to reptiles and amphibians of eastern and central North America. Houghton Mifflin Co., Boston, xvii+429 pp. Dixon, J. R. 1987. The amphibians and reptiles of Texas, with keys, taxonomic synopses, bibliography, and distribution maps. Texas A&M Univ. Press, College Station, in press. Garman, S. 1887. On West Indian reptiles. Iguanidae. Bull. Essex Inst., 19:1-26. Godley, J. S., F. E. Lohrer, J. N. Layne, and J. Rossi. 1981. Distributional status of an introduced lizard in Florida: Anolis sagrei. Herpetol. Rev., 12:84-86. Wilson, L. D., and L. Porras. 1983. The ecological impact of man on the South Florida herpetofauna. Spec. Publ. Mus. Nat. Hist., Univ. Kansas, 9:1-80. FIRST RECORD OF CENTROPAGES TYPICUS KROYER (COPEPODA: CENTROPAGIDAE) IN THE GULF OF MEXICO David C. McAden, George N. Greene, and William B. Baker, Jr. Houston Lighting & Power Company, Environmental Protection Department, P.O. Box 1700, Houston, Texas 77001 On two dates in September, three dates in October, and three dates in November 1985, the calanoid copepod, Centropages typicus Kroyer, previously unknown from the Gulf of Mexico, was collected in plankton samples from the lower Colorado River of Texas, approximately 26.4 kilometers upstream of the mouth of the river and 25.9 kilometers downstream of the Port of Bay City in Matagorda County, Texas. A total of 91 specimens was collected. The Texas Journal of Science, Vol. 39, No. 3, August, 1987 GENERAL NOTES 291 C. typicus is a cool-water species known to occur along the Atlantic coasts of the United States and Europe (Van Engle and Tan, 1965; Sars, 1902). Bowman (1971) stated that C. typicus does not normally occur south of Cape Hatteras, North Carolina. An extensive literature review and personal communications revealed no previous reports of C. typicus from the Gulf of Mexico. Two species of the genus Centropages, C. hamatus and C. velificatus, are known to occur in the neritic waters of Texas (R. D. Kalke, University of Texas Marine Science Institute, Port Aransas, personal communication). NUS Corporation (1976), in an earlier study conducted at the same location and additional locations upstream and downstream, reported no C. typicus , but did record C. hamatus and C. velificatus. C. typicus was collected from the Colorado River in depths ranging from the surface to the bottom (approximately 20 feet). The range of water temperature over which it was found was 19.5 to 29.7° C. Salinity varied from 1.6 parts per thousand (surface) to 30.6 (bottom). Seventy-nine of the 91 specimens collected were taken in salinities above 19.0 parts per thousand. Sixty of the 91 were from a single bottom sample collected on 6 November 1985 (water temperature, 19.5°C; salinity, 21.6 parts per thousand). In samples collected on 31 October and 1 November 1985, which contained a total of 12 individuals of C. typicus, there were also nine specimens of C. velificatus. These were the only samples in which more than one species of Centropages occurred. Exactly when C. typicus first appeared in the lower Colorado River and how it was introduced is not known. It may have been transported via ballast tanks on ships or barges from the Atlantic coast of the United States. Additional sampling is necessary to determine if C. typicus has established a resident population in the lower Colorado River. Thirty specimens were deposited in the United States National Museum of Natural History (USNM no. 231086). The authors wish to acknowledge Dr. Abraham Fleminger of the Scripps Institution of Oceanography and Dr. Thomas Bowman of the Smithsonian Institution for confirming the identification of specimens sent to them. Dr. Frank G. Schlicht of Houston Lighting & Power Company contributed valuable editorial comments and Richard D. Kalke of the University of Texas Marine Science Institute at Port Aransas, Texas, shared with us his knowledge of the genus Centropages from the coastal waters of Texas. Literature Cited Bowman, T. E. 1971. The distribution of calanoid copepods off the southeastern United States between Cape Hatteras and southern Florida. Smithsonian Contrib. Zook, 96:1- 58. NUS Corp. 1976. Final report, Colorado River monitoring program, phase one studies — April, 1975-March, 1976. Doc. no. R-32-00- 12/ 76-676, 147 pp. Sars, G. O. 1902. An account of the Crustacea of Norway, 4. Copepoda Calanoida. Bergen, 171pp. Van Engle, W. A., and E. Tan 1965. Investigations of inner continental shelf waters off lower Chesapeake Bay. Part VI. The Copepods. Chesapeake Sci., 6:183-189. 292 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 EVIDENCE OF COMMUNAL NESTING AND WINTER-KILL IN A POPULATION OF BAIOMYS TAYLORI FROM NORTH-CENTRAL TEXAS Frederick B. Stangl, Jr., and Stephen Kasper Department of Biology, Midwestern State University, Wichita Falls, Texas 76308 The pygmy mouse, Baiomys taylori, may become locally abundant under favorable conditions. Although captive individuals are known to be highly tolerant of each other, we are unaware of any records of communal nesting under natural conditions. On 31 January 1987, a large (25 centimeters in diameter), globular surface nest of fine grasses was found under an inverted wooden crate on the Texas A&M University Agricultural Research Station, 10 mi. N Throckmorton, Throckmorton Co., Texas. The single hollow nest chamber contained seven dead adult pygmy mice, in similarly early stages of decomposition. All but two were partially eaten, but whether from cannabalism or from opportunistic scavenging by other small mammals was impossible to determine. Just previous to our discovery, the research station experienced four days (15-18 January) of continued subfreezing temperatures and frozen precipitation (mostly in the form of freezing rain), forming ice sheets of up to one inch in thickness. The following interval was characterized by cool days and subfreezing nights. The mice apparently huddled together for warmth. The cause of death is not known; possibly the seven animals succumbed to hypothermia, or starved to death because of difficult foraging conditions or by becoming trapped under the ice-covered box. Baiomys taylori is mostly subtropical in distribution, although the expansion of its range northward and westward in Texas (Austin and Kitchens, 1986; Cleveland, 1986; Dalquest and Horner, 1984; Hart, 1972; Hollander et al., 1987; Stangl et al., 1983) and into Oklahoma (Stangl and Dalquest, 1986) is well-documented. However, Stangl and Dalquest (1986) noted that inhospitable weather might periodically halt or reverse the progress of range expansion, giving the decline in numbers of specimens from north-central Texas in recent years as indirect evidence. The lack of similar records elsewhere for communal nesting and cases of winter-kill in Baiomys taylori suggests to us that these occurrences are uncommon, and may be seasonally-induced, geographic phenomena along the northern margins of the range of the species. It would be premature to claim that our observation was more than an isolated incident. However, we note that previous trapping in the vicinity (November 1986) indicated high population densities of the pygmy mouse. Subsequent collecting efforts through April 1987 have failed to produce any additional specimens. We gratefully acknowledge Drs. Rod Heitschmidt and Bill Pinchak for the invitation to collect specimens on the Texas A&M University Research Station. Walter W. Dalquest, Arthur G. Cleveland, and Dawn Morris critically reviewed the manuscript. Weather data was kindly provided by Dorothy Keeter, Throckmorton County weather recorder. Literature Cited Austin, T. A., and J. A. Kitchens. 1986. Expansion of Baiomys taylori into Hardeman County, Texas. Southwestern Nat., 31:547-548. Cleveland, A. G. 1986. First record of Baiomys taylori north of the Red River. Southwestern Nat., 31:547. Dalquest, W. W., and N. V. Horner. 1984. Mammals of north-central Texas. Midwestern State Univ. Press, Wichita Falls, 261 pp. The Texas Journal of Science, Vol. 39, No. 3, August, 1987 GENERAL NOTES 293 Hart, B. J. 1972. Distribution of the pygmy mouse, Baiomys taylori , in north-central Texas. Southwestern Nat., 17:213-214. Hollander, R. R., J. K. Jones, Jr., R. W. Manning, and C. Jones. 1987. Noteworthy records of mammals from the Texas Panhandle. Texas J. Sci., 39:97-102. Stangl, F. B., Jr., and W. W. Dalquest. 1986. Two noteworthy records of Oklahoma mammals. Southwestern Nat., 31:123-124. Stangl, F. B., Jr., B. F. Koop, and C. S. Hood. 1983. Occurrence of Baiomys taylori (Rodentia: Cricetidae) on the Texas High Plains. Occas. Papers Mus., Texas Tech Univ., 85:1-4. NEW WESTERN DISTRIBUTIONAL RECORD OF TERRAPENE CAROLINA TRIUNGUIS Michael R. Gannon Department of Biological Sciences and The Museum, Texas Tech University, Lubbock, Texas 79409 A specimen of the three-toed box turtle, Terrapene Carolina triunguis , was obtained at the Texas Tech University Center at Junction, a mile south of Junction, Kimble Co., Texas (30° 30'N, 99°45'W). The individual, a male, was taken on 25 June 1986 from a trail transversing a wooded area approximately 100 meters north of the South Llano River. It was deposited in the herpetology collection of The Museum, Texas Tech University (TTU 11298). Of the four subspecies of T. Carolina that occur in the United States, T. c. triunguis ranges farthest west, reaching into parts of eastern Texas (Conant, 1975; Ernst and Barbour, 1972; Raun and Gehlbach, 1976). This specimen is the first record from Kimble County and marks a western range extension of nearly 100 miles. I thank M. R. Willig and M. J. van Staaden for reviewing this manuscript. I also thank J. R. Dixon for furnishing current distributional records. Literature Cited Conant, R. 1975. A field guide to reptiles and amphibians. Houghton Mifflin Co., Boston, 429 pp. Ernst, C. H., and R. W. Barbour. 1972. Turtles of the United States. Univ. Kentucky Press, Lexington, 347 pp. Raun, G. G., and F. R. Gehlbach. 1976. Amphibians and reptiles in Texas. Bull. Dallas Mus. Nat. Hist., 2: 1-61+71 pis. FIRST RECORDS OF THE FLAMMULATED OWL ( OTUS FLAMMEOLUS) IN THE CENTRAL TRANS-PECOS OF TEXAS Donna Burt, D. Brent Burt, Terry C. Maxwell, and Ross C. Dawkins Departments of Biology and Chemistry (RCD), Angelo State University, San Angelo, Texas 76909. On 14 September 1986, an adult flammulated owl (Otus flammeolus) in red phase was captured in a mist net in upper Madera Canyon of the Davis Mountains, 26 km. NW Fort The Texas Journal of Science, Vol. 39, No. 3, August, 1987 294 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Davis, Jeff Davis Co., Texas. This locality was at an elevation of 1920 meters in an association of ponderosa pine ( Pinus ponderosa), pinyon pine ( P. cembroides ), and alligator juniper ( Juniperus deppeana ). The owl was captured under ponderosa pines within 100 meters of a stream bed, banded (U.S. Fish and Wildlife Service band no. 1433-13669), photographed, and released. This is the first record of a flammulated owl from the Davis Mountains. On 19 April 1987, an adult female flammulated owl, also in red phase, was collected 16 km. SW Alpine, Brewster Co., Texas. This locality, 58 kilometers southeast of the Madera Canyon site, was at an elevation of 1675 meters. The owl was captured in a mist net placed between buildings of a church encampment in an oak ( Quercus sp.) grove along Toronto Creek. The specimen is deposited in the collection of birds at Angelo State University (ASNHC 504). This is the first record for Brewster County outside of the Chisos Mountains, Big Bend National Park. This migratory species has been reported as a regular but rare and local summer breeding resident (late March through September) at high elevations (above 1900 meters) in the Guadalupe and Chisos mountains (Wauer, 1973; Oberholser, 1974; Newman, 1979; Arnold, 1984), and as irregular in the Chinati Mountains (Oberholser, 1974). Although the dates of our records preclude speculation on the breeding status of the species in the central Trans- Pecos, they do suggest that the species occurs regularly there. Literature Cited Arnold, K. A. 1984. Checklist of the birds of Texas. Texas Ornithol. Soc., 2nd ed., 147 pp. Newman, G. A. 1979. Compositional aspects of breeding avifaunas in selected woodlands