I i+X THE TEXAS JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. For more information regarding membership, student awards, section chairs and vice-chairs, the annual March meeting and author instructions, please access the Academy’s homepage at: www . texasacademy ofscience . org Dues for regular members are $30.00 annually; supporting members, $60.00; sustaining members, $100.00; patron members, $150.00; associate (student) members, $15.00; family members, $35.00; affiliate members, $5.00; emeritus members, $10.00; corporate members, $250.00 annually. Library subscription rate is $50.00 annually. The Texas Journal ofScience is a quarterly publication of The Texas Academy of Science and is sent to most members and all subscribers. Payment of dues, changes of address and inquiries regarding missing or back issues should be sent to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 E-mail: FStevens@schreiner.edu The Texas Journal of Science (ISSN 0040-4403) is published quarterly at Lubbock, Texas, U.S.A. Periodicals postage paid at San Angelo, Texas and additional mailing offices. POSTMASTER: Send address changes and returned copies to The Texas Journal of Science, Dr. Fred Stevens, CMB 5980, Schreiner University, Kerrville, Texas 78028-5697, U.S.A. The known office of publication for The Texas Journal of Science is the Department of Biology, Angelo State University, San Angelo, Texas 76909; Dr. Ned E. Strenth, Managing Editor. COPYRIGHT POLICY All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, recording or otherwise, without the prior permission of the Managing Editor of the Texas Journal ofScience. THE TEXAS JOURNAL OF SCIENCE Volume 56, No. 1 February, 2004 CONTENTS Growth and Survival of Juniperus ashei (Cupressacae) Seedlings in the Presence of Juniperus ashei Litter. By Duncan McKinley and O. W. Van Auken . 3 The Vascular Flora of the Palo Alto National Battlefield Historic Site, Cameron County, Texas. By Robert I. Lonard, Alfred T. Richardson and N. L. Richard . 15 Spatial and Temporal Abiotic Changes along a Canopy to Intercanopy Gradient in Central Texas Juniperus ashei Woodlands. By Rob Wayne and O. W. Van Auken . 35 Reproductive Cycle of the Sidewinder, Crotalus cerastes (Serpentes: Viperidae), from California. By Stephen R. Goldberg . 55 Freshwater Mussels (Bivalvia: Unionidae) of the Village Creek Drainage Basin in Southeast Texas. By Vickie L. Bordelon and Richard C. Harrel . 63 General Notes Noteworthy Records of the Millipeds, Eurymerodesmus angularis and E. mundus (Polydesmida: Eurymerodesmidae), from Northeastern and Westcentral Texas. By Chris T. McAllister, Rowland M. Shelley and Dawn /. Moore . 73 Diet of the White-collared Seedeater Sporophila torqueola (Passeriformes: Emberizidae) in Texas. By Jack C. Eitniear . 77 Reproduction in the Coffee Snake, Ninia maculata (Serpentes: Colubridae), from Costa Rica. By Stephen R. Goldberg . 81 Author Instructions 85 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Managing Editor: Ned E. Strenth, Angelo State University Manuscript Editor: Robert J. Edwards, University of Texas-Pan American Associate Editor for Botany: Janis K. Bush, The University of Texas at San Antonio Associate Editor for Chemistry: John R. Villarreal, The University of Texas-Pan American Associate Editor for Computer Science: Nelson Passos, Midwestern State University Associate Editor for Environmental Science: Thomas LaPoint, University of North Texas Associate Editor for Geology: Ernest L. Lundelius, University of Texas at Austin Associate Editor for Mathematics and Statistics: E. Donice McCune, Stephen F. Austin State University Associate Editor for Physics: Charles W. Myles, Texas Tech University Manuscripts intended for publication in the Journal should be submitted in TRIPLICATE to: Dr. Robert J. Edwards TJS Manuscript Editor Department of Biology University of Texas-Pan American Edinburg, Texas 78541 redwards@panam.edu Scholarly papers reporting original research results 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 Manuscript Editor at the above address. They are also available on the Academy’s homepage at: www . texasacademy ofscience . org AFFILIATED ORGANIZATIONS American Association for the Advancement of Science, Texas Council of Elementary Science Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers Texas Society of Mammalogists TEXAS J. SCI. 56(1):3-14 FEBRUARY, 2004 GROWTH AND SURVIVAL OF JUNIPERUS ASHE1 (CUPRESSACAE) SEEDLINGS IN THE PRESENCE OF JUNIPERUS ASHE1 LITTER Duncan McKinley* and O. W. Van Auken Department of Biology The University of Texas at San Antonio 6900 North Loop 1604 West San Antonio Texas, 78249-0661 * Current address: Division of Biology Kansas State University Manhattan, Kansas 66506 Abstract.— A greenhouse experiment was conducted to determine the effect of Juniperus ashei litter on the growth and survival of J. ashei seedlings. Incremental additions (0-250 g) of J. ashei tree litter or vermiculite (control) were placed on 15 by 15 cm pots, which contained transplanted J. ashei seedlings in 800 g of mineral soil. There were no significant differences in the mean absolute differences in growth of J. ashei seedling considering basal diameter, seedling height and number of branches between the J. ashei tree litter additions and the vermiculite additions, or the amounts of both types of litter. However, there were non-significant positive increases in the seedling growth in the 50 g treatment of both litter types followed by a decrease at higher levels. Mortalities were highest at greater levels of both types of litter, but were still non-significant. The responses of the J. ashei seedlings with respect to growth and survival in the J. ashei litter and vermiculite suggest that there is no allelopathic component in the J. ashei litter affecting seedling growth and survival or if there is, it is transient. Juniperus ashei is an evergreen, aromatic, dioecious, non-sprouting shrub or small tree (Correll & Johnston 1979). It is usually found on calcareous, rocky, shallow soils from southern Missouri and northern Arkansas through Oklahoma, Texas and parts of northern Mexico (Little 1979; Simpson 1988; Hart & Price 1990; Fuhlendorf et al. 1997). Fourteen species of Juniperus have been identified in North America (Little 1979), with over 60 species found worldwide, mostly in semi-arid northern hemisphere ecosystems (Dallimore & Jackson 1967). Various species of Juniperus now cover approximately 10 million hectares in Texas (Gould 1969). Juniperus ashei is a dominant species of many savannahs and woodlands of the Edwards Plateau of central Texas (Van Auken et al. 1980). Estimated density of J. ashei in central Texas ranges from approximately 700 trees ha 1 to 1500 trees ha 1 (Van Auken et al. 1979; Smeins 1990). Evidence suggests that J . ashei, as well as some other species of 4 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Juniperus, have increased in density since European settlement by encroachment into adjacent grasslands (Buechner 1944; Smeins 1980; Fuhlendorf et al. 1996; Van Auken 2000). Historically, 7. ashei was apparently restricted to canyons, rocky outcrops or areas with shallow soils, which were protected from grassland fires (Ellis & Schuster 1968). The most widely cited explanation for woody plant encroachment attributes the shifts in community types to a concomitant reduction in fire frequency and decreased competition from grasses, both of which are promoted by heavy grazing by domestic ungulates (Neilson 1986; Archer et al. 1988; Schlesinger et al. 1990; Bashre 1991; Van Auken 2000). There are many reports of allelopathic effects of litter or litter extracts on various understory species, including woody plant seedlings (Rice 1984). Suppression of understory vegetation by 7. osteosperma is commonly reported in New Mexico and Arizona (Arnold et al. 1964) and 7. virginiana and 7. pinchott may reduce herbaceous cover and diversity (Arnold et al. 1964; Engle et al. 1987; Armentrout & Pieper 1988). Juniperus monosperma litter seems to have a negative effect on the growth of Bouteloua gracilis (blue grama) (Jameson 1966; Jameson 1970b). In addition, reduction of herbaceous vegetation has been reported below Juniperus canopies even after canopy removal (Bonnett 1960; Jameson 1966; Jameson 1970b; Carson 1990; Barnes & Archer 1996). However, the cause of the apparent allelopathic effects is unclear. Juniperus ashei has been observed with a zone of reduced herbaceous cover and diversity beneath the crown near the stem (Blomquist 1990; Fuhlendorf 1992). In closed-canopy stands, J. ashei like other Juniperus sp. can exclude most herbaceous vegetation (Buechner 1944; Johnsen 1962; Burkhart & Tisdale 1969; Yager & Smiens 1999). How¬ ever, there are some places below the canopy that Car ex pianos tacky s (cedar sedge) has high cover (Wayne 2000; Wayne & Van Auken 2002). Juniperus ashei tree litter was demonstrated to have negative effects on seedling recruitment and germination of some herbaceous species including grasses, but negative effects were reduced or absent on a woody plant seedling ( Sophora secundiflora ) by Yager & Smeins (1999). In addition, litter apparently reduced the density of most woody and herbaceous species even after adult 7. ashei canopies were completely removed (Yager & Smeins 1999). However, 7. ashei seedlings have been observed to rapidly establish following the removal of the adult 50 yrs, Eric Lautzenheiser pers. comm.). The site is near the Balconies fault zone and approximately 5 km east of the University of Texas at San Antonio campus. A site was selected representative of a J. ashei woodland with an associated intercanopy 38 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 patch that appeared to be infrequently accessed by humans. Soil is a clayey-skeletal, smectitic, thermic lithic calciustoll (United States Department of Agriculture 2000) in the Tarrant association - rolling - with a slope of 4.5° to 13.5°. Three horizons occur that consist of shallow, clayey, weakly calcareous soil, developed over hard limestone with scattered stones and gravel. The surface horizon ranges from 0 cm to 25 cm in thickness. The subsurface is approximately 20 cm thick, heavily fractured limestone over limestone bedrock (Taylor et al. 1962). Regional climate is classified as subtropical - subhumid with a mean annual temperature of 20°C (Arbingast et al. 1976). Monthly mean temperature ranges from 9.6°C in January to 29.4°C in July (National Oceanic and Atmospheric Administration 1999). Annual precipitation in the study area is 78.7 cm, with two peaks occurring in May and September with monthly means of 10.7 cm and 8.7 cm, respectively. During the study, precipitation was above normal for 1997 at 85.6 cm (National Oceanic and Atmospheric Administration 1999), with a low of 0.0 cm in July, negligible in August, and a high of 18.5 cm in June. The area vegetation is juniper/oak woodland representative of similar woodlands found throughout this region (Van Auken et al. 1981). The predominant woody vegetation is J. ashei and Quercus virginiana (live oak). Other woody species reported from the area are Q. texana (Spanish oak), Celtis laevagata (hackberry), Diospyros texana (Texas persimmon), Berberis trifoliata (agarita) and Rhus virens (evergreen sumac) (Van Auken et al. 1980; 1981; Terletzky & Van Auken 1996). Car ex planostachys (Correll & Johnston 1979) was the dominant herba¬ ceous species below the woodland canopy. The major herbaceous species in the inter canopy patches were Aristida longiseta (red three-awn), Bouteloua curtipendula (side-oats gramma), other C3 and C4 grasses and a variety of herbaceous annuals (Fowler & Dunlap 1986; Van Auken 2000a). Measurements of surface and subsurface soil moisture, soil tempera¬ ture, soil organic content and field capacity were made at each of five positions along six parallel northeasterly transects (41° azimuth). Frequency and time of measurements are indicated for each factor. The surface horizon of the soil was the upper 2 cm of soil and the subsurface horizon was the lowest 2 cm of soil adjacent to the bedrock. Each transect was 15 m in length and at least 3 m from an adjacent transect. A plumb line dropped from the outermost branch of mature 7. ashei trees (2 m above the ground, located directly above each transect) was used to locate the canopy edge (drip line). Surveyor tapes were used to establish the following sampling positions: 10 m inside the canopy WAYNE & VAN AUKEN 39 (canopy), 5 m inside the canopy (mid-canopy), 0 m inside the canopy (canopy edge), 2.5 m outside the canopy (mid- inter canopy) and 5 m outside the canopy (intercanopy). There were 6 transects by 5 sampling positions for the surface horizon and for the subsurface horizon. Significant differences in soil moisture and soil temperature were detected between the surface and subsurface horizons (ANOVA, SAS Institute 1989). Because the overall mean values between the surface and subsurface were small (< 2°C for soil temperature and < 5% for soil moisture) surface measurements will be the main focus of this paper. Soil moisture was determined using the gravimetric procedure and reported as the percent water in the sample on a dry-mass basis (Pearcy 1989; United States Department of Agriculture 1996; Jackson et al. 2000). Soil samples were collected along each transect (n = 6), at each position (n = 5) for the surface and bedrock horizons (n = 2) in April, May, July, August, September, October, and twice in December (n = 8 for a total of 480 samples). Stones and organic litter were removed from the soil surface; soil samples were collected and sealed in plastic bags for transport to the lab. Approximately 40 g of soil was placed in a pre-weighed aluminum planchet, weighed and oven dried at 100°C to a constant mass. Soil temperature was measured within two hours after solar noon on the same dates as soil moisture (with the exception of May and the latter December measurement (n = 6 months for a total of 360 samples) using 15 cm long, probe type, analog soil thermometers (Broadbent 1965; Larcher 1995). Surface temperature was measured by inserting the probe 1 to 2 cm into the soil and recording the temperature after five minutes of equilibration. Subsurface temperature was measured by excavating soil to the bedrock and inserting the probe into the lowest 2 cm of exposed soil. Surface light levels (photosynthetically active photon flux density, X = 400 to 700 run,) were measured at solar noon on cloudless days in July, August, October and December (n = 4 months for a total of 120 samples) with a LI-COR® (LI-COR Inc., Lincoln, Nebraska) LI- 190 SA integrating quantum sensor. Light levels were recorded with a LI-COR® LI- 1000 data logger in instantaneous mode with 60 s averaging at 5 s intervals. No measurements were made April through June 1997 because of overcast conditions. The quantum sensor was placed level on bare ground at each position and no attempt was made to move or disrupt any woody or herbaceous vegetation over the sensor. 40 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 The soil depth profile was measured at the conclusion of this study to minimize potential disturbances to the plants and soil of the study area (Broadbent 1965). Surface litter was removed and measurements were made along each transect at 0.5 m intervals ( n = 186) using a 60 by 1 cm rebar driven vertically into the ground until it would not penetrate any deeper. The distance from the top of the rod to the ground was measured and subtracted from 60 cm to obtain the soil depth. Periodic¬ ally, the rebar was re- measured to ensure the length did not change. Percent soil organic content was determined for the surface and bedrock horizons (n = 2 for a total of 60 samples) using the loss-on- ignition procedure (Broadbent 1965; United States Department of Agri¬ culture 1996). Excess soil collected from the December 1997 soil mois¬ ture sampling was used for the determination of the soil organic content. The soil was air-dried and sieved (#10 mesh), tested for the presence of carbonates (United States Department of Agriculture 1996), oven dried at 90 °C and incinerated in a Fischer Muffle Furnace (Model 58) at 600 °C for 3 hours. The test for presence of carbonates was negative. Determination of percent field capacity (Broadbent 1965) for the surface and bedrock horizon was made using sieved (#10 mesh), air- dried soil, however only four transects were utilized (n = 2 for a total of 40 samples). The soil was placed level into a perforated aluminum planchet lined with # 1 filter paper, thoroughly wetted for 12 h and drained for 20 minutes. The soil was then oven dried to a constant mass at 100°C. The experimental design was factorial for surface light, soil water and soil temperature (position by date). Data were transformed as needed prior to statistical analysis and analyzed with ANOVA (SAS Institute 1989). When significant main effects were detected, data were subset to examine temporal and spatial differences using ANOVA and the Scheffe multiple comparison test (a = 0.05, SAS Institute 1989). Mean surface values were pooled temporal data (all dates) for each transect position to show the overall spatial differences in surface values. Although ANOVA may indicate that a significant difference occurred the Scheffe multiple comparison test may indicate otherwise because of its conservative nature in computing the minimum significant difference (three examples occurred, SAS Institute 1989; Sokal & Rohlf 1995). Results Soil depth was erratic and did not vary significantly from the canopy to the intercanopy patch ( F = 0.69, P = 0.8858, Fig. 1). Mean soil depth (+ SE) ranged from 9.9 ± 2.3 cm under the full canopy to 7. 1 WAYNE & VAN AUKEN 41 O c/> -15 -10 -5 0 5 10 TRANSECT POSITION (m) Figure 1. Mean soil depth profile (surface to bedrock, cm) measured at 0.5 m intervals along the canopy to intercanopy gradient (n = 6 transects) in the Juniperus ashei woodland. Lower bar with dotted line is an example standard error bar. Transect position (x-axis) is in meters from the canopy edge: canopy (-10), mid-canopy (-5), canopy edge (0), mid-intercanopy (2.5) and intercanopy (5). P- value for the AN OVA indicated no significant difference in positions. -15 -10 -5 0 5 10 TRANSECT POSITION (m) Figure 2. Spatial differences in mean (± SE) percent soil organic content and percent field capacity at the surface horizon. P-values indicated are for individual ANOVA’s. Transect position (x-axis) is in meters from the canopy edge: canopy (-10), mid-canopy (-5), canopy edge (0), mid-intercanopy (2.5) and intercanopy (5). Means within a measured parameter with different letters are significantly different (Scheffe multiple comparison test) . ± 2.1 cm at the canopy edge and 10.6 ± 3.1 cm in the intercanopy patch. Soil depth ranged from zero to 40 cm and the overall mean depth was 9.2 ± 2.5 cm. Overall mean soil organic content varied significantly by position (F = 8.59, P = 0.0001) and ranged from 32.0 ± 6.9% under the full canopy (Fig. 2) to 16.8 ± 2.6% at the canopy edge and 12.5 + 0.8% 42 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 1, 2004 TRANSECT POSITION (m) Figure 3. Yearly mean (± SE) surface gradient (n = 6 transects) from below the Juniperus canopy into the intercanopy (n = 5 positions) for (a) surface light levels (junol • m"2 • s'1) and (b) surface soil temperature (°C) and surface soil moisture (%). Light levels were measured at solar noon on cloudless days in July, August, October and December 1997 (n = 4). Soil temperature was measured within two hours after solar noon in April, July, August, September, October and December ( n = 6). Soil moisture was measured in April, May, July, August, September, October and twice in December (n = 8). Transect position (x-axis) is in meters from the canopy edge: canopy (-10), mid-canopy (-5), canopy edge (0), mid-intercanopy (2.5) and intercanopy (5). in the intercanopy patch. The Scheffe multiple comparison test indicated there was a significant difference in mean soil organic content between the canopy position and both patch positions, but no significant differ¬ ence between the mid-canopy and the canopy edge positions. Overall field capacity varied significantly by position (F = 31.90, P = 0.0001) and ranged from 108.5 + 2.8% under the Juniperus wood¬ land canopy (Fig. 2) to 81.3 ± 2.9% at the canopy edge and 82.9 ± 1.6% in the intercanopy. The Scheffe multiple comparison test indicated that there was not a significant difference between the canopy and mid- WAYNE & VAN AUKEN 43 Table 1. F-tables and significance levels from three separate analyses of variance, examining (a) light levels, (b) soil temperature and (c) % soil moisture. Variables examined include the overall model, date (D), transect position (P), soil horizon (H) and the various two and three-way interactions. Transect positions are canopy, mid-canopy, canopy edge, mid-intercanopy patch and intercanopy patch. * = P < 0.05, ** = P < 0.01, *** = P < 0.001, **** = P < 0.0001 and NS = not significantly different. (a) Light levels. (b) Soil temperature. (c) Soil moisture. Source df F Source df F Source df F Model 19 6.67**** Model 59 36.90**** Model 79 23.77**** Date (D) 3 13.92**** Date (D) 5 365.00**** Date (D) 7 218.81**** Position (P) 4 16.37**** Horizon (H) 1 78.55**** Horizon (H) 1 1.32ns D*P 12 1.62NS Position (P) 4 19.80**** Position (P) 4 39 28**** D*H 5 2.69* D*H 7 9.99**** D*P 20 8.38**** D*P 28 5.66*** H*P 4 0.88ns H*P 4 2.86**** H*d*P 20 0.49ns H*D*P 28 0.53ns canopy positions but they differed from all other positions. There was no significant difference between means for the canopy edge and the intercanopy positions. The overall trend in surface light levels, soil temperature and soil moisture are best observed by pooling all surface temporal data for each position (Fig. 3). Mean surface light levels varied significantly by date and position, but the interaction term was not significant (Table la). Spatially, surface light levels (Fig. 3a) were lowest below the canopy and mid-canopy positions, 346 ± 99 /xmol • m"2 • s'1 and 219 ± 77 fxmol • m'2 • s'1 respectively, were intermediate at the canopy edge and highest in the intercanopy (1183 ± 149 fjanol • m2 • s'1) . Mean soil temperature varied significantly by date, horizon, and position, with two significant two-way interactions (Table lb). The significant interactions were date by horizon and date by position, but the three-way interaction was not significant. Spatially, mean yearly surface temperatures (Fig. 3b) were lowest at the canopy edge (27.6 ± 1.4°C), intermediate below the canopy (29.5 ± 1.8°C) and highest in the intercanopy (32.6 ± 2.1°C). Mean soil moisture varied significantly by date and position, with 3 significant two-way interactions (Table lc). The three-way inter¬ action was not significant. The general spatial trend for surface soil moisture (Fig. 3b) was highest values below the canopy (43.4 ± 3.0%), intermediate values at the canopy edge (33.6 ± 2.2%) and lowest values in the intercanopy (30.3 ± 2.1%). Surface light below the canopy did not vary significantly (F = 1.98, 44 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Figure 4. Temporal change in mean (± SE) surface light levels (jnmol«m'2*s‘‘, n = 6 transects) below the canopy, at the canopy edge and in the intercanopy. Surface light was measured at solar noon on cloudless days in July, August, October and December 1997 (i n = 4). Significance levels are indicated to the right of each position in the legend: NS is not significantly different, * is P < 0.05. P > 0.05) and ranged from 675 ± 309 /xmol • m'2 • s'1 in July (Fig. 4) to 39 ± 7 /xmol • m2 • s1 in December. At the canopy edge, surface light varied significantly (F = 3.37, P < 0.05) and ranged from 666 ± 307 /xmol •m‘2«s"1 in July to 78 ± 17 /xmol • m*2 • s"1 in December; however, the Scheffe multiple comparison test did not detect any significant differences between dates. In the intercanopy, surface light varied signi¬ ficantly (F = 6.88, P < 0.05) ranging from 1614 ± 302 /xmol • m2- s"1 in July to 479 + 225 /xmol • m"2 • s"1 in December. The August mean of 1531 ± 243 /xmol • m'2 • s'1 was significantly different from the October and December means (Scheffe multiple comparison test), but not the July mean. Temporal differences in mean surface temperature below the canopy varied significantly (F = 41.37, P = 0.0001) and ranged from 25.6 ± 1 .9°C in May (Fig. 5a) to a high of 46.5 ± 3.3°C in August and a low of 16.0 ± 0.3°Cin December. Mean surface temperature at the canopy edge varied significantly (F = 53.83, P = 0.0001) and ranged from 25.6 ± 0.7°C in May, increased to a high of 39.8 ± 2.2°C in July and a low of 16.3 ± 1.0°C in December. In the intercanopy, mean surface temperature varied significantly (F = 32.66, P = 0.0001) from 31.0 ± 0.7 °C in May to a high of 48.8 ± 1.0°C in July and a low of 18.1 ± 0.8°C in December. Surface soil temperatures followed air tempera¬ tures (with a lag) and were high in July and August, and low in WAYNE & VAN AUKEN 45 o o LU a : D H 2 Ui Q. 2 ui O w 6-Mar 25-Apr 14-Jun 3-Aug 22-Sep 11 -Nov 31 -Dec MONTH-1997 Figure 5. Temporal change in (a) mean (± SE) surface soil temperature (°C, n = 6 transects) and (b) mean (± SE) surface soil moisture (%, n = 6 transects) below the canopy, at the canopy edge and in the intercanopy. Temperature measurements were made within two hours after solar noon in April, July, August, September, October and December 1997 ( n = 6). Soil moisture was measured in April, May, July, August, September, October and twice in December ( n = 8). Significance levels are indicated to the right of each position in the legend: * is P = 0.0001. December. The highest surface soil temperature was 48.8 ± 1.0 °C in July in the intercanopy and the lowest was in December at 16.0 ± 0.3°C under the canopy. A significant decline from the high soil temperatures seen in July and August for all positions occurred in early September (« 12 °C), coinciding with a 0.8 cm precipitation on the day preceding temperature measurements. After September soil temperature continued a significant decline to the low values observed in December for all positions except the intercanopy. Temporal differences in mean surface soil moisture varied signifi- 46 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 cantly below the canopy (F = 16.94, P = 0.0001) and ranged from 68.4 ± 7.0% in May (Fig. 5b) to a low of 18.9 ± 2.1% in July. Fol¬ lowing the September precipitation, soil moisture increased to 52.2 ± 2.1%, followed by a second, but significant decline, and subsequent significant increase to 55.6 ± 6.8% after a late December precipitation event. The canopy edge and intercanopy locations also varied signifi¬ cantly (F = 42.5, P = 0.0001 and F = 84.2, P = 0.0001) and with the same significant decreases and increases seen below the canopy location. The canopy edge was at 47.7 ± 3.8% in May, decreased to a low of 12.5 ± 1.7% in August, increased to 39.5 ± 1.5% in September and was at 51.0 ± 2.4% in December. In the intercanopy, mean soil moisture was 43.2 ± 3.3% in May, declined to 6.8 + 0.4% in July, increased to 38.6 ± 2.3% in September and was at 43.2 ± 2.0% in December. The overall temporal trend was high surface soil moisture in April-May and low surface soil moisture in June-August. Discussion Soil depth in this study did not indicate a gradient from canopy to intercanopy locations. The very erratic soil depth observations from the Juniperus woodland canopy into the intercanopy patch were likely due to numerous surficial bedrock fractures (Davenport et al. 1996). At the northeastern extent of 7. asheV s range, calcareous derived soils are prevalent with rock outcrops common as well as fractures and pockets of deep soil (Quarterman et al. 1993; Ware 2002). These findings in 7. ashei woodlands are not unlike those of Pinus edulus /Juniperus monosperma communities of New Mexico where soil depth fluctuated from 33 to 125 cm over distances of 10 m and without any significant differences between canopy and intercanopy locations (Davenport et al. 1996). Other 7. monosperma communities such as those in Arizona (Johnsen 1962) and 7. pinchotii in north Texas (McPherson et al. 1988) also occur over fractured bedrock. A similar trend of shallow soils over fractured bedrock has been reported for other locations in the Edwards Plateau (Foster 1917; Taylor et al. 1962; Owens & Schreiber 1992). However, gradients of soil depth have been reported in open patch communities in central Texas (Van Auken 2000a) and deeper soils have been confirmed in woodlands compared to intercanopy patches in this same area (Terletzky & Van Auken 1996; Ware 2002). Specific spatial abiotic gradients were found during this study for soil organic content, field capacity, surface light levels, soil temperature and soil water content. The general trend was a decrease in soil organic WAYNE & VAN AUKEN 47 content, field capacity, and soil water content from beneath the Juniperus canopy into the intercanopy patch. Surface light and soil temperature followed a reverse trend with high surface light levels and high soil temperatures in the intercanopy patch and lower values beneath the woodland canopy. Temporal differences in surface light, soil temperature and soil moisture were not presented for the mid-canopy and mid- inter canopy positions. However it was noted when examining individual dates the mid-canopy differed little from the canopy, and the mid-intercanopy differed little from the intercanopy (see Wayne 2000). While surface litter, derived from the overstory, was not measured during this study it does have an influence on soil moisture content as it is incorporated into the soil (Knapp et al. 1993; Breshears et al. 1997b). It was noted that surface litter at the study site was ~ 3 - 5 cm thick below the canopy, thin at the canopy edge, and absent in the inter¬ canopy. In addition, the trend in soil organic content appears to coin¬ cide with areas of litter deposition and greater litter depth. High amounts of organic matter have a direct relationship with the soil water holding capacity and soil field capacity (Bel sky & Canham 1994; Larcher 1995; Jackson et al. 2000). An additional characteristic of surface litter is that it insulates the soil from atmospheric temperature (Knapp et al. 1993; Breshears et al. 1998). It was demonstrated that soil organic content was low or absent in the intercanopy and increased from the canopy edge into the full canopy position. Similar trends in soil organic content and litter have been noted in African savannas with high levels found proximal to overstory trees (Belsky et al. 1989; 1993). In addition, the same has been found in J. pinchotti communities on the northern Edwards Plateau (Dye 1993; Dye et al. 1995) and west Texas (McPherson et al. 1991), pinon/juniper communities in New Mexico (Davenport et al. 1996) and other savanna communities (Belsky & Canham 1994). Surface light levels were reduced beneath the Juniperus woodland likely due to light interception by the overstory canopy. This light reduction has been reported in other J. ashei communities on the Edwards Plateau (Yager & Smeins 1999), in oak savannas on the Edwards Plateau (Anderson et al. 2001), in J. monosperma communities in New Mexico (Breshears et al. 1997b; 1998; Martens et al. 2000) and in J . virginiana communities in the eastern North America (Joy & Young 2002). In pinon/juniper communities, differences in surface light levels are related mainly to canopy/ intercanopy patch variation (i.e., overstory/no overstory) (Breshears et al. 1997b). Differences in light 48 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 levels are not only spatial trends but temporal trends as well; and spatial effects are modified temporally. Light levels in pinon/juniper communities varied less temporally beneath the canopy than in the intercanopy patch, but the observed temporal differences were greatest during summer and least during winter. In J. ashei communities (this study), the spatial /temporal trends in light levels are similar to those reported in the Juniperus communities in New Mexico. Temporally light was highest during summer and reduced in winter. Light levels were higher in the intercanopy patch, intermediate at the canopy edge and lower in the canopy positions, which is consistent with pinon/juniper communities in western North America. Soil temperatures from the canopy to the inter canopy patch followed a trend similar to the surface light gradient, lower soil temperatures below the canopy and highest temperatures in the intercanopy patch. This is consistent with J. monosperma communities in New Mexico (Breshears et al. 1997a) and J. virginiaia communities in eastern North America (Joy & Young 2002). Reduced canopy soil temperature is probably related to the interception of light by the canopy reducing heating of the soil by solar radiation (Helgerson 1990; Belsky et al. 1993; Breshears et al. 1997b). In addition, surface litter probably provides insulation of the soil from atmospheric temperature (Knapp et al. 1993; Breshears et al. 1998). Conversely, the higher soil tempera¬ tures in the intercanopy patch are influenced by the lack of overstory shading and absence of surface litter (Breshears et al. 1998). Soil moisture was also higher below the Juniperus canopy and may also play a role in the reduced canopy soil temperatures. High soil moisture also appears to ameliorate high soil temperatures across the entire gradient as noted following small precipitation events (Berndtsson et al. 1996; Wayne & Van Auken 2002). A specific temporal trend of variable soil temperature was also detected. Peak soil temperatures across the study site were reached in late August; these high temperatures were subsequently modified, ~ 20 °C, by a small precipitation event (0.8 cm) in early September followed by a continued seasonal decline, ~ 10°C, from fall through winter. In addition, during fall and winter there was little difference in mean soil temperature along the gradient (see Wayne 2000). Pinon/ juniper woodlands in New Mexico followed a similar temporal trend where soil temperatures were elevated in the intercanopy patch (relative to the canopy) during the summer and decline fall through winter (Breshears et al. 1998). Differences were attributed to seasonal air WAYNE & VAN AUKEN 49 temperatures and the changing angle of the sun. Trends in soil moisture along the canopy to intercanopy patch gradient were reversed from that described for surface soil temperatures, soil moisture was highest below the canopy and reduced in the intercanopy patch. The exception to this trend was noted after precipitation events when differences between positions were not apparent. Possible causes for differences in soil moisture have been mentioned previously; includ¬ ing the canopy intercepting light resulting in reduced soil temperatures and also the high litter content below the canopy further ameliorating evaporative loss (Yager & Smeins 1999; Anderson et al. 2001; Joy & Young 2002). Some pinon/juniper woodlands (Breshears et al. 1997a; 1997b; 1998) and oak savannas (Anderson et al. 2001) have lower soil moisture below the canopy and canopy edge then the adjacent patch, but it is unclear whether this was due to canopy interception of rainfall and/or evapo- transpiration. With regard to pinon/juniper woodlands the soil moisture trend varies with time such that either patch type, canopy or inter¬ canopy, can have increased soil moisture at some point during the year (Breshears et al. 1997b). Thus, these central Texas Juniperus wood¬ lands were dissimilar from those in New Mexico that had mostly higher soil moisture in the intercanopy. High soil organic content and litter cover below the canopy may account for greater water storage capacity (measured as field capacity, Fig. 2). Runoff during rainfall from small intercanopy areas into canopy areas (Wilcox 1994; Ware 2002) may also increase soil moisture below the canopy and redistribute sediment (and litter) from the intercanopy into the canopy (Reid et al. 1999). Tem¬ porally, soil moisture was found to be decreased from spring into summer after cessation of rainfall (from ~ 53% to 13% soil moisture), but recharge occurred rapidly (from ~ 13% to 44% soil moisture) after small precipitation events (Wayne 2000; Wayne & Van Auken 2002). Throughout most of the year abiotic conditions at the canopy edge are intermediate (see Wayne 2000; Wayne & Van Auken 2002) to the canopy and patch positions. Differences in aboveground canopy cover appear to explain a considerable amount of the heterogeneity detected in abiotic factors along the gradients in these Juniperus woodlands (Breshears et al. 1997b). Soil depth was not significantly different in this study and does not seem to play a role in the abiotic gradients. Higher J . ashei seedling emergence and survival (Jackson & Van Auken 1997; Van Auken et al. 2004), and high predawn xylem water potential 50 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 below the canopy (Wayne & Van Auken 2002) seems related to the reduced stress attributable to slightly lower soil temperature and higher soil moisture. Thus, the canopy likely facilitates J. ashei in the early stages of its growth and development (Callaway et al. 1996; Joy & Young 2002). However, the canopy may also hinder J. ashei seedling growth due to light interception and reduced surface light levels (McKinley & Van Auken 2004), more so below the full canopy position then at the canopy edge. Reduced availability of water and increased soil temperature appears to hinder seedling emergence and survival, while at the same time the increased light likely promotes seedling growth (Van Auken et al. 2004). This anomalous statement appears to explain differences in survival and growth of 7. ashei seedlings in these different positions along the gradient. The intercanopy position exhibited the greatest soil temperature and lowest soil moisture, which seems to explain the low emergence and survival of J . ashei seedlings at this position along the gradient. Small precipitation events during late summer also appears to be important in reducing water stress of J. ashei , and other drought tolerant herbaceous species (see Wayne 2000) in these Juniperus com¬ munities (Fonteyn et al. 1985; Wayne & Van Auken 2002). Acknowledgements The authors wish to thank E. Lautzenheiser and others with the City of San Antonio Parks and Recreation Department for their cooperation, and for permission to carry out this study in Eisenhower Park. The support of W. and L. Collenback, through a generous scholarship to the senior author is most appreciated. In addition, grants provided by the University’s College of Science and Engineering, and the Division of Life Science to the senior author helped make this work possible. Last¬ ly, we wish to thank the Center for Water Research for their support in publishing this work. Literature Cited Anderson, L. J., M. S. Brumbaugh & R. B. Jackson. 2001 . Water and tree-understory interactions: a natural experiment in a savanna with oak wilt. Ecology, 82(l):33-49. Arbingast, S. A., L. G. Kennamer, J. R. Buchanan, W. L. Hezlep, L. T. Ellis, T. G. Jordan, C. T. Granger & C. P. Zlatkovich. 1976. Atlas of Texas. 5th Ed. Bureau of Business Research, University of Texas, Austin, 179 pp. Archer, S., D. S. Schimel & E. A. Holland. 1995. Mechanisms of shrubland expansion: land use, climate or C02. Climatic Change, 29(l):91-99. Baskin, J. M. & C. C. Baskin. 1978. Plant ecology of cedar glades in the Big Barrens WAYNE & VAN AUKEN 51 region of Kentucky. Rhodora, 80:545-557. Baskin, J. M. & C. C. Baskin. 2000. Vegetation of limestone and dolomite glades in the Ozarks and Midwest regions of the United States. Annl. Mo. Bot. Garden, 87:286-294. Berndtsson, R., K. Nodomi, H. Yasuda, T. Perrson, H. Chen & K. Jinno. 1996. Soil water patterns in an arid desert dune sand. J. Hydro., 185 (1-4): 22 1-240. Belsky, A. J. & C. D. Canham. 1994. Forest gaps and isolated savanna trees - An application of patch dynamics in two ecosystems. Bioscience, 44(2):77-84. Belsky, A. J., R. G. Amundson, J. M. Duxbury, S. J. Riha, A. R. Ali & S. M. Mwonga. 1989. The effects of trees on their physical, chemical, and biological environments in a semi-arid savanna in Kenya. J. Appl. Ecol., 26(3): 1005-1024. Belsky, A. J., S. M. Mwonga, R. G. Amundson, J. M. Duxbury & A. R. Ali. 1993. Comparative effects of isolated trees on their undercanopy environments in high- and low-rainfall savannas. J. Appl. Ecol., 30(1): 143-155. Bray, W. L. 1904. The timber of the Edwards Plateau of Texas: It’s relation to climate, water supply, and soil. United States Department of Agriculture, Bureau of Forestry Bulletin No. 47. Breshears, D. D., O. B. Myers, S. R. Johnson, C. W. Meyers & S, N. Martens. 1997a. Differential use of spatially heterogeneous soil moisture two semiarid woody species: Pinus edulus and Juniperus monosperma. J. Ecol., 85(3): 289-299. Breshears, D. D., P. M. Rich, F. J. Barnes & K. Campbell. 1997b. Overstory-imposed heterogeneity in solar radiation and soil moisture in a semiarid woodland. Ecol. Appl., 7(4): 1201-1215. Breshears, D. D., J. W. Nyhan, C. E. Heil & B. P. Wilcox. 1998. Effects of woody plants on microclimate in a semi-arid woodland: Soil temperature and evaporation in canopy and intercanopy patches. Int. J. Plant. Sci., 159(6): 1010-1017. Broadbent, F. E. 1965. Organic matter. Pp. 1397-1400, in Methods of soil analysis, part 2, chemical and microbiological properties, (C. A. Black, ed.). American Society of Agronomy, Madison, Wisconsin, 1572 pp. Brown, J. R. & S. Archer. 1999. Shrub invasion of grassland: recruitment is continuous and not regulated by herbaceous biomass or density. Ecology, 80(7):2385-2396. Callaway, R. M., E. H. DeLucia, D. M. Moore, R. Nowak & W. H. Schlesinger. 1996. Competition and facilitation: contrasting effects of Artemisia tridentate on desert vs. montane pines. Ecology, 77(7):21 30-2141 . Correll, D. S. & M. C. Johnston. 1979. Manual of the vascular plants of Texas. Texas Research Foundation, Renner, Texas, 1881 pp. Davenport, D. W., B. P. Wilcox & D. D. Breshears. 1996. Soil morphology of canopy and intercanopy sites in a pinon-juniper woodland. Soil Sci. Soc. Am. J., 60(6): 1881-1887. Diamond, D. D. 1997. An old-growth definition for western Juniper woodlands: Texas Ashe Juniper dominated or codominated communities. Gen. Tech. Rep. SRS-15. Asheville, NC: U. S. Dept, of Agriculture, Forest Service, Southern Research Station, 10 pp. Diamond, D. D., G. A. Rowell & O. P. Keddy-Hector. 1995. Conservation of Ashe Juniper ( Juniperus ashei Buchholtz) woodlands of the central Texas hill country. Nat. Areas J., 15(2) : 1 89- 1 97. Dye, K. L. 1993. Effect of mature Redberry juniper on associated herbaceous vegetation. MS. Thesis. Texas A&M University, College Station, Texas, 69 pp. Dye, K. L., D. N. Ueckert & S. G. Whisenant. 1995. Redberry juniper-herbaceous understory interactions. J. Range Manage., 48(2): 100-107. Fonteyn, P. J., T. M. McClean & R. E. Akridge. 1985. Xylem pressure potentials of three dominant trees of the Edwards Plateau of Texas. Southwest. Nat., 30(1): 141-146. Foster, J. H. 1917. The spread of timbered areas in central Texas. J. Forestry 15:442-445. 52 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Fowler, N. L. & D. W. Dunlap. 1986. Grassland vegetation of the eastern Edwards plateau. Am. Mid. Nat., 1 15(1): 146-155. Fuhlendorf, S. D., F. E. Smeins & W. E. Grant. 