of the southern Guadalupe Mountains, Texas. Pp. 181-237, in Biological investigations in the Guadalupe Mountains National Park, Texas (H. H. Genoways and R. J. Baker, eds.), Proc. Trans. Ser., Nat. Park Serv., 4:xvii+ 1-442. Oberholser, H. C. 1974. The bird life of Texas. Univ. Texas Press, Austin, 2 vols., 1069 pp. Wauer, R. H. 1973. Birds of Big Bend National Park and vicinity. Univ. Texas Press, Austin, 223 pp. INSTRUCTIONS TO AUTHORS Scholarly manuscripts in any field of science or technology, including science education, will be considered for publication in The Texas Journal of Science. Prior to acceptance, each manuscript will be reviewed by at least two knowledgeable critics, the appropriate Associate Editor, and the Editor. Manuscripts intended for publication in the Journal should be submitted to the Editor, J. Knox Jones, Jr. (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409), in accordance with the following instructions. No manuscript submitted to the Journal is to have been published or submitted elsewhere. Manuscripts must be double-spaced throughout (including tables, legends, and cited literature) and submitted in triplicate on typed or clear machine-produced copies on 8.5 x 11-inch paper, with margins of approximately 1.5 inches. The centered title of the article (usually 10 words or less) should be followed by the name(s) of the author(s) and institutional or business address(es), including zip-code, both also centered on the title page. Each manuscript should have a brief, concise Abstract, terminating with up to five key words. The following text can be subdivided into sections as appropriate (examples follow): introductory information is self evident and thus usually needs no heading; materials and methods (acknowledgments frequently can be placed here as well); results; discussion; summary or conclusions; literature cited. Major sections are centered and capitalized; secondary divisions are italicized (underlined) left flush; tertiary headings are italicized at the beginning of paragraphs. Cite all references in text by author and date in chronological order — Jones (1971); Jones (1971, 1975); (Jones, 1971); (Jones, 1971, 1975); (Jones, 1971; Smith, 1973; Davis, 1975); Jones (1971), Smith (1973), Davis (1975); Smith and Davis (1985); (Smith and Davis, 1985). If more than two authors, use Jones et al. (1976) or (Jones et ah, 1976). Citations to publications by the same author(s) in the same year should be designated alphabetically (1979a, 1979b). Be sure all citations in text are included in the Literature Cited section and vice versa. Hypothetical examples of proper citations are given below. Davis, G. L. 1975. The mammals of the Mexican state of Yucatan. Unpublished Ph.D. dissertation, Texas Tech Univ., Lubbock, 396 pp. Jones, T. L. 1971. Vegetational patterns in the Guadelupe Mountains, Texas. Amer. J. Boh, 76:266-278. - . 1975. An introduction to the study of plants. John Wiley and Sons, New York, xx+386 pp. Jones, T. L., A. L. Bain, and E. C. Burns. 1976. Grasses of Texas. Pp. 205-265, in Native grasses of North America (R. R. Dunn, ed.), Univ. Texas Studies, 205:xx+l- 630. Smith, J. D. 1973. Geographic variation in the Seminole bat, Lasiurus seminolus. J. Mamm., 54:25-38. Smith, J. D., and G. L. Davis. 1985. Bats of the Yucatan Peninsula. Occas. Papers Mus., Texas Tech Univ., 97:1-36. Consecutively-paged journal volumes and other serials should be cited only by volume number and pagination. Serials with more than one number and that are not consecutively paged should be cited by number as well (Smiths. Misc. Coll, 37(3): 1-30). Illustrations are acceptable only as original inked line drawings or photographic prints. They normally should be no larger than 4.5 x 6.5 inches and mounted on 8.5 x 11 paper or backing. Each figure should be marked on the back with the name of the author(s) and figure number. If confusion might result as to arrangement of a figure, label “top.” All legends for figures must be typed (double-spaced) on a separate sheet(s) of paper. All figures must be referred to in text— as “Figure 3” or “(Fig. 3).” All tables are to be typed, double-spaced, and headed by the legend, on a single page(s) for each table. All should be cited at the appropriate place in text as “Table 1” or “(Table 1).” 296 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 3, 1987 Some important specific points for authors: (1) do not break words at the right-hand margin of text; (2) footnotes are to be avoided except as absolutely needed in tables; (3) scales for illustrations should be on the figure, not in the legend, to avoid errors when illustrations are reduced for publication; (4) be sure all lettering or other symbols on illustrations will be clearly evident after reduction of them to Journal page size; (5) the editor should be notified immediately of any change in address of the responsible author, whose telephone number also should appear on correspondence; (6) in order to make papers more readable for the general scientific public, abbreviations are to be avoided in text except for standard mathematical or chemical formulae (where an abbreviation might be used many times to save space, write out the full term the first time used and give the abbreviation, which can be used thereafter, in parentheses); (7) except where fractions are used, write out numbers through nine in text and use numerals for 10 and beyond; (8) consult recent issues of the Journal , beginning with 38(1), for all matters of style. The principal author will receive galley proofs along with edited typescript and a reprint order form. Proofs must be corrected and returned to the editor within five days; failure to return proof promptly will result in delay of publication. Reprint order forms should be returned directly to Texas Tech Press (P.O. Box 4240, Lubbock 79409), not to the editor. Charges of $35 per printed page (or part thereof), or partial payment, strongly are encouraged by members of the Texas Academy of Sciences when grant or institutional funds are available for that purpose. Some contribution, even if modest, is expected for any paper that exceeds 10 printed pages. Nonmembers of the Academy are required to cover all page costs except as rarely excepted by the Treasurer. Authors are provided with page-charge information when their manuscript is accepted for publication. General Notes. — Beginning with volume 39 of the Journal, a section for noteworthy but short contributions may appear at the end of each number. Manuscripts published as General Notes normally will not exceed four or five typed pages. The format is the same as for feature articles except no abstract is included and the only subheading in text is a centered Literature Cited (if needed) unless italicized paragraph subheadings are absolutely essential, as in the case, for example, of more than one account for individual species of plants or aimals. While the decision as to whether a manuscript is best suited for a feature article or a note will be made by the editoral staff, authors are encouraged to indicate their preference at the time a manuscript is submitted to the Editor. THE TEXAS ACADEMY OF SCIENCE, 1986-87 Officers President: President-Elect: Vice-President: Immediate Past President: Executive Secretary: Treasurer: Editor: A AS Council Representative: Directors 1985 George B. McClung, San Angelo Barbara Schreur, Texas A&I University 1986 Caroline P. Benjamin, Southwest Texas State University R. John Prevost, Southwest Research Institute 1987 John P. Fackler, Jr., Texas A&M University David R. Gattis, Freese and Nichols, Inc. Sectional Chairpersons I — Mathematical Sciences : Barbara Schreur, Texas A&I University II — Physical Sciences : C. A. Quarles, Texas Christian University III— Earth Sciences: James L. Carter, University of Texas at Dallas IV — Biological Sciences: Robert S. Baldridge, Baylor University V — Social Sciences: David A. Edwards, San Antonio College VI — Environmental Sciences: Bennett J. Luckens, Austin VII — Chemistry: Rodney Cate, Midwestern State University VIII — Science Education: Dick E. Hammond, Southwest Texas State University IX — Computer Sciences: Ronald King, Baylor University X — Aquatic Sciences: Paula Dehn, University of Texas at San Antonio Counselors Collegiate Academy: Shirley Handler, East Texas Baptist College Helen Oujesky, University of Texas, San Antonio Junior Academy: Ruth Spear, San Marcos Peggy Carnahan, San Antonio Lamar Johanson, Tarleton State University Owen T. Lind, Baylor University Glenn Longley, Southwest Texas State University Billy J. Franklin, Lamar University William J. Clark, Texas A&M University Michael J. Carlo, Angelo State University J. Knox Jones, Jr., Texas Tech University Ann Benham, University of Texas at Arlington 2nd CLASS POSTAGE PAID AT LUBBOCK TEXAS 79401 THE TEXAS JOURNAL OF SCIENCE USPS 616740 Box 4240, Texas Tech University Lubbock, Texas 79409-3151, U.S.A. BOUND PRINTED MATERIAL RETURN POSTAGE GUARANTEED THE TEXAS JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP.— Any person or group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. Dues for members are $20.00 annually; student members, $12.00 annually; sustaining members, at least $30.00 in addition to annual dues; life members, at least $400.00 in one payment; patrons, at least $500.00 in one payment; corporate members, $250.00 annually; corporate life members, $2000.00 in one payment. Library subscription rate is $45.00 annually. Payments should be sent to Dr. Michael J. Carlo, P.O. Box 10986, Angelo State University, San Angelo, Texas 76909. The Texas Journal of Science is a quarterly publication of the Academy and is sent to all members and subscribers. Changes of address and inquiries regarding back issues should be sent to Dr. William J. Clark, Drawer H6, College Station, Texas 77844. AFFILIATED ORGANIZATIONS Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers American Association for the Advancement of Science Texas Society of Mammalogists The Texas Journal of Science (USPS 616740) is published quarterly at Lubbock, Texas U.S.A. Second class postage paid at Post Office, Lubbock, TX 79401. Please send form 3579 and returned copies to Texas Tech Press, Box 4240, Lubbock, TX 79409. ISSN 0040-4403 THE TEXAS JOURNAL OF SCIENCE Volume 39, No. 4 November 1987 CONTENTS Long-period surface-wave phase-velocity partial derivatives with respect to earth parameters. By Doo Jung Jin . . . 299 Distribution and habitat of a new major clone of a parthenogenetic whiptail lizard (genus Cnemidophorus ) in Texas and Mexico. By James M. Walker . 313 The periphytic Cladocera of ponds of Brazos County, Texas. By John M. Campbell and William J. Clark . 335 Comparative in vitro digestive efficiency of cattle, goats, nilgai antelope, and white-tailed deer. By James C. Priebe, Robert D. Brown, and Doreen Swakon . 341 Shallow-water Octocorallia and related submarine lithification, San Andres Island, Colombia. By M. John Kocurko . . .349 The response of woody vegetation to a topographic gradient in eastern Texas. By E. S. Nixon, J. Matos, and R. S. Hansen . 367 General Notes Phaenicia (Diptera: Calliphoridae) myiasis in a three-toed box turtle, Terrapene Carolina triunguis (Reptilia: Emydidae), from Arkansas. By Chris T. McAllister . 377 Clapper rail ( Rallus longirostris) in west-central Texas. By D. Brent Burt, Donna Burt, Terry C. Maxwell, and Delbert G. Tarter . 378 Distributional records for six Texas mammals. By Brian R. Chapman and Sandra G. Spencer . 379 Distibutional notes on four species of Texas mammals. By Jim R. Goetze and Larry L. Choate . 380 Index (including list of authors and reviewers) 383 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Editor: J. Knox Jones, Jr., Texas Tech University Assistant to the Editor: Marijane R. Davis, Texas Tech University Associate Editor for Botany: Chester M. Rowell, Marfa, Texas Associate Editor for Chemistry: Marvin W. Rowe, Texas A&M University Associate Editor for Computer Science: Ronald K. Chesser, Texas Tech University Associate Editor for Mathematics and Statistics: Patrick L. Odell, University of Texas at Dallas Associate Editor for Physics: Charles W. Myles, Texas Tech University Editorial Assistants: Robert R. Hollander, Texas Tech University Richard W. Manning, Texas Tech University Scholarly papers in any field of science, technology, or science education will be considered for publication in The Texas Journal of Science. Instructions to authors are published one or more times each year in the Journal on a space-available basis, and also are available from the Editor (The Museum, Box 4499, Texas Tech University, Lubbock, Texas 79409, 806/742-2487, Tex-an 862-2487). LONG-PERIOD SURFACE- WAVE PHASE-VELOCITY PARTIAL DERIVATIVES WITH RESPECT TO EARTH PARAMETERS Doo Jung Jin cjo Commander, EUSA-TNT (J2), APO, San Francisco, California 96301-0092 Abstract. — Rodi et al. (1975) have developed a fast and accurate method for computing group-velocity partial derivatives with respect to earth parameters when corresponding phase-velocity partial derivatives are known. Novotny (1970) derived an analytical method for phase-velocity partials, but he limited his algorithm to Love waves, which are much simpler than Rayleigh waves. Employing the same approach as Novotny’s, Jin and Herrin (1980) formulated a new exact method for the computation of Rayleigh-wave phase-velocity partial derivatives with respect to earth parameters. Since the publication of the last paper, many inquiries about the details of the method have been received. This paper attempts to provide a set of convenient formulae to be used in the method. Key words-, partial derivatives; phase velocity; surface waves; dispersion. In the inversion of surface-wave dispersion observations to determine the structure of the earth, it is necessary to compute the partial derivatives of phase and/or group velocity with respect to various earth parameters. Because a fast and accurate method has been developed for computing group-velocity partial derivatives when corresponding phase- velocity partial derivatives are known (Rodi et al., 1975), there was a demand for a simple new method for the computation of exact phase- velocity partial derivatives. I have formulated a new convenient analytical method for this purpose and applied it successfully to an inversion problem (Jin and Herrin, 1980). Although the algorithm was described in this cited reference, detailed formulae were not included due to limited space. Since the publication of the referred paper, many inquiries about the details of the method have been received. It seems beneficial to the scientific community to publish the details of the method. The present paper aims to satisfy this need. In the earth model in which the earth consists of many elastic, homogeneous and isotropic parallel layers on a homogeneous and isotropic half-space, earth properties may be defined by longitudinal and transverse wave velocities, density, and thickness of the component layers. Various methods have been proposed in computing the phase- velocity partial derivatives with respect to the above model parameters. Dorman and Ewing (1962), followed by Brune and Dorman (1963), calculated the changes in phase velocity due to perturbation of each physical parameter, while retaining the remaining parameters constant. This method is time-consuming. On the basis of Jeffrey’s (1961) The Texas Journal of Science, Vol. 39, No. 4, November, 1987 300 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 theoretical suggestion that Rayleigh’s principle can be used to find expressions for the effects of small changes of the elastic properties on the phase velocity, Anderson (1964), Takeuchi et al. (1964), Harkrider (1968), and Anderson and Harkrider (1968) approached the problem by using the energy integral technique. McEvilly (1964) has used a combination of the above two methods. Bloch et al. (1969), Der et al. (1970), Der and Landisman (1972), and Knopoff (1972) computed phase- velocity partial derivatives but they have not described the algorithm they employed. Novotny (1970) derived exact expressions for phase-velocity partials in a different way by taking advantage of Thomson- Haskell matrices (Thomson, 1950; Haskell, 1953). But Novotny’s paper limited itself to the derivation of partials for Love waves, which are much simpler than Rayleigh waves. To provide convenience to the reader and to make this paper complete, the new method formulated for Rayleigh waves will first be presented, and then formulae needed in this method will be given. Employing double precision in computer coding, the use of the Haskell layer matrix method has not shown any numerical difficulties (Thrower, 1965; Dunkin, 1965; Gilbert and Backus, 1966; among others) on the commonly investigated long-period band of 10 to 100 seconds. Method Harkrider (1964) derived a matrix equation for the dispersion of Rayleigh waves, which may be expressed as follows: F(c, a >, a*,, 0m, pm, dm) = NK + L*M* - T*(G*N - L*H) (1) where c = c((o, am, (3m, pm, dm ), T* = c po tan Po/ra0, and L = y„ ran An + (yn - 1) A2i - (ra„ A3i - A4i)/(c2p„), K = yn ran An + (yn ~ 1) A22 - (ra„ A32 - A 42)/(c2p„), G yn raw Ai3 T ( yn 1) A23 (raM A33 A43)/(c p«), (2) N = — (yn — 1) An + y„ rpn A2i + (A3i + rpn A4i)/(c2p„), M = — (?„— 1) Ai2 + yn i 'pn A22 + (A32 + rpn A 42)/(c2p„), H = — (yn — 1) Ai3 + yn rPn A23 + (A33 + Tpn A 43)/(c2p„). Here a» is the angular frequency, c phase velocity, a and (3 velocities of compressional and shear waves, p density, d thickness of a layer; y„ = 2 WJc)\ Pm k dm r am •> PARTIAL DERIVATIVES— EARTH PARAMETERS 301 where k represents the wave number: r (c2/a m — l)l/2 c > am T°m= i-iO-c’/aW, c f3m Ifm~ l - i( 1 - C2IP2)V\ c < ft.; where i = (— 1)'/2: A fl/j-l * fli with designating the Haskell’s layer matrix for the rath layer; and subscripts ra, n and 0 refer to the rath solid layer, half-space and liquid surface layer, respectively. In order to minimize typographical requirements, a compound subscript, am, is used where a subscripted subscript is appropriate. An asterisk superscript is used to transform an imaginary quantity into a real one such that Z* = Z/i for an arbitrary imaginary quantity Z. Because T* = 0 for the earth model in which the liquid surface layer does not exist, considering (1) is sufficient for earth models both with and without ocean. From (1), dc = _ <9F / <9F dx dx I dc where x designates one of the earth parameters, a, /?, p and d. If we define two product matrices Am and B,„ such that A rn dm dm- 1 * ’ * d\ , Bw dn-\ dn-2 * ’ ’ dm. then A Bw+i • Qm • A„-i, <9Ay dxm dAl} dc ^Bw+i • r dan~ ! L dc • Am il , ra I 2, 3, •••,/! — 2, dxm J'J n — 2 A„-2 + S Bw+i • — • Aw-i + B2 ra = 2 dc da, "I dc Jij (3) For the particular cases where ra = 1 and ra = n — 1, respectively, the second equation of (3) becomes dAjj dxi dd\ I dxi and 302 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 It is noted from (1) and (2) that in order to compute dF we neecj <9Ajj , and that in * dxm dxm order to compute ^F we need dAli , EL, fogg., dtpn and . It is also noted that in dc dc dc dc dc dc order to compute _EE and EL , in particular, we need dxom_ , E^£l, dyn , EL, 3T dxn dxo dan d(3n djdn dao dpo and dT Inasmuch as dra n , drpn dyn and terms necessary to compute dF an(j <9F ddo dc dc dc dxn arC simple to calculate, they are not discussed here. From the above discussion, it becomes clear that in order to compute phase-velocity partial derivatives with respect to model parameters, we need only to compute d{amh d(am) jj dxm dc , and EL. The formulae needed to compute these terms are presented in the following dc section. Formulae Expressions necessary to compute the phase-velocity partial derivatives are given below. These expressions should give users of the present method a convenience and saving of time required to carry through the tedious calculations. It should be noted that in the case of Rayleigh waves there are three situations in connection with the relation between the phase velocity and body-wave velocities. They are: (i) c> am and c > /3m, that is, both ram and rpm are positive real; (ii) c < am and c > ($m, that is, Tam is negative pure imaginary while rpm is positive real; (iii) c < am and c < /3m, that is, both ram and rpm are negative pure imaginary. In the following expressions a triple sign, that is, three signs put together vertically, applies, in the order from top to bottom, to the above three cases. When the three cases have a common sign, only a single sign appears. It is also understood that whenever ram and r^m are purely imaginary, xam and xpm in the following expressions actually represent r *am and x* pm, respectively; and trigonometric functions become corresponding hyperbolic functions. Because the expressions for the elements of the layer matrix have been published previously — see Haskell (1953: 21, 30) and Harkrider (1964: eqs. 16, 25) — the expressions for only their partial derivatives are presented here. In the following expressions, Qm stands for kdm xpm\ d(am) ii _ ~ c oo dm Jm sin Pm. darn T- tt/fi ram d{Om)u__ 4 Pm (cos pm _ cos Qm) _ c CP dm (7m~l) Sin Qm . dfim C fim rpm d(am) 11 _ 0 ; dpm PARTIAL DERIVATIVES— EARTH PARAMETERS 303 dCflm)1,,1- — kf+7 m ram sin Pm X (7m— 1) rpm sin Qml ; ddm L+ — J d{am) 11 _ _ J_ r27m (cos Pm - cos Qm) + - - - dc C V am r jgm {7 m Qm sill Pm (7m 1) Pm sin Qm}] j d(am) 12_ _ 1 c (7m~ l)(pm cos pm _ sjn pm) . <9«m 4* 0Lm Tam dJ^hl = 2 i [ 2^, / sin_PV, + sin Qmx _ _ 1_ d fim C V, am fim ? fim (sin Qm + Qm COS Qm)] ; d{am)\2 — q . dpm ^uhl = ik [(7„-|) COS Pm + 7m COS Qm] ; ddm d{am) 12 _ / j~ 7m~ (9c L ^2 2 ^ c r«m (C — am) (am Pm COS Pm + c am x mj + 7m (k dm cos Qm- 2 rpm sin Qm + c sin.9.m - 2_smJPm_\ -I “l- (dm r (3m r am d(am) 13 _ _ k dm sin Pw . dam Pm am r am d(am)n k dm sin Qm . dpm Pm (dm rpm d(Gm) 13 1 (cos Pj d Pm 2 2 C Pm d(dm) 13 k ( — r am ddm C Pm + 304 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 d(am) 13 _ _ [ dc ~ ' C Pm . j^2 (COS P m COS Qm) 4~ f am f j8 m P m sin Qm)J \ <9(a'")l4 = + - — (Pm COS Pm - sin Pm) ; doim + Pm OLm f am l9(a'")'4 = - ' _ (sin Qm + Qm cos Qm) ; dp m Pm fim f (3m d(Qm)\4 dpm Pm (3m f /3m /sinPm+ rfjm sin Q \ . C Pm f am — i - - (COS Pm + r2pm COS Qm) ; ddm C pm — S n Pm sin Qm\ ^ P m COS Pm f am 4“ fam d(am)\4 _ _ i r 2 ( ' dc C! Pm ^ f2 sil^ p” ) - g2 p Q” - k dm COS Qm] ; a2m Pm rpm J = i 2 ffm (sin pm + pm cos pm) . ddm dm fam dJ^hl = _ i f Aik (sin Q"»± ram sin P-,) j c' (7m _1) dPm C rpm Pm Pflm (sin Qm Qm COS Qm)J i d(flm) 21 _ Q . dpm d(am)2x — _ /k[(7m-l) COS Qm i 7m fam COS Pm] i ddm ~ d(am)2\__ _ i ryJ J_(k dm COS Prn + 2 r„m sin Pm) + C P™| dc L C + OLm fam 2 7m sin Q, 7m’ Qm cos Q, f3(3m c sin Qm \l . —J2 - )\ ’ pm C ffim C PARTIAL DERIVATIVES— EARTH PARAMETERS 305 d{am) 22 dam c k dm (7m— 1) sin P m am V am djOmhl _ 4 Pm (cos Qm _ cos pm) + 2 k dm Sin Qm dfim C fin: V pm d(am) n ___ q . d Dm d(cim)2,2 ~ __ J^r — {ym 1) yam sin Pm — ym y (im sin Qm"j Jr!.. L_ l J ddm d{am)i2 _ 1 — - — I 2 ym (COS Pm COS Qm) -| - dc C 1 t'arri y$m ^(7m 1) Qm sin Pm ym Pm sin Qm jj j ^ d(am) 23 __ dam Pm GLm /" am (sin Pm “t- Pm COS Pm) \ d(am) 23 _ _ i dfirn + Pm Pm ypm (Qm cos Qm sin Qm) ? d(am) 23 z ^ sin P™ sin Qm ^ . 5pm C2 2 V Pm + ypm 5(flm)23 __ ; ■s 1 + am COS P m "k COS Qm^ j ddm C Pm d(am)i3 * [2( sin Qm Tam Sin Pm^ “ dc ■ y&m c sin Pm ~ 1 (Qm COS Qm c2 sin Qm 01 m r am + rjm - a 2 /j Pm d((lm) 31 _ C Pm k drn ym (7m 1) sin Pm da, am y am COS Pm - 306 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 = 2 pm[2 (2 7m— 1) (COS Pm ~ COS Qm) dpm 1 _ C (7m 1) k dm sin Qw~| . Pm rpm J d^llL =2 Pm (7m— 1) (cos Pm “ COS Qm) ; dpm (9(flm) 31 —2k pm Pm {ym 1) (-|- T am sin Pw _|_ Y sin Qm^ j ddm + — d{(lm) 31 __ £ pw ym |"2 7 m (COS Qm COS Pm) "k — ^ - (Pm sin Qm (9 c V am Yfim Qm sin Pm)] j d(a">)n - 1 ,• C4pm(7m-1)2 (Sin pm - pm cos pm) ; (9am OJm fam d^am^2 — i 4 Pm Pm I" 2[ (7m~ 1) Sin Pm + ^ ^ sin Qw| - _J_ d Pm y am — f(im (sin Qm “1“ Qm COS Qm)] j 3(a'“b± = ic2 r (7m- 1)2 Sin Pm + ^ r/jm sin Q1 . dpm y am = i k C Pm [(7m— l)2 cos Pm + 7 m ^m COS Qm] J ddm — d{am) 32 dc c2 sin P 2 Oim — — i pm [ — — [2 (7m— 1) sin Pm + am £ c (7m-l): yam (p' COS Prn — )} + i#L ( 7m /-0m sin Qm - Sln Qm) ,y + - V + rpm ’ to dm ym COS Qm] <9(am)41 = - j iPmJm (Sin Pm + pm cos pm) . <9«m am r am PARTIAL DERIVATIVES— EARTH PARAMETERS 307 = i c2pm f (7m 1) sin_Qm_ + ym ^ sin p j dj3m C T prn + c (y.n fL-(sin Qm - Q m cos Qm)] ; - pi r\m J =ic2 r (7m-l)2 sin Qm + y2m ram sin p i . d Pm y Pm d{dm) 4j_ — / k C Pm 1) COS Qm — Tm ^am COS PmJ ddm ~ ^gm^41 = I Pm [oj dm 7m COS Pm -f - (7m r„m sin Pm + ffm sin Pm \ <9c L + C v + am ram _1_(2 (7m— 1) Sin Qm -(7m, 0...