1996. Simulation of a fire-sensitive ecological threshold: a case study of Ashe juniper on the Edwards Plateau of Texas. Ecol. Model. 90(3): 245-255. Gould, F. W. 1975. Texas plants - A checklist and ecological summary. Texas Agricultural Experimental Station, College Station, Texas, 121 pp. Hatch, S. L., K. N. Gandhi & L. E. Brown. 1990. Checklist of the vascular plants of Texas. Texas A & M University System, Texas Agricultural Experiment Station Publication MP-1655. College Station, Texas, 158 pp. Helgerson, O. T. 1990. Heat damage in tree seedlings and its prevention. New Forests, 3:333-358. Jackson, J. T. & O. W. Van Auken. 1997. Seedling survival, growth and mortality of Juniperus ashei (Cupressaceae) in the Edwards Plateau region of central Texas. Tex. J. Sci., 49(4):267-278. Jackson, R.B., L.J. Anderson & W.T. Pockman. 2000. Measuring water availability and uptake in ecosystem studies. Pp. 199-214, in Methods in Ecosystem Science, (O.E. Sala, R.B. Jackson, H.A. Mooney, R. Howarth, eds.). Springer- Verlag, New York, 421 pp. Johnsen, T. N. 1962. One-seed juniper invasion of northern Arizona grasslands. Ecol. Monogr., 32:187-207. Johnsen, T. N. & R. A. Alexander. 1974. Juniperus L. Pp. 460-469, in Seeds of woody plants in the United States, (C. S. Schopmeyer, Tech. Coordinator). United States Department of Agriculture, Forest Service, Agricultural Handbook 450, 883 pp. Joy, D. A. & D. R. Young. 2002. Promotion of mid-successional seedling recruitment and establishment by Juniperus virginiana in a coastal environment. Plant Ecol., 160(2): 125-135. Knapp, A. K., J. T. Fahnestock, S. P. Hamburg, L. B. Statland, T. R. Seastedt & D. S. Schimel. 1993. Landscape patterns in soil-plant water relations and primary production in tallgrass prairie. Ecology, 74(2):549-560. Larcher, W. 1995. Physiological plant ecology: Ecophysiology and stress physiology of functional groups. 3rd Ed. Springer- Verlag, New York, 506 pp. Martens, S. N., D. D. Breshears, C. W. Meyer & F. J. Barnes. 1997. Scales of above-ground and below-ground competition in a semi-arid woodland detected from spatial pattern. J. Veg. Sci., 8(5):655-664. Martens, S. N., D. D. Breshears & C. W. Meyer. 2000. Spatial distributions of understory light along the grassland/forest continuum: effects of cover, height, and spatial pattern of tree canopies. Ecol. Model., 126(l):79-93. McKinley, D. & O. W. Van Auken. 2004. Growth and survival of Juniperus ashei (Cupressacae) seedlings in the presence of Juniperus ashei litter. Tex. J. Sci., 56(1 ):3- 14. McPherson, G. R., H. W. Wright & D. B. Wester. 1988. Patterns of shrub invasion in semi-arid Texas grasslands. Am. Mid. Nat., 120(2): 39 1-397. McPherson, G.R., G.A. Rasmussen, D.B. Wester & R.A. Masters. 1991. Vegetation and soil zonation patterns around Juniperus pinchotii plants. Great Basin Nat., 51 (4) :3 16-324. National Oceanic and Atmospheric Administration. 1999. National Climatic Data Center, Asheville, North Carolina, USA. Available: http://www.ncdc.noaa.gov/oa/ncdc.htmi. Owens, M. K. 1996. The role of leaf and canopy-level gas exchange in the replacement of Quercus virginiana (Fagaceae) by Juniperus ashei (Cupressaceae) in semiarid savannas. Am. J. Bot., 83(5):617-623. Owens, M. K. & M. C. Schreiber. 1992. Seasonal gas exchange characteristics of two evergreen trees in a semiarid environment. Photosynthetica, 26(3):389-398. Pearcy, R. W. 1989. Plant physiological ecology: field methods and instrumentation. Pp. WAYNE & VAN AUKEN 53 84-96. Chapman and Hall, New York, 457 pp. Polley, H. W., H. B. Johnson, H. S. Mayeux & C. R. Tischler. 1996. Are some of the recent changes in grassland communities a response to rising C02 concentrations? Pp. 177-195, in Carbon dioxide, populations and communities, (C. Korner & F. A. Bazzaz, eds.). Academic Press, San Diego, California, 465 pp. Quart erman, E. 1950. Major plant communities of Tennessee cedar glades. Ecology, 31:234-254. Quarterman, E., M. B. Burbanck & D. J. Shure. 1993. Rock outcrop communities: limestone, sandstone and granite. Pp. 35-86 in Biodiversity of the southeastern United States: upland terrestrial communities, (W. H. Martin & S. G. Boyce, eds.). John Wiley and sons, New York, 528 pp. Reid, K. D., B. P. Wilcox, D. D. Breshears & L. MacDonald. 1999. Runoff and erosion in a pinon-juniper woodland: influence of vegetation patches. Soil Sci. Soc. Am. J., 63(6): 1869-1879. Riskind, D. H. & D. D. Diamond. 1988. An introduction to environments and vegetation. Pp. 1-16, in Edwards Plateau vegetation: plant ecological studies in central Texas, (B. B. Amos and F. R. Gehlbach, eds.). Edwards Baylor University, Waco, Texas, 144 pp. SAS Institute Inc. 1989. SAS/STAT User’s Guide, Version 6, 4 Ed., Volume 1 and 2, Cary, North Carolina, 1848 pp. Scholes R. J. & S. R. Archer. 1997. Tree-grass interactions in savannas. Annu. Rev. Ecol. Sys., 28(6):517-544. Smeins, F. E. & L. B. Merrill. 1988. Long-term change in semi-arid grassland. Pp. 101-114, in Edwards Plateau vegetation, (B. B. Amos & F. R. Gehlback, eds.). Baylor University Press, Waco, Texas, USA, 144 pp. Sokal, R. & F. J. Rohlf. 1995. Biometry: the principles and practice of statistics in biological research. 3rd Ed. W. H. Freeman and Company, New York, 880 pp. Taylor, F. B., R. B. Hailey & D. L. Richmond. 1962. Soil survey of Bexar County, Texas. United States Department of Agriculture. Soil Conservation Survey. Washington D. C. Terletzky, P. A. & O. W. Van Auken. 1996. Comparison of cedar glades and associated woodlands of the southern Edwards Plateau. Tex. J. Sci., 48(l):55-67. United States Department of Agriculture. 1996. Soil survey laboratory methods manual. Pp. 169-170, in Soil Survey Investigations, Report No. 42, Version 3.0, 716 pp. Available at: http://soils.usda.gov/technical/lmm/ United States Department of Agriculture. 2000. Natural Resources Conservation Service, Soil Survey Division. Official Series Descriptions. Available: http://ortho.ftw.nrcs.usda.gov/osd/osd.htmli2000, November 20]. Van Auken, O. W. 1988. Woody vegetation of the southeastern escarpment and plateau. Pp. 43-55, in Edward’s Plateau vegetation, (B. B. Amos & F. R. Gehlback, eds.). Baylor University Press, Waco, Texas, 144 pp. Van Auken, O. W. 2000a. Characteristics of intercanopy bare patches in Juniperus woodlands of the southern Edwards Plateau, Texas. Southwest. Nat., 45(2):95-l 10. Van Auken, O. W. 2000b. Shrub invasion of North American semiarid grasslands. Annu. Rev. Ecol. Sys., 31:197-215. Van Auken, O. W. , A. L. Ford & J. L. Allen. 1981. An ecological comparison of upland deciduous and evergreen forests of central Texas. Am. J. Bot., 68:1249-1256. Van Auken, O. W., A. L. Ford, A. Stein, & A. G. Stein. 1980. Woody vegetation of upland plant communities in the southern Edwards Plateau. Tex. J. Sci., 32(l):23-35. Van Auken, O. W., J. T. Jackson & P. N. Jurena. 2004. Survival and growth of Juniperus seedlings in Juniperus woodlands. Plant Ecol., in press. Ware, S. 2002. Rock outcrop plant communities (glades) in the Ozarks: a synthesis. Southwest. Nat., 47(4): 585-597. 54 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Wayne, E. R. 2000. Water relations of Juniperus ashei seedlings and changes in biotic and abiotic factors along an environmental gradient on the Edwards Plateau from under a Juniperus woodland canopy into an intercanopy patch. Unpublished M.S. thesis, University of Texas at San Antonio, San Antonio, Texas, 150 pp. Wayne, R. & O. W. Van Auken. 2002. Spatial and temporal patterns of Juniperus ashei seedling xylem water potential. Southwest. Nat., 47(2): 153-161 . Wilcox, B. P. 1994. Runoff and erosion in intercanopy zones of pinon-juniper woodlands. J. Range Manage., 47(4):285-295. Yager, L. Y. & F. E. Smeins. 1999. Ashe juniper (Juniperus ashei: Cupressaceae) canopy and litter effects on understory vegetation in a juniper-oak savanna. Southwest. Nat., 44(1):6-16. RW at: erwayne@utsa.edu TEXAS J. SCI. 56(l):55-62 FEBRUARY, 2004 REPRODUCTIVE CYCLE OF THE SIDEWINDER, CROTALUS CERASTES (SERPENTES: VIPERIDAE), FROM CALIFORNIA Stephen R. Goldberg Department of Biology, Whittier College Whittier, California 90608 Abstract. — Reproductive tissue was examined from 159 museum specimens of Crotalus cerastes from California. Males follow a seasonal testicular cycle with sperm produced June-October; regressed testes were present March-June and October. Timing of this cycle is similar to that of other North American rattlesnakes. Sperm were present in the vasa deferentia March-October. Mean litter size for 26 C. cerastes was 7.96 ± 2.9 SD, range = 3-14. The number of females that were gravid (enlarged follicles > 8 mm or oviductal eggs) during the April to August period of female reproductive activity was 28/53 (53%). The presence of females with early yolk deposition in April and May when other females were gravid suggests more than one reproductive season is needed to complete yolk deposition. The sidewinder, Crotalus cerastes , ranges from southern Nevada, southern California, south-central Arizona and extreme southwestern Utah, south to northeastern Baja California and northwestern Sonora; it occurs from below sea level to around 1830 m and is most common where there are sand hummocks topped with creosote bushes or mes- quite (Stebbins 2003). Information on reproduction in C. cerastes is summarized in Ernst & Ernst (2003). Reiserer (2001) reported on reproduction in C. cerastes but did not perform gonadal histology. The purpose of this paper is to provide information on the reproductive cycle of C. cerastes from California from a histological examination of gonads from museum specimens. Material and Methods Sixty-two female (mean snout- vent length, SVL = 486 mm ± 53 SD, range = 375-592 mm) and 97 male (mean SVL = 446 mm ± 53 SD, range = 331-543 mm) C. cerastes were borrowed from the herpetology collections of the Natural History Museum of Los Angeles County, Los Angeles, California and the San Diego Society of Natural History, San Diego, California. Snakes were collected during 1935-1977. The left testis and part of the vas deferens were removed from males; the left ovary was removed from females for histological examination. Enlarged follicles (> 8 mm length) or oviductal eggs were counted; no histology 56 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 1, 2004 was done on them. Tissues were embedded in paraffin and cut into sections at 5 /xm. Slides were stained with Harris’ hematoxylin, fol¬ lowed by eosin counter stain. Testes slides were examined to determine the stage of the spermatogenic cycle; vasa deferentia were examined for the presence of sperm. Ovary slides were examined for the presence of yolk deposition (= secondary vitellogenesis sensu Aldridge 1979a). Numbers of specimens examined by reproductive tissue were: testis = 97, vas deferens = 75, ovary = 34. The relationship between SVL and litter size was investigated by regression analysis. Unpaired r-tests were used to compare C. cerastes male and female mean body sizes (SVL), mean litter sizes with those from Klauber (1972), and mean litter sizes of northern versus southern populations from Klauber (1972). Material examined.— Specimens of Crotalus cerastes from California (by county) examined from the herpetology collection of the Natural History Museum of Los Angeles County, Los Angeles (LACM) and the San Diego Society of Natural History (SDSNH). IMPERIAL: (LACM) 9202-9204, 52575, 52576, 64024, 104487, 104489, 104490. INYO: 52572, 104491, 116013, 116014; (SDSNH) 3219. KERN: (LACM) 52577, 52578, 63628, 63629, 63631, 63638, 63640-63642, 63644, 69905, 104493, 104495, 137690. LOS ANGELES: (LACM) 28006, 52579, 63447. RIVERSIDE: (LACM) 3025, 19936, 19938, 19942, 19944, 19945, 23235, 27996, 27998, 28000, 28001, 28783, 52582, 104499, 104500, 104507, 104508, 104511, 104512, 104519, 104523, 104542, 104547, 104549, 104552, 104555, 104557, 104560-104565, 104569, 104572, 104578, 104580, 104586, 104589, 104595, 104597, 104601, 104610, 104611, 104619, 104630, 104634, 104641, 104647, 104654, 104665, 104668, 104675, 104677, 104689, 104690, 104692, 104713, 104726, 104735, 104738, 104862, 116002, 1 16004, 1 16007, 116008, 123762, 138215; (SDSNH) 31929, 33096, 39296, 39301, 39302. SAN BERNAR¬ DINO: (LACM) 3018, 19919, 19921, 19922, 19924, 21908, 63632, 63634, 63643, 63645, 63647-63649, 70262, 70265, 70266, 70269, 104750, 104757, 104762, 104768, 104770, 104772, 104776, 104782, 104783, 104785, 104787, 104788, 104790-104793, 104796-104798, 116011, 116012, 125994, 132244; (SDSNH) 25397, 31758. SAN DIEGO: (LACM) 28002-28005, 76300, 104799, 104805, 104806, 104809, 104810, 104813, 125997, 126295. Results and Discussion Testicular histology was similar to that reported by Goldberg & Parker (1975) for two colubrids Masticophis taeniatus and Pituophis catenifer ( = P. melanoleucus) and the viper id Agkistrodon piscivorus by Johnson et al. (1982). In the regressed testis, seminiferous tubules GOLDBERG 57 Table 1. Monthly distribution of reproductive conditions in seasonal testicular cycle of Crotalus cerastes. Values are the numbers of males exhibiting each of the three conditions. Month N Regression Recrudescence Spermiogenesis March 11 7 4 0 April 29 12 17 0 May 31 10 21 0 June 8 2 4 2 July 6 0 2 4 August 5 0 3 2 September 3 0 0 3 October 4 2 0 2 contained spermatogonia and Sertoli cells. There was a proliferation of germ cells; primary and secondary spermatocytes and occasional spermatids were present in testes undergoing recrudescence. During spermiogenesis, seminiferous tubules were lined by spermatozoa. Rows of metamorphosing spermatids were also present. Monthly stages in the testicular cycle are shown in Table 1 . Males undergoing spermiogenesis were present June to October; males with regressed testes were present in March-June and October. Reiserer (2001) found maximum testes sizes of C. cerastes occurred during September. Males with testes in recrudescence were present March to August. The presence of males undergoing spermiogenesis during summer and autumn indicates C. cerastes has a testicular cycle similar to those of other North American rattlesnakes in which sperm formation occurs during this period (Aldridge 1979b; Aldridge & Brown 1995; Goldberg 1999a, 1999b, 1999c, 2000a, 2000b, 2000c, 2002; Goldberg & Holycross 1999; Goldberg & Rosen 2000; Holycross & Goldberg 2001; Goldberg & Beaman 2003). This pattern of spermatogenesis fits the "aestival spermatogenesis" of Saint Girons (1982). Sperm were present in 74/75 (99%) of the vasa deferentia examined: March 8/9, April 27/27, May 21/21, June 3/3, July 4/4, August 4/4, September 3/3, October 4/4. The smallest mature male, LACM 104783 (regressed testis; sperm in vas deferens from previous spermiogenesis) measured 331 mm SVL (360 mm total length, TL). This is less than the smallest male (49.5 mm TL) found in copulation by Secor in Ernest (1992). Field observations have indicated C. cerastes mates both in spring (Klauber 1972; Brown & Lillywhite 1992) and fall (Lowe 1942). 58 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Table 2. Monthly distribution of reproductive conditions in seasonal ovarian cycle of Crotalus cerastes. Values shown are the numbers of females exhibiting each of the four conditions. Month n Inactive Early yolk deposition Enlarged follicles (> 8 mm width) Oviductal eggs January 1 0 1 0 0 February 1 1 0 0 0 March 1 1 0 0 0 April 10 3 4 3 0 May 30 9 1 20* 0 June 4 2 0 1 1 July 5 4 0 0 1 August 4 2 0 0 2 September 5 5 0 0 0 October 1 0 1 0 0 * Includes two females with damaged eggs; litters could not be reliably estimated. Reiserer (2001) reported both spring and fall matings in captive C. cerastes. A captive pair of C. cerastes mated 1 1 October (Klauber 1972). Mean female body size (SVL) was significantly larger than that of males ( t = 4.73, df = 157, P < 0.0001). Reiserer (2001) similarly found female C. cerastes to be generally larger than the same-aged males. Crotalus cerastes may be the only species of North American Crotalus in which females are larger than males (Ernst 1992), however further study will be needed before this is known. Monthly stages in the ovarian cycle are shown in Table 2. Females with enlarged follicles ( > 8 mm length) or oviductal eggs were present April to August. Reiserer (2001) reported ovulation in C. cerastes occurred during late June. The smallest reproductively active female (follicles > 8 mm length) (LACM 104549) measured 383 mm SVL (408 mm TL). This value is less than the 434 mm TL recorded for the smallest gravid C. cerastes female in Klauber (1944). There was no significant difference between the mean litter size (7.96 + 2.9 SD, range = 3-14, n = 26) for C. cerastes in this study and the mean litter size in Klauber (1972) (9.5 ± 3.0 SD , range = 5-18, n = 38 ) t = 2.0, df = 62, P = 0.05. Litters may contain 1-20 young, but typically have 7-12 (Ernst & Ernst 2003). Fitch (1985), using data from Klauber (1972), reported mean litter sizes of 10.8 ± 0.1 SE, range: 7-18 for 10 C. cerastes from the Mohave Desert (northern) and 9.0 + 0.5 SE, range: 5-16 for 28 from the Colorado GOLDBERG 59 Table 3. Litter sizes for Crotalus cerastes from California. Date SVL (mm) Litter size County LACM ft 20 April 1961 485 7 Riverside 104552 27 April 1958 395 4 Riverside 104668 28 April 1962 560 4 Riverside 104738 3 May 1963 421 6 Riverside 104547 3 May 1963 451 6 Riverside 104578 4 May 1968 514 11 Riverside 116004 4 May 1968 528 9 Riverside 104713 5 May 1968 522 7 Kern 63644 5 May 1963 530 12 Riverside 28000 6 May 1961 522 13 Riverside 104619 7 May 1967 445 9 Los Angeles 52579 11 May 1974 435 11 Riverside 138215 16 May 1963 563 14 Kern 69905 16 May 1965 400 9 Riverside 104542 18 May 1966 509 9 Imperial 9203 19 May 1958 560 11 Riverside 104862 20 May 1961 435 7 Riverside 104630 20 May 1961 495 8 Riverside 104508 23 May 1958 438 5 Riverside 104595 24 May 1958 498 8 San Bernardino 104790 24 May 1963 383 3 Riverside 104549 12 June 1961 483 8 Riverside 104511 15 June 1960 463 8* Riverside 104500 27 July 1962 490 3* San Bernardino 21908 5 August 1968 498 6* San Diego 125997 15 August 1954 459 9* Riverside 3025 * Contained oviductal eggs; others contained enlarged follicles > 8 mm length. Desert and Arizona (southern). There was no significant difference between mean litter sizes of these northern versus southern C. cerastes populations (t = 1.7, df = 36, P = 0.10). Examination of additional samples from other areas will be needed to ascertain the degree of geographic variation in C. cerastes litter sizes. Litter sizes for 26 gravid C. cerastes females are given in Table 3. Regression analysis (Fig. 1) revealed a significant positive correlation between In (litter size) and In (SVL) for these 26 litters: (In litter size = 60 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Figure 1 . Linear regression of enlarged follicles ( > 8 mm length) or oviductal eggs on snout- vent length, mm (log transformed variables) for 26 Crotalus cerastes females from California (regression equation in text). -8.34 + 1.68 In SVL) r2 = 0.19, P = 0.024. Back transformed this regression equation describes the allometric relationship via a power function: litter size = e'8 34SVL* 68 (King 2000). The number of gravid females (enlarged follicles > 8 mm or oviductal eggs) during the April to August period of female reproductive activity was 28/53 (53%). The presence of non-reproductive females (Table 2) during the period when other C. cerastes females are gravid indicates that not all females reproduce each year. This has been reported for other western North American rattlesnakes (Goldberg 1999a, 1999b, 1999c, 2000a, 2000b, 2000c, 2002; Goldberg & Holy- cross 1999; Goldberg & Rosen 2000; Holycross & Goldberg 2001; Rosen & Goldberg 2002). The frequency of reproduction in female rattlesnakes is unknown but is likely influenced by available food reserves (Goldberg & Rosen 2000; Rosen & Goldberg 2002). Long¬ term field studies will be required before the frequency of female reproduction can be known for C. cerastes. The presence of C. cerastes females with early yolk deposition in April and May when other females were gravid (Table 2) suggests yolk deposition and ovulation are completed over more than one reproductive season and may be approximately biennial. Biennial reproduction may be "typical" for many species of North American rattlesnakes with the GOLDBERG 61 likelihood of less frequent reproduction during years of low food availability, and the potential of reproduction in successive years when food is abundant. Acknowledgments I thank D. Kizirian (LACM) and B. Hollingsworth (SDSNH) for permission to examine specimens. Literature Cited Aldridge, R. D. 1979a. Female reproductive cycles of the snakes Arizona elegans and Crotalus viridis. Herpetologica, 35(3):256-261. Abridge, R. D. 1979b. Seasonal spermatogenesis in sympatric Crotalus viridis and Arizona elegans in New Mexico. J. Herpetol., 13(2): 187-192. Aldridge, R. D. & W. S. Brown. 1995. Male reproductive cycle, age at maturity, and cost of reproduction in the timber rattlesnake ( Crotalus horridus). J. Herpetol., 29(3): 399-407. Brown, T. W. & H. B. Lilywhite. 1992. Autecology of the Mojave desert sidewinder, Crotalus cerastes cerastes , at Kelso Dunes, Mojave Desert, California, USA. Pp. 279-308, in Biology of the Pitvipers. (J.A. Campbell and E.D. Brodie, Jr., eds.), Selva, Tyler, Texas, xi + 467 pp. Ernst, C. H. 1992. Venomous reptiles of North America. Smithsonian Institution Press, Washington, ix + 236 pp. Ernst, C. H. & E. M Ernst. 2003. Snakes of the United States and Canada. Smithsonian Books, Washington, ix 4- 668 pp. Fitch, H. S. 1985. Variation in clutch and litter size in New World reptiles. Univ. Kansas Mus. Nat. Hist., Misc. Publ., 76:1-76. Goldberg, S. R. 1999a. Reproduction in the tiger rattlesnake, Crotalus tigris (Serpentes: Viperidae). Texas J. Sci., 51(l):31-36. Goldberg, S. R. 1999b. Reproduction in the blacktail rattlesnake, Crotalus molossus (Serpentes: Viperidae). Texas J. Sci., 5 1(4): 323-328. Goldberg, S. R. 1999c. Reproduction in the red diamond rattlesnake in California. Calif. Fish and Game, 85 (4): 177- 180. Goldberg, S. R. 2000a. Reproduction in the twin-spotted rattlesnake, Crotalus pricei (Serpentes: Viperidae). West. North Am. Nat., 60(1):98-100. Goldberg. S. R. 2000b. Reproduction in the rock rattlesnake, Crotalus lepidus (Serpentes: Viperidae). Herpetol. Nat. Hist., 7(l):83-86. Goldberg, S. R. 2000c. Reproduction in the speckled rattlesnake, Crotalus mitchellii (Serpentes: Viperidae). Bull. Southern Calif. Acad. Sci., 99(2): 101-104. Goldberg, S. R. 2002. Reproduction in the Arizona black rattlesnake, Crotalus viridis cerberus (Viperidae). Herp. Nat. Hist., 9(l):75-78. Goldberg, S. R. & A. T. Holycross. 1999. Reproduction in the desert massasauga, Sistrurus catenatus edwardsii, in Arizona and Colorado. Southwestern Nat., 44(4):531-535. Goldberg, S. R. & W. S. Parker. 1975. Seasonal testicular histology of the colubrid snakes, Masticophis taeniatus and Pituophis melanoleucus . Herpetologica, 3 1 (3) :3 17-322. 62 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Goldberg, S. R. & P. C. Rosen. 2000. Reproduction in the Mojave rattlesnake, Crotalus scutulatus (Serpentes: Viperidae). Texas J. Sci., 52(2): 101-109. Goldberg, S. R. & K. R. Beaman. 2003. Reproduction in the Baja California rattlesnake, Crotalus enyo (Serpentes: Viperidae). Bull. Southern Calif. Acad. Sci., 102(1 ): 39-42. Holycross, A. T. & S. R. Goldberg. 2001. Reproduction in northern populations of the ridgenose rattlesnake, Crotalus willardi (Serpentes: Viperidae). Copeia, 2001(2):473-481. Johnson, L. F., J. S. Jacob & P. Torrance. 1982. Annual testicular and androgenic cycles of the cottonmouth (Agkistrodon piscivorous) in Alabama. Herpetologica, 38(1): 16-25. King, R. B. 2000. Analyzing the relationship between clutch size and female body size in reptiles. J. Herpetol., 34(1): 148-150. Klauber, L. M. 1944. The sidewinder, Crotalus cerastes, with description of a new subspecies. Trans. San Diego Soc. Nat. Hist., 10(8):91-126. Klauber, L. M. 1972. Rattlesnakes. Their habits, life histories and influence on mankind. 2nd edit., Vol. 1, University of California Press, Berkeley, xlvi + 740 pp. Lowe, C. H., Jr. 1942. Notes on the mating of desert rattlesnakes. Copeia, 1942(4): 26 1-262. Reiserer, R. S. 2001. Evolution of life histories in rattlesnakes. Unpublished Ph.D. dissertation, Univ. Calif. Berkeley, xxii -I- 256 pp. Rosen, P. C. & S. R. Goldberg. 2002. Female reproduction in the western diamond-backed rattlesnake, Crotalus atrox (Serpentes: Viperidae), from Arizona. Texas J. Sci., 54(4): 347-356. Saint Girons, H. 1982. Reproductive cycles of male snakes and their relationships with climate and female reproductive cycles. Herpetologica, 38(1 ) :5- 1 6. Stebbins, R. C. 2003. A field guide to western reptiles and amphibians, 3rd edit., Houghton Mifflin Company, Boston, Massachusetts, xiii + 533 pp. SRG at: sgoldberg@whittier.edu TEXAS J. SCI. 56(l):63-72 FEBRUARY, 2004 FRESHWATER MUSSELS (BIVALVIA: UNIONIDAE) OF THE VILLAGE CREEK DRAINAGE BASIN IN SOUTHEAST TEXAS Vickie L. Bordelon and Richard C. Harrel Department of Biology Lamar University Beaumont, Texas 77710 Abstract.— A total of 18 species and 2,235 individuals of freshwater mussels were collected from 22 sites in the Village Creek basin. The number of individuals per site ranged from zero to 528 and the number of species per site ranged from zero to 13. Relative abun¬ dance for all collection sites varied from zero to 176 individuals/person-hours. Quadrula mortoni and Fusconaia askewi comprised 60 percent of the total individuals collected and relative abundance was 14.8 and 13.1 individuals/person-hours, respectively. Lampsilis satura, Obovaria jacksoniana and Pleurobema riddellii were collected at several sites and are listed as "of special concern" by the American Fisheries Society. Freshwater mussels are good indicators of water quality and are often the first organism to decline during adverse conditions (Rosenburg & Resh 1993; Howells et al. 1996; Howells 1997). Howells et al. (1997) reported that 52 species of freshwater bivalves occurred in Texas and discussed 18 that were dramatically reduced in abundance. Williams et al. (1993) listed 17 of these 52 species as threatened, endangered, or of special concern. This survey of the freshwater bivalves of the Village Creek drainage basin evaluates the current status of the populations and will serve as a baseline reference for subsequent studies. There has been no extensive study of the bivalves of Village Creek. Strecker (1931) and Parks (1938) listed some bivalves that occurred in Village Creek, but these works are dated and uncertainties in systematics limit their present day use. Vidrine (1990) surveyed one location in Village Creek for his study of parasitic mites of freshwater mussels. Howells et al. (1996) listed some mussels known to have occurred in Village Creek, but in a later paper (Howells 1997) on the status of mussels in the Big Thicket region he mentioned an unsuccessful effort by Texas Parks and Wildlife Department personnel to collect any living mussels from Village Creek. Several studies have been conducted on the physical/chemical condi¬ tions and macrobenthos of Village Creek and its tributaries (Tatum & Commander 1971; Harrel 1977; Kost 1977; Lewis & Harrel 1978; 64 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Commander 1980; Newberry 1982; Harrel 1985; Barclay & Harrel 1985), but the sampling techniques were not adequate to survey the bivalve fauna. Nearby Texas and Louisiana mussel surveys were con¬ ducted by Neck (1986), Feaster (1997), Howells (2000) and Vidrine (2001). Description of the Area Village Creek is a 5th order stream located in Hardin, Tyler and Polk counties in southeast Texas (Figure 1). From its origin, near the city of Livingston in Tyler County, it flows southeasterly into the Neches River. The basin drains an area of approximately 2,883 km2 and has an axial length of 125 km. Land uses in the basin consist of lumber production, several small municipalities (< 10,000 residents) and scat¬ tered residential developments. Some reaches of Village Creek and its tributaries are within the boundaries of the Alabama- Coushatta Indian Reservation, the Big Thicket National Preserve, Roy Larsen Nature Conservancy Sanctuary and Village Creek State Park. The remaining sections of the stem stream of Village Creek, from the Big Thicket National Preserve Big Sandy Creek Unit to the confluence with the Neches River are proposed as additions to the Big Thicket National Preserve (Big Thicket National Preserve 1996). The shallow substrate in the stream channel consisted of fine and coarse sand with pockets of silt, detritus and clay. Sunken logs are abundant. The average gradient is 0.38 m/km and the minimum and maximum daily discharge based on 66 years of record was 1.8 m3/sec and 131.6 m3/sec (USGS 2001). Dominant vegetation along the stream banks consists of Taxodium distichum (bald cypress), Nyssa aquatica (water tupelo), Betula nigra (river birch) and Quercus sp. (water tolerant oaks) . Methods Twenty-two sites were sampled between 9 August 2001 and 25 November 2002 (Figure 1). Seventeen sites were located along the lower stem stream and five sites were in smaller tributaries. Vidrine (1998) reported that small to moderate size streams resulted in low to moderate mussel diversity and larger, downstream reaches often had higher diversity and larger populations. At each site, 1.5 to 3 person- hours were spent hand- searching the substrate for mussels, covering an average of 50 meters of shoreline. Vaughn (1995) and Hornbach & BORDELON & HARREL 65 Figure 1 . The Village Creek drainage basin and locations of sites sampled (in the order in which they were sampled). Deneka (1996) stated that non-quantitative random time search methods are preferred when examining the distribution of freshwater mussels. Sampling was done only during relatively low stream discharge and depth conditions as indicated by the U.S. Geological Survey gauging station 08041500 located near Kountze, Texas (USGS 2001). Mean water depth for all collecting dates was 1.2 m and mean discharge was 5.6 m3/sec. These conditions allowed productive sampling, which could not have occurred at greater depth or discharge. Living mussels collected were identified, counted and measured. Most specimens were returned to the stream, but some were retained in order to confirm identification or to be used as reference specimens. Dead shell material was not documented. Retained specimens were returned to the laboratory and placed in three percent ethyl alcohol to cause the valves to gape, then preserved in 95 percent ethyl alcohol. Identifications were made using the following taxonomic references; Burch (1973), Cummings & Mayer (1992), McMahon (1991), Howells et al. (1996) and Vidrine (2001). Robert Howells (Texas Parks and 66 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 1, 2004 Table 1. Total number of living individuals of each species collected at each site sampled. (Total number of person-hours spent = 48.) Site 1 2 3 4 5 6 7 8 9 10 11 Amblema plicata 13 11 31 4 4 14 5 1 0 0 0 Fusconaia askewi 11 8 212 2 26 1 0 11 0 0 11 Fusconaia flava 2 3 12 0 5 0 0 5 0 0 6 Lampsilis hydiana 5 10 25 2 26 34 9 7 0 0 0 Lampsilis satura 4 3 0 0 0 0 0 0 0 0 0 Lampsilis teres 2 8 20 4 6 4 4 0 0 0 0 Leptodea fragilis 1 2 0 0 0 0 0 0 0 0 0 Obliquaria reflexa 0 0 0 3 1 1 0 1 0 0 0 Obovaria jacksoniana 0 6 0 4 3 0 0 0 0 0 0 Plectomerus dombeyanus 1 1 21 0 1 2 2 0 0 0 0 Pleurobema riddellii 1 5 1 0 0 0 0 0 0 0 0 Potamilus purpuratus 2 0 2 0 0 1 0 0 0 0 0 Quadrula mortoni 61 86 185 82 41 3 54 12 0 3 18 Quadrula nobilis 7 8 15 5 5 1 2 4 0 0 3 Toxolasma texasiensis 0 0 2 3 3 16 0 3 0 0 0 Tritogonia verrucosa 0 0 1 0 0 0 0 0 0 0 3 Uniomerus tetralasmus 0 0 0 0 0 0 0 0 0 0 0 Villosa lienosa 1 4 1 0 9 14 0 3 0 0 0 Total 111 152 528 109 130 91 76 47 0 3 41 Wildlife Department), verified the identifications. Common and scientific names are those of Turgeon et al. (1998). Voucher specimens were placed in a collection at Lamar University. Relative abundance of all mussels for each collection site was calculated by the formula: number of individuals of all species collected/person- hours (48) spent collecting at that site. Relative abundance for each species was determined by the formula: number of individuals of a species collected/ total person-hours (48) for entire study. BORDELON & HARREL 67 Table 1. (Continued) Site 12 13 14 15 16 17 18 19 20 21 22 Ambletna plicata 0 0 7 9 25 4 6 8 15 5 4 Fusconia askewi 37 0 89 10 24 58 13 61 28 15 14 Fusconaia flava 6 0 7 2 16 3 4 3 2 5 0 Lampsilis hydiana 1 0 0 2 4 18 7 3 8 4 5 Lampsilis satura 0 0 0 17 2 3 0 1 0 3 0 Lampsilis teres 0 0 0 1 2 3 4 0 6 4 0 Leptodea fragilis 0 0 0 8 0 0 0 0 0 0 0 Obliquaria reflexa 0 0 0 0 0 0 0 0 0 1 0 Obovaria jacksoniana 0 0 0 0 0 0 0 3 0 0 0 Plectomerus dombeyanus 1 0 0 0 3 0 0 0 0 1 0 Pleurobema riddellii 0 0 2 0 2 0 0 0 0 0 0 Potamilus purpuratus 0 0 0 1 0 0 0 0 0 0 0 Quadrula mortoni 1 0 10 10 33 18 23 9 21 31 7 Quadrula nobilis 1 0 3 5 15 6 8 30 14 10 0 Toxolasma texasiensis 0 0 3 44 0 4 4 0 4 3 0 Tritogonia verrucosa 0 0 5 0 0 0 0 1 0 0 0 Uniomerus tetralasmus 1 1 0 0 0 0 0 0 0 0 0 Villosa lienosa 1 0 1 0 0 0 5 1 4 6 0 Total 49 1 128 109 126 117 74 120 106 88 30 Results and Discussion During the study, 18 species of unionds and 2,235 individuals were collected during a total of 48 person-hours (Table 1). The number of species per collection site ranged from zero at site 9 to 13 at sites 1, 2 and 3 (Table 2). The number of individuals per collection site ranged from zero (site 9) to 528 (site 3). No mussels were found at site 9 after 2.25 person-hours of searching. This was probably due to the unsuitable habitat that was composed of steep cut clay banks and tree roots, which made searching difficult. Relative abundance of all mussels at individual 68 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Table 2. Number of species collected, mussels collected, person-hours spent collecting, and relative abundance for each collecting site. Data indicates living specimens only. Site Species collected Number collected Person-hours Relative abundance 1 13 111 3 37 2 13 152 3 51 3 13 528 3 176 4 9 109 2 55 5 12 130 2 65 6 11 91 2.50 36 7 6 76 2 38 8 9 47 2 24 9 0 0 2.25 0 10 1 3 3 1 11 5 41 2 21 12 8 49 3 16 13 1 1 2.25 0.4 14 9 128 3 43 15 11 109 2 55 16 10 126 1.5 84 17 9 117 1.5 78 18 9 74 1.5 49 19 10 120 2 60 20 9 106 1.5 71 21 12 88 1.5 59 22 4 30 1.5 20 collection sites ranged from zero (site 9) to 176 (site 3) per person-hour (Table 2). Site 3 had a large diversity of microhabitats including substrate types, variations in flow, and a large area of suitable depth for collecting. Site 3 is the location where Vidrine (1990) collected and removed 1,000 individuals for his study of mites associated with mussels. Site 3 is also the location where Texas Parks and Wildlife personnel reported finding no living mussels (Howells et al. 1996). This was probably due to their collecting method. They used a brail, which cannot be effectively utilized in Village Creek due to the amount of sunken trees. Quadrula mortoni and Fusconaia askewi were the most abundant species, representing 31.8 and 28.2 percent, respectively, of the total number of individuals collected during the study (Table 3). Relative abundance of Q . mortoni and F. askewi was 14.8/person-hour and 13.1/person-hour, respectively. Quadrula mortoni occurred at 20 collecting sites and F. askewi occurred at 18 sites. These species are euryecious and were found in all types of substrates and were often the only species found in coarse sand away from the shore. One specimen BORDELON & BARREL 69 Table 3. Total number of sites where species occurred, total number of individuals collected, percentages of all individuals collected, and relative abundance of each species (in order of relative abundance). Data indicates living specimens only. Species Site frequency Number collected % of total collected Relative abundance Quadrula mortoni 20 712 31.8% 14.8 Fusconaia askewi 18 631 28.2% 13.1 Lampsilis hydiana 17 170 7.6% 3.5 Amblema plicata 17 166 7.4% 3.5 Quadrula nobilis 18 135 6.3% 2.9 Fusconaia flava 15 101 4.5% 1.7 Toxolasma texasiensis 11 89 4.0% 1.9 Lampsilis teres 13 68 3.0% 1.4 Villosa lienosa 12 50 2.2% 1.0 Lampsilis satura 7 33 1.4% .7 Plectomerus dombeyanus 9 33 1.4% .7 Obovaria jacksoniana 4 16 <1% .3 Pleurobema riddellii 5 11 <1% .2 Leptodea fragilis 3 11 <1% .2 Tritogonia verrucosa 4 10 <1% .2 Potamilus purpuratus 4 6 <1% .1 Obliquaria reflexa 5 7 <1% .1 Uniomerus tetralasmus 2 2 <1% <.l of F. askewi measured 74 mm in shell length, which exceeds the maximum length recorded for Texas waters (Howells et al. 1996). Uniomerus tetralasmus was the least abundant species and was collected only in two tributary streams; one specimen each in Beech Creek (site 12) and Turkey Creek (site 13). This species is adapted for desiccation, dewatering and stagnant water (Neck & Metcalf 1988; Cummings & Mayer 1992) and was the only species collected only in smaller tributary streams. Three species found during this study are listed as of "special concern" by the American Fisheries Society (Williams et al. 1993). These include Lampsilis satura, Obovaria jacksoniana and Pleurobema riddellii. Only eight living specimens of L. satura had been reported in the Big Thicket region during the past five years (Howells 1997). During this study 33 specimens from seven sites were collected (Tables 1 & 3). Howells (1997) reported that only one dead shell of Obovaria jacksoniana had been found in Texas since 1990. During this study, 16 specimens of O. jacksoniana were collected from four sites (Tables 1 & 3). Since 1987, only two living and two dead specimens of P. riddellii have been reported from the central Neches River in Texas (Howells 1997). During this survey 1 1 specimens from four sites were collected (Tables 1 & 3). 70 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Seven species of mussels were considered to be uncommon or rare and represented less than one percent of the total number collected and their relative abundance was less than 0.5 clams per person-hour (Table 3). Three species of mussels that were previously collected in Village Creek or the nearby Neches River during benthic surveys, but not during this study, include Glebula rotundata, Quadrula apiculata and Megalonaias nervosa. The exotic Asiatic clam, Corbicula fluminea , was noted at sites 1, 2, 4, 5, 8, 9, 11, 12, 13 and 20, but it was abundant only at sites 11 and 13 in Turkey Creek. The results of this study indicate that Village Creek supports a diverse and healthy bivalve fauna. However, Neck (1982), Samad & Stanley (1986), Alderman & Adams (1993), Layzer & Gordon (1993) Neves (1993) and Howells (2000) reported that habitat alterations in and around waterways adversely alter mussel habitats. Within the basin, current and projected residential development and economic growth, together with increased recreational usage of Village Creek, may effect bivalve popu¬ lations. The bivalve fauna should be monitored closely in the future to ensure protection of these organisms. Acknowledgments This study was funded by a student research award from the Texas Academy of Science to V. Bordelon and a Lamar University Scholar award to R. Harrel. Literature Cited Alderman, J. M. & W. F. Adams. 1993. Conservation of critical habitat for freshwater mussels. Pages 81-82, in K.S. Cummings, A. C. Buchanan, & L.M. Koch (eds.), Conservation and Management of Freshwater Mussels. Proceedings of a UMRCC symposium, 12-14 October 1992, St. Louis, Missouri. Upper Mississippi River Conservation Committee, Rock Island, Illinois, 189 pp. Barclay, C. M. & R. C. Harrel. 1985. Effects of pollution effluents on two successive tributaries and Village Creek in Southeastern Texas. Tex. J. Sci., 37(5): 175-188. Big Thicket National Preserve. 1996. Amendment to Land Protection Plan for Big Thicket National Preserve Approved May 7, 1984. Land protection plan Big Thicket National Preserve Addition Act of 1993, 272 pp. Burch, J. B. 1973. Biota of Freshwater Ecosystems Manual 1., U.S. Environmental Protection Agency. Freshwater unionacean clams (Mollusca: Pelecypoda) of North America. Washington, D.C., 177 pp. Commander, S. D. 1980. Physiochemical condition, fecal bacteria, and macrobenthos of streams in the Turkey Creek Unit of the Big Thicket National Preserve. Unpublished M.S. thesis, Lamar University. Beaumont, Texas, 92 pp. Cummings, K. S. & C. A. Mayer. 1992. Field guide to the freshwater mussels of the BORDELON & HARREL 71 midwest. Illinois Natural History Survey, Manuel 5, Champaign, Illinois, 114 pp. Feaster, D. M. 1997. Lotic freshwater mussels (Family Unionidae) of the Angelina and Davy Crockett National Forests of east Texas. Tex. J. Sci., 50(2): 163-170. Harrel, R. C. 1977. Water quality monitoring in the Big Thicket National Preserve. Research Report. Contract No. PX7029-6-0846, 48 pp. Harrel, R. C. 1985. Effects of an oil spill on water quality and macrobenthos of a Southeast Texas stream. Hydrobiologia, 124:223-228. Hornbach, D. J. & T. Deneka. 1996. A comparison of a qualitative and a quantitative collection method for examining freshwater mussel assemblages. J. of the N. A. Benth. Soc., 15:587-596. Howells, R. G. 1997. Status of freshwater mussels (Bivalvia: Unionidae) of the Big Thicket Region of Eastern Texas. Tex. J. Sci., 49(3), Supplement: 21-34. Howells, R. G. 2000. Impacts of dewatering and cold on freshwater mussels (Unionidae) in B. A. Steinhagen Reservoir, Texas. Tex. J. Sci., 52(4), Supplement: 93-104. Howells, R. G., C. M. Mather & J. A. M. Bergmann. 1997. Conservation status of selected freshwater mussels in Texas. Pages 117-128, in K. S. Cummings, A. C. Buchanan, C. A. Mayer & T. J. Naimo (eds.), The Conservation and Management of Freshwater Mussels II: Initiatives for the Future. Upper Mississippi River Conservation Commission, St. Loius, Missouri, 293 pp. Howells, R. G., R. W. Neck & H. D. Murray. 1996. Freshwater mussels of Texas. Texas Parks and Wildlife Press, Austin, Texas, 218 pp. Kost, D. A. 1977. Physicochemical conditions and macrobenthos of streams in the Beech Creek Unit of the Big Thicket National Preserve. Unpublished M.S. thesis, Lamar University, Beaumont, Texas, 93 pp. Layzer, J. B. & M. E. Gordon. 1993. Reintroduction of mussels into the Upper Duck River, Tennessee. Pages 89-92, in K. S. Cummings, A. C. Buchanan, and L. M. Koch (eds.), Conservation and Management of Freshwater Mussels. Proceedings of a UMRCC symposium, 12-14 October 1992. St. Louis, Missouri. Upper Mississippi River Conservation Committee, Rock Island, Illinois, 189 pp. Lewis, S. P. & R. C. Harrel. 1978. Physicochemical conditions and diversity of macrobenthos Village Creek, Texas. Southwest. Nat., 23(3):263-272. McMahon, R. F. 1991. Mollusca: Bivalvia. Pages 315-399, in J. H. Thorp and A. P. Covich, (ed.), Ecology and classification of North American freshwater invertebrates. New York: Academic Press, Inc., 911 pp. Neck, R. W. 1982. A review of interactions between humans and freshwater mussels in Texas. Pages 169-182, in J. R. Davis, (ed.), Proceedings of the Symposium on Recent Benthological Investigations in Texas and Adjacent States. Austin, Texas, 277 pp. Neck, R. W. 1986. Freshwater bivalves of Lake Tawakoni, Sabine River, Texas. Tex. J. Sci., 38(2) :241 -249. Neck, R. W. 1989. Freshwater bivalves of Arrowhead Lake, Texas: apparent lack of extirpation following impoundment. Tex. J. Sci., 4 1(5): 37 1-377. Neck, R. W. & A. L. Metcalf. 