(Qm cos Qm - ^ si" Q,>l )|1 rum1 + r20m V pi ni {dm) 24 — (^m)l3j {dm) 33 — (^m)22j (#m)34 — {dm) 12j (<3m)42 — (#m)3lj (flm)43 — (fifm)2lj (#m)44 — {dm) 11 ^ ^ - PI \w do Oil sec2 P0 + c (c2 - 2 ag) tan P0 j dc c — (xl L ya o J Although Novotny (1970) presented the algorithm for Love waves, his paper did not provide formulae for the partial derivatives. To make this paper more useful and complete, the expressions for the partial derivatives of the elements of the Haskell layer matrix for Love waves are given below. For Love waves, the matrix equation of dispersion is: F(c, CO, Pm, Pm, dm) A21 T Pnfin y/Sn A11, where C C (co, Pm, Pm, dm). In the case of Love waves, we only have to consider two situations in connection with the relation between the phase velocity and shear wave velocity: (i) c > pm and (ii) c < pm. A double sign, that is, two signs put together vertically applies to the above two situations. 308 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 PARTIAL DERIVATIVES -0.12 -0.08 -0.04 C.00 0.0* 008 0.12 0.16 Figure 1. — Group velocity partial derivatives for 25-second period in the Bering Shelf. PARTIAL DERIVATIVES Figure 2. — Group velocity partial derivatives for 40-second period in the Bering Shelf. DEPTH (KM) PARTIAL DERIVATIVES— EARTH PARAMETERS 309 PARTIAL DERIVATIVES -0.06 -0.04 0.00 0.04 0.08 0.12 0.16 Figure 3. — Group velocity partial derivatives for 16-second period in the Aleutian Basin. PARTIAL DERIVATIVES -0.08 -0.04 0.0C 0.04 0.08 0.12 C.I6 0.2C Figure 4.— Group velocity partial derivatives for 50-second period in the Aleutian Basin. 310 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 3(am) ii dfim 3(am) n 3pm 3dm d(am) n 3c 3(am)n 3pm 3{am) 12 dpm d{am) 12 3dm 3{am) 12 (9c d(flm)21 3pm 3(am) 21 3 Pm 3(am) 2i 3dm 3(am) 2i 3c (am)22 = c w dm sin Qm Pm r (im = sin Qw ; = - sin Qm r — L Pm rpm — - - — — — [c Q/n cos Qm “I- sin Qm (c 2/9w)J ■ Pm /? m rpm / sin Qm — ~ 4 - 2 - 2 - ’ pm P m rpm _ / k cos Qm . Pm Pm — - L— - _ - r+C sin Qm — Pm Qm COS Qm"l j c Pm Pm rpm L J _ j — Pm — |"(c 2p m) sin Qm c Qm cos Qm”! *, Pm rpmL J = — i P m rpm sin Qm ■ — — 1 k Pm Pm rpm COS Qm j _ j — Pm — ^ sin Qm “t- Pm Qm COS Qm) j C A* pm (am)n. PARTIAL DERIVATIVES— EARTH PARAMETERS 311 Examples Rayleigh-wave phase-velocity partial derivatives were computed by using the algorithm described in the section on method and the formulae in the previous section. From these phase-velocity partial derivatives, group-velocity partial derivatives were obtained by applying the method of Rodi et al. (1975) to dispersion data from the Bearing Shelf and the Aleutian Basin of the Bering Sea. Those results obtained for periods of 25 and 40 seconds in the Bering Shelf and for periods of 16 and 50 seconds in the Aleutian Basin are shown in Figures 1, 2, 3, and 4, respectively. Those readers further interested in data acquisition and earth models are referred to Jin and Herrin (1980). Literature Cited Anderson, D. L. 1964. Universal dispersion tables I. Love waves across oceans and continents on a spherical earth. Bull. Seism. Soc. Amer., 54:681-726. Anderson, D. L., and D. G. Harkrider. 1968. Universal dispersion tables II. Variational parameters for amplitudes, phase velocity and group velocity for first four Love modes for an oceanic and a continental earth model. Bull. Seism. Soc. Amer., 58:1407-1499. Bloch, S., A. L. Hales, and M. Landisman. 1969. Velocities in the crust and upper mantle of southern Africa from multi-mode surface wave dispersion. Bull. Seism. Soc. Amer., 59:1599-1629. Brune, J., and J. Dorman. 1963. Seismic waves and earth structure in the Canadian Shield. Bull. Seism. Soc. Amer., 53:167-210. Der, Z. A., and M. Landisman. 1972. Theory for errors, resolution, and separation of unknown variables in inverse problems, with application to the mantle and the crust in southern Africa and Scandinavia. Geophys. J., 27:137-178. Der, Z., R. Masse, and M. Landisman. 1970. Effects of observational errors on the resolution of surface waves at intermediate distances. J. Geophys. Res., 75:3399-3409. Dorman, J., and M. Ewing. 1962. Numerical inversion of seismic surface wave dispersion data and crust-mantle structure in the New York-Pennsylvania area. J. Geophys. Res., 67:5227-5241. Dunkin, J. W. 1965. Computation of modal solutions in layered, elastic media at high frequencies. Bull. Seism. Soc. Amer., 55:335-358. Gilbert, F., and G. E. Backus. 1966. Propagator matricaes in elastic wave and vibration problems. Geophysics, 31:326-332. Harkrider, D. G. 1964. Surface waves in multilayered elastic media, I. Rayleigh and Love waves from buried sources in a multilayered elastic half-space. Bull. Seism. Soc. Amer., 54:627-679. - . 1968. The perturbation of Love wave spectra. Bull. Seism. Soc. Amer., 58:861- 880. Haskell, N. A. 1953. The dispersion of surface waves on multilayered media. Bull. Seism. Soc. Amer., 43:17-34. Jeffreys, H. 1961. Small corrections in the theory of surface waves. Geophys. J., 6:115- 117. Jin, D. J., and E. Herrin. 1980. Surface wave studies of the Bering Sea and Alaska area. Bull. Seism. Soc. Amer., 70:2117-2144, Knopoff, L. 1972. Observation and inversion of surface-wave dispersion. Tectonophysics, 13:497-519. 312 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 McEvilly, T. V. 1964. Central U.S. crust-upper mantle structure from Love and Rayleigh wave phase velocity inversion. Bull. Seism. Soc. Amer., 54:1997-2015. Novotny, O. 1970. Partial derivatives of dispersion curves of Love waves in a layered medium. Studia Geophys. Geodaet., 14:36-50. Rodi, W. L., P. Glover, T. M. C. Li, and S. S. Alexander. 1975. A fast, accurate method for computing group-velocity partial derivatives for Rayleigh and Love waves. Bull. Seism. Soc. Amer., 65:1105-1114. Takeuchi, H., J. Dorman, and M. Saito. 1964. Partial derivatives of surface wave phase velocity with respect to parameter changes within the earth. J. Geophys. Res., 69:3429- 3441. Thomson, W. T. 1950. Transmission of elastic waves through a stratified solid medium. J. Appl. Phys., 21:89-93. Thrower, E. N. 1965. The computation of the dispersion of elastic waves in layered media. J. Sound Vib., 2:210-226. DISTRIBUTION AND HABITAT OF A NEW MAJOR CLONE OF A PARTHENOGENETIC WHIPTAIL LIZARD (GENUS CNEMIDOPHORUS) IN TEXAS AND MEXICO James M. Walker Department of Zoology, University of Arkansas, Fayetteville, Arkansas 72701 Abstract. — This report lists five localities in Val Verde, Maverick, and Webb counties, Texas, and seven in Coahuila, Mexico, in the upper Rio Grande Valley for a new major clone (LAR-B) of a parthenogenetic whiptail lizard similar to Cnemidophorus laredoensis. Electrophoretic analyses indicate that all populations tested for selected gene loci are genetically identical to all-female populations discovered at nine localities in Starr, Hidalgo, and Cameron counties, Texas, and five in Tamaulipas, Mexico, in the lower Rio Grande Valley. The paucity of locality records for LAR-B in parts of Val Verde, Kinney, Maverick, and Webb counties, and adjacent parts of Coahuila in the upper Rio Grande Valley and Cameron County and adjacent Tamaulipas in the lower Rio Grande Valley is attributable to the inaccessibility of many areas near the river. The entire range of clone LAR-B is contained within a small part of the range of the gonochoristic species Cnemidophorus gularis, one of the apparent progenitors of the clone. All locality records for LAR-B were found within a narrow zone of less than 10 kilometers in width on both sides of the Rio Grande between Del Rio, Val Verde County, and Palmito Hill, Cameron County (within about 16 kilometers of the Gulf of Mexico). The major habitats occupied by LAR-B in the upper Rio Grande Valley include bunchgrass-weed associations, ecotones between desert scrub and bunchgrass-mesquite, and bunchgrass/weed-mesquite, and in the lower Rio Grande Valley tall or short bunchgrass-weed mixtures, ecotones between thorn shrub and cultivated land, sandy banks along the Rio Grande, and ecotones between mixed forest and bunchgrass/weed-mesquite habitats. Parthenogenesis, in which each individual is a potential founder of a new colony, provides an adaptive reproductive mechanism for the survivor- pioneer responses that are more effective than in gonochoristic populations in those habitats that are subject to major, often catastrophic, changes. Key words : Cnemidophorus ; parthenogenesis; clones; distribution; ecology; lizards of Texas and Mexico. Between 1983 and 1987, an investigation of the distribution and habitats of parthenogenetic whiptail lizards (genus Cnemidophorus) was conducted in the Rio Grande Valley and surrounding areas of Texas and Mexico between Langtry, Val Verde County, and the Gulf of Mexico. The form (LAR-A) described as Cnemidophorus laredoensis by McKinney et al. (1973) was reported from 23 additional localities in Dimmit, LaSalle, Webb, Zapata, Starr, and Hidalgo counties, Texas, and 11 in Tamaulipas, Mexico, by Walker (1987a). Also included were observations on habitats occupied by LAR-A, and on the ecological relationship of the parthenoform to its maternal progenitor Cnemidopho¬ rus gularis. In August 1984, a different parthenogenetic Cnemidophorus (LAR-B) was discovered near the Rio Grande in Bentsen-Rio Grande Valley State Park, Hidalgo County, Texas. Clone LAR-B subsequently was found at The Texas Journal of Science, Vol. 39, No. 4, November, 1987 314 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 localities in Starr, Hidalgo, and Cameron counties, and in Tamaulipas, including some sites where it is syntopic with LAR-A, C. gularis , or both. In August and September 1984, populations of parthenogenetic Cnemidophorus also were discovered near the Rio Grande in Val Verde County, Maverick County, and Coahuila, Mexico, and in northern Webb County in May 1987. Electrophoretic analyses indicate that all populations tested at selected gene loci are genetically identical to LAR-B in the lower Rio Grande Valley (E. D. Parker, Jr., J. M. Walker, E. M. Niklasson, S. E. Trauth, L. J. Lester, and J. E. Cordes, unpublished data). The purpose of this paper is to present data on geographic distribution, habitats, and syntopic relationships of this newly discovered clone of Cnemidophorus from many areas of Texas and Mexico where all-female populations have not been reported previously. Methods Field observations in the Rio Grande Valley were made during 12 field expeditions from 1983 through 1987. Data on activity periods of lizards, soil characteristics, dominant plant forms, and general habitat quality were recorded at each locality. The term “catastrophic alteration” is used in reference to major destructive changes by human intervention (for example, fire, earth-grading, cultivation). The following geographic designations are used: upper Rio Grande Valley (within 10 kilometers of the Rio Grande in Val Verde, Kinney, Maverick, and northern Webb counties, Texas, and from Lake Amistad, Coahuila, to the Nuevo Leon border); middle Rio Grande Valley (near the Rio Grande in Webb, Zapata, and Starr counties, Tamaulipas from about 50 kilometers north of Laredo-Nuevo Laredo to the vicinity of Rio Grande City, and away from the river in Dimmit, LaSalle, and Starr counties); and lower Rio Grande Valley (within 10 kilometers of the Rio Grande in Starr, Hidalgo, and Cameron counties, and Tamaulipas from the vicinity of Rio Grande City to near the Gulf of Mexico). Specimens are deposited in the University of Kansas Museum of Natural History (KU), and University of Arkansas Department of Zoology (UADZ). The coded localities and specimens of Cnemidophorus from each are listed in Results and Discussion. Identification The formal binomial Cnemidophorus laredoensis will not be further used in this report. Rather, I employ the system of informal designations of cloned hybrid populations of Cnemidophorus proposed by Walker (1986). Apomictic Cnemidophorus populations of hybrid origin are not biological species by objective definition. The level of morphological and color and pattern differentiation between clones LAR-B and LAR-A, however, exceeds that between many other all-female populations that have been described as “species” (McCoy and Maslin, 1962; Lowe and Wright, 1964; Wright and Lowe, 1965; Wright, 1967; Scudday, 1973). LAR-B, LAR-A, and C. gularis, which are superficially similar, are distinguishable by distinctive features of the dorsal pattern, ventral body and tail coloration, and postantebrachial scales (Fig. 1; Table 1). Furthermore, LAR-B and LAR-A are interclonally differentiated in granules separating the paravertebral stripes, percent of the granules around midbody lying between the paravertebral stripes, numbers of granules around midbody, granules from occiput to rump, and other characters (Table 2). CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 315 Figure 1. Representative specimens of LAR-B (99, 80 mm), LAR-A (99, 77 mm), and female Cnemidophorus gularis (9(5, 75 mm) from La Grulla (site 1), Starr County (S- 8*+). Results and Discussion More than 100 sites have been searched for the presence of LAR clones. Collection records are presented here in geographical sequence from north-northwest to south-southeast by county in Texas and by state in Mexico. The codes for each locality are consistent with those introduced by Walker (1987a); they are subsequently used in text, figures, and tables. The three whiptail forms are identified in the citations as follows: clone A (22, LAR-A), clone B (22, LAR-B), and C. gularis (25). Whiptail association codes are listed in Table 3. All collection records for all localities considered relevant to interpretation of the distribution and ecology of LAR-B follow (whether the clone was present or not). Characteristics of LAR-B sites are summarized in Table 4. TEXAS (upper Rio Grande Valley). Val Verde County: (V-l) Langtry, site 1 overlooking Rio Grande (9(5, KU 200126-200137); (V-2) Langtry, site 2 in town (9(5, UADZ 1541-1562); (V-3) Comstock on U.S. Hwy. 90 (9$, UADZ 1563-1564); (V-4) 15 km. NW Del Rio, then 2.2 km. S U.S. Hwy. 90 (9(5, UADZ 1143-1149); (V-5) Del Rio, W suburbs (9<5, UADZ 1137-1138); (V-6) Del Rio, city park on San Felipe Creek (9(5, KU 200104, UADZ 1574-1576); (V-7) Del Rio, vicinity of city industrial park (9(5, UADZ 1140- 1142); (V-8) Del Rio, industrial park behind cemetery (9<5, UADZ 1139); (V-9) Del Rio, 4.8 km. W international bridge (9<5, UADZ 1135-1136); (V-10++) Del Rio, 0.8 km. N international bridge on west side of road (99, LAR-B UADZ 1001-1002; 9(5, KU 200113- 316 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Of Of T3 . g O ed o r-v 3 O 3 OfS PQ c * * < 3 £ " o ed \2 Dh © ed C H d cd 60 Oh § >> © 4) X «3 o " •rj © g s s e 5 © © « Of © — ' 5 5 o 3 bo &0 Z, SO .5 3 43 £ vj O It o cd - X -O . ed — f^-| O GO “* So a* cd o w - — *J © 3 -5 c -3 g g> 3 cd «_ > b •!- 3 i T3 > >>• 3 ^ OS 6S •*«* ^ 3 « M Oh 00 GO 3b ! ^ .60 © O- dj % 5 '3 £ 3 e i 1 1 ! H H < 13 "2 „5 o< ^ ^3 t.s g £ ^ CD e* Tj o g a T3 © GO 3 3 © cd _ _ Vi C g GO 60 -?3 ]3 h -S -£ S £ ! 1 3 2 >,>>2 1* _, _, ts s .3 .3 „ o > > c © > > O © -g © 3 3 3 £ * £ >-. 60 >> u b 'u ed 3 cd 4) > 3 13 © 3 ed 3 3 g & 3> 3 g c e 43 3 2 " >, ' o >>o ”S h 3 3 3 <33 § H & § JJ w 3 ^ 60 H © © P 5 u ^ GO cd >, .£ .£ |1 4) _o 4) 3 3 3 >, 3 3 s>> e 3 ^ 2? 3 3 fc: 2 «* 2 ^ -g; •i-»o £ g>g Oco^_,42ed*43e3 >» 6^0 >> g ^ t 5 't S 60 © — C ' £ g 3c^‘-3cd^cd®rt6b ed cd •— Oh H < o <3 &H Dh SC 5 «3 ed 46 GO H > > Cu CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 317 Table 2. Data reductions for characters (= CH) of color and pattern (PV = granules between paravertebral stripes, PV/GAB = percent of granules around midbody lying between paravertebral stripes) and scutellation (GAB = granules around midbody, OR = granules from occiput to rump, FP = femoral pores combined, SDL = subdigital lamellae of the longest toe of the left pes, COS = circumorbital scales combined, LSG = lateral supraocular granules, MS, mesoptychial scales) in Cnemidophorus laredoensis (clones LAR-A and LAR-B) and Cnemidophorus gularis. CH LAR-B (99) Upper Valley LAR-B (99) Lower Valley LAR-A (99) Pooled C. gularis (9<5) Pooled PV 15.5 ± 0.17 15.0 ± 0.12 11.1 ± 0.06 15.5 ±0.15 12-21 (82) 10-17 (106) 9-14 (236) 10-21 (214) PV/GAB 18.3 ± 0.16 18.0 + 0.16 12.0 + 0.07 16.9 ± 0.15 14.6-22.8 (82) 12.1-21.2 (99) 9.6-15.4 (233) 11.2-23.0 (212) GAB 84.8 ± 0.41 83.2 ± 0.23 91.6 ± 0.13 91.6 + 0.39 79-93 (82) 78-88 (99) 84-98 (234) 76-111 (212) OR 210.0 ± 0.99 207.0 ± 0.62 227.4 + 0.33 224.4 ± 0.9 192-229 (80) 194-230 (91) 211-240 (225) 189-266 (198) FP 32.9 ± 0.16 33.4 ± 0.14 32.9 ± 0.08 33.6 ± 0.15 30-36 (79) 30-37 (98) 28-37 (227) 28-40 (210) SDL 31.4 ± 0.11 32.0 ± 0.10 34.0 ± 0.06 31.8 + 0.11 28-33 (78) 30-35 (103) 32-37 (234) 28-38 (210) COS 10.5 ± 0.14 10.3 + 0.09 10.6 ± 0.07 14.5 + 0.25 8-13 (79) 9-14 (104) 8-15 (232) 8-30 (201) LSG 12.4 ± 0.26 12.2 ± 0.19 14.7 ± 0.11 16.1 + 0.31 8-19 (79) 7-18 (103) 10-20 (230) 8-32 (203) MS 16.2 ± 0.18 15.8 + 0.12 17.4 + 0.07 15.1 + 0.11 13-20 (78) 13-19 (97) 15-21 (219) 10-20 (201) 200115, UADZ 1565-1573); (V-l 1) Del Rio, 0.8 km. N international bridge on east side of road (9<5, KU 200105-200112); (V-12) Val Verde Co. side of Sycamore Creek off U.S. Hwy. 277 (9<5, KU 200116-200125). Kinney County: (K-l) Pinto Creek off U.S. Hwy. 277 (9<5, UADZ 1577-1579). Maverick County: (M-l) Eagle Pass, W suburbs (9<5, UADZ 1152- 1153); (M-2) Eagle Pass, just W international bridge near Indian huts (99, LAR-B UADZ 1003-1004); (M-3++) Eagle Pass, 1.6 km from traffic signal nearest international bridge (9$, LAR-B KU 200085-200089, UADZ 1005-1010; 9(5, KU 200098-200103, UADZ 1150-1151); (M-4++) El Indio, 30 km. S Eagle Pass near Rio Grande (99, LAR-B definite field identification; 9(5, definite field identification). Webb County: (W-15++) 96 km. N Laredo (or 79 km. S Eagle Pass) along FM 1472-1054 (= “Mines Road”) (99, LAR-B UADZ 2209-2212, 2217-2222; 9(5, 2213, 2223-2231). Zavala County: (ZA-1) Crystal City, site 1 (9(5, UADZ 1580); (ZA-2) Crystal City, site 2 near school (9(5, UADZ 1581-1588). TEXAS (lower Rio Grande Valley). Starr County: (S-4*) Garceno, S of town (99, LAR-A UADZ 1446-1452, 1660-1670; 9(5, UADZ 1619-1624, 1831-1834); (S-5) Rio Grande City, between international bridge and high school (99, LAR-A KU 199968-199971, UADZ 1365-1375); (S-6+) 7.2 km. E Rio Grande city, then 2 km. S U.S. Hwy. 83 near Rio Grande (99, LAR-B UADZ 1500-1502; 99, LAR-A UADZ 1453-1469); (S-7*+) La Grulla, site 2 (99, LAR-B UADZ 1717; 99, LAR-A UADZ 1671; 9(5, UADZ 1835-1858); (S-8*+) La Grulla, site 1 (99, LAR-B UADZ 1503-1506, 1718-1730; 99, LAR-A UADZ 1470-1472, 1673-1881; 9(5, UADZ 1614-1618, 1859-1878); (S-9*) junction Texas Hwy. 649 and 2686 (99, LAR-A UADZ 1656; 9<5, UADZ 2013-2015); (S-10*) El Sauz (99, LAR-A UADZ 1657-1659; 9(5, UADZ 1625, 1829-1830). Hidalgo County: (H-l) Los Ebanos, at ferry landing (9<5, KU 199960-199964, 200038-200040); (H-2*) Los Ebanos, between town and 318 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Table 3. Explanation of symbols associated with locality codes. Symbol LAR-B 99 LAR-A 99 C. gularis 93 none Present or Present or Present * Absent Present Present + Present Present Absent *+ Present Present Present ++ Present Absent Present Rio Grande (99, LAR-B UADZ 1116-1128, 1329-1345, 1379-1394, 1507-1537, 1731-1781, 1967-1993; 99; LAR-A UADZ 1054-1058, 1307-1314, 1394, 1473-1499, 1682-1716, 1954- 1966; 93, UADZ 1176-1179, 1411, 1608-1613, 1879-1884, 2006-2012); (H-3+) Bentsen-Rio Grande Valley State Park, along last 300 m. of trail to edge of Rio Grande (99, LAR-B KU 199974-199999, 200017-200019, UADZ 1129-1130; 99, LAR-A KU 199972-199973); (H- 4++) Rio Rico Road, levee along Rio Grande (99, LAR-B UADZ 1782; 93, UADZ 1885). Cameron County: (C-l) 3.2 km. S Los Indios and U.S. Hwy. 281, few meters from edge of Rio Grande (99, LAR-B UADZ 1396); (C-2++) El Ranchito, 14.4 km. W Brownsville between U.S. Hwy. 281 and Rio Grande (99, LAR-B KU 200020-200034, UADZ 1783- 1796; 9<$, KU 200049-200084, UADZ 1180-1203, 1348, 1886-1925); (C-3++) Palmito Hill, 13.2 km. E Brownsville or, 16 km. W Gulf of Mexico, then 3.2 km. S Texas Hwy. 4 within 600 m. of Rio Grande (99, LAR-B KU 200035-200037, UADZ 1346-1347, 1397; 93, KU 200042-20048, UADZ 1358-1364, 1404-1409, 1926-1935); (C-4) 21.0 km. E Brownsville, then S Texas Hwy. 4 near Rio Grande (93, UADZ 1936). MEXICO (upper Rio Grande Valley). Coahuila: (CO-1) Lake Amistad, vicinity of dam (93, UADZ 1215-1221); (CO-2) 7.5 km. NW Ciudad Acuna off Mexico Hwy. 349 (93, 1204-1214); (CO-3) Ciudad Acuna, suburbs W international bridge at site 1 (99, LAR-B UADZ 1012-1030, 1291-1292); (CO-4++) Ciudad Acuna, suburbs W international bridge at site 2 (99, LAR-B UADZ 1264-1271; 93, UADZ 1349-1350); (CO-5) Ciudad Acuna, riverfront park just SE international bridge (99, LAR-B KU 200090-200095, UADZ 1011); (CO-6) 17.8 km. NW Jimenez, between Ciudad Acuna and Piedras Negras (99, LAR-B UADZ 1031-1033); (CO-7) Jimenez, between Ciudad Acuna and Piedras Negras (93, UADZ 1222-1226); (CO-8++) Piedras Negras, W suburbs (99, LAR-B UADZ 1035; 93, UADZ 1227); (CO-9++) Piedras Negras, W international bridge (99, UADZ 1036-1041, 1272-1290; 93, UADZ 1228-1234, 1351-1353); (CO-10) Piedras Negras, about 5 km. S international bridge (99, LAR-B KU 200096-200097); (CO-11) Vicente, between Piedras Negras and Nuevo Laredo near Rio Grande (93, UADZ 1228-1234); (CO-12) Hidalgo, NW Nuevo Laredo near Rio Grande (93, UADZ 1245). Nuevo Leon: (NL-1) Colombia, off Hwy. 2 near Rio Grande (93, UADZ 1246-1252). MEXICO (lower Rio Grande Valley). Tamaulipas: (T-3*) Cuidad Miguel Aleman, S side of city (99, LAR-A KU 199933-199934; 93, KU 199965-199968); (T-4) Ciudad Miguel Aleman, N side of city near factory at site 1 (99, LAR-A KU 199935-199940); (T-5*+) Ciudad Miguel Aleman, N side of city near movie theater and baseball field at site 2 (99, LAR-B one 1257; 99, LAR-A UADZ 1076-1100, 1315-1326; 93, UADZ 1259); (T-6+) Ciudad Diaz Ordaz, ferry landing across from Los Ebanos (99, LAR-B UADZ 1398, 1 live; 99, LAR-A one definite field identification); (T-7*+) Reynosa, NW international bridge in suburbs off Hwy. 59 (99, LAR-B UADZ 1131; LAR-A definite field identification; 9$, UADZ 1261-1263); (T-8) Reynosa, NW of city off Hwy. 2 near hydroelectric complex (99, LAR-A UADZ 1101); (T-9*) Reynosa, about 8 km. NW city off Hwy. 2 (99 LAR-A UADZ 1102-1111; 93, UADZ 1260); (T-10+) Reynosa, several km. SE international bridge near canals (99, LAR-B UADZ 1132; 99, LAR-A UADZ 1112-1115); (T-ll+) Nuevo CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 319 S « o y >- o a Ox B •o & + pL, +3 ^ Of p 13S .a «- X o Ox . eS ^ W) C o ^ O o 60 o xO ^ .g jg c X 3 fe 5b j§ -o 2 « 2, o P3 M 3 3 2 o S •S 1 «3 ed & CL) O ^ Ml C* M II < 3 " + II "o «< “ §u s «» 'S' O ^ 4) —I Of '*Ui II 2^ CQ JS> + ^ d 3 < g 2 J ^ >3 II I' -g og S ■S “ ii ll£ ils 4° I 8 -S s 3 ^ ^ O A> — < s' I J 1 3 QQ JS • T3 §2 s a .2 x x M g< e* 2 g 60 O O o Ox H 2 g* & g £ u ii -o >, 2 .ti -d J d g s £ £ co 5b 4) -a Ox o 2 3 2 I 3 3 +j « d o .ti cs il ^ S o aj ^3 §9 C/3 a> T3 I>OQQQCQ!>OQCQ m^-rxrxvovoso^^soir)^-'3-Tiin^fir)inin'^inin»nin»n OOOOOOOOOOOOOSOOOOMOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Os Os Os O' O Os Os Os Os Os Os Os O' Os O O O O Os O O O O O O 2 M M ^ - >22^ + * * A> TPx '•o r--- oo r f r7' M Cfl M K S It m -o i i o o u u u u u t -s N m + + o V) so O© O —< + tt-.