1988. Freshwater bivalves of the lower Rio Grande, Texas. Tex. J. Sci., 40(1) :259-268. Neves, R. J. 1993. A-state-of-the-unionids address. Pages 1-10, in K. S. Cummings, A. C. Buchanan, and L. M. Koch (eds.), Conservation and Management of Freshwater Mussels. Proceedings of a UMRCC Symposium, 12-14 October 1992. St. Louis, Missouri. Upper Mississippi River Conservation Committee, Rock Island, Illinois, 189 pp. Newberry, W. 1982. Physicochemical conditions, fecal bacteria, and benthic macroinverte¬ brates of Big Sandy Creek in the Big Thicket National Preserve. Unpublished M.S. thesis, Lamar University, Beaumont, Texas, 89 pp. 72 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Parks, H. B. 1938. Mollusca. Pages 7-8, in H. B. Parks & V. L. Cory (eds.), Biological Survey of the East Texas Big Thicket Area. Texas Agricultural Experiment Station, College Station, Texas, 22 pp. Rosenberg, D. M. & V. H. Resh. 1993. Introduction to freshwater biomonitoring and benthic macroinvertebrates in freshwater. New York: Chapman and Hall, Inc., 307 pp. Samad, F. & J. G. Stanley. 1986. Loss of freshwater shellfish after water dropdown in Lake Sebasticook, Maine. J. Fresh Ecol., 3:519-523. Strecker, J. 1931. The distribution of naiades or pearly fresh-water mussels of Texas. Baylor University Museum, Bulletin 2, 63 pp. Tatum, J. W. & D. Commander. 1971. Texas Water Quality Board. Water Quality Study of Village Creek, Hardin County, Texas. Ausin, Texas, 142 pp. Turgeon, D. D., A. E. Bogan, E. V. Coan, W. K. Emerson, W. G. Lyons, W. L. Pratt, C. F. E. Roper, A. Scheltema, F. G. Thompson & J. D Williams. 1998. Common and scientific names of aquatic invertebrates from the United States and Canada: mollusks. American Fisheries Society Spec. Publ. 16, Bethesda, Maryland, 277 pp. U.S. Geological Survey. 2001. National Water Information System (NWISWeb). Data available at URL: http://waterdata.usgs.gov/nwis/. Vaughn, C. C. 1995. Freshwater mussel sampling techniques and strategies in Native mussels of Oklahoma: a workshop for field aquatic biologist. U.S. Fish and Wildlife Service symposium conducted at Tulsa Tech. Center, Tulsa, Oklahoma. Vidrine, M. F. 1990. Fresh-water mussel-mite and mussel- Ablabesmyia Associations in Village Creek, Hardin County, Texas. Proc. Louisiana Acad, of Sci., 53:1-4. Vidrine, M. F. 1998. Freshwater mussels of Fort Polk, Louisiana. Pages 228-266, in (C. Allen Ed.) Natural History of Fort Polk, Ft. Polk, Louisiana, 256 pp. Vidrine, M. F. 2001. The historical distributions of freshwater mussels in Louisiana. (Electronic Version 1.0 by C. J. Thibodeaux and B. J. Fontenot). Eunice, Louisiana: Gail Q. Vidrine Collectables, 316 pp. Williams, J. D., M. L. Warren, Jr., K. S. Cummings, J. L. Harris & R. J. Neves. 1993. Conservation status of freshwater mussels of the United States and Canada. Fisheries (Bethesda), 18(9): 6-22. VLB at: VBordelon@aol.com TEXAS J. SCI. 56(1), FEBRUARY, 2004 73 GENERAL NOTES NOTEWORTHY RECORDS OF THE MILLIPEDS, EURYMERODESMUS ANGULARIS AND E. MUNDUS (POLYDESMIDA: EURYMERODESMIDAE), FROM NORTHEASTERN AND WESTCENTRAL TEXAS Chris T. McAllister, Rowland M. Shelley* and Dawn I. Moore Department of Biology, Texas A&M University-Texarkana Texarkana, Texas 75505 and ^Research Laboratory, North Carolina State Museum of Natural Sciences 4301 Reedy Creek Road, Raleigh, North Carolina 27607 The milliped family Eurymerodesmidae occurs from northeastern Nebraska, central Illinois and southeastern North Carolina to the Rio Grande and north Florida, and is the dominant representative of the order Polydesmida in the central United States (Shelley 1990). It is a monotypic genus, but is relatively diverse with 25 known species. Eurymerodesmus mundus Chamberlin has been reported from north¬ eastern Nebraska through eastern Oklahoma and southwestern Arkansas to Cooke, Dallas, Grayson and Johnson counties, Texas, and E. angularis Causey is known from southern Missouri, the Coastal Plain of Arkansas, eastern Mississippi and northern Louisiana (Shelley 1990). This study provides the first report of E. angularis from Texas and four new records for E. mundus that significantly increase its known distribu¬ tion within the state. Between October 2001 and May 2003, locations (primarily in State Parks) within 24 Texas counties (Bosque, Bowie, Brown, Cass, Coryell, Dallas, Delta, Fannin, Freestone, Harrison, Hopkins, Jack, Johnson, Limestone, Marion, Morris, Parker, Red River, Shackleford, Somervell, Taylor, Titus, Tom Green and Travis) and Caddo Parish, Louisiana, were examined for millipeds in general and eurymerodesmids in particu¬ lar. Individuals were encountered primarily in damp spots off park trails by overturning decaying logs and leaf litter with potato rakes. Occasion¬ al specimens were collected by peeling bark off fallen trees and rotting stumps. At each locale, specimens were placed in individually labeled vials containing 70% ethanol and returned to the laboratory for identifi¬ cation. Specimens were identified by examining the male genitalia. In eurymerodesmids both the gonopods and gonopodal apertures in males hold taxonomic utility as do the female cyphopods, which possess 74 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 projections and other unique morphological features. Voucher speci¬ mens were deposited in the invertebrate collection of the North Carolina State Museum of Natural Sciences. Several specimens of E. mundus were found during the study period in Texas; data are as follows: Cass County, 8.1 mi (12.9 km) S Linden, along Yellow Poplar Trail off US Hwy. 59, 56, 39, 12 November 2001 and 26 November 2002. Dallas County, Cedar Hill State Park, DORBA and Talala Trails, 56, 49, 21 January and 16 November 2002. Morris County, Daingerfield State Park, Dogwood Camping Area, 6 , 9, 26 November 2002. Taylor County, Abilene State Park, Elm Creek Nature Trail, 46, 9, 17 November 2001. Titus County, Lake Bob Sandlin State Park, 36, 29, juv., 21 December 2002. Eurymerodesmus mundus is readily recognized by the large, hirsute, clavate lobes on the caudal margin of the gonopodal aperture (Shelley 1990). Shelley speculated that the lobes must alter the millipeds’ posture and locomotion because they are so disproportionately large in relation to the rest of the body that they would otherwise scrape the substrate or become impaled. The published record from Grayson County by Shelley was inadvertently omitted from the text; its data are Grayson County, Sherman, in storm cellar, 46, 79, 3 October 1967, M. Cundliff (Florida State Collection of Arthropods, Gainesville). The sites in Titus and Taylor counties are some 350 miles (563 km) apart, so E. mundus thus occupies the entire breadth of the family’s distribution across northern Texas. The species also inhabits a variety of biotopes as habitats at these locales are quite different. The site in Cass County is a climax forest on acreage owned by International Paper Company that consists primarily of pines, yellow poplar and various oak species, while the sites in Morris and Titus counties are within state parks and comprised of mixed hardwoods. However, at the Dallas and Taylor County sites, the dominant trees are live oak, mesquite and eastern red cedar. In addition, the site in Dallas County includes trails situated near native tall grass prairie habitat. Eurymerodesmus mundus ranges north¬ ward to Nebraska, and in the "Ark-La-Tex" region (Fig. 1). Its occur¬ rence in southwestern Arkansas (McAllister et al. 2002a) and north¬ eastern Texas near the Louisiana state line suggest potential discovery in northwestern Louisiana (perhaps Bossier and/or Caddo parishes), which would constitute a new state record. Interestingly, a large female TEXAS J. SCI. 56(1), FEBRUARY, 2004 75 Figure 1 . Map of the United States with inset of Arkansas and parts of Louisiana, Oklahoma and Texas showing county or parish distributions of Eurymerodesmus angularis (dots) and E. mundus (stars) within these states. County distributions of E. mundus in Kansas and Nebraska not included (see Shelley 1990). Eurymerodesmus resembling E. mundus was collected by the senior author on 6 January 2003 in the vicinity of Oil City, Caddo Parish; however, an authentic male of E. mundus is necessary for specific identification. Specimens of E. angularis were also encountered in three counties in the northeastern corner of Texas, confirming Shelley’s prediction (1990) of discovery in this area. It represents a new species for Texas and the tenth species of Eurymerodesmus in the state. Data are as follows: Bowie Co., 5 mi (8 km) W Texarkana, along County Road 1217 off FM 991, d , juv., 10 October 2001; S of Texarkana (Liberty Eylau) off FM 558 along County Road 1370, lOd, 69, 11 October 2001 and 2d, 19 December 2001; Texarkana, Texas A&M University campus off Robison Rd., 3d, 5 November 2001. Cass Co., Atlanta, Ellington Clinic off U.S. Hwy. 59, 2d, 7 November 2002. Marion Co., Jefferson, 2997 FM 728, Cypress Bend Adventist Elementary School, 3d, 23 October 2002, and d, 4 mi (6.4 km) NW Jefferson, 9 November 76 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 2002. All specimens above represent new county records. Habitat at these sites is typical east Texas piney woods, and specimens were encountered while moving along the ground after brief fall showers. Eurymerodesmus angularis is a highly variable and widely ranging species (Fig. 1), and the most proximate prior record to this current study is that from the vicinity of Myrtis, ca. 30 miles (48.3 km) NNW Shreveport, Caddo Parish, Louisiana (Shelley 1990). Despite several efforts, no specimens of E. angularis were encountered in the vicinity of Caddo Lake State Park in adjacent Harrison County, but its presence is anticipated during the cooler and wetter months of fall and winter. Shelley (1990) depicted four gonopodal variants of E. angularis that he considered to be conspecific, and the northeast Texas form is that found in Caddo Parish, with lightly sinuate gonopodal acropodites and an aperture in which the caudolateral "pouch" flares strongly laterally. To date little milliped sampling has taken place in northeast Texas (Stewart 1969). In addition, northeast Texas likely forms the western distribution boundary for a number of "eastern" diplopods and hence justifies more intensive investigation. Recent studies in proximate parts of Arkansas and Oklahoma produced several important discoveries (McAllister et al. 2002a; 2002b; 2003a; 2003b; Shelley et al. 2003), lending credence to this statement. Focused studies on the northeast corner of Texas may be similarly profitable and are a primary objective of future research. Acknowledgments The senior author thanks TAMU-T, particularly Drs. J. Johnson and G. Mueller for providing Faculty Senate Research Enhancement Grants nos. 140000 and 200900 to fund a portion of this study. We also thank James T. McAllister, III (Brookhaven College, Dallas, Texas), and Nancy Solley (TAMU-T) for assistance in collecting. Literature Cited McAllister, C. T., C. S. Harris, R. M. Shelley & J. T. McAllister, HI. 2002a. Millipeds (Arthropoda: Diplopoda) of the Ark-La-Tex. I. New distributional and state records for seven counties of the West Gulf Coastal Plain of Arkansas. J. Arkansas Acad. Sci., 56:91-94. McAllister, C. T., R. M. Shelley & J. T. McAllister, III. 2002b. Millipeds (Arthropoda: Diplopoda) of the Ark-La-Tex. II. Distributional records for some species of western and central Arkansas and eastern and southeastern Oklahoma. J. Arkansas Acad. Sci., 56:95-98. TEXAS J. SCI. 56(1), FEBRUARY, 2004 77 McAllister, C. T., R. M. Shelley & J. T. McAllister, III. 2003a. Millipeds (Arthropoda: Diplopoda) of the Ark-La-Tex. III. Additional records from Arkansas. J. Arkansas Acad. Sci., 57: (In press). McAllister, C. T., R. M. Shelley & J. T. McAllister, III. 2003b. Millipeds (Arthropoda: Diplopoda) of the Ark-La-Tex. IV. New geographic distribution records from southcentral and southeastern Oklahoma. Proc. Oklahoma Acad. Sci., 83:(In press). Shelley, R. M. 1990. Revision of the milliped family Eurymerodesmidae (Polydesmida: Chelodesmidea) . Mem. Amer. Entomol. Soc., 37:1-112. Shelley, R. M., C. T. McAllister & S. B. Smith. 2003. Discovery of the milliped Pleuroloma flavipes Rafinesque in Texas, with a disjunct record from Louisiana, and new localities from west of the Mississippi River (Polydesmida: Xystodesmidae). Entomol. News 11 4: (In press). Stewart, T. C. 1969. Records of millipeds in twenty five northeast Texas counties. Texas J. Sci., 20(4): 383-385. CTM at: chris.mcallister@tamut.edu * * * * * DIET OF THE WHITE-COLLARED SEEDEATER SPOROPHILA TORQUEOLA (PASSERIFORMES: EMBERIZIDAE) IN TEXAS Jack C. Eitniear Center for the Study of Tropical Birds, Inc. 218 Conway Drive San Antonio, Texas 78209-1716 The white-collared seedeater (Sporophila torqueola), is a very small, black and white finch about 11 cm in total length. The species has a distribution from western Panama to the Rio Grande valley of Texas (American Ornithologists’ Union 1998). Sporophila torqueola sharpei occurs from the Rio Grande of Texas, south along the coastal plain of northeastern Mexico to northern Veracruz, and west to eastern Nuevo Leon and San Luis Potosi (American Ornithologists’ Union 1957). Most papers on temperate subspecies of S. torqueola are taxonomic, with virtually nothing written on its natural history, including diet (Eitniear 1997a). This paper summarizes dietary information collected in Texas from 1995-2000. White-collared seedeaters were studied at two sites in Zapata County, Texas. Site 1 was located on the banks of the Rio Grande River within the city of San Ygnacio (27°02’N 99°26’W) in a black willow ( Salix niger) dominated community, with an understory of barnyardgrass (Echinochloa crus-pavonis) , Louisiana cupgrass ( Eriochloa punctata), spreading panicum {Panicum diffusum ), Bermudagrass ( Cynodon dactylon) and Mexican sprangletop ( Leptochloa uninervia). 78 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 1, 2004 Site 2, a marsh bordering a pond in Zapata County Park (26°54’N 099° 1 6’ W), was located within the city of Zapata. The habitat was characterized by Bermudagrass, buffelgrass ( Cenchrus ciliaris ), Guinea- grass ( Panicum maximum ), Johnson grass {Sorghum halepense), south¬ western bristlegrass {Setaria scheelei ), dock {Rumex chrysocarpus ) and cattail {Typha domingensis) . Trees included sugar hackberry {Celtis laevigata ), black willow, huisache {Acacia minuata) and guajillo {Acacia berlandieri) . Plant identifications follow that of Hatch et al. (1990). Methods and Materials Observations were made from April to August 1995 at Site 1 (Eitniear & Rueckle 1995) and August to October 1994, February 1996, April 1997 and April 2000 at Site 2. Observations began at either 0800 h or 1000 h and continued to about 1800 h or 1900 h. Five birds were captured in mist- nets set at the site. Captured birds were leg banded and placed in a holding cage until a fecal sample was caught on blotting paper placed at the bottom of a small field cage. It was assumed these bird’s fecal contents, although biased by a digestive differential of certain foods, provided a representative sample of recently consumed foods. The white uric acid covering was removed by flushing the sample with water. The remaining fecal mass was stored in 70% ethanol. Food items were identified by comparison to a reference collection of seeds and leaves from all plants at the study sites (Smith 1970; Servat 1993). Observations of foraging birds were conducted using 10 by 50 binoculars. Foraging observations were documented in a field notebook and a botanical specimen, from plants that contained seeds fed on, collected. Plant specimens were later identified by Robert Lonard (UT-Pan American). On occasion seeds were obtained from the mouths of captured birds. No effort, however, was made to flush crops. Results and Discussion Items in the diet of the species are summarized in Table 1. The largest foraging group of seedeaters observed consisted of approximately 10 birds feeding on barnyardgrass and Louisiana cupgrass at Site 1. The birds fed throughout the day, frequently retreating to nearby black willows. Females were observed feeding Louisiana cupgrass seeds to recently fledged young at this location (Eitniear & Rueckle 1995). Fecal samples (five samples from five different birds) contained only barnyard and Louisiana cupgrass seeds, thus supporting the theory that grasses were the principle food resource consumed at this time. Green Louisi- TEXAS J. SCI. 56(1), FEBRUARY, 2004 79 Table 1. Parts of 12 plants consumed by Sporophila torqueola sharpei in Zapata, Zapata County, Texas, 1995-2000. Plant Species Part Eriochloa cruz-pavornis (seeds) Panicum maximum* (seeds) Echinochloa punctata (seeds) Panicum diffiusum (seeds) Dichanthium annulafusum* (seeds) Panicum antidotale* (seeds) Cenchrus cilaris* (seeds) Setaria leucopila (seeds) Setaria scheelei (seeds) Acacia minuata (floral parts) Salix nigra (floral parts) Salix exigua (floral parts) *Non-native species ana cupgrass seeds in the milky stage of development were collected from the mouth and outer portions of the mandible of a female caught in a mist net. Plant succession altered this site significantly during the study. Black willow displaced barnyardgrass along the riverbank, and plains bristlegrass, buffelgrass, Guineagrass and blue panicum became established in open areas. Seedeaters at Site 2 were observed feeding on southwestern bristle- grass, barnyardgrass and Louisiana cupgrass. Bermudagrass, Guinea- grass, Johnsongrass and buffelgrass also were abundant, and contained ripe seeds, but not observed to be utilized as a food resource. Although grass seeds dominated observations of white- collared seedeaters diet, at 1200 h on 25 February 1996 at Site 1, a male foraged on huisache blossoms in a tree near the pond. For 30 minutes it was observed consuming the orange globose clusters of stamens. Subsequent to this observation, seedeaters had been observed feeding on the floral parts of willow (Table 1). Bill morphology of the genus Sporophila favors seed eating (Cody 1985). Observations made during this study, although somewhat limited, support this concept. The greater proportion of barnyardgrass in the diet of the white-collared seedeater may reflect the greater abundance of this species over cupgrass and southwestern bristlegrass at Site 2 (Eitniear 1997b). Despite barnyardgrass growing abundantly on the opposite side of the pond at Site 2, seedeaters were never observed feeding on it; perhaps because no cover existed nearby. 80 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Observations of feeding on the floral parts of willow and huisache in addition to records of its feeding on berries in Costa Rica (Stiles & Skutch 1989) and the pulp of Stemmadenia donnell-smithii in Mexico (McMiarmid et al. 1977) indicates greater plasticity in diet than previous authors have indicated (Cody 1985; Rubenstein et al. 1977). More research is needed to determine dietary shifts in this species in relation to changing seasons, variations in precipitation levels and landscapes. Such research may indicate if the decline of this species from a formerly robust widespread species in south Texas to the current patchily distributed remnant population is principally the result of the use of agrochemicals, habitat loss or some other factors (Eitniear & Rueckle 1996; Woodin et al. 1999). Acknowledgments I wish to thank the numerous field assistants that participated in this study, especially Tom Rueckle. Robin Restall (Phelps Collection: Venezuela), Dr. John T. Baccus (Texas State University), Dr. Robert Lonard (University of Texas- Pan American), Dr. Keith Arnold (Texas A&M University, College Station), Dr. Timothy Brush (Uni-versity of Texas-Pan American) and Dr. Kent Rylander (Texas Tech University, Junction) and two anonymous reviewers contributed valuable suggestions to the study and/or manuscript. All birds were captured under permits from the Texas Parks and Wildlife and the National Biological Survey. Literature Cited American Ornithologists’ Union. 1957. Check-list of North American Birds. The American Ornithologists’ Union, Baltimore, Maryland, 691 pp. American Ornithologists’ Union. 1998. Check-list of North American Birds. The American Ornithologists’ Union, Washington D.C., 877 pp. Cody, M. L. 1985. Habitat selection in birds. Academic Press, Inc., New York, 558 pp. Eitniear, J. C. 1997a. White-collared Seedeater ( Sporophila torqueola ) in The Birds of North America, No. 278 (A. Poole and F. Gill, eds). The Academy of Natural Sciences, Philadelphia, PA, and The American Ornithologists’ Union, Washington, D.C., 12 pp. Eitniear, J. C. 1997b. Diet and habitat preference of the White-collared Seedeater (■ Sporophila torqueola sharpei ) in South Texas. Unpublished Master of Science Thesis, Southwest Texas State University, 31 pp. Eitniear, J. C. & T. Rueckle. 1995. Successful nesting of the White-collared Seedeater in Zapata County, Texas. Bull. Tex. Ornithol. Soc., 28:20-22. Eitniear, J. C. & T. Rueckle. 1996. Noteworthy avian breeding records from Zapata County, Texas. Bull. Tex. Ornithol. Soc., 29:16-17. Hatch, S. L., K. N. Gandi & L. E. Brown. 1990. Checklist of the vascular plants of Texas. Tex. Agri. Exper. Station, College Station, Texas, 402 pp. McDiarmid, R. W., R. E. Ricklefs & M. S. Foster. 1977. Dispersal of Stemmadenia donnell-smithii ( Apocynaceae ) by birds, Biotropica 9:9-25. TEXAS J. SCI. 56(1), FEBRUARY, 2004 81 Servat, G. 1993. A new method of preparation to identify arthropods from stomach contents of birds. J. Field Ornithol., 64:49-54. Smith, H. K. 1970. A method of analyzing fox squirrel stomach contents. Tech Series No. 3, Texas Parks and Wildlife Dept., 75 pp. Stiles, G. F. & A. F. Skutch. 1989. A guide to the birds of Costa Rica. Cornell Univ. Press, Ithaca, NY, 511 pp. Woodin, M. C., M. K. Skoruppa, G. W. Blacklock & G. C. Hickman. 1999. Discovery of a second population of white-collared seedeater, Sporophila torqueola (Passeriformes : Emberizidae) along the Rio Grande of Texas. Southwest. Nat., 44(4):535-538. JCE at: JCE@cstbinc.org ***** REPRODUCTION IN THE COFFEE SNAKE, N1NIA MACULATA (SERPENTES: COLUBRIDAE), FROM COSTA RICA Stephen R. Goldberg Department of Biology, Whittier College Whittier, California 90608 The coffee snake, Ninia maculata is known from Honduras, Nicara¬ gua, Costa Rica and Panama from 36-1800 m (Savage 2002). Fitch (1970) reported N, maculata clutch sizes from Cartago Province, Costa Rica. The purpose of this paper is to provide new information on the reproductive cycle from a histological examination of gonads and kidneys and additional data on clutch sizes. A sample of 41 specimens of N. maculata from Costa Rica (females n — 25, mean snout- vent length [SVL] = 226 mm ± 22 SD, range = 175-275 mm; males n = 16, SVL = 201 mm ± 15 SD, range = 179- 228 mm) was examined from the herpetology collection of the Natural History Museum of Los Angeles County, Los Angeles (LACM). Snakes were collected 1959-1996. Counts were made of enlarged ovarian follicles (> 8 mm length) or oviductal eggs. The left testis, vas deferens and a portion of the kidney were removed from males and the left ovary was removed from females for histological examination. Tissues were embedded in paraffin and sectioned at 5 /xm. Slides were stained with Harris’ hematoxylin followed by eosin counterstain. Histological slides were examined to determine the stage of the testicular cycle and for the presence of yolk deposition (secondary vitellogenesis sensu Aldridge 1979). Not all tissues were available for histological examination due to damage or autolysis. Number of tissues histologi- 82 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 Table 1 . Monthly distribution of stages in the ovarian cycle of Ninia maculata from Costa Rica. Values shown are the numbers of females exhibiting each of the four conditions. Month n Inactive Early yolk deposition Enlarged follicles > 12 mm length Oviductal eggs February 2 1 0 0 1 June 3 2 0 0 1 July 3 2 0 0 1 August 4 0 1 1 2 September 4 2 0 2 0 October 1 0 0 1 0 November 8 1 2 3 2 Table 2. Clutch sizes for Ninia maculata (estimated from counts of yolked follicles > 8 mm length or oviductal eggs*) from Costa Rica. Date SVL (mm) Clutch size Province LACM # 11 February 240 3* Cartago 153798 29 June 230 3* Limon 153808 11 July 215 2* Cartago 153828 2-6 August 220 2* Guanacaste 153788 27 August 210 3* San Jose 153857 30 August 220 3 San Jose 153843 15 September 225 4 San Jose 153851 16 September 246 4 Cartago 153802 13 October 213 2 San Jose 153831 10 November 190 1 San Jose 153849 14 November 225 2 San Jose 153856 20 November 223 3* San Jose 153823 20 November 240 5 San Jose 153821 22 November 233 4 San Jose 153835 cally examined by specimen were: testis = 16, vas deferens = 3, kidney = 1 3 , ovary = 11. Follicles in advanced stages of yolk deposition or oviductal eggs were counted, but were not examined histologically. An unpaired Mest was used to compare body sizes of male and female samples. The relationship between female SVL and clutch size was examined by linear regression analysis. Material examined — The following specimens of Ninia maculata were examined by Costa Rica province: CARTAGO (LACM 153787, 153795, 153798, 153799, 153801-153805, 153828), GUANACASTE (LACM 153788, 153789), LIMON (LACM 153807, 153808, 153812), PUNTARENAS (LACM 153790), SAN JOSE (LACM 38063, 38064, 153818, 153819, 153821, 153823, 153824, 153826, 153829, 153831, 153834, 153835, 153839, 153840, 153843, 153844, 153846, 153848-153852, 153856-153858). Testicular histology of N. maculata was similar to that reported by Goldberg & Parker (1975) for two colubrid snakes, Masticophis TEXAS J. SCI. 56(1), FEBRUARY, 2004 83 LO Figure 1 . Linear regression of female body size (mm) versus clutch size for fourteen Ninia maculata from Costa Rica. taeniatus and Pituophis catenifer. All testes examined exhibited spermiogenesis with metamorphosing spermatids and sperm present. The following numbers of males were undergoing spermiogenesis by month: February (3), April (1), June (3), July (2), August (2), September (1), October (2), November (2). All three vasa deferentia examined contained sperm: April (1), July (1), November (1). All thirteen kidney sexual segments examined were enlarged and contained secretory granules: February (2), April (1), June (2), July (2), August (1), September (1), October (2), November (2). Mating usually coincides with enlargement of the kidney sexual segments (Saint Girons 1982). The smallest spermiogenic males measured 179 mm SVL (LACM 153805, 153858). Males smaller than this size were not examined, therefore the minimum size at which N. maculata begins sperm formation is unknown. Females were significantly larger than males (unpaired f-test, t = 4.01, df = 39, P < 0.001). Females with enlarged follicles (> 8 mm length) or oviductal eggs were observed February, June-November (Table 1). The smallest reproductively active N. maculata female (one oviductal egg) measured 190 mm SVL (Table 2), while the three females undergoing early yolk deposition measured 207 mm SVL (14 November, 84 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 LACM 153818), 240 mm SVL (27 August, LACM 153850), 246 mm SVL (22 November, LACM 153852). The minimum size at which N. maculata females commence reproduction remains to be determined. There was no evidence that females produce more than one clutch of eggs in a reproductive season (i.e., oviductal eggs and yolk deposition in progress in the same female) although the presence of reproductively active females during seven months of the year (Table 2) suggests this might be possible. Fitch (1970) reported gravid female N. maculata from Volcan Turrialba, Cartago Province, Costa Rica that measured 187, 206, 218, 222, 231 and 233 mm SVL respectively. A dissected female contained five eggs. One female was collected 2 June and three were collected 30 August. All clutch sizes are listed in Table 2. Mean clutch size for 14 egg clutches from Costa Rica was 2.9 ± 1.1 SD, range = 1-5. Linear regression analysis revealed a significant positive correlation between female body size and clutch size Y = -9.87 -I- 0.06X, r = 0.77, P = 0.001 (Fig. 1). The preceding observations on the ovarian cycle and the presence of males undergoing spermiogenesis during eight months of the year suggests that N. maculata has a prolonged reproductive cycle. Fitch (1970) similarly concluded that N. maculata reproduced throughout much of the year in Costa Rica, if not all of it. Acknowledgments I thank D. A. Kizirian (LACM) for permission to examine specimens, K. R. Beaman (LACM) for his comments and P. Firth for Fig. 1. Literature Cited Aldridge, R. D. 1979. Female reproductive cycles of the snakes Arizona elegans and Crotalus viridis. Herpetologica, 35(3):256-261. Fitch, H. S. 1970. Reproductive cycles in lizards and snakes. Univ. Kansas, Mus. Nat. Hist., Misc. Publ., 52:1-247. Goldberg, S. R. & W. S. Parker. 1975. Seasonal testicular histology of the colubrid snakes, Masticophis taeniatus and Pituophis melanoleucus . Herpetologica, 31(3):317-322. Saint Girons, H. 1982. Reproductive cycles of male snakes and their relationships with climate and female reproductive cycles. Herpetologica, 1 8(3) :5- 16. Savage, J. M. 2002. The amphibians and reptiles of Costa Rica: A herpetofauna between two continents, between two seas. University of Chicago Press, Chicago, Illinois, 934 pp. SRG at: sgoldberg@whittier.edu AUTHOR GUIDELINES 85 INSTRUCTIONS TO AUTHORS Scholarly manuscripts reporting original research results in any field of science or technology will be considered for publication in The Texas Journal of Science . Prior to acceptance, each manuscript will be reviewed both by knowledgeable peers and by the editorial staff. Authors are encouraged to suggest the names and addresses of two potential reviewers to the Manuscript Editor at the time of submission of their manuscript. No manuscript submitted to the Journal is to have been published or submitted elsewhere. Excess authorship is dis¬ couraged. Manuscripts listing more than four authors will be returned to the corresponding author. Upon completion of the peer review process, the corresponding author is required to submit two letter quality copies of the final revised manuscript as well as a diskette (3.5 inch) copy. Format Except for the corresponding author’s address, manuscripts must be double- spaced throughout (including legends and literature cited) and submitted in TRIPLICATE (typed or photocopied) on 8.5 by 11 inch bond paper, with margins of approximately one inch and pages numbered. Scientific names of species should be placed in italics. Words should not be hyphenated. The text can be subdivided into sections as deemed appropriate by the author(s). Possible examples are: Abstract; Methods and Materials; Results; Discussion; Summary or Conclusions; Acknowledgments; Literature Cited. Major internal headings are centered and capitalized. Computer generated manuscripts must be reproduced as letter quality or laser prints. References References must be cited in the text by author and date in chronological (nor alphabetical) order; Jones (1971); Jones (1971; 1975); (Jones 1971); (Jones 1971; 1975); (Jones 1971; Smith 1973; Davis 1975); Jones (1971); Smith (1973); Davis (1975); Smith & Davis (1985); (Smith & Davis 1985). Reference format for more than two authors is Jones et ah (1976) or (Jones et al. 1976). Citations to publications by the same author(s) in the same year should be designated alphabetically (1979a; 1979b). 86 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 1, 2004 Literature Cited Journal abbreviations in the Literature Cited section should follow those listed in BIOSIS Previews ® Database (ISSN: 1044-4297). All libraries receiving Biological Abstracts have this text; it is available from the interlibrary loan officer or head librarian. Otherwise standard recognized abbreviations in the field of study should be used. All citations in the text must be included in the Literature Cited section and vice versa. Consecutively-paged journal volumes and other serials should be cited by volume, number and pagination. Serials with more than one number and that are not consecutively paged should be cited by number as well. The following are examples of a variety of citations: Journals & Serials. — Jones, T. L. 1971. Vegetational patterns in the Guadalupe Mountains, Texas. Am. J. Bot., 76(3): 266-278. Smith, J. D. 1973. Geographic variation in the Seminole bat, Lasiurus seminolus J. Mammal., 54(l):25-38. Smith, J. D., & G. L. Davis. 1985. Bats of the Yucatan Peninsula. Occas. Pap. Mus., Texas Tech Univ., 97:1-36. Books. — Jones, T. L. 1975. An introduction to the study of plants. John Wiley & Sons, New York, 386 pp. Jones, T. L., A. L. Bain & E. C. Burns. 1976. Grasses of Texas. Pp. 205-265, in Native grasses of North America (R. R. Dunn, ed.), Univ. Texas Studies, 205:630 pp. Unpublished. — Davis, G. L. 1975. The mammals of the Mexican state of Yucatan. Unpublished Ph.D. dissertation, Texas Tech Univ. , Lubbock, 396 pp. In the text of the manuscript, the above unpublished reference should be cited as Davis (1975) or (Davis 1975). Unpublished material that cannot be obtained nor reviewed by other investigators (such as unpublished or unpublished field notes) should not be cited. AUTHOR GUIDELINES 87 Graphics, Figures & Tables Every table must be included as a computer generated addendum or appendix of the manuscript. Computer generated figures and graphics must be laser quality and camera ready, reduced to 5.5 in. (14 cm) in width and no more than 8.5 in. (20.5 cm) in height. Shading is unacceptable. Instead, different and contrasting styles of crosshatching, grids, line tints, dot size, or other suitable matrix can denote differences in graphics or figures. Figures, maps and graphs should be reduced to the above graphic measurements by a photographic method. A high contrast black and white process known as a PMT or Camera Copy Print is recommended. Authors unable to provide reduced PMT’s should submit their originals. Do not affix originals or PMT’s to a cardboard backing. Figures and graphs which are too wide to be reduced to the above measurements may be positioned sideways. They should then be reduced to 9 in. (23 cm) wide and 5 in. (12.5 cm) in height. Black and white photographs of specimens, study sites, etc. should comply with the above dimensions for figures. Color photographs cannot be processed at this time. Each figure should be marked on the back with the name of the author(s) and figure number and top of figure. All legends for figures and tables must be typed (double-spaced) on a sheet(s) of paper separate from the text. All figures must be referred to in text as "Figure 3" or "(Fig. 3)"; all tables as "Table 3" or "(Table 3)". Galley Proofs & Reprints The corresponding author will receive galley proofs prior to the final publishing of the manuscript. Proofs must be corrected and returned to the Managing Editor within five days; failure to return corrected galley proofs promptly will result in delay of publication. The Academy will provide 100 reprints without charge for each feature article or note published in the Journal. These will be mailed to the corresponding author or the designated contact person following the publishing of each issue of the Journal. The distribution of reprints among coauthors is the responsibility of the cor¬ responding author. Authors will have the opportunity to purchase additional reprints (in lots of 100) at the time that the corrected galley proofs are returned to the Managing Editor. Voucher Specimens When appropriate, such as new records, noteworthy range extensions, or faunal or floral listings for an area, the author(s) should provide proper information (to include accession numbers) relative to the deposition of voucher specimens. Specimens should be placed with the holdings of a recognized regional or national museum or herbarium. The name(s) and designated initials used by the museum should be given as part of the introduction or methods section. Do not site the deposition of voucher specimens in personal collections. 88 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 1, 2004 The Editorial Staff is very aware that many members of the Academy work with organisms that are protected by state or federal regulations. As such, it may not be possible to collect nor deposit these specimens as vouchers. In the interest of maintaining credibility, authors are expected to provide some alternate means of verification such as black and white photographs, list of weights or measurements, etc. The Editorial Staff retains the option to determine the validity of a record or report in the absence of documentation with a voucher specimen. Page Charges Authors are required to pay $50 per printed page. While members of the Academy are allowed four published pages per year free of charge on publications with one or two authors, all authors with means or institutional support are requested to pay full page charges. Full payment is required for all pages in excess of four. All publications authored by three or four persons will incur full page charges. Nonmembers of the Academy are required to pay full page charges for all pages. The Academy, upon written request, will subsidize a limited number of pages per volume. These exceptions are, however, generally limited to students without financial support. Should a problem arise relative to page charges, please contact: Dr. Ned E. Strenth TJS Managing Editor Department of Biology Angelo State University San Angelo, Texas 76909 E-mail: ned.strenth@angelo.edu Additional Guidelines An expanded version of the above author guidelines which includes instructions on style, title and abstract preparation, deposition of voucher specimens, and a fisting of standardized abbreviations is available on the Academy’s homepage at: www . texasacademy ofscience . org A hard copy is also available upon request from: Dr. Robert J. Edwards TJS Manuscript Editor Department of Biology University of Texas-Pan American Edinburg, Texas 78541 E-mail: redwards@panam.edu THE TEXAS ACADEMY OF SCIENCE, 2003-2004 OFFICERS President : President Elect : Vice-President : Immediate Past President : Executive Secretary : Corresponding Secretary : Managing Editor: Manuscript Editor : Treasurer : Council Representative : John T. Sieben, Texas Lutheran University John A. Ward, Brook Army Medical Center Damon E. Waitt, Lady Bird Johnson Wildflower Center Larry D. McKinney, Texas Parks and Wildlife Department Fred Stevens, Schreiner University Deborah D. Hettinger, Texas Lutheran University Ned E. Strenth, Angelo State University Robert J. Edwards, University of Texas-Pan American James W. Westgate, Lamar University Sandra S. West, Southwest Texas State University DIRECTORS 2001 David S. Marsh, Angelo State University Felipe Chavez-Ramirez, International Crane Foundation 2002 Sushma Krishnamurthy, Texas A&M International University Raymond D. Mathews, Jr., Texas Water Development Board 2003 Hudson R. DeYoe, University of Texas-Pan American Cynthia Contreras, Texas Parks and Wildlife Department SECTIONAL CHAIRPERSONS Anthropology : Roy B. Brown, Instituto Nacional de Antropologia y Historia Biological Science : David S. Marsh, Angelo State University Botany : Joan E. N. Hudson, Sam Houston State University Chemistry : Benny E. Arney, Jr., Sam Houston State University Computer Science: Laura J. Baker, St. Edwards University Conservation and Management: Andrew C. Kasner, Lamar University Environmental Science: William Thomann, University of Incarnate Word Freshwater and Marine Science: Thomas Whelan III, University of Texas-Pan American Geology and Geography: Joe Satterfield, Angelo State University Mathematics: Patrick L. Odell, Baylor University Physics: David Bixler, Angelo State University Science Education: Jimmy Hand, Austin, Texas Systematics and Evolutionary Biology: Allan Hook, St. Edward’s University Terrestrial Ecology: Jerry Cook, Sam Houston State University Threatened or Endangered Species: Flo M. Oxley, Lady Bird Johnson Wildflower Center COUNSELORS Collegiate Academy: Jim Mills, St. Edward’s University Junior Academy: Vince Schielack, Texas A&M University Nancy Magnussen, Texas A&M University THE TEXAS JOURNAL OF SCIENCE PrinTech, Box 43151 Lubbock, Texas 79409-3151 PERIODICAL POSTAGE PAID AT LUBBOCK TEXAS 79402 RETURN SERVICE REQUESTED SMITHSONIAN INSTITUTION LIBRARIES NHB25MRC 154 PO BOX 37012 WASHINGTON, DC 20013-7012 -t ‘f-yC A Iff THE TEXAS JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. For more information regarding membership, student awards, section chairs and vice-chairs, the annual March meeting and author instructions, please access the Academy’s homepage at: www . texasacademyofscience . org Dues for regular members are $30.00 annually; supporting members, $60.00; sustaining members, $100.00; patron members, $150.00; associate (student) members, $15.00; family members, $35.00; affiliate members, $5.00; emeritus members, $10.00; corporate members, $250.00 annually. Library subscription rate is $50.00 annually. The Texas Journal of Science is a quarterly publication of The Texas Academy of Science and is sent to most members and all subscribers. Payment of dues, changes of address and inquiries regarding missing or back issues should be sent to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 E-mail : FStevens@schreiner . edu The Texas Journal of Science (ISSN 0040-4403) is published quarterly at Lubbock, Texas, U.S.A. Periodicals postage paid at San Angelo, Texas and additional mailing offices. POSTMASTER: Send address changes and returned copies to The Texas Journal of Science, Dr. Fred Stevens, CMB 5980, Schreiner University, Kerrville, Texas 78028-5697, U.S.A. The known office of publication for The Texas Journal of Science is the Department of Biology, Angelo State University, San Angelo, Texas 76909; Dr. Ned E. Strenth, Managing Editor. COPYRIGHT POLICY All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, recording or otherwise, without the prior permission of the Managing Editor of the Texas Journal of Science. THE TEXAS JOURNAL OF SCIENCE Volume 56, No. 2 May, 2004 CONTENTS Natural Source of Arsenic in East Texas Lake Sediments. By Kathy Judy, E. B . Ledger and C. A. Barker . 91 Community Ecology of Freshwater, Brackish and Salt Marshes of the Rio Grande Delta. By Frank W. Judd and Robert /. Lonard . 103 Physiological Tolerance Ranges of Larval Caenis latipennis (Ephemeroptera: Caenidae) in Response to Fluctuations in Dissolved Oxygen Concentration, pH and Temperature. By Robert T. Puckett and Jerry L. Cook . 123 Natural History of the Southern Plains Woodrat Neotoma micropus (Rodentia: Muridae) from Southern Texas. By John R. Suchecki, Donald C. Ruthven, Ill, Charles F. Fulhorst and Robert D. Bradley . 131 Adult Foraging Behavior in Meams’ Grasshopper Mouse, Onychomys arenicola (Rodentia: Muridae) is Influenced by Early Olfactory Experience. By Fred Punzo . 141 Robotics Repeatability and Accuracy: Another Approach. By Jan Brink, Bill Hinds and Alan Haney . 149 Historical Population Dynamics of Red Snapper ( Lutjanus campechanus) in the Northern Gulf of Mexico. By J. R. Gold and C. P. Burridge . 157 General Notes Notes on Reproduction in the False Coral Snakes, Erythrolamprus bizona and Erythrolamprus mimus (Serpentes: Colubridae) from Costa Rica. By Stephen R. Goldberg . 