+ i i i i i * + * o — OOOOOiris0r"'“"“ UUUUUHHHHH 320 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Progreso, NW international bridge ($$, LAR-B UADZ 1133-1134; 9$, LAR-A UADZ 1327-1328); (T-12) Matamoros, several sites ($$, UADZ 1354-1357, two uncatalogued). No specimens of either of the LAR clones have been collected upriver from Ciudad Acuna, Mexico (CO-1, CO-2), or Del Rio, Texas (V-l, V-2, V-3, V-4), as the river there is entrenched in a deep valley surrounded by rocky roughlands with desert shrub vegetation. However, C. gularis was abundant at all of these localities and C. tigris was present at CO-3, V-l, and V-2. Investigations between Eagle Pass and Laredo and Piedras Negras and Nuevo Laredo have been hampered by lack of access to the Rio Grande. In May 1985, sites along the river reached from Hwy. 2 in Mexico yielded only specimens of C. gularis at CO- 11 and CO- 12 in Coahuila and NL-1 in Nuevo Leon. In May 1987, LAR-B was observed in syntopy with C. gularis near the Rio Grande at El Indio (M-4++), Maverick County, and both were collected in northern Webb County 96 kilometers north of Laredo (or 49 kilometers south of the El Indio record). Numerous visits have been made to sites in the middle Rio Grande Valley but have not resulted in the discovery of LAR-B. LAR-A occurs alone or with C. gularis (usually) at many sites between Laredo-Nuevo Laredo and Rio Grande City and at a few sites well away from the River in parts of Dimmit, LaSalle, and Starr counties (Fig. 2). In the lower Rio Grande Valley, LAR-B apparently has a patchy distribution from about 7.2 km. E Rio Grande City and from Ciudad Miguel Aleman to within about 16 kilometers of the Gulf of Mexico at Palmito Hill (C-3++). Most records for LAR-B are situated within a few hundred meters of the Rio Grande; no records for the clone farther than 10 kilometers from the river are known. Conant (1975) depicted the range of C. sexlineatus, an apparent progenitor of the LAR clones (McKinney et al., 1973; Bickham et al., 1976; Wright et al., 1983), as extending to the Rio Grande in much of southern Texas. Intensive searches in 1986 produced no records for C. sexlineatus closer to the river than Bruni (66 km. E Laredo — UADZ 1538, 2018-2019) and 10.2 km. S Mirando City (53 km. E Laredo — UADZ 2020-2021), Webb County, and La Gloria (51 km. NE Rio Grande City — UADZ 2022-2023), Starr County. Although C. sexlineatus occurs on South Padre Island to the southern tip (UADZ 1399, 1798- 1805), no sites for the species have been discovered in mainland Cameron County (only C. gularis was collected at the bridge at Port Isabel, about 3.2 kilometers from South Padre Island — UADZ 1400-1403). The distance between the closest records for LAR-A at S-9* and C. sexlineatus at La Gloria (both in Starr County; Fig. 2) is about 32 kilometers; the distance between the locality for LAR-B at C-3++ and C. sexlineatus on South Padre Island is about 20 kilometers. CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 321 Figure 2. Map depicting the known ranges of the parthenogenetic clones LAR-B ($$) and LAR-A ($$) in Texas and Mexico. Letters and numbers correspond to locality codes in text. The entire range of LAR-B is contained within a small part of the range of the nominal subspecies of C. gularis (Walker, 1981a, 1981b). Syntopy between LAR-B and C. gularis in the upper Rio Grande Valley was observed at four sites in Texas and three in Coahuila (Table 4). LAR-B occurs in patches of altered habitat with sandy to loamy soil in Ciudad Acuna and Piedras Negras, Coahuila, where C. gularis frequently is absent. In the lower Rio Grande Valley, the two forms are syntopic at six sites in Texas and two in Tamaulipas. LAR-B is syntopic with both LAR-A and C. gularis at three localities in Texas and two in Tamaulipas (Table 4). Further efforts are expected to reveal that C. gularis occurs at several sites where brief visits have produced only specimens of LAR-B and LAR-A (T-6+, T-10+, T-ll+). 322 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 The zone of syntopy between LAR-B and LAR-A is on both sides of the Rio Grande from about Ciudad Miguel Aleman to Nuevo Progreso, Tamaulipas, a distance of about 90 kilometers by road. Distributional Ecology in the Lower Rio Grande Valley All locality records for LAR-B are in an intensively utilized region. Extensive urban, suburban, and agricultural developments have made large areas between Mission and Brownsville, Texas, unavailable to whiptails. The adjacent part of Mexico also is extensively altered and few areas are available. The original habitats of the lower Rio Grande Valley included the following sequence of plant communities from the eastern boundary of Zapata County to the Gulf of Mexico: Chihuahuan thorn forest, upper valley flood forest, barretal, upland thorn shrub, mid-valley riparian woodland, sabal palm, and loma/ tidal flats (Rio Grande Valley National Wildlife Protection Plan, Department of the Interior). Protected areas under state (Bentsen-Rio Grande Valley State Park) or federal (Santa Ana Wildlife Refuge) management, patches around cultivated land, and scattered tracts for grazing livestock provide most of the surviving examples of these formations between Rio Grande City and Brownsville. Whether LAR-B is present or absent at a site in the lower Rio Grande Valley seems most affected by distance of the site from the river, soil characteristics, and vegetation. Whereas LAR-A is associated with sandy habitats over much of its range in Dimmit, LaSalle, Webb, Zapata, and Starr counties, and adjacent parts of Tamaulipas, all sites of syntopy between LAR-B and LAR-A in Starr County, Hidalgo County, and Tamaulipas are characterized by loamy soils with varying mixtures of sand, silt, and mud (Geologic Atlas of Texas: McAllen-Brownsville Sheet). At a typical locality for LAR-A (S-5), the presence of the clone is apparent by large numbers of burrows and tail tracks. Whereas sandy soil drains rapidly, even small amounts of rainfall can result in a quagmire at the kinds of loamy sites where LAR-B and LAR-A are syntopic (S-7*+, S-8*+, H-2*+, T-ll++). Upon drying, some loamy soils develop deep cracks or become compacted; whiptail burrows and tracks are seldom apparent at such sites (Fig. 3A). Soil characteristics (for example, texture, exposure, drainage, pene¬ trance, origin) clearly are not as important in the distribution of C. gularis as they are in the distribution of LAR-B and LAR-A. C. gularis is successful on caliche, rock surfaces (including sandstone and granite in Oklahoma), “ersatz” substrates in Texas (including cement, asphalt, gravel), loose soils (sandy to loamy), and heavily eroded rocky to gravelly soils (Walker, 1981a, 1981b; Walker et al., 1986a; Walker and Cordes, 1988). The inability of both LAR clones to become established in areas CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 323 devoid of sandy to loamy soils appears to be directly related to the absence of suitable burrowing sites (Walker et al., 1986b). At sites of syntopy between LAR-B, LAR-A, and C. gularis , the relative numbers of each form can be correlated with the geographical location of the site along a northwest-southeast transect between Starr and Hidalgo counties, use of the land in the recent past (undisturbed, moderate alteration, steady degradation, catastrophic alteration), location relative to thorn shrub tracts (modified or undisturbed), and vegetation (affected by soil, exposure, successional stage). C. gularis inhabits all Chihuahuan thorn shrub, barretal, and upland thorn shrub climax formations bordering or entering the Rio Grande entrenchment, and from these it enters all other types of habitat available to whiptails along the river. LAR-B and LAR-A have been found only in the edge of climax formations, and always in areas where mixed forest (Fig. 3B) or bunchgrass-weed (Fig. 4A) associations form narrow ecotones with such habitats. Weedy habitats with insular characteristics (located well away from thorn shrub tracts) often support larger numbers of parthenogens than C. gularis (H-2*+), whereas the reverse is usually true for weedy habitats adjacent to climax formations (S-7*+). LAR-A reaches highest densities in predominantly bunchgrass/weed-mesquite habitats with sandy to loamy open spaces, both in near absence of C. gularis (T-5*+) and with sizeable numbers of the species present (S-4*). LAR-B is rare (T-5*+) to absent (S-4*, S-5) in bunchgrass-mesquite habitats in the lower Rio Grande Valley; the clone reaches greatest densities in riparian bunchgrass- weed habitats in old fields (H-2*+, C-2++) and mixed forest and bunchgrass-mesquite ecotones (H-3+). The rare occurrence or absence of LAR-B upriver from S-6+ reflects the paucity of these habitats in the middle Rio Grande Valley. Four sites in Starr and Hidalgo counties, each with significantly different relative numbers of LAR-B, LAR-A, and C. gularis were studied. At S-6+, large numbers of LAR-A were found in a strip of bunchgrass-weed habitat a few meters in width along the Rio Grande, the adjacent part of a small melon field, and part way up a small slope with buffelgrass and weeds. Of 20 lizards collected in about three hours at S- 6+, 17 were LAR-A and only three were LAR-B (all within 75 meters of the river). Although C. gularis was present in a nearby thorn shrub tract, none was collected along the river at S-6+, possibly because of the patchy and disturbed character of the site. Two sites only a few hundred meters apart near La Grulla supported quite different assemblages of whiptails. S-7*+ consisted of a fenced tract of heavily grazed thorn shrub beside a large overgrown field crossed by an unsurfaced road. The virtual absence of lizards in the field seemed to be directly related to the absence of suitable open spaces; the few whiptails present were found along the access road (four C. gularis and 324 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 one LAR-A collected). As expected from the appearance of the habitat, large numbers of C. gularis (20 collected) and almost no parthenogens (one LAR-B and no LAR-A obtained) were present in the grazed thorn shrub tract with huge clumps of Opuntia scattered within the compacted loamy to gravelly soil. Site S-8*+, located only a few kilometers from S-7*+ consisted of a tract of land out of cultivation for several years and located between narrow borders of mixed forest-mesquite. The vegetation of the overgrown field consisted of bunchgrass mixed with tall weeds (Russian thistle, sunflower, and so forth). Totals of 25 C. gularis , 17 LAR-B, and 12 LAR-A were captured at S-8*+ in a morning and two late afternoon visits. The striking difference in the composition of the whiptail communities at S-7*+ and S-8*+ was apparently related to the friable soil and open-structured grass-weed vegetation at S-8*+ and their absence at S-7*+. Site S-8*+ is near the northwestern distributional limits for LAR-B in the lower Rio Grande Valley. Northwest of S-8*+, LAR-B made up only 3.7 percent (one of 27) of specimens at S-7*+, 15.0 percent (three of 20) at S-6+, and 2.6 percent (one of 38) at T-5*+. The most remarkable site studied was an “insular” tract a few hundred meters south of Los Ebanos, Hidalgo County (Fig. 3A). H-2*+ consisted of scattered mesquite, huisache, and paloverde in a riparian site with ground vines, bunchgrasses, and weeds in an old field (currently part of Los Ebanos Unit, Rio Grande Valley National Wildlife Refuge). The site, which was originally covered with mixed forest, is surrounded on three sides by cultivated land and by a border of mixed forest along the Rio Grande to the east. Remnants of thorn shrub formations with compacted gravelly soil occur above the floodplain in Los Ebanos (H-l), which is separated by several hundred meters of cultivated land from H-2*+. In April 1986, whiptails were widely spread through the open-structured vegetation, along an access road, and along roads around the periphery at H-2*+. In May 1986, many of the small open spaces in the overgrown field were filled with lush growths of grasses and weeds. However, many lizards still were present in areas away from the road. By mid-July 1986, the area was almost completely overgrown except for scattered open spaces along the unsurfaced roads, where lizard activity was most intense. Although H-2*+ initially appeared to be marginally suitable for whiptails when first visited in 1985, collecting data indicate that no other site except C-2++ in Cameron County supported such large numbers of whiptails. In seven visits to H-2*+ in 1985 and 1986, totals of 157 LAR- B, 92 LAR-A, and 23 C. gularis were collected out of an estimated 1000 whiptails observed. The relative numbers of LAR-B, LAR-A, and C gularis observed at H-2*+ are consistent with expectations developed from studies elsewhere in the lower Rio Grande Valley. The low numbers of C. gularis at H-2*+ CLONE OF PARTHENOGENETIC WHIPTAIL LIZARD 325 Figure 3. A) Example of riparian bunchgrass-weed habitat on abandoned agricultural tract at Los Ebanos, between town and Rio Grande, Hidalgo County (H-2*+), which supports large numbers of LAR-B ($$) and LAR-A ($$) and a few Cnemidophorus gularis ($$)• B) Example of mixed forest and bunchgrass-cactus-mesquite ecotone in Bentsen-Rio Grande Valley State Park, along last 300 meters of trail to edge of Rio Grande, Hidalgo County (H-3+), which supports large numbers of LAR-B ($$), a few LAR-A ($$), and no Cnemidophorus gularis (9 cd ^ "T3 'd- — > 5 ,A> 3 ?3 O C 1 = C> 3 c 3 2 o 3 Su ;! O tq s I 1 3 § § 3 =3 ;so ^ “3 $ rs 3 6 si 03 bo © 3 -© 3 -S ^ bo Q 03 3 Dh 3 ^ O .