171 A New Distribution Record and Notes on the Biology of the Brittle Star Ophiactis simplex (Echinodermata: Ophiuroidea) in Texas. By Ana Beardsley Christensen . 175 First Definitive Record of more than Two Nesting Attempts by Wild White- winged Doves in a Single Breeding Season. By Cynthia L. Schaefer, Michael F. Small, John T. Baccus and Roy D. Welch . 179 Annual Meeting Notice for 2005 . . . . 183 Membership Application . 184 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Managing Editor: Ned E. Strenth, Angelo State University Manuscript Editor: Robert J. Edwards, University of Texas-Pan American Associate Editor for Botany: Janis K. Bush, The University of Texas at San Antonio Associate Editor for Chemistry: John R. Villarreal, The University of Texas-Pan American Associate Editor for Computer Science: Nelson Passos, Midwestern State University Associate Editor for Environmental Science: Thomas LaPoint, University of North Texas Associate Editor for Geology: Ernest L. Lundelius, University of Texas at Austin Associate Editor for Mathematics and Statistics: E. Donice McCune, Stephen F. Austin State University Associate Editor for Physics: Charles W. Myles, Texas Tech University Manuscripts intended for publication in the Journal should be submitted in TRIPLICATE to: Dr. Robert J. Edwards TJS Manuscript Editor Department of Biology University of Texas-Pan American Edinburg, Texas 78541 red wards@panam . edu Scholarly papers reporting original research results 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 Manuscript Editor at the above address. They are also available on the Academy’s homepage at: www . texasacademy ofscience . org AFFILIATED ORGANIZATIONS American Association for the Advancement of Science, Texas Council of Elementary Science Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers Texas Society of Mammalogists TEXAS J. SCI. 56(2):91-102 MAY, 2004 NATURAL SOURCE OF ARSENIC IN EAST TEXAS LAKE SEDIMENTS Kathy Judy*, E. B. Ledger and C. A. Barker * Department of Geology, Blinn College Bryan, Texas 77805 and Department of Geology, Stephen F. Austin State University Nacogdoches , Texas 75962 Abstract.— Elevated arsenic levels occur in the sediment of several east Texas reservoirs. Eight reservoirs exceed the statewide 85th percentile of 17 mg/kg dry weight for arsenic in lake sediment. Average arsenic concentrations in the sediments of these lakes ranges from 19.5-83.5 mg/kg. The source of the arsenic is the marine mudstone formations which crop out in east Texas. Arsenic is common in marine mudstone where it substitutes for sulfur in the mineral pyrite. Unusually high levels of arsenic (up to 122 mg/kg compared to a global average of 13 rag/kg) are known to occur in the Weches Formation in east Texas. Other east Texas marine mudstone formations have not been analyzed for arsenic content. Oxidation of arsenic-bearing pyrite produces acid sulfate conditions, precipitated Fe(OH)3 and oxidized arsenic species. Arsenic species readily adsorb to Fe(OH)3 which is transported to reservoirs by streams and incorporated into the sediment. Arsenic has recently been found to occur at elevated levels in some east Texas rock units (Ledger & Judy 2003). It probably substitutes for sulfur in the ubiquitous mineral pyrite. Pyrite occurs in a variety of geologic settings, including marine mudstone formations in which iron and sulfur were both present and conditions were sufficiently anaerobic to reduce them. This type of depositional environment was present at times in east Texas during the Eocene. Present day exposure of py rite¬ bearing mudstone formations to oxygenated surface and ground water oxidizes the pyrite and releases arsenic into the environment. Monitor¬ ing of streams and lakes by the Texas Commission on Environmental Quality (TCEQ) generally shows levels of arsenic in lake water well below the MCL (Maximum Contaminant Level) established by the EPA. However, elevated arsenic levels occur in the sediment of several east Texas reservoirs. Geologic Setting The Claiborne Group consists of a thick series of cyclic transgressive/ regressive sedimentary strata deposited in east Texas during the middle Eocene (Deussen 1911; Dumble 1918; Berg 1970; Collins 1980; Collins 1982). The Queen City Sand, Sparta Sand, Carrizo Sand and Yegua Formations are composed of fine to medium grained sand deposited in 92 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 a nearshore environment. The Reklaw Formation, Weches Formation and Cook Mountain Formation are composed primarily of mudstone deposited in a quiet marine environment such as a lagoon or shelf (Figure 1). There are few data available, but the pyrite content of the mudstone formations appears to vary laterally and can be appreciable. Selected hand specimens from the southern part of the Weches Formation contain as much as 10% pyrite with some crystals being up to a few millimeters in diameter. Further north, pyrite is rare, while siderite (FeC03) is abundant. The arsenic content is virtually unknown, but likely to be high where pyrite is abundant. Eight samples from a road cut near Nacogdoches, Texas average almost 100 mg/kg arsenic (Ledger & Judy 2003) com¬ pared to a global average shale value of 13 mg/kg. Present day weathering of the mudstones occurs most rapidly where the formations crop out or are near the surface. This process releases soluble arsenic oxides into ground and surface water. Past structural events have affected the outcrop patterns, stream patterns, and even the deposition of east Texas rock units. Most of the rock layers in the eastern half of Texas dip gently to the southeast, toward the Gulf of Mexico. However, the dip rate flattens out and then reverses to north¬ west or west dip on the Texas side of the Sabine Uplift, a circular regional structure located in northeast Texas and northwest Louisiana over a basement high (Nicolas & Waddell 1989). An uplift is an area where deep rocks have been pushed upward. The zone of flat to re¬ versed dip on the flank of the Sabine Uplift causes the Weches, and other possibly arsenic bearing formations, to have a much wider outcrop area than they would have otherwise. Jackson & Laubach (1991) con¬ cluded that the Sabine area was uplifted about 170m during the middle of the Cretaceous, and that a second episode of uplift occurred early in the Eocene. Three major fault systems also affect east Texas rock outcrops: the Mt. Enterprise, Mexia and Talco fault zones. These fault systems consist of down-dropped grabens bounded by normal faults which formed when overloading of sedimentary rock deposits above the unstable low-density Louann Salt caused the salt to flow and intrude upward into areas of lesser pressure (Jackson & Wilson 1982). The Mt. JUDY, LEDGER & BARKER 93 Figure 1. Stratigraphic column of the Middle Eocene Claiborne Group of east Texas (modified from Satin & Brooks 1977). Enterprise fault system is a linear zone of grabens trending east north¬ east that are bounded by growth faults that were active during the time of sediment deposition (Ferguson 1984). Structural control of stream drainage patterns shows up on detailed maps as stream segments aligned with faults and grabens (Baumgardner 1987). Fault and joint fracture planes are primary conduits for movement of ground water through otherwise impermeable mudstone layers and thus may exert significant control on the localization of arsenic, iron and other elements. Results of Weathering Oxidation of pyrite produces Fe(III) and acid sulfate conditions. Fe(III) is mobile below about pH 3-4. At higher pH, Fe(III) quickly hydrolyzes to precipitate as amorphous Fe(OH)3, a red, colloidal gel. 94 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 This is easily transported by streams as suspended or bed load and settles out in calmer lake settings. Initial breakdown of pyrite underground: 2FeS2 + 2H20 + 702 = > 2Fe2+ + 4S042 + 4H+ Oxidation and hydrolysis of Fe2+ in contact with atmosphere: 4Fe2+ + 02 + 10H2O = > 4Fe(OH)3 + 8H+ Under strong oxidizing conditions, As(V) is thermodynamically stable, but the As(III)/As(V) transformation occurs at such a slow rate that both species are usually present. H2As03 and H2As04' are the most abundant species in well oxygenated surface water between pH 3-7. Arsenic species readily adsorb to Fe-oxides and clay minerals and become incorporated in the sediment of streams and lakes. Rates of Weathering The rate at which pyrite oxidizes in natural environmental systems is usually accelerated by the action of sulfur and iron oxidizing bacteria such as Thiobacillus sp. , Ferrobacillus sp. , Gallionella , Sphaerotilus and others (Langmuir 1997). Rates of oxidation caused by bacterial catalysis vary greatly depending on pH, surface area of pyrite, dissolved oxygen concentration and other factors. However, the rate increase is com¬ monly in orders of magnitude (Olson 1991; Stumm & Morgan 1996; Edwards et. al. 1998). Such rapid oxidation results in pH levels low enough that Fe(OH)3 does not form and arsenic species are mobile in ground or surface waters. Judy (1999) measured pH as low as 3.95 in distilled water mixed with dried samples of the Reklaw formation. Screening Levels for Arsenic Currently, no federal or state standards for allowable levels of arsenic in lake sediments exist. The National Oceanic and Atmospheric Admin¬ istration (NOAA 1999) has established probable effects levels (PELs) for substances at which they are likely to be toxic. For arsenic in lake sediment, the PEL is 32.7 mg/kg. To identify water bodies with ele¬ vated sediment metals concentrations, the TCEQ uses a statewide 85th percentile. These are derived from long-term monitoring data and indicate concentrations below which 85 % of measurements occur. State- JUDY, LEDGER & BARKER 95 Table 1 . Average concentration of arsenic in sediment (mg/kg) for twenty-one lakes in east Texas (data provided by the TCEQ, 1985-003). Reservoir Average Concentration of Arsenic in Sediment (mg/kg) Number of Samples Lake Nacogdoches 83.5 2 Lake Jacksonville 53.8 4 Sam Rayburn Reservoir 34.1 22 Lake Cherokee 31.0 2 Ellison Creek Reservoir 30.3 7 Pinkston Reservoir 28.0 1 Lake Tyler East 24.1 4 Lake Tyler 20.0 4 Lake Palestine 10.3 6 Lake O’ the Pines 8.9 6 Martin Lake 8.8 3 Caddo Lake 8.7 6 Wright Patman Lake 6.1 10 Lake Monticello 5.8 4 Houston County Lake 5.8 1 Lake Bob Sandlin 5.3 13 Lake Cypress Springs 4.2 16 Toledo Bend Reservoir 3.3 9 Lake Fork Reservoir 2.7 5 Lake Murvaul 2.5 2 Lake Tawakoni 1.7 3 wide 85th percentiles indicate areas where metals concentrations are elevated and are not based on negative biological effects. For arsenic in sediment in reservoirs, the statewide 85th percentile is 17 mg/kg, close to the global average of 13 ppm for shale. Methods All data for arsenic levels in lake sediments were provided by the Texas Commission on Environmental Quality (TCEQ) and are available to the public. If available, data collected between 1 January 2000 and 1 April 2003 were used. Some lakes were not monitored for arsenic in sediment during this time period. For these, data acquired between 1985 and 2000 were used. Surface outcrops of the Weches Formation, Reklaw Formation and Cook Mountain Formation are those shown on the Geologic Atlas of Texas Texarkana Sheet (Barnes 1979), Palestine Sheet (Barnes 1993) and Tyler Sheet (Barnes 1975). Stream drainage patterns were illus¬ trated based on the Geologic Atlas of Texas and USGS topographic maps. 96 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Results and Discussion In general, lakes receiving substantial discharge from streams flowing through mudstone formations have elevated levels of arsenic in their sediments. Eight of the twenty-one lakes for which arsenic in sediment data are available exceed the statewide 85th percentile of 17 mg/kg and three exceed the PEL of 32.7 mg/kg (Table 1). Four of these: Lake Nacogdoches (Figure 2), Lake Jacksonville (Figure 3), Lake Tyler (Figure 4) and Ellison Creek Reservoir (Figure 5) are near outcrops of the Weches Formation and are fed by streams which flow through it. Lake Cherokee (Figure 6) is fed by discharge from streams flowing through outcrops of the Reklaw Formation which may contain elevated arsenic levels. Sam Rayburn Reservoir (Figure 7) is fed by large streams which flow across the Weches, Reklaw and Cook Mountain Formations. Lake Tyler East (Figure 4) and Pinkston Reservoir have elevated sediment arsenic levels but do not have a source that is apparent on the geologic map. The remaining thirteen lakes are all well below the statewide 85th percentile. Ten of these are fed by streams which flow primarily across sand formations. The remaining three: Lake Palestine (Figure 8); Lake O’ the Pines (Figure 5); and Houston County Lake receive some stream drainage from mudstone outcrops, but do not show elevated levels of arsenic in their sediment. Individual study of the Eve lakes which appear to be anomalous is likely to reveal a simple explanation for the levels of arsenic present. For example, Lake O’ the Pines (Figure 5) is near the northern Weches in which siderite formed and pyrite is rare. Field research by the authors found that surface outcrops in this area are very thin, only a few feet in some locations. Also, small reservoirs are present on the two major streams flowing across the Weches Formation into Lake O’ the Pines. These would trap sediment before it gets to the lake. Therefore, it is seems that arsenic is either not present, not abundant, or is being trapped in the smaller reservoirs. The proximity of a reservoir to mudstone outcrops is not a perfect predictor of elevated arsenic levels in lake sediments. However, the correlation observed here suggests that this would be useful in deciding which lakes to most closely monitor. JUDY, LEDGER & BARKER 97 Miles Ew Weches Formation Figure 2. Lake Nacogdoches, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrow with number indicates average concentration of arsenic in sediments in mg/kg dry weight at sampling sites. Figure 3. Lake Jacksonville, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. 98 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Evv Weches Formation Figure 4. Lake Tyler and Lake Tyler East, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. Figure 5. Ellison Creek Reservoir and Lake O’ the Pines, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. JUDY, LEDGER & BARKER 99 Er Reklaw Formation Figure 6. Lake Cherokee, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. Figure 7. Sam Rayburn Reservoir, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. 100 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Figure 8. Lake Palestine, Texas. Arsenic-bearing formation outcrop is shown in dark gray. Arrows with numbers indicate average concentrations of arsenic in sediments in mg/kg dry weight at sampling sites. Arsenic in lake sediments is not bioavailable to pelagic organisms or organisms that drink the lake water. Its possible effects on benthic organisms may be a field of future study. An interesting and un¬ answered question is whether or not arsenic is bioavailable at any time between the initial weathering of arsenic-bearing pyrite and the deposi¬ tion of Fe(OH)3 with adsorbed arsenic species. JUDY, LEDGER & BARKER 101 Acknowledgments We thank Ken Farrish and Chris Mathewson for comments on an earlier draft of the manuscript. We thank the TCEQ for their very well organized system for managing information and making it available to the public. Literature Cited Barnes, V. E. 1993. Geologic atlas of Texas: Palestine Sheet. Austin, Texas, Bureau of Economic Geology, The University of Texas at Austin, 1:250 000, 1 sheet. Barnes, V. E. 1979. Geologic atlas of Texas: Texarkana Sheet. Austin, Texas, Bureau of Economic Geology, The University of Texas at Austin, 1:250 000, 1 sheet. Barnes, V. E. 1975. Geologic atlas of Texas: Tyler Sheet. Austin, Texas, Bureau of Economic Geology, The University of Texas at Austin, 1:250 000, 1 sheet. Baumgardner, R. W., Jr. 1987. Landsat-based lineament analysis, east Texas basin and Sabine Uplift area. The University of Texas at Austin, Bureau of Economic Geology, 167:1-26. Berg, R. R. 1970. Outcrops of the Claiborne Group in the Brazos Valley, Southeast Texas. Guidebook, Texas A&M Univ. Dept, of Geology, College Station, 54 pp. Collins, A. M. 1982. Petrology of the Eocene Marquez Shale Member of the Reklaw Formation, Bastrop County, Texas. Unpublished M.S. Thesis, The Univ. of Texas, Austin, 142 pp. Collins, E. W. 1980. The Reklaw Formation of east Texas. Pp. 67-70, in Middle Eocene coastal and nearshore deposits of east Texas, a field guide to the Queen City Formation and related papers (B.F. Perkins, and D.K. Hobday, eds.), SEPM, Tulsa, Oklahoma, 95 pp. Deussen, A. 1911. Notes on some clay from Texas. US Geological Survey Bulletin, 470:302-351. Dumble, E. T. 1918. The geology of east Texas. University of Texas Bulletin, 1869, 388 pp. Edwards, K. J., M. O. Schrenk, R. Hamers & J. Banfield. 1998. Microbial oxidation of pyrite: Experiments using microorganisms from an extreme acidic environment. Am. Min., 83:1444-1453. Ferguson, J. D. 1984. Jurassic age salt tectonism within the Mt. Enterprise Fault System, Rusk County, Texas. Pp. 157-161, in The Jurassic of East Texas (M. W. Presley, ed.), East Texas Geological Society, 304 pp. Jackson, M. L. W. & S. E. Laubach. 1991. Structural history and origin of the Sabine arch, east Texas and northwest Louisiana. The University of Texas at Austin, Bureau of Economic Geology, Geological Circular, 91-3:1-47. Jackson, M. P. A. & B. D. Wilson. 1982. Fault tectonics of the east Texas basin. The University of Texas at Austin, Bureau of Economic Geology, 31 pp. Judy, K. 1999. Clay mineralogy and electrical conductivity of the Claiborne Group, Eastern Texas. Unpublished M.S. Thesis, Stephen F. Austin State Univ., Nacogdoches, 124 pp. Langmuir, D. 1997. Aqueous environmental geochemistry. Prentice Hall, New Jersey, vii-f 600 pp. Ledger, E. B. & K. Judy. 2003. Elevated arsenic levels in the Weches Formation, Nacogdoches County, Texas, GSAGS Transactions, in press. Nicolas, R. L., & D. E. Waddell. 1989. The Ouachita system in the subsurface of Texas, 102 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Arkansas, and Louisiana. Pp. 661-672 in The Geology of North America, Vol. F-2, The Appalachian-Ouachita Orogen in the United States (R.D. Hatcher, Jr. , W. A. Thomas, and G.W. Viele, eds.), Geological Soci ety of America, xiii-t-767 pp. NOAA. 1999. NOAA Screening Quick Reference Tables (SquiRTs). NOAA, Seattle, Washington. Olson, G. J. 1991. Rate of pyrite bioleaching by Thiobacillus ferrooxidans : Results of an interlaboratory comparison. Applied and Environmental Microbiology, 57:642-644. Sartin, A. A. & E. C. Brooks. 1977. Heavy mineral analysis of Queen City and Sparta Formations (Eocene) in east Texas, The Compass of Sigma Gamma Epsilon, 54:72-77 Stumm, W. & J. J. Morgan. 1996. Aquatic Chemistry 3d ed. Wiley-Interscience, New York, xvi + 1022 pp. EBL at: eledger@sfasu.edu TEXAS J. SCI. 56(2): 103-122 MAY, 2004 COMMUNITY ECOLOGY OF FRESHWATER, BRACKISH AND SALT MARSHES OF THE RIO GRANDE DELTA Frank W. Judd and Robert I. Lonard Department of Biology University of Texas-Pan American Edinburg, Texas 78541-2999 Abstract.-— Species composition and importance, species diversity and evenness, species richness, and community similarity are compared among 6 freshwater, 9 brackish and 1 1 salt marshes in the Rio Grande Delta. Community similarity is generally low among marshes, but salt marshes have a greater mean coefficient of similarity than brackish marshes. Species richness per marsh ranges from 15 to 31 for freshwater marshes, 7 to 24 for brackish marshes and 7 to 26 for salt marshes. Each freshwater marsh has a different dominant species. The first six species in importance in all three kinds of marshes contribute from 72.6 to 99.8% of the relative cover. Thus, most species are of low importance. There is no significant difference in species richness, species diversity or evenness among the three kinds of marshes. The generalization of the relationships found in this study awaits additional information on marshes from other areas of the Texas coast. The physiography of southern Texas is characterized by offshore barrier islands, an enclosed lagoon (Laguna Madre), and the delta of the Rio Grande on the Texas mainland. The base of the delta is about 46 km long extending from Port Mansfield in Willacy County to the mouth of the Rio Grande in Cameron County. The apex of the delta is located approximately 66 km inland from the Gulf of Mexico (Brown et al. 1980). Prior to the construction of dams, floodways and levees, the Rio Grande overflowed its banks annually depositing new sediment and moving water into a variety of meander channels in the delta. These flood waters constituted significant freshwater input into the wetlands of the Rio Grande Delta. However, in the past 50 years dams and flood control projects have eliminated this source of freshwater (Jahrsdoerfer & Leslie 1988) and the wetlands are now dependent on rainfall alone for freshwater input. Unlike streams of the upper and central Texas coast, the Rio Grande does not have associated swamps or freshwater marshes (White et al. 1986). Rather, there is a gradational array of infrequently to permanent¬ ly inundated wetlands in the Rio Grande Delta. Brackish marshes are common because: (1) evaporation exceeds precipitation, (2) prevailing southeasterly winds carry salt spray inland from the Laguna Madre and, (3) extremely high storm tides flow inland along drainage courses during 104 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 hurricanes (Brown et al. 1980). Salt marshes are less common and less extensive because wind-tidal flats occupy the areas of the delta that are typically occupied by salt marshes on the central and upper Texas Coast (Brown et al. 1980). Freshwater marshes are even more uncommon because of the absence of freshwater input by river overflow and low annual rainfall. Little information is available on the marshes of the Rio Grande Delta. White et al. (1986) used color- infrared photographs to identify and classify wetlands in the delta. They recognized seven major kinds of wetlands including freshwater, brackish and salt marshes. Kinds of marshes were distinguished based on elevation, vegetation and soil and surface moisture. Lists of species characteristic of each type of marsh were provided, but many of the species used to characterize the vegeta¬ tion of a given kind of marsh were also listed as characteristic of one or both of the other types of marsh. There was no quantification of species abundance or diversity. Johnston (1955) recognized differing marsh communities along an elevation gradient. He reported that at low elevations a community comprised of Bat is maritima , Salicomia virginica and Suaeda linearis graded almost imperceptibly into slightly higher elevations characterized by Borrichia frutescens, B. maritima and Monanthochloe littoralis, which in turn graded into a community of Spartina spartinae. Judd et al. (1997a) used multispectral videography to distinguish the pattern of zonation and species composition in a brackish marsh at Laguna Atascosa National Wildlife Refuge (LANWR), Cameron County, Texas. At the lowest elevations there was a distinct zone dominated by maritime saltwort, B. maritima. Intermediate elevations supported a zone dominated by shoregrass, M. littoralis. At the highest elevations the third zone was dominated by Gulf cordgrass, S. spartinae. The upper margin of this zone graded into a shrub-grassland community that occurred on lomas (clay dunes). A salt marsh also was organized into three zones along an elevation gradient and had the same dominant species in each zone (Judd et al. 1997b). Judd & Lonard (2002) compared species richness and diversity in a brackish and salt marsh at LANWR. Forty-seven species were present in the two marshes, but only 15 were common to both. Monanthochloe littoralis and B. maritima were the dominant species in the brackish marsh and S . spartinae was dominant in the salt marsh. In both marshes, four species contributed from 73% to 86% of the cover. Consequently, most species contributed little to vegetation abundance and community structure. There were no significant differences in species diversity within marshes JUDD & LONARD 105 between years or between marshes within a year. Lonard & Judd (1999) catalogued the vascular plant species found in fresh, brackish and salt marshes in the Rio Grande Delta based on a survey of 27 marshes. They found 84 species representing 27 families were present. Thirty-five species were limited to freshwater marshes and 12 species were limited to salt marshes. No species were unique to brackish marshes. Occurrence in fresh, brackish and salt marshes was provided for each species, but there was no quantification of abundance or comparison of species richness or community similarity among the kinds of marshes. Marshes of the Rio Grande Delta provide critical habitat for numerous waterfowl species and several threatened and endangered mammalian species. It is important to know the composition, structure, species diversity and fidelity of marsh communities in the Rio Grande Delta to facilitate re-establishment of native vegetation at disturbed sites and to facilitate wise management decisions relative to providing appropriate habitat for marsh fauna. To date, quantified information on species abundance, diversity and community similarity are available for only one brackish and one salt marsh in the Rio Grande Delta. Herein, this study reports on the species composition, species diversity and species richness of 6 freshwater, 9 brackish and 1 1 salt marshes in the Rio Grande Delta. Community similarity, dominant species, species richness, species diversity and evenness are compared among these marshes. Materials and Methods The locations of marshes studied are given in Table 1. The line intercept method (Canfield 1941) was used to quantify species abun¬ dance. The number of transects sampled at each site was dependent upon the size and configuration of the wetland basin. A minimum of two and a maximum of 10 transects were sampled at the marshes. Transects were established along an elevation gradient extending from the low point in the marsh until an interval with upland vegetation (trees and shrubs) was encountered. Each transect was divided into 10 m intervals and readings were taken along the length of each interval. Each species intercepted by the line was rated individually and was recorded without separation into strata (i.e., tree, shrub and ground layers). Species and foliage cover were recorded and from these data the frequency of occurrence, relative frequency, relative cover and an importance value which is the sum of relative frequency and relative cover were calculated. The importance value was used to determine dominant species. 106 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 Table 1. Marshes studied, their locations and mean salinities. NWR = National Wildlife Refuge, TPWD = Texas Parks and Wildlife Department. LANWR = Laguna Atascosa National Wildlife Refuge, NPS = National Park Service. Marsh Location Mean Salinity Freshwater Marshes Paso Real, TPWD 26° 18’55.56" N, 97°31’27.48" W 0.5 Russelltown 26°04’51 .25" N, 97°34’52.50" W 0.5 Resaca de la Palma, TPWD 25°58’32.86" N, 97°34’00.76" W 0.0 Audubon Sabal Palm Sanctuary 25”5r00.76" N, 97°25’07. 15" W 0.0 Cattail Lake, Santa Ana NWR 26°04’32.41" N, 98°09’14. 15" W 0.0 Willow Lake, Santa Ana NWR 26°05’00.72" N, 98°08’ 18.79" W 0.5 Brackish Marshes Palo Alto #1, NPS 26°0ri7.43" N, 97°28’ 12.26" W 6.0 Palo Alto #2, NPS 26°00’ 18.04" N, 97°27’18.55" W 6.0 Laguna Atascosa NWR Resaca 26° 10’21 .00" N, 97° 19’53.55" W 17.0 Olmito Resaca 26°00’48.75" N, 97°32’30.14" W 2.3 Tio Cano #1, NWR 26°12’37.01" N, 97°48’50.43 " W 4.5 Tio Cano #2, NWR 26° 12’39.36" N, 97°48’47.82" W 3.8 Bay view Resaca #1 26°07’57.80" N, 97°22’56.08" W 6.2 Bayview Resaca #2 26° 10’31 .67" N, 97°22’59.75" W 9.0 Willamar 26°23’16.56" N, 97°34’59.66" W dry Salt Marshes Stover Point, LANWR 26° 13’01 .00" N, 97° 19’00.00" W 44.8 Spillway Crossing, LANWR 26° 16’00.00" N, 97°23’44.09" W 22.0 Large Marsh, LANWR 26° 12’50.79" N, 97° 19’52.06" W 20.5 Dry Marsh, LANWR 26° 13’00.39" N, 97° 19’02.46" W dry Osprey Point, LANWR 26° 13’58.32" N, 97°21’01.97" W 51.0 Laguna Atascosa Cayo, LANWR 26° 14’45.55" N, 97°25’13.12" W 22.0 Redhead Ridge, LANWR 26° 10’27.74" N, 97° 18’15.67" W 55.6 Rangerville #1, TPWD 26°05’ 17.22" N, 97°44’25.02" W 25.0 Rangerville *2, TPWD 26°05’08.78" N, 97°44’41 .65" W 22.0 Bayview Dry Marsh 26° 10’20.32" N, 97°22’55.51" W 33.0 Bayview Brine Marsh 26° 10’19.51" N, 97°23’59.73" W 67.5 Similarity of species composition among marshes was calculated using Sorensen’s Coefficient of Community (Krebs 1999). Species importance value was used as the measure of abundance for calculating species diversity indices. Species diversity was assessed using the Shannon diversity index (Brower et al. 1998; Krebs 1999). Evenness was deter¬ mined as the ratio of heterogeneity (H') to maximum heterogeneity (H' max) (Brower et al.; Krebs 1999). One-way analysis of variance was used to compare species richness, species diversity and evenness among the three kinds of marshes (Sokal & Rohlf 1981). Nomenclature follows Jones et al. (1997). Common names follow Hatch et al. (1999). When surface water was present, salinity readings were obtained with a temperature compensated hand-held refractometer (Table 1). Marshes were classified as freshwater (0.0 to 0.5 ppt), brackish water (0.5 to 17.0 ppt) or saltwater (> 17.0 ppt). JUDD & LONARD 107 Results Freshwater marshes. — A total of 81 species were present in the six marshes (Table 2). Species richness per marsh ranged from 15 to 31. No species occurred in all of the marshes, but five species, Cy perns articulatus (jointed flatsedge), Urochloa maxima (Guineagrass), Paspalum lividum (longtom), Polygonum pensilvanicum (pink smart- weed) and Typha domingensis (narrow-leaf cattail) were present in five marshes. The introduced grass, U. maxima , was found only in the last interval of transects where the marsh graded into an upland shrub- grassland community. Tree seedlings and scattered shrubs including Acacia famesiana (huisache), Celtis laevigata (sugar hackberry), Ipomoea camea (shrubby morningglory) , Mimosa asperata (black mimosa), Parkinsonia aculeata (retama), Salix exigua (sandbar willow) and S. nigra (blackwillow) were present occasionally in the marshes. There was a low degree of community similarity among the marshes (Table 3). Coefficients of similarity ranged from 0.103 to 0.525. Resaca de la Palma and Cattail Lake at Santa Ana National Wildlife Refuge (SANWR) were the only marshes that had a coefficient of similarity greater than 0.500. The mean of 15 coefficients of similarity was 0.322 ( SD = 0.116). Clearly, there were marked differences in species composition of freshwater marsh communities. Each of the freshwater marshes had a different dominant species (Table 4) and only a few species were responsible for most of the cover. Indeed, the first six species in importance contributed from 72.6% to 96.4% of the relative cover. As with the flora in general, there was low similarity among the marshes in the species making up the six most important species. If each of the six most important species was different in the six marshes, a total of 36 different species was possible; however, 24 different species or 67% of the maximum diversity were found. Nineteen of the 24 species occurred in two or more marshes and 12 occurred in three or more marshes. Brackish water marshes. — Eighty-one species were present in nine brackish marshes (Table 5). Species richness per marsh ranged from 7 to 24. No species occurred in all of the marshes, but Borrichia frutescens occurred in eight marshes (all but Tio Cano #2). No other species occurred in more than six of the marshes (Table 5). There was a low degree of species similarity among most of the marshes (Table 6). The exception was the two resacas at Palo Alto National Battlefield, which had 66.7% of their species in common. These two sites were separated by less than 0.5 km of coastal prairie. Thus, the similarity of 108 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 2. Species present in freshwater marshes in the Rio Grande Delta. 1 = Paso Real, 2 = Russelltown, 3 = Resaca de la Palma, 4 = Audubon Sabal Palm Sanctuary, 5 = Cattail Lake and 6 = Willow Lake. Species 2 3 4 5 Acacia farnesiana Alternanthera paronychioides Amaranthus sp. Ambrosia psilostachya Ammania coccinea Bacopa monnieri Bothriochloa laguroides Cardiospermum halicacabum Celtis laevigata Chlorocantha spinosa Chromolaena odorata Clematis drummondii Cocculus diversifolius Commelina erecta Croton sp. Cucumis melo Cynodon dactylon Cyperaceae: unidentified Cyperus articulatus Cyperus digitatus Cyperus elegans Cyperus ochraceus Cyperus odoratus Cyperus rotundas Cyperus virens Cyperus sp. (1) Cyperus sp. (2) Dichanthium annulatum Dichanthium aristatum Dichanthium sp. Eclipta prostrata Echinochloa colona Echinochloa muricata Echinodorus beteroi Eleocharis austrotexana Eleocharis interstincta Eleocharis parvula Eleocharis sp. Eragrostis reptans Eriochloa punctata Helianthus annuus Heteranthera dubia Ipomoea amnicola Ipomoea carnea Iva annua Lemna sp. Leptochloa fusca Leptochloa nealleyi Leptochloa panicea Ludwigia octovalvis Ludwig ia repens Malachra capitata Malvastrum coromandelianum Marsilea vestita Mikania scandens X X X X X X X X X X X X X X X XXX X X X X X X X X X X X X X X X X 6 X XX X XXX XX X XX XX X XX JUDD & LONARD 109 Table 2. Continued. Species 1 2 3 4 5 6 Mimosa asperata X X Panicum hirsutum X X X Parkinsonia aculeata X Paspalum denticulatum X X X X X Phyla nodiflora X X Physalis sp. X Pluchea purpurascens X X Poaceae: unidentified X X Polygonum densiflorum X Polygonum pensylvanicum X X X X X Prosopis reptans X Ricinus communis X Rubus riograndis X Salix exigua X Salix nigra X X Schoenoplectus californicus X X X Sesbania herbacea X X Sida sp. X Solanum americanum X Solanum campechiense X X X Sorghum halepense X Spermacoce glabra X Symphyotrichum divaricatum X Typha domingensis X X X X X Urochloa maxima X X X X X Vigna luteola X Table 3. Comparison of Sorensen’s community similarity coefficients among freshwater marshes in the Rio Grande Delta. 1 = Paso Real, 2 = Russelltown, 3 = Resaca de la Palma, 4 = Audubon Sabal Palm Sanctuary, 5 = Cattail Lake and 6 = Willow Lake. 1 2 Sites 3 4 5 2 3 Sites 4 0.370 0.300 0.370 0.178 0.103 0.311 5 0.361 0.218 0.525 0.217 6 0.491 0.298 0.415 0.263 0.407 their vegetation is not surprising. Coefficients of similarity for brackish marshes ranged from 0.098 to 0.667 (Table 6). The mean of 36 coeffi¬ cients was 0.258 ( SD = 0. 123). Thus, the mean similarity for brackish marshes was even less than for freshwater marshes. Typha domingensis was the dominant species in three brackish marshes (Table 7) and it was a co-dominant in a fourth. Bads tnaridma was the dominant species in two brackish marshes. The six most important species accounted for most of the cover (Table 7). Indeed, the six most important species accounted for 88.0 to 99.8% of the 110 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 4. Comparison of species importance among freshwater marshes of the Rio Grande Delta. Freq. = frequency, Rel. Freq. = relative frequency, Rel. Cov. = relative cover, Imp. Val. = importance value (sum of relative frequency and relative cover). Marsh Species Freq. Rel. Freq. % Cover Rel. Cov. Imp. Val. Paso Real Cyperus ochraceus 85 15.2 21.36 25.4 40.6 Eleocharis austrotexana 65 11.6 17.21 20.5 32.1 Polygonum densiflorum 60 10.7 14.83 17.7 28.4 Heteranthia dubia 45 8.0 11.23 13.4 21.4 Leptochloa fusca 60 10.7 7.05 8.4 19.1 Schoenoplectus californicus 24 additional species 15 2.7 Total 4.77 83.95 5.7 8.4 Russelltown Urochloa maxima 70 9.9 21.99 24.0 33.9 Cyperus odoratus 60 8.5 20.87 22.8 31.3 Typha domingensis 80 11.3 12.62 13.8 25.3 Paspalum denticulatum 50 7.0 7.37 8.0 15.0 Mikania scandens 60 8.5 3.53 3.9 12.4 Eriochloa punctata 18 additional species 40 5.6 Total 3.59 91.37 3.9 9.5 Resaca de Panicum hirsutum 84.8 17.9 36.11 37.5 55.4 la Palma Typha domingensis 72.7 15.4 20.87 21.7 37.1 Cardiospermum halicacabum 45.5 9.6 5.20 5.4 15.0 Paspalum denticulatum 24.2 5.1 7.38 7.7 12.8 Sesbania herbacea 27.3 5.8 4.75 4.9 10.7 Solanum campechiense 24 additional species 24.2 5.1 Total 3.93 96.25 4.1 9.2 Sabal Palm Malachra capitata 100.0 22.4 24.02 28.7 51.1 Sanctuary Panicum hirsutum 72.7 16.3 21.82 26.1 42.4 Echinodorus beteroi 54.5 12.2 14.18 16.9 29.1 Eleocharis sp. 36.4 8.2 10.27 12.3 20.5 Heteranthera dubia 45.5 10.2 4.05 4.8 15.0 Lemna sp. 9 additional species 27.3 6.1 Total 6.36 83.73 7.6 13.7 Cattail Lake Typha domingensis 66.7 8.8 16.48 18.0 26.8 Malachra capitata 66.7 8.8 12.90 14.1 22.9 Schoenoplectus californicus 46.7 6.1 13.58 14.9 21.0 Paspalum denticulatum 66.7 8.8 11.07 12.1 20.9 Phyla nodiflora 66.7 8.8 8.55 9.4 18.2 Cucumis melo 25 additional species 66.7 8.8 Total 3.75 91.31 4.1 12.9 Willow Lake Paspalum denticulatum 75.0 16.0 55.26 62.5 78.5 Malachra capitata 56.3 12.0 8.59 9.7 21.7 Bacopa monnieri 37.5 8.0 7.91 8.4 16.4 Cyperus ochraceus 43.8 9.3 3.61 4.1 13.4 Eleocharis parvula 43.8 9.3 0.28 0.3 9.6 Schoenoplectus californicus 17 additional species 18.8 4.0 Total 4.91 88.48 5.5 9.5 JUDD & LONARD Table 5. Species present in brackish marshes in the Rio Grande Delta. 1 = Palo Alto #1 , 2 = Palo Alto #2, 3 = LANWR Resaca, 4 = Olmito Resaca, 5 — Tio Cano #1,6 = Tio Cano #2, 7 = Bay view Resaca #1,8 = Bay view Resaca #2 and 9 = Willamar. Species 2 3 4 5 6 7 8 Ambrosia psilostachya Andropogon glomeratus Atriplex pentandra Bacopa monnieri Batis maritima Bolboschoenus maritimus Borrichia frutescens X Chamaesyce serpens Chara sp. X Chlorophyta filaments X Chromolaena odorata Cissus incisa Citharexylum berlandieri Conoclinium betonicifolium Cynodon dactylon Cyperus articulatus X Cyperus ochraceus Cyperus sp. Dalea scandens Dichanthium sp. Distichlis spicata Echinodorus beteroi X Eclipta prostrata Eleocharis austrotexana X Eleocharis interstincta Eleocharis sp. Eriochloa punctata Eustoma exaltatum Forestiera angustifolia Funastrum cynanchoides Gossypianthus lanuginosus Havardia pallens Heliotropium curassavicum Heteranthera dubia Hydrocotyle bonariensis Ipomoea amnicola Ipomoea sagittata Isocoma drummondii Iva annua Karwinskia humboldtiana Lemna sp. X Leptochloa fusca Leucophyllum frutescens Limonium carolinianum Lycium carolinianum X Machaeranthera phyllocephala Malachra capitata Marsilea vestita X Maytenus phyllanthoides Melothria pendula Mikania scandens Mimosa asperata X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X XX X XX X X X X X X XXX X X X X X X X X X X X X X X X X X X X 9 X X X X X X X X X X X 112 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 5. Continued. Species 1 2 3 4 5 6 7 8 9 Mimosa strigillosa Monanthochloe littoralis X X X Monocotyledon: unidentified Opuntia engelmannii Panicum hirsutum X X X Parkinsonia aculeata X X Paspalum denticulatum X X X X X X Phyla nodiflora Pluchea purpurascens X X X X X X Poaceae: unidentified X Prosopis glandulosa Prosopis reptans X X X X X X Rumex chrysocarpus Salix nigra Schoenoplectus californicus X X X X X Schoenoplectus pungens X X Seshania drummondii Seshania herbacea X X X Sesuvium maritimum X X X Sesuvium sessile Sesuvium verrucosum Solanum elaeagnifolium X X X X Spartina spartinae X X X X Sporobolus virginicus Sporobolus wrightii X X Suaeda linearis X X X X X Symphyotrichum divaricatum X X X Typha domingensis X X X X X Urochloa maxima X X X Table 6. Comparison of Sorensen’s community similarity coefficients among brackish marshes in the Rio Grande Delta. 1 = Palo Alto #1,2 = Palo Alto #2, 3 = LANWR Resaca, 4 : = Olmito Resaca, 5 = Tio Cano #1,6 = Tio Cano #2, 7 = Bayview Resaca #1, 8 = Bayview Resaca #2 and 9 = Willamar. Sites 1 2 3 4 5 6 7 8 2 0.667 3 0.216 0.268 4 0.154 0.216 0.217 Sites 5 0.211 0.167 0.133 0.468 6 0.235 0.250 0.098 0.279 0.333 7 0.182 0.129 0.300 0.333 0.293 0.222 8 0.273 0.300 0.276 0.258 0.200 0.154 0.480 9 0.111 0.176 0.176 0.489 0.450 0.250 0.205 0.143 relative cover except in the Olmito marsh where the top six species contributed only 66.7% of the relative cover. If each of the six most important species was different in the nine marshes, a total of 54 different species was possible. Thirty-one different species or 57.4% of the maximum diversity were found. JUDD & LONARD 113 Table 7. Comparison of species importance among brackish marshes of the Rio Grande Delta. Freq. = frequency, Rel. Freq. = relative frequency, Rel. Cov. = relative cover, Imp. Val. = importance value (sum of relative frequency and relative cover). Marsh Species Freq. Rel. Freq. % Cover Rel. Cov. Imp. Val. Palo Alto Typha domingensis 94.4 18.8 41.14 50.4 69.2 #1 Borrichia frutescens 100.0 19.8 18.66 22.8 42.6 Eleocharis austrotexana 77.8 15.3 9.37 11.0 26.3 P asp alum denticulatum 38.9 7.7 6.06 7.2 14.9 Lycium carolinianum 55.5 11.1 1.53 1.9 13.0 Marsilea vestita 8 additional species 44.5 8.6 Total 1.95 82.66 2.4 11.0 Palo Alto Eleocharis austrotexana 93.3 22.2 44.89 59.7 81.9 n Spartina spartinae 40.0 9.5 13.99 18.8 28.3 Borrichia frutescens 53.3 12.7 6.05 8.6 21.3 Marsilea vestita 53.3 12.7 5.11 7.2 19.9 Echinodorus beteroi 53.3 12.7 2.85 3.5 16.2 Heteranthera dubia 7 additional species 26.7 6.3 Total 0.49 74.49 0.8 7.1 LANWR Batis maritima 71.4 17.0 24.00 26.5 43.5 Resaca Monanthochloe littoralis 48.6 11.6 26.37 29.2 40.8 Borrichia frutescens 68.6 16.3 12.28 13.6 29.9 Spartina spartinae 25.7 6.1 14.98 16.6 22.7 Sporobolus virginicus 40.0 9.5 3.06 3.4 12.9 Schoenoplectus californicus 16 additional species 22.9 5.4 Total 3.21 90.39 3.6 9.0 Olmito Leptochloa fusca 83.3 13.2 12.90 20.9 34.1 Sesuvium sessile 66.7 10.5 10.04 16.3 26.8 Pluchea purpurascens 61.1 9.6 5.06 8.2 17.8 Paspalum denticulatum 33.3 5.3 4.87 7.8 13.1 Parkinsonia aculeata 55.6 8.8 2.59 4.2 13.0 Sesuvium maritimum 18 additional species 22.2 3.5 Total 5.76 61.77 9.3 12.8 Tio Cano Typha domingensis 97.4 22.8 32.44 32.9 55.7 n Schoenoplectus pungens 39.5 9.3 21.31 21.6 30.9 Iva annua 36.8 8.6 13.53 13.7 22.3 Lycium carolinianum 71.1 16.7 2.76 2.8 19.5 Leptochloa fusca 36.8 8.6 7.85 8.0 16.6 Borrichia frutescens 17 additional species 26.3 6.2 Total 8.90 98.70 9.0 15.2 Tio Cano Typha domingensis 91.4 17.4 41.94 35.1 52.5 n Eleocharis interstincta 82.9 15.8 32.22 27.0 42.8 Distichlis spicata 77.1 14.7 24.08 20.1 34.8 Lycium carolinianum 82.9 15.8 3.53 3.0 18.8 Eleocharis sp. 28.6 5.4 6.68 5.6 11.0 Symphyotrichum divaricatum 13 additional species 34.3 6.5 Total 1.75 119.55 1.5 8.0 Bayview Batis maritima 87.5 20.0 21.79 29.0 49.0 Resaca #1 Borrichia frutescens 68.8 15.7 19.35 25.8 41.5 Suaeda linearis 56.3 12.9 8.10 10.8 23.7 Eriochloa punctata 12.5 2.9 9.13 12.2 15.1 Distichlis spicata 31.3 7.1 5.70 7.7 14.8 Pluchea purpurascens 12 additional species 43.8 10.0 Total 2.48 74.96 3.3 13.3 114 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Table 7. Continued Marsh Species Freq. Rel. Freq. % Cover Rel. Cov. Imp. Val. Bayview Borichia frutescens 83.3 25.0 37.43 33.3 58.3 Resaca #2 Typha domingensis 66.7 20.0 41.07 36.5 56.5 Distichlis spicata 67.3 20.0 17.05 15.2 35.2 Spartinci spartinae 33.3 10.0 6.28 5.6 15.6 Bolboschoenus maritimus 33.3 10.0 3.45 3.1 13.1 Urochloa maxima 1 additional species 16.7 5.0 Total 6.88 112.48 6.1 11.1 Willamar Sesuvium maritimum 60.0 23.1 22.51 44.4 67.5 Sesbania herbacea 22.0 8.3 8.17 16.1 24.4 Heliotropium curassavicum 34.1 13.0 4.62 9.1 22.1 Bacopa monnieri 14.6 5.6 5.17 10.2 15.8 Borrichia frutescens 17.1 6.5 2.95 5.8 12.3 Pluchea purpurascens 15 additional species 19.5 7.4 Total 1.21 50.66 2.4 9.8 Salt water marshes.— Seventy-three species were present in 11 salt marshes (Table 8). Species richness per marsh ranged from 7 to 26. No species occurred in all the marshes, but B. frutescens was present in 10. Batis maritima and Prosopis reptans occurred in nine marshes and Sporobolus virginicus was present in eight. Coefficients of similarity ranged from 0.049 to 0.690 (Table 9). The mean of 55 coefficients was 0.372 ( SD = .147). One-way analysis of variance of coefficients of similarity among freshwater, brackish and salt marshes showed signifi¬ cant variation among the kinds of marshes (i.e., among groups), F = 7.994, 2 & 103 df P < 0.001. Pairwise comparisons revealed only one significant difference; the mean coefficient of similarity for salt marshes was significantly greater than that for brackish marshes, t — 3.851, 89 df,P< 0.001. The first six species in importance (Table 10) contributed from 82.2 to 99.4% of the relative cover. Borrichia frutescens and Paspalum vaginatum each was a dominant species in three marshes and S. spartinae and S. virginicus each was the dominant species in two marshes (Table 10). There was greater similarity in the important species of salt marshes than in freshwater or brackish marshes. A list of the six most important species included only 23 different species or 34.8% of the maximum diversity of 66 different species. Freshwater and salt marshes had no dominant species in common (Tables 4 and 10), but brackish and salt marshes shared two dominant species, B. frutescens and Sesuvium maritimum (Tables 7 and 10). Freshwater and brackish marshes shared one dominant species, T. domingensis (Tables 4 and 7). JUDD & LONARD 115 Table 8. Species present in salt marshes in the Rio Grande Delta. 1 = Stover Point, 2 = Spillway Crossing, 3 = Large Marsh, 4 = Dry Marsh, 5 = Osprey Point, 6 = Laguna Atascosa Cayo, 7 = Redhead Ridge, 8 = Rangerville #1,9 = Rangerville #2, 10 = Bay view Dry Marsh, 11 = Bay view Brine Marsh. Species 123456789 10 11 Abutilon sp. Allowissadula lozanii Ambrosia psilostachya Atriplex pentandra Bacopa monnieri Batis maritima Bolboschoenus maritimus Borrichia frutescens Bothriochloa laguroides Char a sp. Chromolaena odorata Clappia suaedifolia Cressa nudicaulis Croton sp. Cynanchum barbigerum Cynodon dactylon Cyperus articulatus Desmanthus virgatus Dichanthium annulatum Dichanthium aristatum Dichanthium sericeum Distichlis spicata Echinocereus sp. Eriochloa punctata Gaillardia pulchella Hcliotropium angiospermum Heliotropium curassavicum Ibervillea lindheimeri Isocoma drummondii Jatropha dioica Leptochloa uninerva Limonium carolinianum Lycium carolinianum Machaeranthera phyllocephala Malvastrum amcricanum Malvastrum coromandelianum Maytenus phyllanthoides Monanthochloe littoralis Opuntia engelmannii Opuntia leptocaulis Panicum hallii Paspalum vaginatum Passiflora foetida Pennisetum ciliare Phyla nodiflora Portulaca pilosa Pluchea purpurascens Prosopis glandulosa Prosopis reptans Rhynchosia americana Rhynchosia senna Ruppia maritima Salicornia virginica Sesuvium maritimum Sesuvium portulacastrum X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X XXX X X 116 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 8. Continued. Species i 2 3 4 5 6 7 8 9 10 11 Sesuvium sessile X Sesuvium verrucosum X X X X Setaria leucopila Setaria parviflora Sida sp. X X X X Solarium americanum Solarium eleagnifolium X X Solarium triquetrum X X Spartina spartinae X X X X X X Sporobolus pyramidatus X Sporobolus virginicus X X X X X X X X Sporobolus wrightii X X X X X X X Suaeda linearis X X X X X X Trixis inula X Typha domingensis X X X X X X Urochloa maxima Xylothamia palmeri Yucca treculeana X X X X X Table 9. Comparison of Sorensen’s community similarity coefficients among salt marshes in the Rio Grande Delta. 1 = Stover Point, 2 = Spillway Crossing, 3 = Large Marsh, 4 = Dry Marsh, 5 = Osprey Point, 6 = Laguna Atascosa Cayo, 7 = Redhead Ridge, 8 = Rangerville #1,9 = Rangerville #2, 10 = Bayview Dry Marsh, 11 = Bayview Brine Marsh. 1 2 3 4 Site 5 6 7 8 9 10 2 0.483 3 0.458 0.571 4 0.593 0.458 0.526 5 0.478 0.500 0.533 0.556 Site 6 0.429 0.520 0.300 0.435 0.421 7 0.440 0.500 0.529 0.450 0.500 0.476 8 0.049 0.286 0.160 0.065 0.087 0.364 0.074 9 0.231 0.304 0.278 0.286 0.353 0.409 0.211 0.414 10 0.298 0.439 0.387 0.324 0.690 0.410 0.364 0.333 0.400 11 0.205 0.364 0.261 0.207 0.381 0.387 0.320 0.125 0.222 0.636 Comparison of species richness , species diversity and evenness among marshes.— \ alues for species richness, species diversity, and evenness are provided for each freshwater, brackish and salt marsh in Table 11. One-way ANOVAs for each of these parameters showed no significant differences among the kinds of marshes (Table 12). Freshwater and brackish marshes shared 35 species (coefficient of similarity = 0.216). Brackish and salt marshes had 30 species in common (coefficient of similarity = 0.195), while freshwater and salt marshes shared only 19 species (coefficient of similarity = 0. 123). Freshwater and salt marshes had only two important species in common, U. maxima and T. domingensis. JUDD & LONARD 117 Table 10. Comparison of species importance among salt marshes of the Rio Grande Delta. Freq. = frequency, Rel. Freq. = relative frequency, Rel. Cov. = relative cover, Imp. Val. = importance value (sum of relative frequency and relative cover). Marsh Species Freq. Rel. Freq. % Cover Rel. Cov. Imp. Val. Stover Spartina spartinae 31.1 6.8 23.28 36.7 43.5 Point Borrichia frutescens 60.6 13.3 5.46 8.6 21.9 Monanthochloe littoralis 36.1 7.9 8.51 13.4 21.3 Sporobolus virginicus 27.9 6.1 9.26 14.6 20.7 Prosopis reptans 45.9 10.0 0.94 1.5 11.5 Bothriochloa laguroides 26 additional species 14.7 3.2 Total 4.70 63.38 7.4 10.6 Spillway Paspalum vaginatum 65.8 18.7 25.42 29.8 48.5 Crossing Borrichia frutescens 44.7 12.7 12.30 14.4 27.1 Sporobolus virginicus 28.9 8.2 14.46 17.0 25.2 Satis maritima 42.1 11.9 8.78 10.3 22.2 Distichlis spicata 34.2 9.7 8.79 10.3 20.0 Bolboschoenus maritimus 20 additional species 28.9 8.2 Total 3.50 85.28 4.1 12.3 Large Sporobolus virginicus 55.7 13.9 36.15 36.1 50.0 Marsh Batis maritima 85.2 21.3 19.98 20.0 41.3 Monanthochloe littoralis 78.7 19.7 20.46 20.4 40.1 Borrichia frutescens 75.4 18.9 18.66 18.6 37.5 Lycium carolinianum 50.8 12.7 0.52 0.5 13.2 Sesuvium portulacastrum 10 additional species 18.0 4.5 Total 0.51 100.06 0.5 5.0 Dry Salt Spartina spartinae 84.4 19.3 76.52 76.3 95.6 Marsh Borrichia frutescens 87.5 20.0 10.24 10.2 30.2 Prosopis reptans 84.4 19.3 1.47 1.5 20.8 Monanthochloe littoralis 31.3 7.1 4.22 4.2 11.3 Salicornia virginica 21.9 5.0 1.70 1.7 6.7 Cressa nudicaulis 16 additional species 18.8 4.3 Total 0.66 100.27 0.7 5.0 Osprey Borrichia frutescens 90.0 20.0 33.64 41.1 61.1 Point Sporobolus virginicus 70.0 15.6 21.46 26.2 41.8 Monanthochloe littoralis 50.0 11.1 6.88 8.4 19.5 Typha domingensis 50.0 11.1 5.70 7.0 18.1 Batis maritima 60.0 13.3 2.14 2.6 15.9 Char a sp. 8 additional species 30.0 6.7 Total 1.82 81.81 2.2 8.9 Laguna Paspalum vaginatum 65.4 14.8 31.5 31.7 46.5 Atascosa Borrichia frutescens 57.7 13.0 16.38 16.5 29.5 Cayo Bolboschoenus maritimus 61.5 13.9 15.02 15.1 29.0 Distichlis spicata 57.7 13.0 13.47 13.6 26.6 Suaeda linearis 23.1 5.2 7.95 8.0 13.2 Sporobolus wrightii 18 additional species 15.4 3.5 Total 4.26 99.25 4.3 7.8 Redhead Sporobolus virginicus 75.0 16.2 35.18 39.2 55.4 Ridge Sporobolus wrightii 54.2 11.7 20.68 23.1 34.8 Borrichia frutescens 75.0 16.2 14.15 15.8 32.0 Char a sp. 29.2 6.3 7.87 8.8 15.1 Batis maritima 45.8 9.9 4.50 5.0 14.9 Prosopis reptans 12 additional species 45.8 9.9 Total 1.77 89.72 2.0 11.9 118 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 10. Continued. Marsh Species Freq. Rel. Freq. % Cover Rel. Cov. Imp. Val. Rangerville Paspalum vaginatum 69.0 37.7 29.63 53.1 90.8 n Sesuvium maritimum 51.7 28.3 16.14 28.9 57.2 Sesuvium verrucosum 17.2 9.4 4.63 8.3 17.7 Pluchea purpurascens 13.8 7.5 2.82 5.1 12.6 Typha domingensis 13.8 7.5 0.87 1.6 9.1 Urochloa maxima 3 additional species 6.9 3.8 Total 1.36 55.80 2.4 6.2 Rangerville Sesuvium maritimum 66.7 18.5 20.05 31.9 50.4 #2 Sesuvium verrucosum 50.0 13.9 8.75 13.9 27.8 Borrichia frutescens 23.3 6.5 10.78 17.2 23.7 Sporobolus virginicus 30.0 8.3 6.97 11.1 19.4 Suaeda linearis 30.0 8.3 4.03 6.4 14.7 Typha domingensis 14 additional species 23.3 6.5 Total 4.50 62.80 7.2 13.7 Bayview Borrichia frutescens 87.5 24.1 32.51 32.2 56.3 Brine Distichlis spicata 62.5 17.2 27.46 27.2 44.4 Marsh Ruppia maritima 50.0 13.8 25.11 24.9 38.7 Batis maritima 87.5 24.1 6.40 6.3 30.4 Sporobolus wrightii 37.5 10.3 8.41 8.3 18.6 Prosopis reptans 1 additional species 25.0 6.9 Total 0.26 100.98 0.3 7.2 Bayview Borrichia frutescens 76.9 18.9 20.8 21.2 40.1 Dry Marsh Distichlis spicata 69.2 17.0 19.36 19.7 36.7 Batis maritima 53.8 13.2 15.53 15.8 29.0 Char a sp. 30.8 7.5 14.42 14.7 22.2 Sporobolus wrightii 38.5 9.4 9.85 10.0 19.4 Typha domingensis 9 additional species 23.1 5.7 Total 7.20 98.26 7.3 13.0 Discussion Only Judd & Lonard (2002) have provided information on species diversity and evenness of Rio Grande Delta marshes and this is for only one salt marsh and one brackish marsh. The marshes they studied are included in the data set of this investigation. Information on species richness is meager. White & Schmedes (1986) identified species “typical” of each of the three marsh types rather than providing a list of all species occurring in each kind of marsh. Thus, they do not provide a measure of species richness. However, if one compares their list of “typical” species with our group of important species (the number of different species in the list of the six most important species), the numbers are similar. For example, White & Schmedes (1986) identified 18 species typical of salt marshes and this study found 23 important species. They report 26 typical species in brackish marshes and this JUDD & LONARD 119 Table 11. Comparison of species richness (N), species diversity (H'), and Evenness (J') among freshwater, brackish and salt marshes of the Rio Grande Delta. Marsh N H' J' Freshwater Marshes Paso Real 30 1.477 0.755 Russelltown 24 1.380 0.857 Resaca de la Palma 30 1.477 0.753 Audubon Sabal Palm Sanctuary 15 1.176 0.788 Cattail Lake 31 1.491 0.839 Willow Lake 22 1.362 0.735 Brackish Marshes Palo Alto #1 14 1.461 0.729 Palo Alto #2 13 1.114 0.749 LANWR Resaca 22 1.342 0.760 Olmito Resaca 24 1.380 0.870 Tio Cano #1 23 1.362 0.753 Tio Cano #2 19 1.279 0.735 Bay view Resaca #1 18 1.255 0.803 Bayview Resaca #2 7 0.845 0.880 Willamar 21 1.322 0.766 Salt Marshes Stover Point 32 1.505 0.779 Spillway Crossing 26 1.415 0.741 Large Salt Marsh 16 1.204 0.674 Dry Salt Marsh 22 1.342 0.621 Osprey Point 14 1.146 0.801 Laguna Atascosa Cayo 24 1.380 0.774 Redhead Ridge 18 1.255 0.763 Rangerville #\ 9 0.954 0.681 Rangerville #2 20 1.301 0.798 Bayview Brine Marsh 7 0.845 0.884 Bayview Dry Marsh 15 1.176 0.837 Table 12. Analysis of variance for species richness, species diversity, freshwater, brackish and salt marshes of the Rio Grande Delta. and evenness among Parameter & Source of Variation DF SS MS F (Probability) Species Richness (N) Among Marshes 2 236.425 118.213 2.749 (P > 0.05) Within Marshes 23 988.960 43.000 Total 25 1,225.385 Species Diversity (H') Among Marshes 2 0.110 0.055 1.719 (P > 0.1) Within Marshes 23 0.727 0.032 Total 25 0.837 Evenness (J') Among Marshes 2 0.004 0.002 0.500 (P > 0.5) Within Marshes 23 0.096 0.004 Total 25 0.100 120 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 study found 31 important species. White & Schmedes identify 26 species typical of freshwater marshes and this study found 24 are important. Clover (1937) identified 44 species associated with freshwater habitats in the Lower Rio Grande Valley of Texas, but she did not list species associated with brackish or salt marshes. The number of species she lists for freshwater habitats is far greater than the number of typical species for freshwater marshes reported by White & Schmedes (1986), but far less than the 81 species this study found in freshwater marshes. Only 14 of the 44 freshwater species Clover (1937) identified were found in this study. Conversely, this study found 13 of the 26 species White & Schmedes (1986) listed as occurring in freshwater marshes in freshwater marshes and three others in brackish marshes and this study found 17 of the 26 species they listed for brackish marshes in brackish marshes and four others in salt marshes. This study found 13 of the 18 species they listed for salt marshes in salt marshes. Perhaps this study found a lower percentage of the freshwater species identified by Clover (1937) because there has been a longer time for changes in the flora since her study than there has been since White & Schmedes’ (1986) study. Species composition among marshes of a given type such as fresh¬ water marshes is highly variable even within a relatively small area such as the Rio Grande Delta. Jacobson & Jacobson (1989) found a similar relationship among 18 salt marshes of the Maine coast. Despite the variability in species composition, in most cases one can separate freshwater marshes from salt marshes by the important species present (especially the dominant species) . Only two important/dominant species, T. domingensis and U. maxima , were common to freshwater and salt marshes. Typha domingensis clearly exhibits a broad range of salinity tolerance for the species was found in freshwater, brackish and salt marshes. White & Schmedes (1986) list T. domingensis as a species characteristic of freshwater marshes and they also found it in brackish marshes, but they do not list it as one of the species occurring in salt marshes in the Rio Grande Delta area. White & Schmedes (1986) do not list U. maxima as a species associated with any of the three kinds of marshes. This is likely because the species was uncommon in the Rio Grande Delta area when they did their field investigations, i.e., prior to 1986. Today, U. maxima is found on the margins of freshwater, brack¬ ish and salt marshes and it invades freshwater and brackish marshes when they begin to dry. White & Schmedes (1986) noted that brackish marshes are transitional JUDD & LONARD 121 between freshwater and salt marshes and contain some species typical of both marsh types. This current study found that this was certainly so. Of the 32 important species occurring in brackish marshes, 12 also were important in freshwater marshes and 13 were important in salt marshes. Typha domingensis and U. maxima were important in all three kinds of marshes. Species richness that was observed in Rio Grande Delta marshes appears to be similar to species diversity in marshes distant from the area. For example, Jacobson & Jacobson (1989) reported that species richness of 1 8 salt marshes along the Maine coast ranged from 11 to 25 ( x = 17.22, SD — 4.37). This study found that species richness in 11 Rio Grande Delta salt marshes ranged from 7 to 32 (x = 18.45, SD = 7.38). There was no significant difference in the means ( t = 0.569, 27 df,P> 0.5). Testing the general izability of the marsh species richness, species diversity and evenness values obtained in this study awaits the reporting of additional information from other areas of the Texas coast. Acknowledgments Financial support was provided, in part, by Texas Higher Education Coordinating Board Advanced Technology Grant No. 003599-009-1997, which is gratefully acknowledged. Literature Cited Brower, J. E., J. H. Zar & C. N. Von Ende. 1998. Field and Laboratory Methods for General Ecology. WCB/McGraw-Hill. Boston, Massachusetts. U.S.A., 273 pp. Brown, L. F., Jr., J. L. Brewton, T. J. Evans, J. H. McGowen, W. A. White, C. G. Groat & W. L. Fisher. 1980. Environmental Geologic Atlas of the Texas Coastal Zone - Brownsville - Harlingen Area. The University of Texas at Austin, Bureau of Economic Geology, 140 pp 4- 9 maps. Canfield, R. H. 1941. Application of the line interception method in sampling range vegetation. Journal of Forestry, 39:388-394. Clover, E. U. 1937. Vegetational survey of the lower Rio Grande Valley, Texas. Madrono, 4:41-72, 77-100. Hatch, S. L., J. L. Schuster & D. L. Drawe. 1999. Grasses of the Texas Gulf Prairies and Marshes. Texas A&M University Press, College Station, 355 pp. Jacobson, H. A. & G. L. Jacobson, Jr. 1989. Variability of vegetation in tidal marshes of Maine. Canadian Journal of Botany, 67:230-238. Jahrsdoerfer, S. E. & D. M. Leslie, Jr. 1988. Tamaulipan brushland of the Lower Rio Grande Valley of South Texas: description, human impacts, and management options. U. S. Fish and Wildlife Service, Biological Report 88 (36), 63 pp. Johnston, M. C. 1955. Vegetation of the eolian plain and associated features of southern Texas. Unpubl. Ph.D. dissertation. The Univ. of Texas at Austin, Texas, 167 pp. Jones, S. D., J. K. Wipff & P. M. Montgomery. 1997. Vascular plants of Texas: a comprehensive checklist including synonymy, bibliography, and index. Univ. Texas Press, Austin, 404 pp. 122 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Judd, F. W. & R. I. Lonard. 2002. Species richness and diversity of brackish and salt marshes in the Rio Grande Delta. Journal of Coastal Research, 18:751-759. Judd, F. W., R. I. Lonard, J. H. Everitt, D. E. Escobar & M. R. Davis. 1997a. Using multispectral videography to distinguish the pattern of zonation and plant species composition in brackish water marshes of the Rio Grande Delta. Proceedings Fourth International Conference on Remote Sensing for Marine and Coastal Environments (Orlando, Florida, 17-19 March, 1997), pp. 621-629. Judd, F. W., R. I. Lonard, J. H. Everitt, D. E. Escobar & M. R. Davis. 1997b. Using multispectral videography to compare the pattern of zonation between brackish water marshes and saltwater marshes of the Rio Grande Delta. Proceedings 16th Biennial Workshop of Videography and Color Photography in Resource Assessment (Weslaco, Texas, April 29 - May 1, 1997), pp. 394-405. Krebs, C. J. 1999. Ecological Methodology, Menlo Park, California. Addison Wesley Longman, 620 pp. Lonard, R. I. & F. W. Judd. 1999. Vascular plants of the Rio Grande Delta marshes. Proceedings 1999 Symposium of the Native Plant Society of Texas (October 14-17, 1999, Harlingen, Texas), pp. 22-29. Sokal, R. R. & F. J. Rohlf. 1981. Biometry. The Principles and Practice of Statistics in Biological Research. 2nd Ed. W. H. Freeman and Company. New York, New York, 859 pp. White, W. A., T. R. Calnan, R. A. Morton, R. W. Kimble, T. G. Littleton, J. H. McGowen, H. S. Nance & K. E. Schmedes. 1986. Submerged Lands of Texas, Brownsville - Harlingen Area: Sediments, Geochemistry, Benthic Macroinvertebrates, and Associated Wetlands. The University of Texas at Austin, Bureau of Economic Geology, 138 pp. -I- 6 maps. White, W. A. & K. E. Schmedes. 1986. Wetlands. Pp. 67-93, in White, W. A., T. R. Calnan, R. A. Morton, R. W. Kimble, T. G. Littleton, J. H. McGowen, H. S. Nance, and K. E. Schmedes (eds.), Submerged Lands of Texas, Brownsville - Harlingen Area: Sediments, Geochemistry, Benthic Macroinvertebrates, and Associated Wetlands. The University of Texas at Austin, Bureau of Economic Geology, 138 pp. 4- 6 maps. FWJ at: ljudd@panam.edu TEXAS J. SCI. 56(2): 123-130 MAY, 2004 PHYSIOLOGICAL TOLERANCE RANGES OF LARVAL CAEN1S LAT1PENNIS (EPHEMEROPTERA: CAENIDAE) IN RESPONSE TO FLUCTUATIONS IN DISSOLVED OXYGEN CONCENTRATION, pH AND TEMPERATURE Robert T. Puckett* and Jerry L. Cook Department of Biological Sciences, Sam Houston State University Huntsville, Texas 77341 * Current address : Department of Entomology , Texas A&M University College Station, Texas 77843-2475 Abstract. — Laboratory experiments were conducted to investigate the physiological tolerance ranges of the mayfly Caenis latipennis (Ephemeroptera: Caenidae) from Tanyard Branch Creek in Walker County, Texas in response to stepwise fluctuations in dissolved oxygen concentrations, temperature and pH. Survivorship decreased slightly at a dissolved oxygen concentration of 7.0 mg/L, while trial groups suffered a dramatic decrease in survivorship at a dissolved oxygen concentration of 4.5 mg/L. Mean CTMax (Critical Thermal Maximum) for 10 individuals was 37.8°C with a range from 36.7°C to 38.5°C. Mean critical lower pH for three trials of 10 individuals was 2.56 and mean critical upper pH for three trials of 10 individuals was 12.5. Assessments of benthic macroinvertebrate communities provide general information regarding the water quality of the streams that support them once baseline information regarding specific streams has been gathered (Edmunds et al. 1976; Hilsenhoff 1977; Barbour et al. 1999; Rabeni et al. 1999; Lydy et al. 2000). However, the ultimate goal of managing stream quality through the practice of bioassessment is the ability to make stream management decisions based on reference data (chemical, physical and biological). These data are typically gathered from a specific region to bypass the expense and time of developing baseline information from each regional stream (Barbour et al. 1999). The cost effectiveness of stream bioassessment versus physical /chemical monitoring is realized only after this baseline informa¬ tion is gathered (Barbour et al. 1999). A critical requirement of a regionally specific bioassessment program is an understanding of the physiological tolerance ranges of the species comprising the resident benthic macroinvertebrate community. While information exists regarding species specific tolerance ranges, this in¬ formation is typically anecdotal and not empirically derived (Hilsenhoff 1977; 1982). In addition, many species have large geographical ranges 124 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 raising the possibility that a continuum of intraspecific physiological tolerance ranges occur. This stresses the necessity for determining regionally specific species tolerance ranges. Caenis latipennis occurs throughout North America north of Mexico, including south central Canada to extreme southern Texas with a disjunct population in southern Mexico (Provonsha 1990). In a previous study, streams from two neighboring counties in southeast Texas (Walker and San Jacinto counties) were monitored monthly for a period of one year regarding their ephemeropteran community diversity responses to fluctuating physical /chemical parameters. Regression analysis of mayfly diversity against fluctuation of stream quality values indicated that of the eight parameters sampled throughout the period, dissolved oxygen, temperature and pH show the greatest correlation with fluctuating mayfly diversity (Puckett 2003). The goal of this study was to determine the range of dissolved oxygen, temperature and pH that C. latipennis can tolerate with the hope that this information can be used in stream bioassessment practices specific to Walker County streams. The techniques used here may provide a model for further investigations into species specific tolerance ranges. Although this is not an investigation into the potential intra¬ specific geographical physiological tolerance gradient mentioned above, the data presented here could serve for comparison to similar values obtained for C. latipennis in other areas of its distribution. Materials and Methods Caenis latipennis larvae were collected from Tanyard Branch Creek, taken to the laboratory at Sam Houston State University and allowed to acclimate to laboratory conditions over a period of approximately one week. Mayflies were collected using a standard 0.8 m by 0.8 m kick screen and were transferred to the laboratory in 4 dram vials containing stream water. Larvae were housed in mesh bottomed containers that were submerged in water from the stream in which they were collected. Of the thirty individuals housed in each container, twenty were selected (10 per trial and 10 per control) for both dissolved oxygen and pH experiments. Individuals were selected from the remaining laboratory population for critical thermal maximum (CTMax) experiments. Dissolved Oxygen Tolerance— A 2 liter beaker was capped with a 1 .5 cm styrofoam disk that was cut to precisely fit the beaker mouth. Holes PUCKETT & COOK 125 were then cut in the disk to accommodate the container that housed the mayflies, the connector hose from a N2 cylinder and dissolved oxygen meter (YSI® Dissolved Oxygen Meter-Model 55/12FT). The containers that housed the mayflies during the trials were made by first removing the bottoms of two 100 mL plastic cups. A 7.6 cm by 7.6 cm piece of fine mesh was then stretched around the bottom opening of one cup and forced into the second cup. Once taut, this mesh provided an artificial substrate and allowed for a homogenous mixing of water inside and outside of the container. The conical shape of the cups also allowed for a tight fit into the hole in the styrofoam disk which diminished the amount of diffusion of atmospheric oxygen. A plunger to seal off the original opening of this container was built by attaching a 12 cm section of Pyrex® glass cylinder to the center of the removed cup bottom. During trials this plunger was placed into the cup so that it fit snugly beneath the water line, again with the goal of reducing atmospheric oxygen diffusion into the trial beaker. The entire apparatus was placed on a Corning® stirrer /hot-plate. During trials the stir bar revolved at approximately 68 rpm. Stirring the water during trials was essential for the operation of the dissolved oxygen meter. De-ionized water was used in all trials. Mayflies were placed in DI water three hours before the start of each trial. During the trials dissolved oxygen was removed by purging the water slowly with gaseous nitrogen to lower oxygen levels by 0.5 mg/L increments. Each 02 level was held for 45 min. until lethal 02 levels were met. The time interval of 45 min. was determined after subjecting a pre-trial group of ten individuals directly to a dissolved oxygen concentration of 0.5 mg/L. After 40 min. all individuals were dead. Control groups were setup in an identical fashion excluding only the N2 purge. Ten individuals each in trial and control groups were monitored. All other water parameters remained constant during trials. Thermal Tolerance.— Determination of lethal maximum temperature levels was carried out in a similar apparatus as that described for dissolved oxygen trials. However, in the temperature trials an aquarium heater and oxygen pump/bubbler were added to the apparatus and the nitrogen component removed. Additionally, the plunger described in the dissolved oxygen trials was removed. Critical thermal maximum (CTMax) trials rely on the observation of a trial endpoint that is specific to the organism being studied (Lutter schmidt & Hutchison 1997). For 126 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Caenis latipennis , observation of lack of righting response followed by the mayfly’s inability to cling to the artificial substrate was always followed immediately by death. Inability to cling to the artificial substrate was used as an endpoint in these trials. Temperature was raised 1.5°C/min. until the endpoint was observed. A total of ten individuals were subjected to these trials. Each trial was performed on one individual per trial while controls were simultaneously run and held at room temperature. As in dissolved oxygen trials, de-ionized water was used. Trial and control individuals were allowed the same acclimation period of approximately 3 hours. All other water parameters remained constant during trials. pH Tolerance.— pH trials were also carried out in closed beakers. However, in these trials 1 liter beakers were used to minimize chemicals necessary to accomplish stepwise manipulation of pH. Mayflies were housed as described above. Three trials were run in which a group of 10 individuals were subjected to stepwise fluctuations of pH (both up and down) starting at a pH value of 8.0. Separate trial groups were used for each trial. pH levels were manipulated by titration with 2nHC1 (pH decrease) and 2NNaOH (pH increase). Levels were raised or lowered by half a pH unit per hour. The time period of one hour was decided upon after subjecting a pre-trial group of 10 individuals to water with a pH value of 2. In just under an hour all individuals were dead. VWR Scientific Products® benchtop pH meters (Model SB21) were used to monitor pH levels during trials. Stream water was used in these trials rather than de-ionized water as a result of discrepancy between the pH levels of stream and de-ionized water. Death was signaled by individuals bending at the first abdominal segment accompanied by an inability to remain attached to the artificial substrate. Control groups of ten individuals were run simultaneously. All other water parameters remained constant during trials. Results Dissolved Oxygen Tolerance. — When exposed to stepwise reduction of dissolved oxygen, survivorship of Caenis latipennis showed a subtle decrease once a dissolved oxygen concentration of 7.0 mg/L was reached. However, a dramatic decrease in survivorship was observed after dissolved oxygen concentration levels were reduced to 4.5 mg/L (Fig. la). Mortality continued to increase with relative dissolved PUCKETT & COOK 127 D.O. Concentration (mg/I) pH pH Figure 1 . Survivorship of three Caenis latipennis (a) dissolved oxygen tolerance threshold trials, (b) pH decrease trials (One-way ANOVA on ranks [P=0.795]) and (c) pH increase trials (One-way ANOVA on ranks [P= 1 .0001). 128 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 oxygen reduction with no individuals surviving below 1.5 mg/L. Percent survivorship of the control groups during trials 1 , 2 and 3 were 80%, 90% and 100% respectively. Thermal Tolerance.— CTMax trials show that the average upper critical temperature for Caenis latipennis is 37.8 °C. All ten individuals subjected to CTMax trials died between 36.7°C and 38.5 °C. The critical thermal maximum temperature of individuals in these trials was well above the maximum temperature value recorded in the stream during the monitoring period (22.3°C). Controls were run simul¬ taneously at a temperature of 24.5 °C with no mortality. pH Trials.— The critical lower pH level under which Caenis latipennis could not survive was 2.5 (Fig. lb). In two of the three trials all individuals were alive after being exposed to stepwise decrease of pH to a level of 3.0 with 100% mortality after exposure to the same water at a pH of 2.5. During the third trial 40% of the individuals died at pH of 3.0 with the remaining individuals dying at a pH of 2.5. The lowest pH value recorded from a stream during the monitoring period was 7.7. Controls groups were run during the trials in a sample of the same water that was used for trial groups. This water maintained a pH of 8.2 from collection through the end of trials. No mortality was recorded in the control groups. The critical upper pH level above which Caenis latipennis could not survive was 12.5 (Fig. lc). All individuals in each of three trials were alive after being exposed to stepwise increase of pH to a level of 12.0, after which at a pH value of 12.5 all three groups experienced 100% mortality. The highest pH value recorded during the monitoring period was 8.6. Control groups were run during the trials at a pH of 8.2 in which no mortality was recorded. Discussion Caenis latipennis can cope with dramatic fluctuations in pH, dissolved oxygen, and temperature. It is very unlikely that under natural conditions C. latipennis larvae would be exposed to water quality parameter values that would fall outside of the tolerance values determined in the laboratory. This suggests a species that should be considered extremely tolerant of a wide range of values pertaining to the water quality parameters investigated in this study. This information is in agreement with previously published pollution tolerance values regarding C. latipennis by Hilsenhoff (1987). PUCKETT & COOK 129 The unlikelihood that the values of the parameters investigated here should, in natural systems, fall outside of this species range of tolerance suggests that the utility of C. latipennis as an indicator of stream quality is limited. However, when found in systems of low mayfly diversity this species and others found to be similarly tolerant could serve as valuable predictors of acute stream perturbation. At best, C. latipennis should be assigned little weight when included in stream assessments based on some biological index such as Hilsenhoffs Biotic Index. The relative ease with which the range of tolerance values regarding the parameters investigated were obtained suggests that empirically derived tolerance ranges for most Ephemeropteran species can be determined. Due to general similarities in morphology, life history, and ecological requirements, it is likely that these laboratory methods could also be used to gather data regarding physiological requirements of other stream macroinvertebrates such as the orders Plecoptera and Trichop- tera. With specific data regarding true tolerance ranges of these insects and other stream invertebrates, bioassessment practices can be ap¬ proached and interpreted with greater accuracy and relied upon with greater confidence. Acknowledgments We thank The Texas Academy of Science for partial funding of this project through the 2002 student research award. For use of equipment we thank Dr. Bill Lutter schmidt, Dr. Andrew Dewees and Dr. Jack Turner. Special thanks to Brandon Lowery for help in specimen collection. Literature Cited Barbour, M. T., J. Gerritsen, B. D. Snyder & J. B. Stribling. 1999. Rapid Bioassessment Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates and Fish, Second Edition. EPA 841-B-99-002. U.S. Environmental Protection Agency; Office of Water; Washington D.C., 339 pp. Edmunds, G. F., Jr., S. L. Jensen & L. Berner. 1976. The mayflies of North and Central America. Univ. Minnesota Press, Minneapolis. 330 pp. Hilsenhoff, W. H. 1977. Use of arthropods to evaluate water quality of streams. Technical Bulletin Wisconsin Department of Natural Resources 100:1-15. Hilsenhoff, W. H. 1982. Using a biotic index to evaluate water quality in streams. Technical Bulletin Wisconsin Department of Natural Resources, 132:1-22. Hilsenhoff, W. H. 1987. An improved biotic index of organic stream pollution. Great Lakes Entomol . , 20:31-39. Lutterschmidt, W. I. & V. H. Hutchison. 1997. The critical thermal maximum: history and critique. Can. J. of Zool., 75:1561-1574. 130 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 Lydy, M. J., C. G. Crawford & J. W. Frey. 2000. A comparison of selected diversity, similarity, and biotic indices for detecting changes in benthic-invertebrate community structure and stream quality. Arch, of Environ. Contam. Toxicol., 39:469-479. Provonsha, A. V. 1990. A revision of the genus Caenis in North America (Ephemeroptera: Caenidae). Trans. Am. Entomol. Soc., 116:801-884. Puckett, R. T. 2003. Bioassessment potential and water quality tolerance thresholds of larval ephemeroptera in southeast Texas streams. Unpublished M.S. thesis, Sam Houston State Univ., Huntsville, Texas, 74 pp. Rabeni, C. F., N. Wang & R. J. Sarver. 1999. Evaluating adequacy of the representative stream reach used in invertebrate monitoring programs. J. N. Am. Benth. Soc., 18:284-291. RTP at rpuck@tamu.edu TEXAS J. SCI. 56(2): 131-140 MAY, 2004 NATURAL HISTORY OF THE SOUTHERN PLAINS WOODRAT NEOTOMA MICROPUS (RODENTIA: MURID AE) FROM SOUTHERN TEXAS John R. Suchecki*, Donald C. Ruthven, III, Charles F. Fulhorst and Robert D. Bradley* * Department of Biological Sciences Texas Tech University, Lubbock, Texas 79409-3131 , Chaparral Wildlife Management Area, P.O. Box 115 Artesia Wells, Texas 78001 and Department of Pathology UT Medical Branch, Galveston, Texas 77555 Abstract. — One hundred forty-eight middens of the southern plains woodrat {Neotoma micropus ) were excavated from eight study sites on the Chaparral Wildlife Management Area in southern Texas. Several parameters were examined within and between study sites, including sex and age of individuals, demographics of occupancy, and distance between middens. One hundred seventy-seven individuals were captured, with significantly more adult woodrats represented than any other age category. Ninety males and 87 females were captured indicating an equal sex ratio. Analyses revealed that no difference existed in distances between male middens or in distances between female middens. Together, the data suggest no apparent patterns of social structure in woodrats at this study site. The southern plains woodrat {Neotoma micropus) is distributed from southeastern Colorado and southwestern Kansas through western Texas into northern Mexico (Hall 1981; Wilson & Reeder 1993). In Texas, N. micropus occupies the western two-thirds of the state, and generally is associated with brushlands of the semi-arid region between the eastern timberlands and the arid deserts to the west (Davis & Schmidly 1994). Woodrats construct middens (nests) from sticks, cactus, and other debris that are arranged into an above ground pile (Finley 1958; Birney 1973). It is common to find aluminum cans, spent ammunition casings, trash, and livestock dung on or within a midden, giving woodrats the nickname "packrat." Below ground (if soil composition/texture permits excava¬ tion), a midden usually contains an elaborate tunnel system. In this tun¬ nel system, woodrats store food and nest material, and avoid predation. In areas where non- friable soils do not permit the excavation of tunnels, woodrats often rely on crevices in rocks, decaying timber and canopies of trees for housing. Virtually all middens, whether in trees or below ground, have the characteristic mound of sticks over the opening. Several studies have been conducted on the systematics and phylo¬ genetic relationships of woodrats (see Edwards & Bradley 2002). How- 132 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 ever, only a few studies have examined natural history parameters. These indicate that woodrats are solitary and territorial animals (Braun 1989; Conditt & Ribble 1997; Johnson 1952; Raun 1966) with size of territories or home ranges most likely depending on density of individu¬ als and availability of food. Among the most detailed study, to date, is the study by Conditt & Ribble (1997) conducted on N. micropus at a research site in south central Texas. From March 2001 to January 2003, woodrat middens were excavated and occupants captured as part of an ongoing study examining the ecology of the White Water Arroyo arenavirus. Although woodrats were collected under a destructive sampling design, natural history parameters and other life history traits were recorded during the study. The objective of this study was to compare and contrast these natural history attributes (density, distance between middens, sex ratio, number of young, number of animals per midden, and age class distribution) to that available from other studies of N. micropus , especially to those of Conditt & Ribble (1997) and Henke & Smith (2000) whose study sites were located approximately 160 km northeast and 175 km southeast, respectively, of the study site examined during this study. Materials & Methods Study sites for this project were located on the Chaparral Wildlife Management Area (CWMA; 28° 20’ N, 99° 25’ W) that consists of 6,500 ha in the Rio Grande Plains of southern Texas (Ruthven & Synatzske 2002). The CWMA is located approximately 160 km south of San Antonio, between Catarina and Artesia Wells, Texas on Highway 133. The CWMA occurs within Dimmit and La Salle counties with the county border approximately bisecting the property. Soils typically are classified as Duval Fine Sandy Loam (DYB) and Dilley Fine Sandy Loam (DFC) (Stevens & Arriaga 1985). Average annual precipitation is 55 cm with most precipitation occurring between the months of April and September (Stevens & Arriaga 1985). Vegetation (McLendon 1991 ; Ruthven & Synatzske 2002) includes woody species such as mesquite {Prosopis glandulosa ) and granjeno ( Celtis pallida ), herbaceous species such as Lehmann lovegrass ( Eragrostis lehmanniana) , fringed singal- grass ( Brachiaria cilliatissima ) , and hairy grama ( Bouteloua hirsuta) as well as a wide array of cactus species ( Opuntia sp.). Dominant plant species coupled with climatic factors results in classification as a semi-arid acacia-grassland or mesquite-grassland. SUCHECKI ET AL. 133 Figure 1. Map depicting the locations of the eight midden sites examined in this study. Dashed line depicts the county line separating Dimmit and La Salle counties and the heavy black line depicts the boundaries of the Chaparral Wildlife Management Area. The soil composition and availability of food and cover on CWMA provide habitats capable of supporting large populations of woodrats (Finley 1958; Raun 1966). The northern half of CWMA is relatively more open and contains a higher concentration of grassland habitat, whereas the southern half contains a greater concentration of brush. Rotational grazing with cattle occurs yearly during the period October through April. Fire is used throughout the property to control brush and provide livestock and native species with food resources and cover. Woodrats were captured (by hand) during the excavation of middens located at eight different sites (Fig. 1). Sites were defined as an area possessing suitable habitat for maintaining a high density of woodrats (a high density was required for aspects of the arenavirus study). Sites were selected using a predetermined protocol to provide a relative means of providing a uniform density (high) among sites. Sites for this study (0.2 ha) were circular with a 25 m radius. Sites were not located closer than 500 m from any other site. Once a suitable area was selected, a center point was determined and middens visible along a 25 m transect (in each cardinal direction) were counted. If the number of middens observed along transects was equal to or greater than 10, the site was deemed suitable for excavation. If the number of middens was less than 10, a new site was selected and the protocol repeated. Excavation was 134 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 1. Comparison of age and sex across the eight midden sites by collecting date. Roman numbers refer to midden sites. Number of individuals for each age class and sex are in parentheses (males, females). Site Date Number Age Class Adult Subadult Juvenile Pup I Mar 2001 31 (15, 16) 10 (4, 6) 0 (0, 0) 7 (4, 3) 14 (7, 7) II Jun 2001 25 (13, 12) 13 (6, 7) 11 (6, 5) 1 (1,0) 0 (0, 0) III Oct 2001 23 (10, 13) 13 (7, 6) 5 (1,4) 5 (2, 3) 0 (0, 0) IV Jan 2002 19 (8, 11) 14 (6, 8) 3 (2, 1) 2 (0, 2) 0 (0, 0) V Mar 2002 21 (10, 11) 13 (6, 7) 1 (1,0) 5 (2, 3) 2(1, 1) VI Jun 2002 20(10, 10) 13 (6, 7) 3 (2, 1) 4 (2, 2) 0 (0, 0) VII Oct 2002 19 (11,8) 12 (6, 6) 4 (4, 0) 3 (1, 2) 0 (0, 0) VIII Jan 2003 19 (13, 6) 16 (10, 6) 2 (2, 0) 1 (1,0) 0 (0, 0) Total 177 (90, 87) 104 (51, 53) 29(18, 11) 28(13, 15) 16(8, 8) conducted four times per year (January, March, June and October) over a two-year period (Table 1). Each site was excavated only once during the study; and all sites were excavated during a single trip to circumvent seasonal biases. Every midden within the boundaries of each site was excavated, regardless of appearance. Because of the potential for an extensive tunnel system within a midden, every tunnel was excavated to its termi¬ nation point to ensure that all individuals were captured from the midden or to determine if the midden truly was uninhabited. Universal Trans¬ verse Mercator (UTM) coordinates were recorded with a hand-held GPS unit for each midden excavated regardless if midden was inhabited or vacant. These coordinates were later used to map each site to establish a geographical perspective (Fig. 2). If an individual woodrat escaped during the excavation of a midden, an immediate effort was made to recapture it. Excavation activities were conducted during daylight hours when rodent activity was lowest (wood- rats are nocturnal). Each captured woodrat was assigned a TK number (Museum of Texas Tech University identification number), weighed, sexed, aged, reproductive status determined and locality (UTM) record¬ ed. Ages were catagorized as adult, subadult, juvenile, and pup based on molting pattern (adult versus subadult), size/mass (subadult versus juvenile), and attachment to mammae (juvenile versus pup). Animals were either sacrificed (voucher specimens deposited in the Museum at Texas Tech University) or transported to the University of Texas Medi¬ cal Branch at Galveston, Texas for inclusion in a prospective study on the biology of arenaviruses in N. micropus. SUCHECKI ET AL. 135 Midden Site III Figure 2. Map of Midden Site III. Distance between grid lines is 5 meters. Spatial relationships between middens were constructed using UTM coordinates collected in the field (maps are oriented using north and east corrdinants). The labels include age class, gender, and museum identification number (TK) of the woodrats collected from the middens. Abbreviations include: AM = adult male, SAM = subadult male, AF = adult female, SAF = subadult female, and J = juvenile. The Chi-square test (x2) and Student’s f-test were performed, among midden sites and within midden sites, to test for statistically significant differences in age, sex, distance from other middens, etc. For examin¬ ing differences between the distribution of adult male and female wood- rats, a gender- specific centroid was calculated for each midden site using UTM coordinates collected in the field. The distance of each midden from the corresponding centroid was measured in meters, and the mean of the adult male woodrat-centroid distances was compared to the mean of the adult female woodrat-centroid distances using a RankSum test. Middens that were co-occupied by adults of different sexes were not included in this study. Results One hundred forty-eight middens were excavated and 177 individuals were captured (Table 1). Five escapees, that were not recaptured, 136 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Table 2. Mean distances between middens occupied by adult male or adult female woodrats, by site. Ranges are in parentheses. Site Number of middens Male Mean distance Number of middens Female Mean distance I 4 19.5(16.8-24.1) 5 14.2 (9.8-20.5) II 7 7.9 (1.4-15.5) 7 8.2 (5.0-15.6) III 6 8.8 (7.2-11.0) 7 11.4 (2.0-23.2) IV 5 14.0(10.6-17.2) 9 12.9 (5.0-17.1) V 7 14.7(12.5-11.1) 7 9.9 (2.8-16.4) VI 7 7.3 (5.0-7.2) 6 7.7 (3.0-28.2) VII 8 5.8 (4. 