« -a "* 5 "5 © ~s © "3 § -s: 3 -2 ©, cu ^ 5f 03 r^J Cj ft* O co ■is b 23 2 Sh o T3 >, -§ I © © II a, cd a Macrothrix laticornis Moinidae Moinodaphnia macleayi Sididae Latonopsis occidentalis Sida crystallina 338 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Species in the family Macrothricidae are specialized for living in benthic deposits (Fryer, 1974), but occurred frequently on plants that had developed thick coatings of flocculent organic materials. Five cladoceran species that occurred occasionally in samples from the periphytic habitat included the bosminid Bosmina longirostris, daphniids Ceriodaphnia lacustris, Daphnia ambigua, and Daphnia parvula , and the sidid Diaphanosoma brachyurum. These species are not listed in Table 2 because they are planktonic specialists and probably occurred on the plants by accident. Scapholeberis sp. is hyponeustonic (Cole, 1975) but was found in this study in association with floating leaves. Periphytic species that occurred on many dates in at least three of the four main ponds and were widespread in the 20-pond surveys included A Iona setulosa , Alonella hamulata , Chydorus brevilabris , Simocephalus serrulatus , Ilyocryptus spinifer , and Latonopsis occidentalis (Table 2). These species may be tolerant of a broad range of environmental conditions or are able to colonize and exploit new habitats quite readily, or both. Alona affinis , Alonella hamulata , and Simocephalus serrulatus appeared to be cool-season species, occurring in more ponds during the March survey than during the survey in July. Ceriodaphnia quadrangula , Macrothrix laticornis , and Latonopsis occidentalis were apparently warm-season species, occurring in about one-third of the ponds during the July survey but not occurring in any ponds during the March survey. Absence of these species in samples collected during their “non-preferred” season, however, does not mean that individuals were not present in a pond. On the contrary, our observations on the reproductive strategies of these animals suggested that most species persisted through periods of stress as low-density parthenogenetic populations. Evidence of this phenomenon is discussed at length in Campbell (1983). The total number of periphytic Cladoceran species found in the four ponds sampled intensively throughout the seasons ranged from 11 to 16 (Table 2). The average number of species per pond was 13.5. In the 20- pond surveys (both surveys combined), three of the ponds sampled contained zero or one species, but these three ponds were polluted. One received stormwater runoff directly from a major street on the Texas A&M campus; the second received poorly treated wastewater from a rural community’s sewage treatment facility; and the third pond contained sediments contaminated with industrial waste. The average number of species per pond found in each of the 20-pond surveys was 3.85 and 3.60 species in March and July, respectively. The average for the two surveys combined was 6.1 species per pond. This indicates that sampling on a single date from a pond yielded a poor estimate of the total number of species actually occurring there, which is not surprising because several periphytic cladoceran species appeared to PERIPHYTIC CLADOCERA OF BRAZOS COUNTY, TEXAS 339 U / r i i r i i \ i i i i iiii — i — n — r 5 10 15 20 Number of Ponds Sampled Figure 1. Species-pond curve showing the number of different periphytic cladoceran species accumulated from additional ponds in order as the ponds were actually sampled. The species lists were based upon five samples collected from each pond on each of two dates — 2 March and 22 July 1982. have cool- or warm-season “preferences.” The average number of species per pond for the combined surveys (6.1) was much less than the average number of species per pond found with frequent sampling in the basic four-pond study (13.5) (Table 2), suggesting that sampling a pond on only two dates in two different seasons is probably not adequate to detect all of the species that occur in a particular pond. Regardless of the inadequacy of our 20-pond surveys for characterizing the total species composition of individual ponds, we have evidence that two 20-pond surveys at different seasons provide a fairly accurate picture of the total pool of periphytic cladoceran species occurring in a particular region. A species-area curve (actually a species-pond curve) constructed for the combined March and July 20-pond surveys (Fig. 1) (with the number of species accumulated from additional ponds shown in order as the ponds were actually sampled) indicates that no additional new species were collected beyond the fourteenth pond. The only species that we know we missed in the 20-pond surveys was Oxyurella tenuicaudis , which was represented by only a single individual collected on one date from Central Park Pond. Therefore, we are confident that the list of periphytic cladoceran species presented in Table 2 is quite complete for this region for the time period during which the study was conducted. Literature Cited Allard, D. W. 1982. Littoral microcrustacean population dynamics in Post Oak Lake. Unpublished Ph.D. dissertation, Texas A&M Univ. College Station, 122 pp. 340 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Campbell, J. M. 1983. Interpond and intrapond variation in populations of periphytic cladoceran microcrustacea. Unpublished Ph.D. dissertation, Texas A&M Univ. College Station, 275 pp. Campbell, J. M., W. J. Clark, and R. J. Kosinski. 1982. A technique for examining microspatial distribution of Cladocera associated with shallow water macrophytes. Hydrobiol., 97:225-232. Cole, G. A. 1975. Textbook of limnology. The C. V. Mosby Co., Saint Louis, Missouri, 283 pp. Fryer, G. 1968. Evolution and adaptive radiation in the Chydoridae (Crustacea: Cladocera): a study in comparative functional morphology and ecology. Royal Soc. London Phil. Trans., ser. B, 254:221-385. - . 1974. Evolution and adaptive radiation in the Macrothricidae (Crustacea: Cladocera): a study in comparative functional morphology and ecology. Royal Soc. London Phil. Trans., ser. B, 269:137-274. Present address of Campbell: Biology Department, Mercyhurst College, Erie, Pennsylva¬ nia 16546. COMPARATIVE IN VITRO DIGESTIVE EFFICIENCY OF CATTLE, GOATS, NILGAI ANTELOPE, AND WHITE-TAILED DEER James C. Priebe, Robert D. Brown, and Doreen Swakon Caesar Kleberg Wildlife Research Institute and College of Agriculture and Home Economics (DS), Texas A&I University, Kingsville, Texas 78363. Abstract. — The digestive efficiencies of rumen inocula from domestic cattle and goats, nilgai antelope ( Boselaphus tragocamelus ), and white-tailed deer ( Odocoileus virginianus ) were compared to test the effect of the donor ruminant species maintained on a diet of commercial wildlife pellets on in vitro digestion. A modification of the two-stage Tilley and Terry (1963) technique was used to evaluate their digestion of five substrates: the pelleted ration, cellulose, spiney hackberry browse (Celt is pallida ), alfalfa hay (Medicago sativa ), and coastal Bermada hay ( Cynodon dactylon ). The similar in vitro dry and organic matter digestion (IVDMD, IVOMD) across substrates of cattle (69.5 and 68.3 percent), goats (68.3 and 67.4), and nilgai (70.9 and 69.4) were greater than (P = 0.008, P = 0.03) those of deer (57.9 and 57.4). The IVDMD of the three forages by all inocula decreased as cell wall content increased, whereas in vitro cell wall digestion (IVCWD) did not. Overall mean IVCWD across forages of cattle (33.6), goats (33.8), and nilgai (34.6) exceeded (P < 0.05) those of deer (20.2). We concluded that donor ruminant species may influence in vitro digestion. Key Words : forage; digestion; cattle; goats; Odocoileus virginianus ; Boselaphus tragocameleus. One of the most perplexing problems for wildlife managers is maintaining large ruminants within the carrying capacity of the habitats they occupy. Estimating available nutrient resources facilitates solving this problem. Estimates of these resources for a given animal species depends on the production, nutrient composition, intake, and digestibility of preferred forages, and their relation to nutrient requirements of the animals. In the Texas Coastal Bend, cattle, goats, nilgai antelope (. Boselaphus tragocamelus ), and white-tailed deer ( Odocoileus virginianus) must share habitats. Before effective management of these species can be implemented, a technique that accurately determines the nutrient availability of forages to the different species of ruminants must be found. In vitro analysis is the standard method for estimating the in vivo digestibility and digestible nutrient content of forages for livestock. However, the inoculum of cattle that have been fed a hay-grain ration often is used and may not account for the fact that the natural diets consumed by these ruminants vary (Sheffield at ah, 1983). While the diet of the inoculum donor is known to affect in vitro digestion (Bezeau, 1965; Horton et al., 1980; Campa et al., 1984), the effect of donor species is unclear. The purpose of this investigation was to compare the in vitro The Texas Journal of Science, Vol. 39, No. 4, November, 1987 342 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 digestive efficiency of inocula from cattle, goats, nilgai and deer fed a common diet. Methods Substrates We used commercial wildlife pellets (P&M Products, San Antonio, Texas), cellulose (solka floe), spiny hackberry browse ( Celtis pallida ), alfalfa hay ( Medicago sativa), and coastal Bermuda hay ( Cynodon dactylori) as substrates for digestion. The three forages were selected because they are from different forage classes and provide a wide range of cell wall content (Table 1). Duplicate samples of substrates were air dried at 60° C and ground in a Wiley mill through a one-millimeter screen. Dry matter (DM), organic matter (OM), ash, and crude protein (CP) were determined by proximate analysis (A.O.A.C., 1980). Cell wall content (CWC) was determined by the Van Soest technique (Van Soest and Wine, 1967). Inocula Sources Inocula from two animals of each ruminant species were used to digest substrates. All animals were maintained on a diet of commercial wildlife pellets (Wheaton and Brown, 1983). Feed and water were offered ad libitum for at least one month prior to inocula collection. Inocula were collected from 0800 to 1200 hours. Cattle and nilgai were sedated mildly with xylazine hydrochloride, and goats were restrained physically before inocula collection via stomach tube. White-tailed deer were sacrificed for the collection. In Vitro Procedure Approximately 400 milliliters of forestomach fluid was collected from each animal and transferred to a plastic jug. The jugs were placed in a beaker filled with water at 37° C, and CO2 from a portable tank was bubbled through the contents until they reached the laboratory. Transport time ranged from five to nine minutes with a mean of 6.8 minutes. The in vitro procedure was the two-stage method of Tilley and Terry (1963), as modified by Moore (Harris, 1970). Percent in vitro dry and organic matter digestion (IVDMD, IVOMD) were determined on all substrates for each inoculum source. Percent in vitro cell wall digestion (IVCWD) (Van Soest et al., 1966) was determined only on the three forages. All analyses were run in duplicate. Statistical Methods The results are presented in tables as percentages, but were arcsine transformed for statistical analysis. In vitro DM, OM, and CW digestion among ruminant species across substrates was analyzed by analysis of variance (ANOVA) for a split-plot design with repeated measures (Gill, 1978; SAS Institute, Inc., 1982). The model used was: Yijk = /x + A' + D(i)j + Bk+ (AB)ik + (DB)(i)jk, Where i = 1, 2, 3, 4 ruminant species; j = 1, 2 animals per species; and k = 1, 2, 3, 4, 5 forage substrates. Tukey’s studentized range (HSD) test was used to determine significant differences in IVCWD means between ruminants (SAS Institute, Inc., 1982). Results Collection and processing times of inocula were far below the critical limits stated by Schwartz and Nagy (1972) for maintaining microbial viability. The IVDMD values of the commercial wildlife pellets by all inocula, (Table 2) were higher than those observed in vivo for cattle (63.3 COMPARATIVE IN VITRO DIGESTION 343 Table I. Dry matter, organic matter, ash, crude protein, and cell wall content of substrates used in in vitro digestibility comparison. Substrate DM OM1 ASH1 CP1 CWC1 Wildlife pellets 93.3 85.1 14.9 15.7 47.5 Cellulose 95.5 99.8 0.2 0.0 99.8 Spiny hackberry 90.9 81.8 18.2 17.8 27.6 Alfalfa hay 87.7 90.4 9.6 17.1 41.0 Coastal Bermuda hay 89.2 91.5 8.5 9.1 72.4 'Percent of dry matter. percent) and goats (59.1) (Swakon, unpublished data), nilgai antelope (57.0) (Priebe, 1985), and deer (54.4) (Wheaton and Brown, 1983). Pelleted diets often have a faster rate of passage in vivo than natural feedstuffs (Mautz and Petrides, 1971), and this reduces fermentation time and lowers digestibility (O’Dell et al., 1963). Interspecies variation among inocula was greatest in the IVDMD and IVOMD of the cellulose substrate. Inocula that had higher cellulose digestion had greater overall digestion across substrates. Variation was contributed by substrates ( P = 0.0001), ruminant species [IVDMD ( P 0.008 )„ IVOMD ( P — 0.03)], and substrate times ruminant species interaction [IVDMD ( P = 0.002), IVOMD ( P = 0.003)]. When inocula were compared across the three forage substrates only, variation due to interaction was insignificant [IVDMD (P = 0.19), IVOMD (P = 0.31)]. Substrate variation remained highly significant (P = 0.0001), whereas ruminant species variation was less significant [IVDMD (P = 0.03), IVOMD (P= 0.08)]. The IVDMD of the three forages by all inocula (Table 2) appeared to decrease with an increase in CWC (Table 2), whereas IVDWD seemed to be irrespective of CWC (Table 3). No statistical analyses were made with regards to these relationships because they were beyond the scope of this study. However, they agree with previous researchers (Herschberger et al. 1959; Donefer et al., 1960; Tomlin et al., 1961; Baumgardt et al. 1962; Barnes, 1967; Short et al., 1974; Robbins et al., 1975), who found that CWC greatly effect DM digestion, but not CW digestion. Specific physical and chemical properties of forage cell walls determine their degradation. Variation in IVCWD was contributed by both forage substrate (P = 0.001) and ruminant species (P = 0.006) (Table 3). The ability of deer inoculum to digest cell walls was significantly lower (P = 0.05) than that of the other ruminants. Discussion and Conclusions Species-specific differences in the digestive efficiency of inocula can be explained on the basis of the ability of ruminal microbes to digest fiber. 344 THE TEXAS JOURNAL OF SCIENCE— VOL. 39, NO. 4, 1987 Digestive physiology and ruminal microflora vary depending on forage and niche selection of ruminants (Hofmann, 1968; Gieseche and Van Goylswyk, 1975; Kay et al., 1980). Consequently, rumens of cattle, goats, and nilgai may have allowed for maintenance of greater numbers of cellulolytic bacteria than that of deer. The greater relative size of the rumen measured as the mean wet-weight of rumen contents as a percent of body weight in cattle, goats, and nilgai compared to white-tailed deer (Teer and Sheffield, 1972; Kay et al., 1980) may have allowed the digesta to be retained longer, thereby supporting a larger microbial population. In vitro cellulose and cell wall digestion and in vivo digestive efficiency comparisons of Arman and Hopcraft (1975), Huston (1976), and Mould and Robbins (1982) support this conclusion. Discussions of the effect of donor ruminant species on in vitro digestion from previous research appear contradictory. Studies where no differences were found among species compared either animals with similar dietary habits (LeFevre and Kamstra, 1960; Newman, 1974) or those with varying habits on different diets (Welch et al., 1983). One exception was that of Palmer et al. (1976), but they found significant substrate-animal species interaction. Our results generally agree with those of Short (1963), Ward (1971), Robbins et al. (1975), and Blankenship et al. (1982). Short (1963) and Robbins et al. (1975) found slight differences between cattle and deer in the in vitro digestion of browse. Ward (1971) reported significant differences between cattle and elk inocula in the digestion of grasses, but not in forbs and shrubs. Blankenship et al. (1982) maintained inocula donors (with the exception of deer) on a pelleted diet and found species differences in in vitro digestion of forages. Goat and cattle inocula were the most efficient digestors of grasses, whereas deer inoculum had the highest digestion of forbs, shrubs, and cactus. The deer were the most efficient digestor overall, but this may have been due to their maintenance diet. Interaction seemed evident in most cases, but was not tested. In conclusion, inocula obtained from cattle, goats, and nilgai fed a pelleted ration appeared to be of similar digestive efficiency. Deer inocula seemed less efficient than other ruminant inocula, and this difference could be attributed to differences in ability to digest fiber. The species of donor animal also may affect the in vitro digestion of forages. The use of results acquired from inoculum of one animal species, where these factors may be different, to predict in vivo digestibility in another may be erroneous. Because diet selection and digestive physiology of free-ranging ruminants vary, modification of procedures used in this technique may be necessary to obtain accurate estimates. Modifications, such as the adjustment of in vitro fermentation time, selection of an inocula donor with similar dietary habits and digestive physiology, and maintenance of Table 2.— Mean (+ SE) in vitro dry (DMD) and organic matter digestion (OMD) percent of five substrates by inocula obtained from cattle, goats, nilgai COMPARATIVE IN VITRO DIGESTION 345 h_i c 3 3 VO n d — ' — — d? m VO — 1 q q q o OO Ov d di d‘ d vd- 3" vd- -3- vt || rn oo Ov oo — < i/5 c5 © d ' — ' Q m o oo vd- q O d od r5 OO O ;> in V~) ■3- vd vt d vd- q q q d © — — — ' oo O © q q vd; C/3 — o — d — I ‘o J3 O T3 © d in Vi 3 — V 3 (U 3 V- 3 3 vq oo di q od q in q oo 3 c/3 >v dl o q £ o C/3 X5 3 C/3 O o _c Ov Ov ov vO oo _o o > vq 3; OO q >n _o i— _ i o 3 3 — ' -4 d •"d- — .2 e o C/3 c Q Q q _ q q ‘C cd 3 cd £ O O VO vo vo in vo <-> (U CD T3 T3 3 O .£ 3 C 3 C 3 .— id C/3 (Z) f S< William J . Clark Texas Academy of Science Drawer H6 College Station, Texas 77844 KNOWN BONDHOLDERS, MORTGAGEES, AND OTHER SECURITY HOLDERS OWNING OR HOLDING 1 PERCENT OR MORE OF TOTAL AMOUNT OF BONDS, MORTGAGES OR OTHER SECURITIES (If there are none, so state) COMPLETE MAILING ADDRESS FOR COMPLETION BY NONPROFIT ORGANIZATIONS AUTHORIZED TO MAI L AT SPECIAL RATES (Section 423.1 2 DMM only) The purpose, function, and nonprofit status of this organization and the exempt status for Federal income tax purposes ( Check one) □ HAS NOT CHANGED DURING PRECEDING 12 MONTHS □ HAS CHANGED DURING PRECEDING 12 MONTHS (If changed, publisher must submit explanation of change with this statement. ) EXTENT AND NATURE OF CIRCULATION (See Instructions on reverse side) AVERAGE NO. COPIES EACH ISSUE DURING PRECEDING 12 MONTHS ACTUAL NO. COPIES OF SINGLE ISSUE PUBLISHED NEAREST TO FILING DATE A. TOTAL NO. COPIES (Net Press Run) 1 , 100 1,100 B. PAID AND/OR REQUESTED CIRCULATION 1 . Sales through dealers and carriers, street vendors and counter sales 2. Mail Subscription (Paid and/or requested) 792 C. TOTAL PAID AND/OR REQUESTED CIRCULATION (Sum of 1 0B1 and 10B2) 792 D. FREE DISTRIBUTION BY MAIL, CARRIER OR OTHER MEANS SAMPLES. COMPLIMENTARY. AND OTHER FREE COPIES E. TOTAL DISTRIBUTION (Sum of C and D) 797 797 F. COPIES NOT DISTRIBUTED 1. Office use, left over, unaccounted, spoiled after printing 2. Return from News Agents G. TOTAL (Sum of E, FI and 2 -should equal net press run shown in A ) I certify that the statements made by me above are correct and complete PS Form 3526, Dec. 1985 1,100 THE TEXAS ACADEMY OF SCIENCE, 1986-87 Officers President: President-Elect: Vice-President: Immediate Past President: Executive Secretary: Treasurer: Editor: A AS Council Representative: Lamar Johanson, Tarleton State University Owen T. Lind, Baylor University Glenn Longley, Southwest Texas State University Billy J. Franklin, Lamar University William J. Clark, Texas A&M University Michael J. Carlo, Angelo State University J. Knox Jones, Jr., Texas Tech University Ann Benham, University of Texas at Arlington Directors 1985 George B. McClung, San Angelo Barbara Schreur, Texas A&I University 1986 Caroline P. Benjamin, Southwest Texas State University R. John Prevost, Southwest Research Institute 1987 John P. Fackler, Jr., Texas A&M University David R. Gattis, Freese and Nichols, Inc. Sectional Chairpersons Bio- Medical Science: David Eldridge, Baylor University Botany. O. W. Van Auken, University of Texas at San Antonio Cellular and Molecular Biology: R. R. Eller, Tarleton State University Chemistry: Jerry Darsey, Tarleton State University Computers in Science and Teaching: John A. Ward, Incarnate Word College Conservation: Dennis B. Fenn, Texas A&M University Environmental Science: Thomas A. Driscoll, U.S. Environmental Protection Agency Freshwater and Marine Sciences: Augustine de la Cruz, Texas Water Commission Geology: John L. Russell, Texas A&I University Mathematics: Philip J. Morey, Jr., Texas A&I University Physics: Joseph Pizzo, Lamar University Science Education: Ann Benham, University of Texas at Arlington Sociology: Ronald Cox, San Antonio College Systematics and Evolutionary Biology: Kenneth T. Wilkins, Baylor University Terrestrial Ecology and Management: George Newman, Hardin-Simmons University Counselors Collegiate Academy. Shirley Handler, East Texas Baptist College Helen Oujesky, University of Texas, San Antonio Junior Academy. Ruth Spear, San Marcos Peggy Carnahan, San Antonio THE TEXAS JOURNAL OF SCIENCE USPS 616740 Box 4240, Texas Tech University Lubbock, Texas 79409-3151, U.S.A. 2nd CLASS POSTAGE PAID AT LUBBOCK TEXAS 79401 BOUND PRINTED MATERIAL RETURN POSTAGE GUARANTEED 3024 I a? Library Acquisitions Smithsonian Institute Washington DC 20560 iiyvnan libraries smithsonian institution NoiiniusNi nvinoshiiw rn — xiouxs^x m CO — CO — CO ST1TUTION NOlinillSNI NVINOSH1IWS S3iava8ll LIBRARIES SMITHSONIAI m z 09 z > v 2 CO Z CO navaan libraries smithsonian institution NoiiniusNi nvinoshhiai co :z co STITUTION NOlinillSNI NVIN0SH1IIAIS S3iavaail LIBRARIES SMITHSONIA oo 70 > m VW ^ XlVASV^r zz co £ co n 3iavaan libraries smithsonian institution NoiiniusNi nvinoshim Z _ z CO z > ' ^ _ CO CO Z CO STITUTION NOlinillSNI _ NVIN0SH1I INS S3iava8ll LIBRARIES SMITHSONIA CO -» \ co _ ~ CO co 3 lava a n libraries smithsonian institution NoiiniusNi nvinoshiiw r- z r- _ Z r~ o £ Ip ^STITUTION^ NOIinillSNI~NVINOSHlllNS S3 i a V a a ll~L I B r a r i es^smithsonia CO Z _ CO z 2 > '\ UJ CO jvaan libraries smithsonian institution NoiinmsNi nvinoshiiims m ITUTION °° NOlinillSNl” NVINOSHIIIMS S3 1 H Vd 8 I "1~L 1 B R AR 1 E S ^SMITHSONIAN CO Z ^ zE /• ^ rE < 21 -h / a/ ^v> <£- A\\ \\ *h 2 1 x . ^ * ■''/#, o i ($£ry%$\ o co x/ o ^ 5 > '^|Pr 2 ^ 2 ‘ ■ z co z co * 2! oo jvaan libraries smithsonian institution NouniiiSNi nvinoshiiims — co =; _ co _ — CO v^v^rx co CO q: < XI ^£7 m k? m X^yos^z m co — co _ co IVdaH LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIMS 1 z co _ _ z * co 2 c/> A < ^ 1 < * I X co O z W > x<^>- 2 • \X' > CO Z CO ^ Z CO TUTION NOlinillSNI NVINOSHIIIMS S3IBVBail LIBRARIES SMITHSONIAN “ < O _ - co o; < CD 'X o X^osvavz — O Z _j 2 _j Z dvaan libraries smithsonian institution NouniiiSNi nvinoshiiims 2 r‘ z m r; x^uu^jx rn co ™ co ± co ITUTION NOI X_“SN! NVINOSHIIIMS S3iava9ll LIBRARIES SMITHSONIAN CO Z x i fi * > > 'W** S X\A > Bvaan libraries smithsonian institution NouniiiSNi nvinoshiiims^ “ CO X CO — CO ~ ^ tili £