2-7. 2) 6 9.3 (3.6-13.6) VIII 11 7.9 (3.6-13.9) 6 14.8(11.0-19.7) occurred during the study. Number of captures by site ranged from 19 to 31 and number of middens per site ranged from 11 to 23. Age— Individuals were separated into four age classes (adult, sub¬ adult, juvenile and pup) resulting in 104 adults (58.8% of the total population), 29 subadults (16.4%), 28 juveniles (15.8%), and 16 pups (9.0%) being captured. Comparison of age classes (Table 1) revealed a difference in the number of individuals within age classes across sites (x2= 90.39, df = 21, P < 0.001), with adults typically being more numerous than either subadults, juveniles or pups. However, in Site I, adults and pups were more numerous than subadults or juveniles and in Site II, adults and subadults were more numerous than juveniles or pups. Sex. — Ninety males (50.8% of the population) and 87 females (49.2%) were captured (Table 1). A /-test revealed no difference between the number of males and females across the eight middens ( t = 0.32, df = 15, P > 0.05) or in a comparison of sex by age class ( t = 0.35, df=3,P> 0.05). No differences (x2 = 100.49, df = 45, P > 0.05) were found between sexes by age class over the eight midden sites. Distances between middens. — Calculation of distances between mid¬ dens (nearest-neighbor distance) were calculated from UTM coordinates as shown in (Fig. 2). Estimates from all study sites (Table 2) resulted in a mean of 6.58 m (range: 1.70 - 14.12 m). Mean distance between male middens for the eight study sites was 10.75 m (range: 7.23 - 16.40 m), whereas mean distance between female middens was 11.05 m (range: 5.16 - 19.49 m). No significant difference in distance was detected between each midden among the eight sites (x2, P > 0.05 for each of the eight sites), between sexes within sites (/-test, P > 0.05 for each of the eight sites) , or in mean differences between sexes among the SUCHECKI ET AL. 137 Table 3. Average distances between middens occupied by adult woodrats and gender-specific centroids, by midden site. The number of males or females captured is in parentheses. Midden site Gender I II III IV V VI VII VIII Overall Male Female 16.9 (4) 15.7 (5) 15.3 (6) 17.8 (7) 20.7 (7) 20.3 (6) 18.0 (6) 22.6 (8) 20.2 (6) 18.1 (7) 10.9 (6) 16.6 (7) 16.0 (7) 14.8 (5) 16.3 (10) 20.2 (6) 18.5 (52) 16.8 (51) eight sites (/ = 0.22, df = 7, P > 0.05). Middens containing both adult males and adult females were excluded from this analysis, as it was impossible to determine whether the male or female was the primary occupant of the midden. The means of midden-centroid distances of male and female woodrats were 18.5 m (range: 3.9 - 32.6 m) and 16.8 m (range: 5.6 - 31.9 m), respectively. The results of a RankSum test (Table 3) indicated that there was no statistically significant difference (Type I error = 0.10) between the mean midden-centroid distance of male woodrats and the mean midden- centroid distance of the female woodrats. Middens /site. — Average number of middens per site was 18.37. No differences were identified between number of middens per site ( t = 0.00, df = 7, P > 0.05), number of male middens versus female middens among sites (x2 = 3.13, df — 7, P > 0.05), or number of male middens versus female middens within sites ( t = 0.27, df = xx, P > 0.05). Site VIII contained the greatest number of middens (23), whereas Site I had the fewest (11). Site I had the largest number of individuals (31) and Sites IV, VII and VIII had the fewest (19). Occupancy per midden— One hundred six of the 148 excavated middens (71.6%) were occupied. Calculations of multiple occupancy, (how many individuals of each age class and sex occupy the same midden), indicated that adult females and their young were found together on 27 (18.2%) occasions. Using number of pups as a baseline, the average litter size is two (27 females with 54 pups). Adult females and adult males were found in the same midden 13 (8.7%) times. The greatest number of individuals found in a single midden was six and five middens contained five individuals. Discussion Several parameters were examined and only age class structure varied statistically by site, season or between years. The adult age class ( n — 104) contained the highest number of individuals and the pup age class 138 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 contained the least ( n = 16). Interestingly, pups were collected only in March indicating a peak reproductive effort in late winter or early spring; however, the presence of juveniles during other months suggests that some reproduction occurs throughout the year. In addition, this time frame corresponded to the population peak in February reported by Conditt & RIbble (1997). However, reproductive efforts appeared to taper off more rapidly in this study than reported in Conditt & Ribble (1997), where the number of lactating or pregnant females peaked at 50% in October. Because of the idea that the woodrats might have a polygynous or promiscuous mating system (Conditt & Ribble 1997), it was assumed that the number of females collected would be more numerous than males. As indicated, there was no significant difference in the overall number of males (90) compared to females (87) or in any age class. The ratio of adult males to females was 1:0.97; whereas, the study by Conditt & Ribble (1997) reported a ratio of 1:1.16. The social structure within the midden itself was another aspect of the study that did not hold true with assumptions pertaining to woodrat habits. The most surprising finding was that adult males and adult females being captured within the same midden. Conditt & Ribble (1997) never observed more than one adult woodrat in a midden at the same time. However, during this study, an adult male and an adult female were observed in 13 middens. There are at least two possibilities to explain this. The simplest would be that the male was there solely for mating purposes. Although this may be true, all middens were exca¬ vated during daylight hours, and N. micropus is a nocturnal species. Because of this, several questions arise as to the social habits of N. micropus. How long does courtship take place, perhaps they stay "over¬ day. " Second because of high densities of woodrat middens on CWMA, perhaps adult males and females cohabitate. Parameters of this study do not provide significant conclusions to these questions. Because of the direct capture of all individuals throughout all midden sites, one aspect of their natural history that could not be measured is home range. Studies by Henke & Smith (2000) and Conditt & Ribble (1997) that examined home range within N. micropus found the home range of males to be 1696 m2 and 1829.2 m2, respectively. Female home range was found to be significantly less at 188 m2 and 258.2 m2, respectively. Although one could not calculate home ranges due to the destructive sampling design of this study, the data are not consistent with a harem mating system. Instead, maps of each midden site revealed no SUCHECKI ET AL. 139 visible patterns that would support social structure regarding midden placement or midden selection by males or females. In addition, if a polygynous or promiscuous mating system existed, average distances between male middens and average distances between female middens should differ. For example, there should be a “standard” distance between male middens and to a lesser degree for female midden dis¬ tances. Statistical tests failed to support this hypothesis. In addition, the number of woodrats per hectare in this study was 110.6 and the number of middens per hectare was 92.5. The number of adult males was 31.9 and the number of adult females was 33. 1 per hectare. These numbers are much greater than that found (2.0 woodrats per hectare in October to 5.5 per hectare in February) by Conditt & Ribble (1997). One possible explanation for the large increase is that this study was biased for high densities of woodrat middens; these numbers obviously would be lower if sites had been selected at random. Due to the large numbers of woodrats per hectare and abundance of resources, home ranges of woodrats on CWMA are most likely not that large. When superimposed (not shown) on a map of the midden sites (produced in this study) , the home ranges reported by Conditt & Ribble (1997) and Henke & Smith (2000) for a single individual would extend well beyond the boundaries of the entire midden site. This is somewhat surprising given the similarities in habitats and geographic proximity of the three studies. This study answered several questions regarding the natural history of N. micropus. Because of suitable habitat conditions, CWMA is ideal for sustaining large populations of woodrats. The large amount of food and cover resources available to woodrats on CWMA enable populations to not only survive but do so in such close proximity with each other that early predictions on habits and social structure simply do not apply. Acknowledgments We thank D. S. Carroll, B. R. Amman, J. D. Hanson, F. M. Mendez-Harclerode, S. A. Reeder, M. L. Haynie, N. D. Durish, L. K. Longhofer, L. R. McAiley, A. Vestal, B. D. Cabbiness and J. G. Brant (Texas Tech University) and C. Milazzo, Jr., M. L. Milazzo, M. Cajimat, S. Gardner, J. Comer (University of Texas Medical Branch) for assistance in field work. Special thanks to J. D. Hanson for assistance with data analysis. D. R. Synatzske and other members of Texas Parks and Wildlife Department at the CWMA provided important logistical help. R. J. Baker and the staff at the Natural Science 140 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Research Laboratory, Museum Texas Tech University provided assis¬ tance with specimen deposition. This research was supported by the National Institutes of Health (grant DHHS A 14 1435-01) entitled "Ecology of emerging arenaviruses in the southwestern U.S.". Literature Cited Birney, E. C. 1973. Systematics of three species of woodrats (genus Neotoma) in central North America. Misc. Publ., Mus. Nat. Hist., Univ. Kansas Publ., 58:1-173. Braun, J. K. 1989. Neotoma micropus. Mammalian Species, 330:1-9. Conditt, S. A. & D. O. Ribble. 1997. Social organization of Neotoma micropus , the southern plains woodrat. Am. Mid. Nat, 137(2): 290-297. Davis, W. B., & D. J. Schmidly. 1994. The mammals of Texas. Texas Parks And Wildlife Press, Austin, 338 pp. Edwards, C. W. & R. D. Bradley. 2002. Molecular systematics of the genus Neotoma. Mol. Phylo. Evol., 25(3): 489-500. Finley, R. B., Jr. 1958. The wood rats of Colorado: distribution and ecology. Mus. Nat. Hist., Univ. Kansas Publ., 10:213-552. Hall, E. R. 1981. The mammals of North America. 2nd ed. John Wiley & Sons, New York, vi + 601-1181 + 90. Henke, S. E. & S. A. Smith. 2000. Use of aluminum foil balls to determine home ranges of woodrats. Southwest. Nat., 45(2):352-355. 71(4):510-519. Johnson, C. W. 1952. The ecological life history of the packrat, Neotoma micropus, in the brushlands of Southwest Texas. Unpubl. M.S. Thesis, Univ. Texas, Austin. 115 p. McLendon, T. 1991 . Preliminary description of the vegetation of south Texas exclusive of coastal saline zones. Texas J. Sci., 43(1): 13-32. Raun, G. G. 1966. A population of woodrats {Neotoma micropus) in southern Texas. Bulletin of Texas Memorial Museum, 11:1-62. Ruthven, D. C., Ill & D. R. Synatzske. 2002. Response of herbaceous vegetation to summer fire in the western south Texas Plains. Texas J. Sci., 54(2): 195-210. Stevens, J. W. & D. Arriaga. 1985. Soil Survey of Dimmit and Zavala Counties, Texas. United States Department of Agriculture, Washington D.C. Wilson, D. E. & D. M. Reeder. 1993. Mammal species of the world. 2nd ed. Smithsonian Institution Press, Washington D.C., 1206 pp. RDB at: robert.bradley@ttu.edu TEXAS J. SCI. 56(2): 141-148 MAY, 2004 ADULT FORAGING BEHAVIOR IN MEARNS’ GRASSHOPPER MOUSE, ONYCHOMYS AREN1COLA (RODENTIA: MURID AE) IS INFLUENCED BY EARLY OLFACTORY EXPERIENCE Fred Punzo Department of Biology University of Tampa Tampa, Florida 33606 Abstract.— Studies were conducted to assess the effects of early exposure to food-borne olfactory cues and subsequent searching behavior and odor preferences in adult males of the grasshopper mouse, Onychomys arenicola. Twenty-day old mice were randomly assigned to 1 of 3 treatment groups: a control group (CG) was fed on crickets (Acheta domesticus ) and mealworms ( Tenebrio molitor). Another group (EG) received an enriched diet of crickets, mealworms, roaches ( Periplaneta americana), and commercial dog and cat chow. The IG group received an impoverished diet consisting only of crickets. These feeding regimes continued for 80 days. Mice were then presented with odor choice tests in a Y-maze olfactometer. Mice from each treatment group were tested for their choices between known and novel prey odors (NPO), and between known odors and a novel pure chemical odor (NCO). Control mice exhibited a preference of 70% for the known prey odor (cricket) and only 30% for the NPO (wolf spider, Hogna carolinensis). In contrast, EG mice showed a significantly higher preference (70%) toward the NPO. Only 20% of the IG animals chose the NPO. In addition, EG mice made decisions on which odor to investigate significantly faster than CG or IG animals. These results indicate that O. arenicola relies on olfactory cues when making decisions concerning prey choice during foraging bouts. They also suggest that knowledge of olfactory cues associated with prey is not innate in this species, but is acquired during early sensitive periods of development (olfactory imprinting). This is the first demonstration of olfactory imprinting in a murid rodent within the genus Onychomys. Previous studies have shown that early olfactory experience can affect the subsequent foraging behaviors or prey choice of adult predators including insects (Chapman et al. 1987), spiders (Punzo & Kukoyi 1997), rock crabs (Rebach 1996), turtles (Punzo & Alton 2002), lizards (Punzo 2003a), polecats (Apfelbach 1973), ferrets and other mustelids (Apfelbach 1992), murid rodents (Berdoy & Macdonald 1991), shrews (Churchfield 1990; Punzo 2003b) and canids (Weldon 1990). Further¬ more, a study on the ferret Mustela putorius showed that olfactory imprinting may be involved because certain odors encountered by young animals during sensitive periods can serve as acquired sign stimuli for 142 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 subsequent prey identification and selection (Apfelbach 1992). How¬ ever, little information is available on the effects, if any, of early olfactory experiences on subsequent foraging behavior in murid rodents (Frank & Heske 1992). Mearns’ grasshopper mouse, Onychomys arenicola (Rodentia: Muridae) is an inhabitant of low desert areas in west Texas. They prefer foothills, xeric flats and mesquite-covered mesas with sandy soils (Whitaker 1996), and feed primarily on a variety of arthropods and small vertebrates as well as seeds (Horner et al. 1965; Brown & Zeng 1989; Punzo 2000). The purpose of the present study was to assess the influence of early olfactory experience on subsequent searching behavior and odor preferences of adults of O. arenicola. Materials and Methods All animals used in these experiments were the second or third generation offspring of adults originally collected from several localities within a 4 km radius of Redford, Texas (Presidio County) in 1999 and 2000. This area lies within the northern region of the Chihuahuan Desert. The experimental protocol used in this study was similar to that employed by Apfelbach (1978). To summarize, 10 newly weaned mice were randomly assigned to each of three groups, all of which were fed a diet of crickets {Acheta domesticus ) and mealworms ( Tenebrio molitor) until they were 20 days old. After this time, each group was fed on a different diet regime until the age of 80 days. A control group (CG) continued to receive crickets and mealworms; another group (EG) was fed an "enriched" diet consisting of crickets, mealworms, roaches (Periplaneta americana) and commercial cat and dog chow (Ralston Purina, St. Louis, MO). An impoverished group (IG) received only crickets. In addition, to enhance olfactory deprivation, the IG group was exposed to an artificial olfactory environment saturated with the odor of geraniol. It has been reported that the continuous exposure to a single predominant odor can mask the ability of an animal to experi¬ ence other environmental odors resulting in what has been termed a state of olfactory deprivation (Weldon 1990). Behavioral studies were conducted on adult males from the three PUNZO 143 Figure 1. Diagram of the Y-maze olfactometer used in odor choice experiments. GT = glass tubing; F = flowmeters; V = valves. Arrows indicate direction of air flow. See text for details. groups (n — 10/group) when they reached 7 months of age. These animals were tested in a Y-maze olfactometer to determine if there was any preference shown toward certain odor cues. Two tests were given to each animal: one in which the subject was given a choice between a known prey odor and a novel prey odor, and another test where the choice was between a known prey odor and a novel pure chemical odor. There was a 10 min delay between tests. The general procedure was similar to that employed by Apfelbach (1992). To summarize, the olfactometer consisted of a Y-maze constructed of plexiglass (Fig. 1) connected to sources of odor via glass tubing (GT). The air and odor flow were controlled through the use of flowmeters (F) located before and after the odor saturators (odors). Teflon valves (V), located at each end of the Y-maze, were used to control the direction of flow of the odors. Test odors were randomly introduced into the left or right end of the maze before a mouse was allowed to leave the start box. Test odors consisted of a known prey odor (cricket) , a novel prey odor (wolf spider, Hogna carolinensis) and a novel pure chemical odor (oil of wintergreen) . At the start of each trial an individual mouse, food- deprived for 72 hr, was placed into the start box and allowed to remain 144 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 90 n 80: N P O NCO Figure 2. Percent choice (percent of trials in which animals explored the novel odors) of adult males of Onychomys arenicola toward novel prey (NPO) and novel chemical (NCO) odors. Data show that EG animals (experienced an enriched diet) exhibited a greater tendency to explore novel odors as compared to the control (CG) and impoverished (IG) groups. Black bars = CG (control) group; stippled bars = enriched group (EG); unshaded bars = impoverished group (IG). there for a period of 10 min. The start box door was then lifted, and the mouse was allowed to enter the maze. A record was made of which arm of the maze was chosen (% choice) for each trial, as well as the time (sec) needed for a mouse to make its decision. An arm was considered chosen if the animal moved into it at least as far as point C or F. All observations were made behind a one-way mirror to minimize disturbance to the animals. Data on odor preference tests and time needed to make a decision were analyzed using Chi-Squared and Kruskal-Wallis tests (Sokal & Rohlf 1995). Results The results of the odor preference tests are shown in Fig. 2. In the choice condition of known prey odor (crickets) versus novel prey odor (NPO; wolf spiders), control animals (CG) exhibited a preference of 30% toward the NPO, and 70% to the cricket odor. In contrast, PUNZO 145 30n NPO NCO Figure 3. Amount of time (sec) required for males of Onychomys arenicola to make a decision as to which odor to choose. Data are expressed as means + SD (n = 10/group). Black bars = CG (control group); stippled bars = enriched diet group (EG); unshaded bars = impoverished group (IG); NPO = novel prey odor; NCO = novel pure chemical odor. animals exposed to an enriched diet (EG) showed a preference of 70% toward the NPO, whereas only 20% of the IG animals chose the NPO. The differences between the CG and EG, and between CG and IG were significant (P < 0.01). In addition, novel chemical odors (NCO) were less attractive to these mice than were novel prey odors. The time needed by these animals to make a decision as to which odor to investigate is shown in Fig. 3. In the choice condition of known prey odor vs. NPO, mice exposed to the enriched diet (EG) made decisions significantly faster than controls (P < 0.01) and IG (P < 0.001) animals. Similar results were obtained when a novel chemical odor (NCO) was presented rather than a NPO. Discussion These results indicate that the cricetid rodent Onychomys arenicola utilizes olfactory cues when making decisions during foraging bouts. They also suggest that knowledge of olfactory cues associated with prey is not innate in this species, but is acquired during early periods of development. This type of olfactory imprinting on cues associated with prey or other food items has been reported for animals from a diversity 146 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 of taxa including insects (Thorpe 1939; Chapman et al. 1987), spiders (Punzo & Kukoyi 1997), turtles (Burghardt & Hess 1966; Punzo & Alton 2002), lizards (Punzo 2003a), polecats (Apfelbach 1973), ferrets (Apfelbach 1992), and murid rodents (Berdoy & Macdonald 1991). To the author’s knowledge, this is the first demonstration of olfactory imprinting in a murid rodent within the genus Onychomys. It has been argued that the ability to imprint on specific environmental cues during an early sensitive maturational period would allow an animal to combine the advantages of hardwired specialist feeders with those of generalists who rely to a greater extent on learning (Johnston 1982; Stephens 1991). The situation whereby a predator is exposed to only a small number of prey items during some early sensitive period of life might facilitate the formation of an olfactory search image, thereby focusing food searching behavior for specific prey (Burghardt 1993). Thus, even though an animal may have the capacity to feed on a variety of food types (broad trophic niche), by concentrating on a single, abundant and reliable food encountered early in life, individuals would minimize energy costs associated with trial-and-error learning while benefiting from the increased foraging efficiency associated with having a single search image to facilitate hunting. In these experiments, grasshopper mice that were exposed to only a small number of prey objects early in life, did not respond strongly to novel prey odors and even less to a novel chemical odor, both of which convey less important olfactory information. Onychomys arenicola is found in xeric habitats, where seasonal fluctuations in prey availability are common (Punzo 2000). Although it is a generalist predator that feeds on a variety of arthropods, small vertebrates and seeds (Horner et al. 1965; Brown & Zeng 1989; Whitaker 1996), the ability to form an early search image associated with one or a few prey types that may be more locally abundant and available, would contribute to its overall fitness. Acknowledgments I thank R. J. Edwards, S. Jenkins, C. Lowell, G. Price and anonymous reviewers for comments on an earlier draft of the manuscript, L. Hane for assistance in procuring some of the research literature, and A. Nardelli for assistance in maintaining the animals in PUNZO 147 captivity. This study was supported by a Faculty Development Grant from the University of Tampa. Literature Cited Apfelbach, R. 1973. Olfactory sign stimulus for prey choice in polecats (Putorius putorius) . Zeitschrift Tierpsychologie, 33:273-281. Apfelbach, R. 1978. A sensitive phase for the development of olfactory preference in ferrets ( Mustela putorius). Zeitschrift Saugertierkiinde, 43:289-294. Apfelbach, R. 1992. Ontogenetic olfactory experience and adult searching behavior in the carnivorous ferret. Pp. 155-165, in Chemical signals in vertebrates. Vol. 6 (R. L. Doty & D. Muller-Schwarze, eds.), Plenum Press, New York, 598 pp. Berdoy, M. & D. W. Macdonald. 1991. Factors affecting feeding in rats. Acta Zoologica, 12:261-279. Brown, J. H. & Z. Zeng. 1989. Comparative population ecology of eleven species of rodents in the Chihuahuan Desert. Ecology, 70:1507-1525. Burghardt, G. M. 1993. The comparative imperative: genetics and ontogeny of chemorecpetive responses in natricine snakes. Brain, Behavior and Evolution, 41:138- 146. Burghardt, G. M. & E. H. Hess. 1966. Food imprinting in the snapping turtle. Science, 151:108-109. Chapman, R. F., E. Bemays & J. G. Stoffaland. 1987. Perspectives in chemoreception and behavior. Springer, New York, 466 pp. Churchfield, S. 1990. The natural history of shrews. Cornell University Press, Ithaca, New York, 178 pp. Frank, D. H. & E. J. Heske. 1992. Seasonal changes in space use patterns in the southern grasshopper mouse, Onychomys torridus. Journal of Mammalogy, 73:292-298. Horner, B. E., J. M. Taylor & H. Padykula. 1965. Food habits and gastromorphology of the grasshopper mouse. Journal of Mammalogy, 45:513-535. Johnston, T. D. 1982. The selective costs and benefits of learning: an evolutionary analysis. Advances in the Study of Behavior, 12:65-106. Punzo, F. 2000. Desert arthropods: life history variations. Springer, New York, 311 pp. Punzo, F. 2003a. The effects of early experience on subsequent feeding responses in the tegu, Tupinambis teguixin (Squamata: Teiidae). Journal of Environmental Biology, 24:23-27. Punzo, F. 2003b. The response of the least shrew ( Cryptotis parva) to olfactory cues associated with prey. Prairie Naturalist, 35:213-221. Punzo, F. & L. Alton. 2002. Evidence for the use of chemosensory cues by the alligator snapping turtle Macroclemys temminckii to detect the presence of musk and mud turtles. Florida Scientist, 65:134-139. Punzo, F. & O. Kukoyi. 1997. The effects of prey chemical cues on patch residence time in the wolf spider Trochosa parthenus (Cahmberlin) (Lycosidae) and the lynx spider Oxyopes salticus Hentz (Oxyopidae). Bulletin british Arachnological Society, 10:323- 326. Rebach, S. 1996. Role of prey odor in food recognition by rock crabs, Cancer irroratus Say. Journal of Chemical Ecology, 22:2197-2207. Sokal, R. R. & F. J. Rohlf. 1995. Biometry. 3rd ed. W. H. Freeman, New York, 887 pp. Stephens, D. W. 1991. Change, regularity, and value in the evolution of animal learning. 148 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Behavioral Ecology, 2:77-89. Thorpe, W. H. 1939. Further studies on pre-imaginal olfactory conditioning in insects. Proceedings Royal Society of London B, 127:424-433. Weldon, P. J. 1990. Responses by vertebrates to chemicals from predators. Pp. 500-521 in Chemical signals in vertebrates. Vol. 5 (D. Macdonald, D. Muller-Schwarze & S. E. Natynczuk, eds.), Oxford University Press, New York, 672 pp. Whitaker, J. O., Jr. 1996. Field guide to North American mammals. Alfred Knopf, New York, 936 pp. FP at: tpunzo@ut.edu TEXAS J. SCI. 56(2): 149-156 MAY, 2004 ROBOTICS REPEAT ABILITY AND ACCURACY: ANOTHER APPROACH Jan Brink, Bill Hinds* and Alan Haney Department of Manufacturing Engineering Technology and * Department of Mathematics , Midwestern State University Wichita Falls, Texas 76308 Abstract.— Repeatability is one characteristic of a robot, which is of tremendous importance. In this paper the concept of repeatability is clearly defined in terms of the standard deviation of the random component of the error of a robot in returning to a taught position and accuracy is defined in terms of the mean error as a function of three important variables. Data used to estimate repeatability and accuracy were obtained from a full- factorial experiment in which speed, payload and amount of axis movement were used as independent variables. The robot used to furnish data for this research was a PUMA 560. A regression model was developed to estimate the accuracy at various factor levels and the repeatability was determined to be 0.0036 inches. The statistical analysis clearly indicated that all three factors, as well as their interactions, affect the accuracy of the robot. The regression model indicated that approximately 35% of the radial error variability was explained by the linear model and 65% of the radial error was due to repeatability. The performance of a robot is highly dependent upon both the repeatability and the accuracy of the robot. Repeatability is the robot’s ability to return to a previously taught point (Rehg 1985). Repeatability is especially important in assembly applications of robots and has a critical effect on product quality since product tolerances are decreasing (Khouja & Kumar 1999). Specifications on robots are often obtained from robot vendors, but the problem with the use of these data is that the user does not know the conditions under which they are tenable. It is therefore necessary to investigate the interaction among various robot process variables and determine the conditions under which a given mix of values can be achieved (Offodile & Ugwu 1991). Repeatability and accuracy are often confused and rarely defined in a clear and unambiguous way. Necessarily, both accuracy and repeata¬ bility must be estimated by using the error made by the robot when trying to return to a previously taught point. This error is defined to be the radial distance from the previously taught point to the point at which the end effector comes to rest. The method for estimating accuracy and repeatability in this research will depend upon errors obtained experi¬ mentally by varying the speed, the weight of the payload and the amount of axis movement. More specifically, for any combination of the three variables the accuracy will be defined as the mean of the distribution of 150 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 errors for that combination. As a result of this definition, accuracy is constant for any fixed combination of speed, weight and amount of axis movement. Therefore, accuracy has no connection to the variability of the distribution of errors. The repeatability of a robot does depend on the variability of the distribution of errors. In fact, repeatability will be defined to be equal to three times the standard deviation of the distribu¬ tion of errors. The definitions of accuracy and repeatability in the above paragraph indicate that the mean of the distribution of errors depends on the values of the variables while the standard deviation does not. Therefore, the variation in the errors due to changes in the mean as a result of changes in the three input variables must be removed in order to estimate accuracy and repeatability. The standard mechanism for this task is a model for the means developed by using statistical techniques on data obtained from a designed experiment. Materials and Methods The parameters speed, payload and percentage of axis movements were varied on a PUMA 560 robot using different combinations to estimate accuracy and repeatability. A conventional X-Y-Z Cartesian coordinate measurement system was used. The points of movements to the X, Y and Z gauges were found by driving each of the six axes to different percentages of axis movements. Errors were measured using precision gauges for the X, Y and Z coordinates. A test stand was constructed for this experiment similar to the one discussed by Warnecke & Schraft (1982). The test stand was securely clamped down to a table that was leveled. The test stand allowed measurement of X, Y and Z deviations using three Mitutoyo dial indicator gauges. The three gauges used have flat faced contact plates. The resolution of the Mitutoyo gauges used is 0.0001 inches with a 0.25 inch stroke. The deviations were expected to be in the 0.001-0.004 inches range. The rule of “10” was therefore applied. This means the gages have a resolution 10 times the expected reading. The temperature in the laboratory was kept at a constant 71° F which is very close to the desired 68° F for precision measurements (DeGarmo et al. 1997). The three parameters weight, speed and percent of movement in each axis were varied at three different levels designated low, medium and high. A total of 27 different combinations were used. The PUMA robot used had six different axes. BRINKS, HINDS & HANEY 151 Weight.— The payload of the PUMA robot used was 2.5 kg (5.5 lbs). This did not include the gripper. Four “one” lb weights and two “0.5” lb weights were used. A special designed fixture that can be attached to the wrist was used for varying the weight. It included a precision ground 0.5000 inch diameter + /- 0.0001 inch precision tooling ball. The tooling ball probe has a small “negligible” weight. The probe was locked in position so no movement was available in the X, Y and Z direction. The following loads were used in this experiment: low (1.5 lbs * 30% of the payload), medium (3.0 lbs ~ 60% of the payload) and high (4.5 lbs « 90% of the payload). Speed. — Maximum speed of this robot was 0.5m/sec, which is equiva¬ lent to an external program speed of 100. The following speeds were used: low (30% of the maximum speed), medium (60% of the maximum speed) and high (90% of the maximum speed). Percent of range in each axis.— The maximum range of motion for each of the axes was as follows: Joint 1: 320 degrees (waist), Joint 2: 250 degrees (shoulder), Joint 3: 270 degrees (elbow), Joint 4: 280 degrees (wrist 1), Joint 5: 200 degrees (wrist 2) and Joint 6: 520 degrees (wrist 3). The following ranges of motion were used: low (10% of the total range), medium (30% of the total range) and high (50% of the total range) . The three ranges of the total motion used in this study are given in Table 1. The 50% axis movement was not exceeded because the return approach of the robot would have been unpredictable. The robot was operated for a 15 minute warm up period before the data gathering began. The point called GAUGE to which the end effector was programmed to return was located near one of the extreme points of the axis system. This extreme point was determined by rotating joints 1, 3 and 5 of the robot the maximum amount in the negative direction and joints 2, 4 and 6 the maximum in the positive direction. The fixture with the three gauges was located at the point called GAUGE and contact was made with the tooling ball to accurately zero the three gauges. The PUMA 560 Victor Assembly Language was used to create a program that drove the end effector to one of the three locations determined by the chosen values for the variable, amount of axis movement, and returned it to the point GAUGE. This movement was repeated ten times for each of the twenty-seven combinations for levels of speed, weight and amount of axis movement. The radial error 152 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Table 1 . Ranges of motion used for each of the six axes for each of the three levels of axis movement. Low (10%) Medium (30%) High (50%) Joint 1 : 32 degrees Joint 1 : 96 degrees Joint 1 : 160 degrees Joint 2: 25 degrees Joint 2: 75 degrees Joint 2: 125 degrees Joint 3: 27 degrees Joint 3: 81 degrees Joint 3: 135 degrees Joint 4: 28 degrees Joint 4: 84 degrees Joint 4: 140 degrees Joint 5: 20 degrees Joint 5: 60 degrees Joint 5: 100 degrees Joint 6: 52 degrees Joint 6: 156 degrees Joint 6: 260 degrees was measured each time the tooling ball returned to the point GAUGE. The total of 270 data measurements met the minimum for the twenty- seven factor level combinations according to the ANSI/RIA R 15. 02-2 standard (ANSI 1992). Results and Discussion There were 10 measurements taken at each of the 27 factor level combinations. Therefore, the experiment is considered a full -factorial experiment with 10 replications. The response variable was the radial distance R from the point gauge to the location of the center of the tooling ball. This distance was computed from the errors in the X, Y and Z directions by R = (X2 + Y2 + Z2),/2. After the experiment was designed and the 270 data points were obtained, data analysis was performed to determine if the three variables used in the experiment were all significant in determining the mean of error R. The analysis of the data using the Minitab software package yielded the main effects plots shown in Figure 1 . These main effects plots indicate that each of the three variables was significant in determining the mean error. In general, the mean error was increased when any of the variables were changed from their medium or zero setting which indicated a quadratic relationship. Further evidence of the influence of the variables can be seen in Figure 2 which shows a graph of the error data in groups of ten replicates. This graph clearly indicates that the replicates produced tightly grouped errors but changes in levels of the three factors produced large changes in the magnitudes of the errors. Much of the variability in the values of the errors reflects changes in the factor levels. In order BRINKS, HINDS & HANEY 153 Figure 1 . Main Effects Plot for R. The mean radial error is given as a function of each of the three factors at three levels as used in the study. Order of Observation Figure 2. Radial Errors in Replicate Groups. The factor levels were changed after each group of ten observations. The data represent 270 observations with 27 different factor level combinations. to get to the component of the data that reflects the repeatability of the robot, regression techniques with a linear model were used to remove the variation due to changes in factor levels. Figure 3 reveals the distribution of the random components of the data that determines the repeatability characteristic of the robot. This graph indicates an approximately normally distributed random pattern of error variation about the mean for the particular factor level combination at which the 154 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 Residuals (.0001 inch) Figure 3. Histogram of Residuals Errors. Residual errors were calculated by subtracting the predicted radial error or accuracy from each measured radial error. readings were taken. The computations in the analysis of variance ( ANOVA ) and the linear regression yielded the following equation to predict the accuracy: RMEAN(W,S,A) = 11.8 - 3.25 S + 1.44 A + 1.85 W - 3.32 W*? - 3.02 W*A + 4.85 S*A + 2.23 W2 + 8.59 S2 + 4.21 A2 Where: W = weight, S = speed and A = percent of axis movement. The computations confirmed that the three factors, as well as their interactions, are statistically significant (P < 0.05) in the mean of the radial error values. Analysis of variance (ANOVA) computations produced a computed value of 98.4 for the variance of the random component of the radial error values. The square root of the variance yields the standard deviation of the random component to be 9.9. The repeatability of a robot was defined to be three times the standard deviation of the random component of the radial error. Therefore, the estimate for the repeatability of the Puma 560 turns out to be 29.7. Since measurements were in 0.0001 inch units, the repeatability estimate would be stated as 0.00297 inch. The estimate is somewhat smaller than the ±0.004 inch specified by the manufacturer. If regression techniques had not been used to remove the variability due to the changes in the factor levels, the standard deviation of the raw data would be 12.12. This standard deviation yields 0.001212 when the units are changed to BRINKS, HINDS & HANEY 155 inches and a corresponding repeatability estimate of 0.0036 inch. When rounded to the nearest thousandth of an inch, this estimate agrees with the manufacturer’s estimate. The adequacy of such a model is usually judged by R 2, the coefficient of determination, because it gives the fraction of the total variation in the radial error data explained by the model. This model developed for predicting the accuracy of the robot had an F? value of 35.2%. The statistical analysis clearly indicates that all three factors, as well as their interactions, affect the accuracy of the robot. However, the relationship between these factors and the accuracy is such that the standard linear regression techniques will not produce models which account for more than approximately 35% of the radial error variability, leaving approxi¬ mately 65% of the radial error variability due to repeatability. When using a robot, the accuracy of the robot at a particular setting of the parameters can be determined by the regression model and adjustments can be made to compensate for the predicted mean radial error. How¬ ever, the portion of the radial error which is due to repeatability must be tolerated without recourse. Manufacturers should therefore concen¬ trate on giving more information about the accuracy of a robot. Since they have extensive test data for each model of robot, the manufacturer could provide a linear model for the purposes of predicting accuracy of the robot as well as an estimate of the constant repeatability. Acknowledgments We thank the administration of Midwestern State University and especially Dr. Norman Horner in providing the funds to perform this research. We further like to thank Mr. Andy Webb for the construction of the table and the testing fixture. We also thank Mrs. Lois Moore and Dr. Michael Shipley for their advice. This paper would not have been possible without the help of all these people and MSU institutional support. Literature Cited ANSI. 1992. American National Standard for Industrial Robots and Robot Systems- Path-Related and Dynamic Performance Characteristics-Evaluation-ANSI/RIA 15.05-2- 1992. American National Standards Institute, New York, 45 pp. DeGarmo, P., J. T. Black & R. Kohser. 1997. Materials and Processes in Manufacturing, 8th ed. Prentice Hall, Upper Saddle River, New Jersey, 1259 pp. 156 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Khouja M. J. & R. L. Kumar. 1999. An options view of robot performance in a dynamic environment. Int. J. Prod Res., 37 (6): 1244-1250. Offodile, F. & K. Ugwu. 1991. Evaluating the effect of Speed and Payload on Robot Repeatability. Robot. Comput.-Integr. Manuf., 8(l):27-28. Rehg, J. 1985. Introduction to Robotics: A Systems Approach. Prentice Hall, Inc., Englewood Cliffs, New Jersey, 230 pp. Warnecke, H. J. & R. D. Schraft. 1982. Industrial Robots, Application Experience. I.F.S. Publications Ltd., Kempston, Bedford, England, 298 pp. JB at: jan.brink@mwsu.edu TEXAS J. SCI. 56(2): 157-170 MAY, 2004 HISTORICAL POPULATION DYNAMICS OF RED SNAPPER (LUTJANUS CAMPECHANUS) IN THE NORTHERN GULF OF MEXICO J. R. Gold and C. P. Burridge Center for Biosystematics and Biodiversity Texas A&M University, College Station, Texas 77843-2258 Abstract.— A total of 313 young-of-the-year red snapper {Lutjanus campechanus ) belonging to the 1999 year class were sampled from three geographic regions in the northern Gulf of Mexico and assayed for haplotype variation in mitochondrial (mt)DNA. Analysis of molecular variance revealed that only a small proportion (0.24%) of the genetic variance was distributed among regions; accordingly, the corresponding dPST value did not differ significantly from zero. Exact tests of homogeneity of haplotype distributions also were non-significant. Tests for departure from a neutral Wright-Fisher model of genetic polymorphism, however, were significant, and a ‘mismatch’ distribution of nucleotide-site differences in mtDNA indicated that the departure from neutrality could be due to population expansion. Estimates of the time since expansion ranged from ==270,000 to =420,000 years before present. The latter is consistent with the hypothesis that red snapper likely colonized the continental shelf in the northern Gulf following a glacial retreat. The observed departure from a neutral Wright-Fisher model also may suggest that insufficient time has lapsed for red snapper in the northern Gulf to attain equilibrium between mutation and genetic drift. However, the temporal signature provided by the ‘mismatch’ distribution is far older than the last glacial retreat which began = 18,000 years ago. If the departure from neutrality reflects events occurring after the last glacial retreat, tests of present-day population or stock structure may well be compromised. The same may be true for other marine fish species in the northern Gulf. Red snapper ( Lutjanus campechanus) is an important, highly exploited marine fish distributed primarily along the continental shelf in the Gulf of Mexico from the Yucatan Peninsula in Mexico to the northeastern Florida coast (Hoese & Moore 1998). Although the species has pro¬ vided an important fishery since the early 1900s, red snapper in U.S. waters have declined by an estimated 90% since the 1970s (Goodyear & Phares 1990). Factors impacting red snapper abundance include overexploitation by directed commercial and recreational fisheries, juvenile mortality associated with bycatch in the shrimp fishery, and habitat change (Gallaway et al. 1999; Ortiz et al. 2000). Management of red snapper resources in U.S. waters is currently based on a unit stock hypothesis (GMFMC 1989). Whether red snapper in fact com¬ prise a single stock across the northern Gulf, however, remains an issue. Separate management of regional stocks, if they exist, would be a de¬ sirable goal to avoid regional over-exploitation and to conserve adaptive genetic variation (Carvalho & Hauser 1995; Hauser & Ward 1998). 158 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Previous genetic work generally has been consistent with the existence of a single stock of red snapper in the northern Gulf (Camper et al. 1993; Gold et al. 2001) and with the hypothesis that significant gene flow occurs at one or more life-history stages (Goodyear 1995; Gold & Richardson 1998a). The hypothesis of significant gene flow is not consistent with a number of tagging studies that have shown adult red snapper to be sedentary and exhibit high site fidelity (Szedlmayer & Shipp 1994; Szedlmayer 1997). However, Patterson et al. (2001) recently documented extensive movement of adult red snapper in the northeastern Gulf and suggested that movement of adults might be sufficient to facilitate mixing across the northern Gulf. A second hypothesis is that observed genetic homogeneity reflects historic rather than contemporary gene flow, and that present-day red snapper could be isolated yet have been in sufficient genetic contact in the past to remain genetically indistinguishable (Camper et al. 1993; Gold & Richardson 1998a). In such situations, populations may not have reached equi¬ librium between mutation and genetic drift, and if so, would be expected to depart from expectations of the neutral Wright- Fisher model of genetic polymorphism (Fu 1997). This study examined the alternate hypothesis by assessing patterns of mitochondrial (mt)DNA variation among red snapper sampled from three geographic regions in the northern Gulf and asking whether mtDNA haplotype distributions deviated from those expected under mutation-drift equilibrium. Populations that are expanding or declining typically are not in mutation-drift equilibrium (Fu 1997), and in such situations may leave a characteristic ‘mismatch’ distribution signature (Rogers & Harpending 1992). Consequently, this study also examined the ‘mismatch’ distribution of nucleotide site differences in mtDNA between pairs of individuals in order to assess whether red snapper in the northern Gulf had expanded or declined demographically. Red snapper were likely precluded from occupying most of the contemporary continental shelf in the northern Gulf during Pleistocene glacial advance (Gold & Richardson 1998a), and colonization of shelf waters following glacial retreat could have generated conditions conducive to population expansion. Materials and Methods Young-of-the-year red snapper were procured in the fall of 1999 during a demersal trawl survey of the northern Gulf carried out by the GOLD & BURRIDGE 159 National Marine Fisheries Service (NMFS). Individual fish were sampled from the catch of a 12 m shrimp-trawl net, frozen onboard and returned to College Station where tissues were removed and stored at -80°C. Specimens were obtained from different offshore localities corresponding to three geographic regions (Fig. 1) representing the northwestern Gulf (south Texas coast, 14 trawls, n = 127, range/trawl = 4-12, mode = 8), the northcentral Gulf (Louisiana coast, 14 trawls, n = 123, range/trawl = 1-20, mode = 10), and the northeastern Gulf (Mississippi- Alabama coast, 9 trawls, n = 63, range/trawl = 1-13, mode = 10). Genomic DNA was isolated from frozen tissues as described in Gold & Richardson (1991). Assay of mtDNA employed single strand conformational polymorphism or SSCP (Orita et al. 1989). Regions within the NADH-4 (ND-4) and NADH-6 (ND-6) protein-coding genes were sequenced and the Lasergene software package Primer Select was used to design polymerase-chain-reaction (PCR) primers that amplified mtDNA fragments less than 250 base pairs (bp) in size. The fragments were 163 bp from ND-4 and 122 bp from ND-6. PCR primers (forward primer first, then reverse primer) were as follows: ND-4 (5’ - CAAAACCTTAATCTTCTACAATGCT - 3’; 5’ - CAGGGGGTCTGTTGCTAT - 3’) and ND-6 (5’ - CGAAGCGTCCCCCGACT - 3’; 5’ - CGGTTGATGAACTAGGTGATTTTTC - 3’). PCR conditions followed those used for red snapper microsatellites (Gold et al. 2001), except that annealing was carried out at 58 °C and both primers for each fragment amplified were radioactively labelled. Following PCR, 5/xL of stop solution (95% formamide, 0.05% bromophenol blue and xylene cyanol, 10 mM NaOH) was added to 10/xL of PCR product. This solution was heat denatured at 100° C for 10 min and then snap-chilled in ice water. Varying gel composition and electrophoresis conditions optimized resolution of electromorphs. Adequate resolution was provided by electrophoresing PCR products at 500 V for 16 h on 8% non-denaturing polyacrylamide gels (37.5:1 acrylamide: bis- acrylamide, 0.5X TBE), supplemented with 5.0% glycerol (4.0% for NADH6) and run in 0.5X TBE buffer. The ND-4 and ND-6 electromorphs were best resolved by electrophoresis at 12 °C. Efficiency of SSCP procedures to identify sequence variants was assessed by sequencing multiple representatives of each electromorph and comparing patterns of sequence divergence among them. Representatives of each electromorph were run on subsequent SSCP gels as reference controls. 160 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Figure 1 . Collection localities of young-of-the-year red snapper ( Lutjanus campechanus) from the northern Gulf of Mexico: northwestern Gulf {n = 127), northcentral Gulf ( n = 123, and northeastern Gulf ( n = 63). MtDNA haplotype (nucleon) and nucleotide diversity were estimated after Nei (1987). The former represents the probability that any two individuals drawn at random will differ in mtDNA haplotype, whereas the latter represents the number of nucleotide differences per site between two randomly chosen sequences. Private haplotypes were tabulated and a V test (DeSalle et al. 1987) was used to test whether the proportion of private haplotypes differed significantly among regional samples. Homogeneity of mtDNA haplotype distributions among regions was assessed via analysis of molecular variance and exact tests (based on a Markov-chain procedure). For Amova, significance of the variance among samples and of Osx was assessed by permutation (10,000 replicates). Both tests of homogeneity were carried out using Arlequin (Schneider et al. 2000). Deviation from mutation-drift equilibrium was assessed via Fu & Li’s (1993) D* and F* and Fu’s (1997) Fs measures of selective neutrality. Tests of significance of Fu and Li’s D* and F* and Fu’s Fs statistics were performed using DNAsp (Rozas et al. 2003) and Arlequin, respectively, and were based on 1,000 (£>* and F*) and 10,000 ( Fs ) randomizations. Mismatch-distribution analysis (Rogers & Harpending GOLD & BURRIDGE 161 1992) was used to assess population expansion. As populations at mutation-drift equilibrium are expected to have ragged mismatch distributions (Rogers & Harpending 1992), the r measure of 4 ragged¬ ness ’ (Harpending 1994) was calculated using Arlequin; tests of r — 0 were carried out by parametric bootstrapping (10,000 replicates), also using Arlequin. Results Twelve electromorphs (A-L) of the 163 bp ND-4 fragment and fourteen electromorphs (A-N) of the 122 bp ND-6 fragment were identified via SSCP. Sequences of all electromorphs may be found in Table 1. All electromorphs of the ND-4 fragment differed by no more than a single nucleotide substitution from the most common electro- morph (designated ‘A’); for the ND-6 fragment, two electromorphs (‘F’ and ‘G’) differed by more than one nucleotide substitution from any other electromorph. Multiple representatives of each electromorph (both fragments) were sequenced but no variation within an electromorph type was detected. A total of 32 composite mtDNA haplotypes were identified (Table 2). Haplotypes AA, BB, and AC were the most common, occurring at frequencies within regions of >0.300 (AA), 0.190 - 0.331 (BB), and 0. 134 - 0.238 (AC). Twenty-one private haplotypes were observed; the number of private haplotypes per regional locality was 8 (Texas), 10 (Louisiana), and 3 (Mississippi/ Alabama). None of the private alleles occurred at a frequency greater than 0.017, and the proportion of private haplotypes did not differ significantly among regions (V[2J = 0.657, P > 0.05). Nucleon and nucleotide diversities among regions were 0.770 (Texas), 0.776 (Louisiana), and 0.798 (Mississippi/ Alabama), and 0.006 (Texas), 0.007 (Louisiana), and 0.006 (Mississippi/ Alabama), respec¬ tively. Analysis of molecular variance revealed that only 0.24% of the molecular variation was distributed among samples rather than within samples; the <4?ST value of 0.002 did not differ significantly ( P = 0.253) from zero. An exact test of homogeneity in mtDNA haplotype distri¬ bution among regions also was non- significant ( P = 0.307). Given the absence of heterogeneity in the distribution of mtDNA haplotypes among samples, all mtDNA haplotypes were pooled into a single sample for all subsequent analysis. 162 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 §•8 SC T) .« O ^ o s e ag U CO M § s ■-I la •S .2 8 8 V3 - a) co &s O - .3 J < .s §■8 i>t>i>i>i>i>i>i>i>i>i>i jJ4->4J4-)4-)4-J4J4-)4J4->4J4-> oooooooooooo I — I r— I I — 1< — I i — |i — I i — I i — I I — ) r — ) i — ) r — I aaaaaaaa.acua.a, fCn5fd(0f0fU(dnJ(Uf0f0 • • • 0 0 • • U • • • 0 . . . U • . . 0 . . . S • • : I : : i U . . . u • • • u • • • u • • • 0 • • • 0 . . • g! : i 0 . . . 0 • • • Si ; i Si i i < • • • Eh . . . 0 . . . g : : : I i i i Eh • • • 0 . . • 0 . . . 0 . • . 0 . . . 0 • • • 0 • • • 0 . . . 0 . . • 0 . . . I i i i 0 • • • g : : : 0 Eh 0 0 < < 0QWfcJ0KHh)^^SS (DCL)CU(L>a)a) -P-P-P-P-P-P-P-P-P-P-P-P-P OOOOOOOOOOOOO «— I 1 — I 1— I 1 — Ii — 1< — 1< — I 1— I 1 — 1| — 1| — 1> — Ii — I aaaaaaaaaaaaa fdftSfC(On3fO(Tjn3(OfdrCfOrO Haplotype 164 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Table 2. Frequencies of mitochondrial (mt)DNA haplotypes from age 0-1 red snapper ( Lutjanus campechanus ) sampled from three regions in the northern Gulf of Mexico. Sample region and number of individuals are northwestern Gulf (TX, n '= 127), northcentral Gulf (LA, n = 123), and northeastern Gulf (MS-AL, n = 63). First letter (A-L) represents sequence eletromorphs at ND-4; second letter (A-N) represents sequence electromorphs at ND-6. Electromorph sequences may be found in Table 1. MtDNA haplotype TX LA MS-AL MtDNA haplotype TX LA MS-AL AA 0.306 0.328 0.333 AJ 0.008 BB 0.331 0.319 0.190 AK 0.008 AC 0.165 0.134 0.238 AL 0.008 AB 0.066 0.017 0.079 AM 0.008 BA 0.008 0.017 0.048 BH 0.008 AD 0.008 0.017 0.032 BN 0.008 AE 0.016 0.017 CC 0.016 CA 0.008 0.017 0.016 FC 0.008 AF 0.008 0.008 GC 0.008 AG 0.008 0.016 GH 0.008 AH 0.008 0.016 HA 0.008 BC 0.016 IA 0.008 BD 0.016 IC 0.008 DA 0.017 JF 0.008 EA 0.017 KA 0.008 AI 0.008 LA 0.016 Fu & Li’s (1993) D* and F* and Fu’s (1997) Fs measures of selective neutrality were negative and significant for the pooled samples (D* — -2.85, P = 0.019; F* = -2.73, P = 0.007; Fs = -22.59, P = 0.000), consistent with demographic growth of a population (Fu 1997). Popula¬ tion growth (expansion) also was indicated by the unimodal mismatch distribution (Fig. 2) and by Harpending’s (1994) raggedness index (r) which was non-significant (r = 0.107, P = 0.070). The time at which demographic expansion in red snapper might have occurred was estimated via the relationship r = 2ut (Rogers & Harpending 1992). The value r is the crest or peak of a unimodal mismatch distribution (measured in units of 1 Hu generations) , u is the mutation rate/generation of the region under study, and t is time in generations. The estimate of r (2.412) was obtained from Arlequin; u was estimated as the product of trijjx , where mT is the number of nucleotides assayed (285) and fx is an estimate of the mutation rate per nucleotide. For estimate(s) of /x, the molecular-clock calibrations for mitochondrial protein-coding genes developed by Bermingham et al. (1997) were used and employed two rates (1.0% /106 yr and 1.5%/106 yr) for the (combined) ND-4 and ND-6 sequences from red snapper. For generation time, 15 and 20 years were used, framing the hypothesized generation time in red snapper of 17-19 years (J. Cowan, Louisiana State University, pers. comm.). Estimates GOLD & BURRIDGE 165 Figure 2. Mismatch distribution observed for mitochondrial DNA sequences (haplotypes) of young-of-the-year red snapper ( Lutjanus campechanus ) from the northern Gulf of Mexico. Bars represent observed frequency of differences between sequences; line represents the expected distribution assuming demographic expansion. of u ranged from 1.5 x 10'7/generation (/x = 1 . 0 % / 1 06 yr, 15 yr/ generation) to 3.0 x 107/generation (/x = 1 . 5 % / 1 06 yr, 20 yr/ generation). Estimates of the time when demographic expansion in red snapper could have occurred ranged from « 200,000 yr (u = 3.0 x 107/generation) to — 540,000 yr (w = 1.5 x 107/generation). Despite uncertainties surrounding appropriateness of the molecular clock calibrations (Martin & Palumbi 1993; Rand 1994), and issues with use of pairwise-difference parameters such as r (Felsenstein 1992), estimates of the time since demographic expansion in red snapper fit well within the Pleistocene epoch. Discussion The observed homogeneity of mtDNA-SSCP haplotype frequencies among sample localities is consistent with the hypothesis that red snapper constitute a single stock in the northern Gulf. Similar findings were reported by Camper et al. (1993) based on restriction-site analysis of whole mtDNA and by Gold et al. (2001) based on analysis of micro¬ satellites. Because genetic homogeneity typically implies sufficient gene flow to offset genetic divergence, continuous movement of red snapper at various life-history stages has been hypothesized (Goodyear 1995; Gold & Richardson 1998a; Patterson et al. 2001). The significant departure of mtDNA variation from expectations of 166 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 the neutral Wright-Fisher model of genetic polymorphism indicates that red snapper in the northern Gulf have not attained equilibrium between mutation and genetic drift. Moreover, the negative values for the ‘neutrality’ indices, particularly Fu’s (1997) Fs index, suggest that the departure from neutrality stems from population growth. However, in addition to population growth, the D* and F* indices of Fu and Li (1993) and the Fs index of Fu (1997) also can signify either background selection or genetic hitchhiking, respectively (Fu 1997). Neither seems plausible in this case, in part because data are from mtDNA which is inherited as a single gene and independently from all nuclear genes, and in part because the mismatch distribution and Harpending’s (1994) raggedness index were consistent with the hypothesis of historical population expansion. In addition, because red snapper were precluded from occupying much of the contemporary continental shelf in the Gulf when sea levels during Pleistocene glaciations were at least 100 m lower than they are today (CLIMAP 1976; Rezak et al. 1985), colonization of shelf waters and opening of favourable habitat following glacial retreat would be expected to generate conditions conducive to population expansion. This scenario is consistent with the estimated time of * 200,000 - 540,000 years ago, given that the Pleistocene Epoch began approximately 1.8 million years ago (http://vulcan.wr.usgs.gov/ Glossary /geotimescale . html) . Camper et al. (1993) and Gold et al. (2001) suggested that the genetic homogeneity observed among present-day red snapper in the northern Gulf might reflect historical rather than current gene flow. Briefly, genetic homogeneity among putatively isolated, present-day populations could be sustained provided there has been both insufficient time since colonization of continental-shelf waters and sufficiently large effective population sizes such that allele frequency differences arising via mutation have not reached mutation-drift equilibrium. However, the time since expansion indicated from the mismatch distribution ( — 200,000 - 450,000 years ago) would seem too long for genetic divergence not to have arisen, assuming there has been no gene flow among localities and that effective population sizes are even one-tenth to one-hundredth of the current estimated census size of 7 - 20 million individuals. Unfortunately, estimating approximately how long it would take for genetic divergence to arise in this situation is problematic, given the absence of estimates of the effective (female) size of red snapper populations in the northern Gulf and the possibly unrealistic assumptions that red snapper form ‘idealized’ populations that exhibit an infinite- GOLD & BURRIDGE 167 island model of population structure. On the other hand, the last glacial retreat and the (re)opening of the continental shelf in the northern Gulf was only within the last 18,000 years (Rezak et al. 1985), a time period that is potentially too short for genetic divergence to occur if effective (female) sizes are only 1-2 orders of magnitude smaller than current census size and particularly if there is periodic gene flow among (semi-) isolated stocks. There are a number of caveats to the above inferences. The first is that immigration of rare, genetically distinct mtDNA haplotypes also could generate negative D*, F*, and Fs values (Skibinski 2000). How¬ ever, such immigration would be expected to lead to multimodal mismatch distributions (Marjoram & Donelley 1994), unlike the unimodal distributions observed here. A second caveat is that declining rather than expanding populations also can produce unimodal mismatch distributions. However, the ‘wave’ of a unimodal distribution of a declining population is expected to have an extremely steep leading edge, often with several secondary peaks that have large values (Rogers & Harpending 1992), a pattern not observed in the mismatch distribution generated from mtDNA sequences. Finally, the tests of neutrality may not necessarily measure the same temporal period as the mismatch distribution. The latter indicated a period of population expansion that occurred between —200,000 and 450,000 years ago, whereas the tests of neutrality could reflect an expansion dating to the last glacial retreat. At present, there is no way to distinguish among these alternatives. Assuming red snapper in the northern Gulf deviate from mutation- drift equilibrium because of demographic expansion following the last glacial retreat, the question arises as to how prevalent are the same genetic patterns and demographic histories in other marine fishes in the northern Gulf. Grant & Bowen (1998) hypothesized that the combina¬ tion of high haplotype diversity and low nucleotide diversity for mtDNA was indicative of a population bottleneck followed by rapid growth (their Category 2), and assigned two species that are common in the northern Gulf (red drum, Sciaenops ocellatus, and greater amberjack, Seriola dumerili) to this category. They erroneously assigned red snapper to Category 1 (low haplotype diversity and low nucleotide diversity) based on an error in reading Table 3 in Camper et al. (1993). Given the range of haplotype (0.770 - 0.798) and nucleotide (0.006 - 0.007) diversity found here, red snapper clearly belong in Category 2. A review of the literature reveals that many other fishes in the northern Gulf also appear 168 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 to belong to Grant and Bowen’s Category 2: Gulf toadfish, Opsanus beta (cf. A vise et al. 1987); Spanish sardine, Sardine lla aurita (cf. Tringali & Wilson 1993); common snook, Centropomus undecimalis (cf. Tringali & Bert 1996), and black drum, Pogonias chromis , spotted seatrout, Cy noscion nebulosus , and king mackerel, Scomberomorus cavalla (synopsized in Gold & Richardson 1998b). Analysis of selective neutrality and of mismatch distributions of mtDNA datasets may demonstrate that these species also have undergone demographic expansions that could be dated approximately to changes in habitat availability during or following Pleistocene glaciation. Consequently, it may be that the (spatial) genetic homogeneity observed for many fishes in the northern Gulf of Mexico owes more to historical than contemporary gene flow, and that stocks meriting independent manage¬ ment may have gone unnoticed. A final important point to note that these current results do not necessarily reflect contemporary trends or contradict the documented decline of present-day red snapper stocks (Goodyear & Phares 1990), as evidence of historic demographic expansion is not necessarily affected by even severe bottlenecks that occur subsequent to population expansion (Rogers 1995; Lavery et al. 1996). Acknowledgments We thank W. Patterson for assistance in procuring samples, T. Dowling for carrying out the V tests, and E. Saillant and T. Turner for constructive comments on the manuscript. Research was supported by a grant (NA87FF0426) from the MARFIN Program of the National Marine Fisheries Service (Department of Commerce) and by the Texas Agricultural Experiment Station under Project H-6703 . Views expressed in the paper are those of the authors and do not necessarily reflect the views of the sponsoring grant agencies. This paper is number 41 in the series ‘Genetic Studies in Marine Fishes’ and Contribution 124 of the Center for Biosystematics and Biodiversity at Texas A&M University. Literature Cited Avise, J. C.,C. A. Reeb & N. C. Saunders. 1987. Geographic population structure and species differences in mitochondrial DNA of mouthbrooding marine catfishes (Ariidae) and demersal spawning toadfishes (Batrachoidae). Evolution, 41:991-1002. Bermingham, E, S. S. McCafferty & A. P. Martin. 1997. Fish biogeography and molecular clocks: perspectives from the Panamanian Isthmus. Pp. 113-128, in Molecular Systematics of Fishes (T. D. Kocher & C. A. Stepien, eds), Academic Press, San Diego, GOLD & BURRIDGE 169 CA., 314 pp. Camper, J. D., R. C. Barber, L. R. Richardson & J. R. Gold. 1993. Mitochondrial DNA variation among red snapper ( Lutjanus campechanus) from the Gulf of Mexico. Mol. Mar. Biol. Biotechnol., 3:154-161. Carvalho, G. R. & L. Hauser. 1995. Molecular genetics and the stock concept in fisheries. Pp. 55-80, in Molecular Genetics in Fisheries (G. R. Carvalho & T. T. Pitcher, eds.), Chapman and Hall, London, 141 pp. Clark, P .U., J. X. Mitrovica, G. A. Milne & M. E. Tamisiea. 2002. Sea-level fingerprinting as a direct test for the source of global meltwater pulse IA. Science, 295:2438-2441. CLIMAP, 1976. The surface of the ice-age earth. Science, 191:1131-1137. DeSalle, R., A. Templeton, I. Mori, S. Pletscher & J. S. Johnston. 1987. Temporal and spatial heterogeneity of mtDNA polymorphisms in natural populations of Drosophila mercatorum. Genetics, 116:215-233. Felsenstein, J. 1992. Estimating effective population size from samples of sequences: inefficiency of pairwise and segregating sites as compared to phylogenetic estimates. Genet. Res., 59:139-147. Fu, Y.X. 1997. Statistical tests of neutrality of mutations against population growth, hitchhiking, and background selection. Genetics, 147:915-925. Fu, Y-X & W.-H. Li. 1993 Statistical tests of neutral mutations. Genetics, 133:693-709. Gallaway, B. J., J. C. Cole, R. Meyer & P. Roscigno. 1999. Delineation of essential habitat for juvenile red snapper in the northwestern Gulf of Mexico. Trans. Am. Fish. Soc., 128:713-726. GMFMC (Gulf of Mexico Fishery Management Council). 1989. Amendment number 1 to the Reef Fish Fishery Management Plan. Gulf of Mexico Fishery Management Council, Tampa, FL. Gold, J. R. & L. R. Richardson. 1991. Genetic studies in marine fishes. IV. An analysis of population structure in the red drum (Sciaenops ocellatus) using mitochondrial DNA. Fish. Res., 12:213-241. Gold, J. R. & L. R. Richardson. 1998a. Genetic homogeneity among geographic samples of snappers and groupers: evidence of continuous gene flow? Proc. Gulf Carib. Res. Inst., 50:709-726. Gold, J. R. & L. R. Richardson. 1998b. Mitochondrial DNA diversification and population structure in fishes from the Gulf of Mexico and western Atlantic. J. Hered. , 89:404-414. Gold, J. R., E. Pak & L. R. Richardson. 2001. Microsatellite variation among red snapper {Lutjanus campechanus) from the Gulf of Mexico. Mar. Biotechnol., 3:293-304. Goodyear, C. P. & P. Phares. 1990. Status of red snapper stocks of the Gulf of Mexico: report for 1990. National Marine Fisheries Service, Southeast Fisheries Centre, Miami Laboratory, CRD 89/90-05, Miami, FL, 72 pp. Goodyear, C. P. 1995. Red snapper stocks in U.S. waters of the Gulf of Mexico. National Marine Fisheries Service, Southeast Fisheries Centre, Miami Laboratory, CRD 95/96-05, Miami, FL, 125 pp. Grant, W. S. & B. W. Bowen. 1998. Shallow population histories in deep evolutionary lineages of marine fishes: insights from sardines and anchovies and lessons for conservation. J. Hered., 89:415-426. Harpending, H., 1994. Signature of ancient population growth in a low-resolution mitochondrial mismatch analysis. Human BioL, 66:591-600. Hauser, L. & R. D. Ward. 1998. Population identification in pelagic fish: the limits of molecular markers. Pp. 191-224, in Advances in Molecular Ecology (G. R. Carvalho, ed.), IOS Press, Amsterdam, 313 pp. Hoese, H. D. & R. H. Moore. 1998. Fishes of the Gulf of Mexico. 2nd Edition. Texas 170 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 A&M University Press, College Station, Texas, 422 pp. Lavery, S., C. Moritz & D. R. Fielder. 1996. Genetic patterns suggest exponential population growth in a declining species. Mol. Bio. Evol., 13:1106-1113. Marjoram, P. & P. Donnelly. 1994. Pairwise comparisons of mitochondrial DNA sequences in subdivided populations and implications for early human evolution. Genetics, 136:673-683. Martin, A. P. & S. R. Palumbi. 1993. Body size, metabolic rate, generation time, and the molecular clock. Proc. Natl. Acad. Sci. USA, 90:4087-4091. Nei, M., 1987. Molecular Evolutionary Genetics. Columbia Univ. Press, New York, NY, 512 pp. Nesbo, C. L., E. K. Rueness, S. A. Iversen, D. W. Skagen & K. S. Jakobsen. 2000. Phylogeography and population history of Atlantic mackerel {Scomber scomber L.): a genealogical approach reveals genetic structuring among the eastern Atlantic stocks. Proc. Royal Soc. Lond., B, 267:281-292. Orita, M., Y. Suzuki, T. Sekiya and K. Hayashi. 1989. Rapid and sensitive detection of point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics, 5:874-879. Ortiz, M., C. M. Legault & N. M. Ehrhardt. 2000. An alternative method for estimating bycatch from the US shrimp trawl fishery in the Gulf of Mexico, 1972-1995. Fish. Bull. US, 98:583-599. Patterson, W. F., J. C. Watterson, R. L. Shipp & J. H. Cowan. 2001. Movement of tagged red snapper in the northern Gulf of Mexico. Trans. Am. Fish. Soc., 130:533-545. Rand, D. M. 1994. Thermal habit, metabolic rate and the evolution of mitochondrial DNA. Trends Ecol. Evol., 9:125-131. Rezak, R., T. J. Bright & D. W. McGrail. 1985. Reefs and banks of the northwestern Gulf of Mexico. John Wiley & Sons, New York, NY, 259 pp. Rogers, A. R. 1995. Genetic evidence for a Pleistocene population expansion. Evolution, 49:608-615. Rogers, A. R. & H. Harpending. 1992. Population growth makes waves in the distribution of pair-wise genetic differences. Mol. Biol. Evol., 9:552-569. Rozas J, J. C. Sanchez-Delbarrio, X. Messeguer & R. Rozas. 2003. DNASP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics, 19:2496-2497. Schneider, S., D. Roessli & L. Excoffier. 2000. ARLEQUIN ver. 2000: a software for population genetic analysis. Genetics & Biometry Lab., Univ. Geneva, Switzerland. Skibinski, D. O. F. 2000. DNA tests of neutral theory: applications in marine genetics. Hydrobiologia, 420:137-152. Szedlmayer, S. T., 1997. Ultrasonic telemetry of red snapper, Lutjanus campechanus, at artificial reef sites in the northeast Gulf of Mexico. Copeia, 1997:846-850. Szedlmayer, S. T. & R. L. Shipp. 1994. Movement and growth of red snapper, Lutjanus campechanus , from an artificial reef area in the northeastern Gulf of Mexico. Bull. Mar. Sci., 55:887-896. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics, 123:585-595. Tringali, M. D. & T. M. Bert. 1996. The genetic stock structure of common snook {Centropomus undecimalis. Can. J. Fish. Aquat. Sci., 53:974-984. Tringali, M. D. & R. R. Wilson. 1993. Differences in haplotype frequencies of mtDNA of the Spanish sardine Sardinella aurita between specimens from the eastern Gulf of Mexico and southern Brazil. Fish. Bull. US, 91:362-370. JRG at: goldfish@tamu.edu TEXAS J. SCI. 56(2), MAY, 2004 171 GENERAL NOTES NOTES ON REPRODUCTION IN THE FALSE CORAL SNAKES, ERYTHROLAMPRUS BIZONA AND ERYTHROLAMPRUS MIMUS (SERPENTES: COLUBRIDAE) FROM COSTA RICA Stephen R. Goldberg Department of Biology, Whittier College Whittier, California 90608 Erythrolamprus bizona ranges from Costa Rica, south to Colombia and northern Venezuela and occurs from 8-1450 m in Costa Rica; Erythrolamprus mimus ranges from Honduras through Panama, western Colombia, Ecuador and northwestern Venezuela and occurs from 1-1200 m in Costa Rica (Savage 2002). Both are uncommon diurnal, secretive snakes that are oviparous (Savage 2002) . The purpose of this note is to provide information on reproduction from a histological examination of gonadal material from museum specimens. A sample of 40 specimens of E. bizona (females n = 25 , mean snout- vent length [SVL] = 702 mm ± 83 SD , range = 545-835 mm; males n = 15, SVL = 614 mm ± 54 SD, range = 535-715 mm) and a sample of 13 specimens of E. mimus (females n = 7 , SVL = 557 mm ± 37 SD, range = 504-615 mm; males n = 6, SVL = 482 mm + 105 SD, range = 288-553 mm) from Costa Rica were examined from the herpetology collection of the Natural History Museum of Los Angeles County, Los Angeles (LACM). Erythrolamprus bizona were collected 1959-1980; E . mimus were collected 1966-1982. Counts were made of enlarged ovarian follicles ( > 12 mm length) or oviductal eggs. The left testis, vas deferens and a portion of the kidney were removed from males and the left ovary was removed from females for histological examination. Tissues were embedded in paraffin and sectioned at 5pm. Slides were stained with Harris’ hematoxylin followed by eosin counter¬ stain. Histological slides were examined to determine the stage of the testicular cycle and for the presence of yolk deposition (secondary vitellogenesis sensu Aldridge 1979). Number of tissues histologically examined by species were: E. bizona testis = 15, vas deferens = 15, kidney = 15, ovary = 12; E . mimus testis = 6, vas deferens = 6, kidney 6, ovary = 4. Follicles in advanced stages of yolk deposition or oviductal eggs were counted, but not histologically examined. An unpaired r-test was used to compare body sizes of male and female E. bizona samples. 172 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Material examined.— The following specimens of Ery thro lamp rus bizona were examined by Costa Rica province: ALAJUELA (LACM 145932, 150656), CART AGO (LACM 145843, 145847, 145954, 147512, 150643, 150644, 150650-150653, 150657-150660, 150703, 150706-150708, 150710), GUANACASTE (LACM 150654), PUNT ARENAS (LACM 145792, 150704), SAN JOSE (LACM 67258, 145549, 145785, 145786, 145791, 145845, 145846, 145851, 145933, 145934, 145977, 147510, 150641, 150655, 150705, 150711). The following specimens of Ery thro lamp rus mimus were examined by Costa Rica province: ALAJUELA (LACM 150714, 150715, 150723, 150725, 150728), HEREDIA (LACM 150716, 150717, 150719), LIMON (LACM 150720), PUNT AREN AS (LACM 150718, 150724), PROVINCE DATA MISSING (LACM 150721, 150722). All testes examined from E. bizona and E. mimus were undergoing spermiogenesis (= sperm formation) with metamorphosing spermatids and sperm present. The following numbers of males were undergoing spermiogenesis: E. bizona February (1), April (1), June (2), July (1), August (1), September (1), October (5), November (1), December (2); E. mimus March (3), October (1), December (1). One E. mimus male (LACM 150728, SVL 288 mm) from February exhibited testicular recru¬ descence with spermatogonia and primary spermatocytes present. The size at which this snake would have undergone spermiogenesis is un¬ known. All vasa deferentia contained sperm and all kidney sexual segments from E. bizona and E. mimus were enlarged and contained secretory granules. Mating usually coincides with enlargement of the kidney sexual segments (Saint Girons 1982). The smallest E. bizona male to undergo spermiogenesis (LACM 150659) measured 535 mm SVL; the smallest E. mimus male to undergo spermiogenesis (LACM 150720) measured 432 mm SVL. It will be necessary to examine additional males to ascertain the minimum sizes at which E. bizona and E . mimus begin sperm formation. Female E . bizona were significantly larger than males ( t = 3.67, df = 38, P < 0.01). Samples of E. mimus were too small to make valid size comparisons between males and females. Females of E. bizona with oviductal eggs or enlarged follicles > 12 mm length were found in January-March and September-November (Table 1). One female from June (LACM 150650, SVL 111 mm) and one from October (LACM 150643, SVL 730 mm) were undergoing moderate yolk deposition and contained follicles 5-6 mm in length. It was not possible to predict the clutch size as other follicles might have undergone yolk deposition. Three females were undergoing early yolk deposition (secondary vitello¬ genesis sensu Aldridge 1979): June (LACM 145785, SVL = 821 mm), TEXAS J. SCI. 56(2), MAY, 2004 173 Table 1. Monthly distribution of stages in the seasonal ovarian cycle of Erythrolamprus bizona from Costa Rica. Values shown are the numbers of females exhibiting each of the five conditions. Month n Inactive Early yolk deposition Moderate yolk deposition* Enlarged follicles > 12 mm length Oviductal eggs January 5 1 0 0 2 2 February 2 0 0 0 1 1 March 2 1 0 0 1 0 May 2 2 0 0 0 0 June 3 1 1 1 0 0 July 1 1 0 0 0 0 September 2 0 1 0 1 0 October 5 2 0 1 2 0 November 1 0 0 0 1 0 December 2 1 1 0 0 0 *follicles 5-6 mm length; one cannot predict final clutch size. September (LACM 150653, SVL = 720 mm), December (LACM 150707, SVL = 645 mm). The smallest reproductively active female E. bizona (LACM 145932) measured 602 mm SVL (Table 2). The minimum size at which female E. bizona commence reproduction remains to be determined. Clutch sizes are listed in Table 2. Mean clutch size for 1 1 E. bizona clutches was 5.5 ± 1.8 SD , range = 3-9. Mean clutch size for 4 E. tnimus clutches was 3.8 ± 0.50 SD, range = 3-4. Body sizes, collection dates and locations are in Table 2. The smallest reproductively active female (oviductal eggs) measured 504 mm SVL (Table 2). The minimum size at which E. mimus females begin reproduction remains to be determined. One female from March (LACM 150718, SVL = 615 mm) and one female from October (LACM 150722, SVL = 575 mm) were not undergoing yolk deposition. One female E. mimus from December (LACM 150714, SVL = 563 mm) was undergoing early yolk deposition (secondary yolk deposition sensu Aldridge 1979). There was no evidence that females of either E. bizona or E. mimus produce more than one clutch per year (i.e., oviductal eggs and yolk deposition in progress in the same female). However, in view of the extended period in which males undergo spermiogenesis and reproduc¬ tively active females were found (Table 2), more than one clutch per year might be possible. Erythrolamprus bizona deposits its eggs in rotten logs or decomposed litter (Hardy & Boos 1995). Amaral (1978) reported the congener Erythrolamprus aesculapii from Brazil produced 6-9 eggs. Marques (1996) reported reproduction occurred throughout the year in E. aesculapii from southeastern Brazil and multiple clutches 174 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 2, 2004 Table 2. Clutch sizes for Erythrolamprus bizona and E. mimus (estimated from counts of enlarged follicles > 12 mm length or oviductal eggs*) from Costa Rica. Date SVL (mm) Clutch size Province LACM # 3 January 827 Erythrolamprus bizona 6 Cartago 150651 24 January 760 5* Puntarenas 145792 26 January 670 3* San Jose 147510 27 January 700 6 Cartago 150708 1 February 835 9* San Jose 145549 16 February 783 8 San Jose 145786 16 March 667 4 San Jose 150641 2 September 650 5 San Jose 145845 12 October 695 5 Cartago 145847 17 October 750 5* Cartago 150706 24 November 602 4 Alajuela 145932 12 February 580 Erythrolamprus mimus 4* Alajuela 150715 8 March 533 4* Alajuela 150723 1 April 504 4* Puntarenas 150724 6 Sept 531 3 Heredia 150719 were recorded from captive snakes. Clutch sizes ranged from one to eight eggs. Additional monthly samples of E. bizona and E. mimus will need to be examined to obtain further information on the reproductive biology of these two species. Acknowledgment I thank D. A. Kizirian (LACM) for permission to examine specimens. Literature Cited Aldridge, R. D. 1979. Female reproductive cycles of the snakes Arizona elegans and Crotalus viridis. Herpetologica, 35(3):256-261 . Amarial, A. do. 1978. Serpentes do Brasil: iconografia colorida. Brazilian snakes: a color iconography. 2nd Edit., Edit. Melhoramentos, Edit.Univ. Sao Paulo, Brasil, 246 pp. Hardy, J. D., Jr., & H. A. E. Boos. 1995. Snakes of the genus Erythrolamprus (Serpentes: Colubridae) from Trinidad and Tobago, West Indies. Bull. Maryland Herpetol. Soc., 3 1 (3) : 158-190. Marques, O. A. V. 1996. Biologia reprodutiva da cobra-coral Erythrolamprus aesculapii Linnaeus (Colubridae), no sudeste do Brasil. Revta. Bras. Zool., 13(3):747-753. Saint Girons, H. 1982. Reproductive cycles of male snakes and their relationships with climate and female reproductive cycles. Herpetologica, 1 8(3) :5- 16. Savage, J. M. 2002. The amphibians and reptiles of Costa Rica: A herpetofauna between two continents, between two seas. Univ. Chicago Press, Chicago, Illinois, 934 pp. SRG at: sgoldberg@whittier.edu TEXAS J. SCI. 56(2), MAY, 2004 175 A NEW DISTRIBUTION RECORD AND NOTES ON THE BIOLOGY OF THE BRITTLE STAR OPHIACTIS SIMPLEX (ECHINODERMATA: OPHIUROIDEA) IN TEXAS Ana Beardsley Christensen Department of Biology, PO Box 10037 Lamar University, Beaumont, Texas 77710 Brittle stars (Echinodermata: Ophiuroidea) are a common component of marine communities and often make up a significant portion of the biomass. Identification, however, can be problematic, particularly in the small fissiparous species. Fissiparity, asexual reproduction in which an individual divides in two and regenerates missing parts, occurs in 34 of the 2,000 species of brittle star (Emson & Wilkie 1980). One of these is Ophiactis simplex, an eastern Pacific species, with distribution from the Channel Islands to Panama and the Galapagos Islands (Neilsen 1932; Lonhart & Tupen 2001). Like other fissiparous brittle stars most specimens have six arms and are asymmetrical, with three long arms and three shorter arms. However, individuals with Eve and seven arms are not uncommon; the author has collected one with nine arms. One of the distinguishing characteristics of this species is the red tube feet. The red color is due to the presence of hemoglobin containing coelomocytes (RBCs) present in the water vascular system (Christensen 1999). In late May 2001, five specimens of O. simplex were collected in a tide trap located on the research pier at the University of Texas Marine Science Institute, in Port Aransas, Texas. The specimens were found on algae caught in the net and were very small (disc diameter < 2 mm) . Later that week approximately 200 specimens were collected from algae and other fouling material scraped from the rocks of the south jetty at Port Aransas. This represents a first report of this species along the Texas coast. Official counts were not made at this time. Voucher specimens were sent to Dr. Gordon Hendler at Museum of Natural History of Los Angeles for positive identification. Several subsequent collections have been made from the south jetty to determine habitat preference and population structure. In January 2002, various species of algae, sponge, hy droid and tunicate colonies were scraped from the south jetty during an extremely low tide. Brittle stars were removed from the substrate, counted, and the volume of the substrate was estimated by measuring displacement 176 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 volume. The brittle stars were sorted by disk diameter (large > 3 mm; medium 2-3 mm; small < 2 mm), regeneration state (recently split [2 or more arms < 2 mm], regenerating [2 or more arms of unequal length] and whole [all arms of equal length]) and redness of tube feet (bright red, medium red and colorless). The redness of the tube feet is a crude measure of the hematocrit (proportion of RBCs to water vascular system fluid). It is noted that individual hematocrit is variable in the Texas population: individuals possessing bright red tube feet have large numbers of RBCs in the water vascular system while others have color¬ less tube feet due to the scarcity of RBCs in the water vascular system. Actual hematocrits were not measured but were inferred from micro¬ scopic examination of several dissected individuals. Collections were made again in June 2002, January and July 2003, primarily from colonies of the tunicate Eudistoma carolinense . The densest aggregations of Ophiactis simplex were found in colonies of the sandy lobed tunicate, Eudistoma carolinense (75 individuals per 100 mL) (Table 1). Other substrates in which O. simplex were found included fire sponge ( Tedania ignis), eroded sponge ( Haliclona loosanoffi) and brown ribbed algae (. Dictyopteris sp.) (Table 1). In January 2002, a total of 537 individuals was collected. Medium size individuals (2-3 mm disc diameter) were dominant (67%) and 58% of the individuals were nearly full or fully regenerated (Table 2). In contrast, the June 2002 collection yielded 414 individuals, 70.8% belonging to the small size class (< 2 mm disc diameter) and 82.6% of the individuals were in some stage of regeneration (Table 2). These animals were not sorted by tube feet color as significant mortality occurred before sorting. In July 2003, 229 individuals were collected, 88.2% belonging to the small size category and 83.4% were in some stage of regeneration. Fission appears to be an important means of reproduction in the small and medium size classes, as most collected were in some stage of regeneration. Only two of the 27 large individuals collected were regenerating. The large size class also appears to be fairly uncommon; the largest individual collected had a disc diameter of 4.8 mm. Sexual reproduction also plays a role in this population of O. simplex. In the June 2002 collection, a large proportion (186 individuals) of the small size class was < 1 mm. The high number of small individuals indicates larval recruitment into the area (Mladenov & Emson 1984). Although TEXAS J. SCI. 56(2), MAY, 2004 177 Table 1. List of substrates and densities from which Ophiactis simplex was collected. The different numbers associated with Eudistoma carolinense represent different colonies of the tunicate. Species Density Tedania ignis 15/100 mL Haliclona loosanoffi 17/100 mL Dictyopteris sp. 8/100 mL Eudistoma 35/100 mL Eudistoma #2 41/100 mL Eudistoma #3 79/100 mL Eudistoma #4 28/100 mL Eudistoma #5 61/100 mL Eudistoma #6 25/100 mL Table 2. Results of sorting the collections on the basis of size (small: disc diameter < 2 mm; medium: disc diameter 2-3 mm; and large: disc diameter > 3 mm); regeneration state (recently split: half disc and 2 or more arms < 2 mm; regenerating: 2 or more arms of unequal length; and whole: all arms of equal length), and color of tube feet (indication of hematocrit). January, 2002 June, 2002 July, 2003 Size Small Medium Large Regeneration state Recently split Regenerating Whole Color of tube feet Bright red Medium red Colorless 157 293 202 359 117 25 21 4 2 34 67 31 191 275 160 312 72 38 199 * 94 196 * 128 142 * 7 * June, 2002, sample not sorted for color of tube feet due to significant mortality before sorting. The red color fades with death. nothing has been reported on the reproductive periodicity of O. simplex , the appearance of so many extremely small individuals in the summer suggests an early spring spawn period. The July 2003 sample also yielded many very small animals but the exact numbers were not quantified. There does not appear to be any relationship between regeneration state and color of tube feet. However, there does appear to be a weak relationship between size and color. There was only one large indi- 178 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Figure 1. Aboral (surface) views of Ophiactis savignyii (left) and Ophiactis simplex (right). vidual (disc diameter 3 mm) with colorless tube feet; all other large individuals possessed either medium or bright red tube feet. The larger individuals may be dependent upon hemoglobin for oxygen transport due to their reduced surface area to volume ratio whereas smaller individuals are likely small enough to obtain sufficient oxygen needed for aerobic metabolism by simple diffusion. Differences in the numbers of RBCs among individuals of the same size class may be due to oxygen availa¬ bility in the microhabitat: those with bright red tube feet may inhabit areas with a lower oxygen tension than those with colorless tube feet. This possibility will be investigated further. It is not known if this population of O. simplex is a recent introduc¬ tion (e.g. , through ballast water or drift algae) or if it has been present, but misidentified. A closely related species, Ophiactis savignyii , appears on collection lists for the area. Both are small and fissiparous, but O. savignyii does not possess hemoglobin. As mentioned earlier, not all specimens of the Texas population possess large amounts of hemoglobin and the red color disappears upon preservation with alcohol or formalin. Even with the small size, the two species are morphologically different. The radial shields (two at the base of each arm) of O. savignyii are very large; the length often exceeds half the disc radius, while those of O. simplex are much smaller (Hendler et al 1995) (Figure 1). The arm spines are also markedly different: 4-5 long thin spines in O. simplex and 5-6 shorter, stubby spines in O. savignyi. Acknowledgments I would like to thank the following: Dr. Gordon Hendler for the morphological identification; Dr. David Hicks for his aid in collections; TEXAS J. SCI. 56(2), MAY, 2004 179 Denise Dean for assistance in counting and sorting brittle stars; and Jay Carroll at Tenera Environmental for collection of California O. simplex for comparisons; Drs. Richard Harrel and Andy Kasner for their comments on the manuscript. Literature Cited Christensen, A. B. 1998. The properties of the hemoglobins of Ophiactis simplex (Echinodermata, Ophiuroidea). Am. Zool., 38:12. Emson, R. H. & I. C. Wilkie. 1980. Fission and autotomy in echinoderms. Oceanogr. Mar. Biol. Ann. Rev., 18: 155-250. Hendler, G., J. E. Miller, D. L. Pawson & P. M. Kier. 1995. Sea stars, sea urchins, and allies: Echinoderms of Florida and the Caribbean. Smithsonian Institution Press, Washington. 390 pp. Lonhart, S. I. & J. W. Tupen. 2001. New range records of 12 marine invertebrates: The role of El Nino. Bull. Southern California Acad. Sci., 100:238-248. Mladenov, P. V. & R. H. Emson. 1984. Divide and broadcast: sexual reproduction in the West Indian brittle star Ophiocomella ophiactoides and its relationship to fissiaprity. Mar. Biol., 81:273-282. Nielsen, E. 1932. Ophiurans from the Gulf of Panama, California, and the Strait of Georgia. Vidensk. Medd. fra Dansk naturh. Foren., 91: 241-346 [pp. 257-60]. ABC at: christenab@hal.lamar.edu 5)« * * FIRST DEFINITIVE RECORD OF MORE THAN TWO NESTING ATTEMPTS BY WILD WHITE- WINGED DOVES IN A SINGLE BREEDING SEASON Cynthia L. Schaefer, Michael F. Small, John T. Baccus and Roy D. Welch* Department of Biology, Texas State University -San Marcos San Marcos, Texas 78666 and *Texas Parks and Wildlife Department, 1601 East Crest Drive Waco, Texas 76705 The historical breeding range and recruitment of white-winged doves ( Zenaida asiatica) in Texas was primarily restricted to a four-county region in the lower Rio Grande Valley (Cottam & Trefethen 1968). Recruitment in peripheral populations in adjacent south Texas counties and the Trans-Pecos region have been considered negligible (Gray 2002). In recent years, white- winged dove nesting chronology data have shown a geographic shift in nesting to include urban areas (Small & Waggerman 2000). This shift in nesting range occurred concurrent with 180 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 a substantial northward range expansion of breeding white-winged doves, colonization of urban areas, and establishment of year-round populations over the last three decades (George et al. 1997; Schwertner et al. 2002). As white- winged doves continue expanding their range and congregat¬ ing in urban habitats, accurate measurement of annual recruitment is fundamental to understanding the ecology of this dynamic species. White-winged doves can nest twice in a single breeding season with speculation by some biologists of a greater number of nesting attempts (Cottam & Trefethen 1968, Alamia 1970, Swanson 1989). However, definitive records of more than two nesting attempts have not been documented prior to our account. Two studies of breeding white- winged doves were conducted using surgically implanted radio transmitters. In 2000, breeding white- winged doves were monitored in Kingsville, Texas and in 2002-2003 in Waco, Texas. All white- winged doves were trapped locally in standard wire funnel traps (Reeves 1968) and implanted with subcutaneous radio transmitters in the field at trap sites (Small et al. 2004). In 2000, 40 doves (24 males, 16 females) were trapped and implanted between 19 May and 9 June. All doves were located to source once/ week until onset of nesting. Nests were then monitored every four days using a mirror on an extendable pole and nest status recorded. In 2002, 39 doves (16 males, 23 females) were trapped and implanted with transmitters in June and in 2003, 40 doves (17 males, 16 females, six unknown) were trapped and implanted in February and March. All doves were monitored as in 2000, for the life of the transmitter, up to but not exceeding 90 days. During 2000, three male white- winged doves participated in three nesting attempts with unmarked females. Each attempt resulted in new nest construction. In each case, two nesting attempts proved successful with 1 failure. Young fledged on the first and second nesting but failed on the third for two nesting pairs. The other fledged young on the first and third attempts with the second failing. During 2002, one white¬ winged dove (sex unknown) made three nesting attempts. Two attempts fledged young, nests 1 and 2, with nest 3 failing. During 2003, one female white-winged dove made four nesting attempts with the first and fourth attempts fledging young. The second attempt resulted in nest TEXAS J. SCI. 56(2), MAY, 2004 181 Table 1. Observations for an individual white-winged dove attempting four successive nestings. Nest Tree Attempt Date Success Height (m) Distance from last nest (m) Species Height (m) Same/ Different 1 04/08/03 2 fledged 2.32 NA Pecan 6.67 NA 2 05/23/03 abandoned 2.90 7.0 Pecan 6.67 same 3 06/11/03 nest failed 8.06 7.0 Live Oak 16.64 different 4 06/18/03 2 fledged 2.33 7.0 Pecan 6.67 different abandonment and the third nest failed. In all multiple nesting attempts, no doves reused a nest. Doves built new nests either in the same tree or a nearby tree = 100 m from the old nest. Because of its uniqueness, additional information for the individual with four nesting attempts is presented (Table 1). Although some anecdotal evidence of > two nesting attempts by white- winged doves exists, radio telemetric methodology allowed us to report the first definitive occurrence of > two nesting attempts. Whether this is a unique occurrence or a fundamental aspect of white- winged dove natural history is unknown. Because of the dynamic range expansion, urbanization, and proportional residency shifts of white- winged doves over the last 30 - 50 years, frequency of > two nesting attempts in historic populations will probably never be known. The availability of anthropogenic food and water resources and habitat associated with urbanization have the potential to extend the breeding season (Hayslette & Hayslette 1999) which could represent a shift in the reproductive strategy for white- winged doves. During 2002, one pair of doves with radio transmitters pair bonded, but both batteries failed after 1 successful nesting. Consequently, the issue of monogamy in wild populations of white- winged doves remains unanswered in this study. Further research is fundamental to understanding the dynamics of multiple nesting, monogamy and an extended breeding season on recruitment. This report was part of a study funded by the Texas Parks and Wildlife Department’s white-winged dove stamp research fund. 182 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 2, 2004 Literature Cited Alamia, L. A. 1970. Renesting activity and breeding biology of the white-winged dove {Zenaida asiatica ) in the lower Rio Grande Valley of Texas. Unpubl. M.S. thesis, Texas A&M University, College Station, Texas, USA, 126 pp. Cottam, C. & J. B. Trefethen. 1968. Whitewings: the life history, status, and management of the white-winged dove. D. Van Nostrand Inc., New York, New York, USA, 348 pp. George, R. R., R. E. Tomlinson, R. W. Engel-Wilson, G. L. Waggerman, & A. G. Spratt. 1994. White-winged dove. Pages 28-50, in T. C. Tacha and C. E. Braun, editors. Migratory shore and upland game bird management in North America, Allen Press, Lawrence, Kansas, USA, 223 pp. Gray, M. G. 2002. Breeding biology of White- winged Doves {Zenaida asiatica ) with subcutaneously implanted transmitters in Kingsville, Texas. Unpubl. M.S. thesis. Southwest Texas State University, San Marcos, Texas. 51 pp. Hayslette, S. E. & B. E. Hayslette. 1999. Late and early season reproduction of urban white-winged doves in southern Texas. Texas Journal of Science, 51 (2): 173-180. Reeves, H. M., A. D. Geis, & F. C. Kniffin. 1968. Mourning dove capture and banding. United States Fish and Wildlife Service, Special Scientific Report 117, Washington, D. C., USA, 63 pp. Schwertner, T. W., H. A. Mathewson, J. A. Roberson, M. Small, & G. L. Waggerman. 2002. White-winged Dove {Zenaida asiatica), in A. Poole & F. Gill, editors. The Birds of North America, No. 710. The Birds of North America, Inc., Philadelphia, Pennsylvania, USA, 28 pp. Small, M. F. & G. L. Waggerman. 1999. Geographic redistribution of breeding white-winged doves in the lower Rio Grande Valley of Texas: 1976-1997. Texas Journal of Science, 51(1): 15-19. Small, M. F., J. T. Baccus & G. L. Waggerman. 2004. Mobile anesthesia unit for implanting radio transmitters in birds in the field. The Southwestern Naturalist, 49(2): 279-282. Swanson, D. A. 1989. Breeding biology of the white- winged dove {Zenaida asiatica) in south Texas. Unpubl. M.S. thesis, Texas A&I University, Kingsville, Texas, USA, 121 pp. SLS at: cyndyschaefer@yahoo.com THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 2, 2004 183 Plan Now for the 108th Annual Meeting of the Texas Academy of Science March 3 - 5, 2005 University of Texas-Pan American Local Host Program Chair Damon Waitt Lady Bird Johnson Wildflower Center 4801 LaCrosse Ave. Austin, Texas 78739 Phone: 512.292.4200 E-mail: dwaitt@wildflower.org Hudson DeYoe Dept, of Biology and Center for Subtropical Studies University of Texas-Pan American 1201 West University Dr. Edinburg, Texas 78541 Phone: 956.381.3538 FAX: 956.381.3657 E-mail: hdeyoe@panam.edu For additional information relative to the Annual Meeting, please access the Academy homepage at: www . texasacademyofscience.org Future Academy Meetings 2006 - Lamar University 2007 - Baylor University THE TEXAS ACADEMY OF SCIENCE www.texasacademyofscience.org Membership Information and Application MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. (Please print or type) Name Last First Middle MailingAddress City State Zip Ph FAX E-mail: Regular Member $30.00 Supporting Member $60.00 Sustaining Member $100.00 Patron Member $150.00 Student-Undergraduate $15.00 Student-Graduate $15.00 Joint $35.00 Emeritus $10.00 Corporate Member $150.00 Affiliate $5.00 (list name of organization) Contribution AMOUNT REMITTED SECTIONAL INTEREST AREAS: A. Biological Science B. Botany C. Chemistry D. Computer Science E. Conservation & Management F. Environmental Science G. Freshwater & Marine Sciences H. Geography I. Geology K. Mathematics L. Physics M. Science Education O. Systematics & Evolutionary Biology P. Terrestrial Ecology Q. Anthropology R. Threatened or Endangered Species Please indicate your Sectional interest(s) below: 1. _ 2. _ 3. _ Send Application Form and Check or Money Order to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 Please photocopy this Application Form THE TEXAS ACADEMY OF SCIENCE, 2004-2005 OFFICERS President : President Elect : Vice-President : Immediate Past President : Executive Secretary. Corresponding Secretary : Managing Editor : Manuscript Editor : Treasurer : Council Representative : DIRECTORS 2002 Sushma Krishnamurthy, Texas A&M International University Raymond D. Mathews, Jr., Texas Water Development Board 200J Hudson R. DeYoe, University of Texas-Pan American Cynthia Contreras, Texas Parks and Wildlife Department 2004 Benjamin A. Pierce, Baylor University Donald L. Koehler, Balcones Canyonlands Preserve Program SECTIONAL CHAIRPERSONS Anthropology : Roy B. Brown, Instituto Nacional de Antropologia y Historia Biological Science : Francis R. Horne, Texas State University Botany : Herbert D. Grover, Harden-Simmons University Chemistry: Mary Kipecki-Fjetland, St. Edward’s University Computer Science : Laura J. Baker, St. Edwards University Conservation and Management: Felipe Chavez-Ramirez, Platte River Whooping Crane Trust Environmental Science: William Thomann, University of the Incarnate Word Freshwater and Marine Science: Sharon Conry, Baylor University Geology and Geography: Carol Thompson, Tarleton State University Mathematics: Hueytzen J. Wu, Texas A&M University-Kingsville Physics: David Bixler, Angelo State University Science Education: Jimmy Hand, Austin, Texas Systematics and Evolutionary Biology: Kathryn Perez, University of Alabama Terrestrial Ecology: Jerry Cook, Sam Houston State University Threatened or Endangered Species: Alice L. Hempel, Texas A&M University-Kingsville COUNSELORS Collegiate Academy: William J. Quinn, St. Edward’s University Junior Academy: Vince Schielack, Texas A&M University Nancy Magnussen, Texas A&M University John A. Ward, Brook Army Medical Center Damon E. Waitt, Lady Bird Johnson Wildflower Center David S. Marsh, Angelo State University John T. Sieben, Texas Lutheran University Fred Stevens, Schreiner University Deborah D. Hettinger, Texas Lutheran University Ned E. Strenth, Angelo State University Robert J. Edwards, University of Texas-Pan American James W. Westgate, Lamar University Sandra S. West, Southwest Texas State University SMITHSONIAN INSTITUTION LIBRARIES 3 9088 01 15 33 THE TEXAS JOURNAL OF SCIENCE PrinTech, P. O. Box 43151 Lubbock, Texas 79409-3151 PERIODICAL POSTAGE PAID AT LUBBOCK TEXAS 79402 RETURN SERVICE REQUESTED 6543 SMITHSONIAN INSTITUTION LIBRARIES NHB25MRC 154 PO BOX 37012 WASHINGTON, DC 20013-7012 -npx tUf THE TEXAS JOURNAL OF SCIENCE SivUTmsq PUBLISHED QUARTERLY BY THE TEXAS ACADEMY OF SCIENCE Volume 56 Number 3 August 2004 GENERAL INFORMATION MEMBERSHIP. — Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. For more information regarding membership, student awards, section chairs and vice-chairs, the annual March meeting and author instructions, please access the Academy’s homepage at: www . texasacademy ofscience . org Dues for regular members are $30.00 annually; supporting members, $60.00; sustaining members, $100.00; patron members, $150.00; associate (student) members, $15.00; family members, $35.00; affiliate members, $5.00; emeritus members, $10.00; corporate members, $250.00 annually. Library subscription rate is $50.00 annually. The Texas Journal ofScience is a quarterly publication of The Texas Academy of Science and is sent to most members and all subscribers. Payment of dues, changes of address and inquiries regarding missing or back issues should be sent to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 E-mail: FStevens@schreiner.edu The Texas Journal of Science (ISSN 0040-4403) is published quarterly at Lubbock, Texas, U.S.A. Periodicals postage paid at San Angelo, Texas and additional mailing offices. POSTMASTER: Send address changes and returned copies to The Texas Journal of Science, Dr. Fred Stevens, CMB 5980, Schreiner University, Kerrville, Texas 78028-5697, U.S.A. The known office of publication for The Texas Journal of Science is the Department of Biology, Angelo State University, San Angelo, Texas 76909; Dr. Ned E. Strenth, Managing Editor. COPYRIGHT POLICY All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, recording or otherwise, without the prior permission of the Managing Editor of the Texas Journal of Science. THE TEXAS JOURNAL OF SCIENCE Volume 56, No. 3 August, 2004 CONTENTS Observations of Bird Communities in Relation to Reservoir Impoundment. By Dean Ransom, Jr. and R. Douglas Slack . 187 Seasonal and Ecological Associations of the Avifauna from Sierra San Antonio-Pena Nevada, Zaragoza, Nuevo Leon, Mexico. By Irene Ruvalcaba- Ortega, Jose I. Gonzalez-Rojas, Armando J. Contreras-Balderas and Alina Olalla-Kerstupp . 197 Mate Guarding in Northern Mockingbirds {Mimus polyglottos). By Rebecca Y. Bodily and Diane L. H. Neudorf . 207 A Late Cretaceous Durophagus Shark, Ptychodus martini Williston, from Texas. By Shawn A. Hamm and Kenshu Shimada . 215 New Records of the Texas Homshell Popenaias popeii (Bivalvia: Unionidae) from Texas and Northern Mexico. By Ned E. Strenth, Robert G. Howells and Alfonso Correa-Sandoval . 223 Paraboloids for Maximum Solar Energy Collection. By Ali R. Amir-Moez . 231 Characteristics of Peripheral Populations of Parthenogenetic Cnemidophorus laredoensis A (Squamata: Teiidae), in Southern Texas. By James M. Walker, James E. Cordes and Mark A. Paulissen . 237 Comparison of Branch Elongation among Four Acacia Species and Texas Ebony in the Lower Rio Grande Valley of Texas. By Melissa R. Eddy and Frank W. Judd . 253 general Notes Systematic and Ecological Notes on Tubificoides heterochaetus (Oligochaeta: Tubificidae) from the Neches River Estuary, Texas. By Richard C. Harrel . 263 Reproduction in the Western Hognose Snake, Heterodon nasicus (Serpentes: Colubridae) from the Southwestern Part of its Range. By Stephen R. Goldberg . 267 Endoparasites of the Sequoyah Slimy Salamander, Plethodon sequoyah (Caudata: Plethodontidae), from McCurtain County, Oklahoma. By Chris T. McAllister and Charles R. Bursey . 273 Annual Meeting Notice for 2005 . 278 Recognition of Member Support . . . . . . . 279 Membership Application . 280 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Managing Editor: Ned E. Strenth, Angelo State University Manuscript Editor: Robert J. Edwards, University of Texas-Pan American Associate Editor for Botany: Janis K. Bush, The University of Texas at San Antonio Associate Editor for Chemistry: John R. Villarreal, The University of Texas-Pan American Associate Editor for Computer Science: Nelson Passos, Midwestern State University Associate Editor for Environmental Science: Thomas LaPoint, University of North Texas Associate Editor for Geology: Ernest L. Lundelius, University of Texas at Austin Associate Editor for Mathematics and Statistics: E. Donice McCune, Stephen F. Austin State University Associate Editor for Physics: Charles W. Myles, Texas Tech University Manuscripts intended for publication in the Journal should be submitted in TRIPLICATE to: Dr. Robert J. Edwards TJS Manuscript Editor Department of Biology University of Texas-Pan American Edinburg, Texas 78541 redwards@panam .edu Scholarly papers reporting original research results 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 Manuscript Editor at the above address. They are also available on the Academy’s homepage at: www . texasacademy ofscience . org AFFILIATED ORGANIZATIONS American Association for the Advancement of Science, Texas Council of Elementary Science Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers Texas Society of Mammalogists TEXAS J. SCI. 56(3): 187-196 AUGUST, 2004 OBSERVATIONS OF BIRD COMMUNITIES IN RELATION TO RESERVOIR IMPOUNDMENT Dean Ransom, Jr. and R. Douglas Slack Texas A&M University Agricultural Experiment Station P. O. Box 1658, Vernon, Texas 76384 and Department of Wildlife and Fisheries Sciences, 210 Nagle Hall Texas A&M University, College Station, Texas 778433-2258 Abstract.— This study describes trends in terrestrial avian communities in response to construction of Aquilla Lake in north-central Texas. Reservoir construction and filling resulted in substantial loss of area in each of four major habitat types. Pre-impoundment surveys began in 1979, with follow up post-impoundment surveys in 1984, 1987 and 1992. Mean bird density, species richness and species diversity were highest among all seasons during the pre-impoundment survey, but declined markedly by the first post-impoundment study. Similarity in bird species composition was greatest among the post-impoundment avian communities. Northern cardinal ( Cardinalis cardinalis ) and Carolina chickadee ( Poecile carolinensis ) were the two most common species encountered in all seasons across study phases. Comparisons with data from two adjacent North American Breeding Bird Survey routes suggest that declines among six species may have been related to reservoir construction. Over time, post-impoundment bird communities on Aquilla Lake had fewer bird numbers, had fewer bird species, and were more similar to one another in species composition. Riparian habitats are productive, diverse and structurally complex habitats that support large aggregations of breeding and riparian dependent bird species (Carothers & Johnson 1975). These habitats also provide critical resources to more vertebrate species than any other habitat type, yet less than 2% of the United States (US) land area is comprised of this habitat type (Sedgwick & Knopf 1987; Douglas et al. 1992; Naiman et al. 1993). Further, > 89% of riparian habitat in the US has been lost over the last 200 years, primarily due to logging, agricultural practices and development (Douglas et al. 1992; Croonquist & Brooks 1993). The damming of stream and river systems for reser¬ voir construction has also resulted in substantial loss of riparian habitats. Reservoirs are created for a variety of uses that include flood control, recreation and municipal water supply. As human populations continue to grow, the demand for water resources will continue to increase with greater emphasis on reservoir construction to supply that need. In Texas, for example, there are currently 440 reservoirs with greater than 400 ha of conservation storage capacity; 211 of these have greater than 2,000 ha of conservation storage capacity (Texas Water Development 188 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Board 2002). Construction of an additional eight major and 10 minor reservoirs has been recommended to meet the future water needs of a growing Texas population beyond 2002, as mandated by the state water plan (Texas Water Development Board 2002). Also, 33 sites uniquely suited for reservoir development have been identified for future development by water board planning groups. Reservoir construction can have negative impacts on habitat for terrestrial wildlife species. Impoundment of natural watercourses results in direct loss of species- rich riparian habitats, and fragmentation of remaining forest patches. The proposed construction of 44 reservoirs in Texas during the early 1990s, for example, would have directly impacted an estimated 344,399 ha of wildlife habitat (Frye & Curtis 1990). Habitat loss and fragmentation effects on terrestrial bird communities have been well studied in numerous environments (Ambuel & Temple 1983; Terbourgh 1989; Hill & Hagen 1991; James et al. 1992; Sauer & Droege 1992; Andren 1994; Herkert 1994; Winter & Faaborg 1999; Coppedge et al. 2001). The impacts on terrestrial avian communities resulting from construction and subsequent filling of reservoirs have largely been ignored by avian ecologists. This is surprising in light of the many reservoirs that exist throughout the southern US, and Texas in particular. This study describes the changes in terrestrial avian com¬ munities in context to reservoir construction in north-central Texas over a 14 year time frame. Methods Study area. — The project study site was located in Hill County, approximately 1 1 .2 km southwest of Hillsboro, Texas. The project area was defined as all lands purchased in fee and/or easement necessary for reservoir construction, as well as all lands within the flood pool eleva¬ tion of 169.5 m. The 4,133.2 ha study site was located within the Black-land Prairie and eastern Cross Timbers and Prairies vegetation zones (Gould 1975; Slack et al. 1996). The Blackland Prairie region has alkaline black clay soils with high organic content overlying parent Cretaceous limestone. Prior to agricultural conversion, the dominant herbaceous vegetation was little bluestem (Schizachrium scoparium ); currently it is confined to small scattered areas in the eastern part of the county. The Eastern Cross Timbers consists of a belt of post oak ( Quercus stellata ) and blackjack oak ( Q> . marilandica) extending from the Red River into southern Hill County. The terrain of the study site RANSOM & SLACK 189 was nearly level to rolling, and was dissected by Aquilla, Little Aquilla and Hackberry Creeks. Impoundment of Aquilla Lake by the U.S. Army Corps of Engineers (USACOE) began on 29 April 1983 and reached conservation pool level (163.9 m) two years later on 21 March 1985. The dam site was located in Hill County (97°13’24"W, 31°054’44"N) on Aquilla Creek at river mile 23.6 (km 38). Habitat mapping and bird surveys. — Major habitat types within the project boundaries were mapped and their post- impoundment areal changes quantified from color aerial photographs using ARCINFO Geographic Information System beginning with the pre- impoundment phase I (1979), and each post- impoundment phase: II (1984), III (1987), IV (1991). The avian community was surveyed using three 40 m wide belt transects established prior to lake construction; transects were placed in a manner that would sample the major habitat types in proximity to the projected reservoir basin. Each transect differed in length and sampled habitat types to varying degrees. Transect one was initially 3.7 km long, 37%, 53% and 9.9% of which was represented by forest parkland, old field and riparian woodland habitat types, respectively. Transect one was reduced in length by rising water levels to 2.8 km and 2.5 km in 1987 and 1984, respectively. Transect two was 2.8 km long, 97% of which was in the old field habitat type. Transect three was 1.7 km long and was comprised of 38% forest parkland, 16% riparian woodland and 46% old field habitat. Lengths of transect two and three were unaffected by the filling of the reservoir. Initially, a transect was established in riparian woodland habitat off the reservoir acquisition site as a control to evaluate reservoir impacts. In the winter of 1984, this site was cleared and converted to tame pasture, negating its use as a true control; results from this transect are not reported in this study. To establish some context for interpreting reservoir effects, data from two North American Breeding Bird Survey (BBS) routes located near Aquilla Lake over the same time period (Sauer et al. 2001) was compared. Abundance data for the 11 most abundant species encountered during June surveys on Aquilla Lake were obtained from the Osage BBS route (TEX-050, 97°33’27MW, 31°22’23nN) and the Pidcoke BBS route (TEX-051, 97°52,29,,W, 31°20’29"N), pooled (n = 27) and regressed against time (1979-1992). The hypothesis that the slope of the regression line (fy) for each species did not differ using 95 % confidence intervals (Johnson 1999) was tested. If a particular bird species declined on the study site post- impoundment and also was 190 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Table 1. Area (ha) and percent change (%) over time of habitat types on Aquilla Lake reservoir site. Years correspond to pre-impoundment (1979-80), and post-impoundment I (1984-85), II (1987-88) and III (1991-92) surveys. Habitat Type 1979-80 1984-85 % 1987-88 % 1991-92 % Forest Parkland 428.8 331.8 -22.6 235.3 -29.0 235.3 0 Scrub/Shrub 484.6 29.3 -93.9 25.2 -14.0 25.2 0 Riparian Woodland 1633.8 614.3 -62.4 57.2 -90.7 57.2 0 Old-field 1 735.5 660.4 -10.2 435.4 -34.0 435.4 0 1 Includes area of crop, pasture and old-field habitats pre-impoundment. declining on BBS routes for the same time frame, this would suggest that reservoir construction had little or no effect on the changing numbers for that species. Transects were walked once per quarter during the first three hours of daylight. All birds seen within 20 m on either side of the transect line were identified and recorded. Each of the post- impoundment studies employed a different observer in conducting transect counts. Bird density was calculated seasonally on each transect by dividing the number of birds seen by the area covered (transect length x 40 m); transect density estimates were averaged to obtain a mean bird density (birds/ha ± SE) across the study area. Species richness (r), Simpson’s D, Shannon’s diversity (H’) and Morisita’s index of similarity (Krebs 1989) were computed seasonally for pre-impoundment and post¬ impoundment surveys to compare seasonal bird communities across all phases of this study. Results Four major habitat types were classified from pre- impoundment aerial photographs: forest parkland, riparian woodland, scrub/shrub and old field. All four habitats types were reduced in area due to reservoir construction (Table 1). Riparian woodland was the largest habitat type prior to impoundment and experienced the most rapid rate of loss over the course of the study (Table 1). Mean bird density and species richness was higher in the pre¬ impoundment phase across seasons than in all post-impoundment phases (Table 2); pre- impoundment bird densities were highest during fall and summer. Bird density then declined in all seasons (< 10 birds/ha) between the pre- impoundment and the first post- impoundment phase RANSOM & SLACK 191 Table 2. Species richness (r), Simpson’s D (S[D]) and Shannon-Weiner H’ (SW[H’]) diversity values, and mean density (D, birds/ha) and standard error (SE) for land-bird communities by year and season on Aquilla Lake. Years correspond to pre-impoundment (1979-80), and post-impoundment I (1984-85), II (1987-88) and III (1991-92) surveys. Season Year r S(D) SW(H’) D D(SE) Winter 1979-80 44 0.931 3.003 20.2 (6.1) 1984-85 25 0.914 2.731 8.7 (2.4) 1987-88 13 0.789 1.913 3.5 (0.9) 1991-92 17 0.870 2.269 5.3 (1.7) Spring 1979-80 45 0.918 2.987 24.7 (5.4) 1984-85 17 0.858 2.252 6.3 (0.8) 1987-88 15 0.891 2.309 2.4 (0.8) 1991-92 17 0.915 2.504 2.8 (1.0) Summer 1979-80 64 0.907 2.978 59.6 (9.8) 1984-85 12 0.872 2.187 2.2 (1.1) 1987-88 14 0.766 1.864 2.8 (0.5) 1991-92 18 0.840 2.226 2.9 (0.7) Fall 1979-80 84 0.907 3.037 77.4 (19.6) 1984-85 14 0.851 2.137 2.8 (1.3) 1987-88 11 0.783 1.865 2.6 (0.8) 1991-92 11 0.868 2.113 2.1 (0.8) (Table 2). Species richness values also declined >50% across seasons between the pre- impoundment and first post-impoundment sampling periods (Table 2); pre- impoundment richness values were highest during winter and summer. Morisita’s index of similarity revealed a reduction in similarity between the pre- impoundment survey and all post- impoundment surveys during the fall and winter seasons (Table 3). Collectively, post¬ impoundment bird communities were most similar to the pre-impound¬ ment values during the summer (Table 3). In all seasons but winter, there was greater similarity among post-impoundment surveys than between pre- impoundment and post- impoundment comparisons (Table 3). The similarity between pre- impoundment and post- impoundment bird communities exceeded 50% only in the summer survey periods. Forty-eight percent (n = 19), 51% (n = 23), 54% ( n — 23) and 79% ( n = 66) of the birds recorded during the pre- impoundment surveys during winter, spring, summer and fall, respectively, were never recorded in the subsequent post-impoundment surveys. The two most abundant species encountered in all seasons and surveys were northern cardinal and Carolina chickadees; American robins and eastern meadow¬ larks were most abundant during the winter and spring surveys. 192 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Table 3. Morisita’s community similarity values of seasonal pre-impoundment and post-impoundment land bird communities over time on Aquilla Lake, Hill County, Texas. Years correspond to pre-impoundment (1979-80), post-impoundment I (1984-85), II (1987-88) and III (1991-92) surveys. Season Year 1984-85 1987-88 1991-92 Winter 1979-80 0.685 0.485 0.404 1984-85 0.631 0.255 1987-88 0.267 Spring 1979-80 0.332 0.385 0.339 1984-85 0.667 0.648 1987-88 0.674 Summer 1979-80 0.448 0.571 0.607 1984-85 0.829 0.858 1987-88 0.967 Fall 1979-80 0.310 0.269 0.224 1984-85 0.514 0.505 1987-88 0.843 Twenty-four species of neotropical migrants were observed during the summer pre- impoundment phase of the study. Yellow-billed cuckoos (Coccyzus americanus ) and dickcissels (, Spiza americana ) were the most abundant neotropical migrants in all four phases of summer surveys, and both exhibited the most marked decline in post- impoundment surveys. The 1 1 most abundant birds seen during summer surveys on Aquilla Lake included northern bob white ( Colinus virginianus) , northern cardinal, Carolina chickadee, yellow-billed cuckoo, dickcissel, killdeer ( Charadrius vociferus), lark sparrow ( Chondestes grammacus ), eastern meadowlark, northern mockingbird (Mimus polyglottos), mourning dove and painted bunting ( Passerina ciris). BBS data from the Osage and Pidcoke route were pooled for each of these species to achieve better representation of the area around Aquilla Lake. Confidence interval tests of =0 for northern cardinal, Carolina chickadee, mourning dove, painted bunting, yellow-billed cuckoo and dickcissel showed no significant decline for the time frame of this study (P > 0.05, n = 27, df = 25) . Negative trends in abundance were found for eastern meadow larks, northern bobwhite, killdeer, lark sparrows and northern mocking¬ birds (P < 0.05, n = 27, df =25). Discussion The decline of terrestrial birds in the Aquilla Lake area over the course of this study was apparent in density, species diversity and species richness values. The greatest reduction in bird abundance RANSOM & SLACK 193 occurred between pre- impoundment and the first post- impoundment phase. Bird densities leveled off after Aquilla Lake reached conserva¬ tion pool level in 1985. The decline in bird density was mirrored by declines in species richness and species diversity. Results from this study showed post- impoundment bird communities on Aquilla Lake had fewer bird numbers, had lower species diversity and richness, and were more similar to one another in species composition when compared to the pre- impoundment surveys. Analysis of BBS route data suggest that there were changes among bird species at Aquilla Lake that were not occurring in the surrounding region. Northern cardinals and Carolina chickadees were the two most abundant residents during the pre- impoundment survey of 1980, and both declined to 5 and 16% of their pre- impoundment abundance, respectively, by 1984; this was somewhat surprising given that these two species were not habitat specialists or forest interior obligates. Indeed, the combined BBS data showed no significant trend in the abundance of these two species in the surrounding region for the time period of our study. Likewise, mourning doves, painted buntings, yellow-billed cuckoos and dickcissels showed no trend on BBS routes, but all declined on the Aquilla Lake study site. The change in bird density coincided with the loss of habitat area on the Aquilla Lake site. This apparent cause and effect relationship has been documented by numerous studies of habitat fragmentation effects on bird communities (Forman et al. 1976; Galli et al. 1976; Whitcomb et al. 1977; Robbins 1980; Terbourgh 1989). Loss of habitat area alone, however, has not always explained downward trends in songbird populations (Ambuel & Temple 1983). Sauer & Droege (1992) reported that over the long term, more species of neotropical migrants were increasing than were decreasing with no association between short term declines and changes in forest acreage. James et al. (1992) also reported results that were not consistent with the view that neotropical migrant warblers occupying forest habitats were declining. Hill & Hagen (1991) analyzed population trends of North American birds and found that many species were declining, but that declines in the past 20 years might be in part a result of normal short-term population fluctuations. Plant succession could also account for some of the change in abundance among species at Aquilla Lake, especially in the old-field and scrub/shrub habitat types. There was no active habitat management 194 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 (e.g., prescribed fire) on the USACOE property surrounding Aquilla Lake. Over the course of this study, old-field and scrub/shrub habitats likely changed in floristics and structure with subsequent effects on the avian community. This might explain some of the declines seen in eastern meadowlarks, northern bobwhites and lark sparrows. The ability to detect reservoir impacts was hampered by several factors. First, land use impacts on a control area off the reservoir site precluded direct evaluation of reservoir effects. A true control site would have been difficult to obtain for the length of the study period, since most of the surrounding property was privately owned and sub¬ jected to various agricultural land use practices; such practices did affect the initial control site early in this study. Second, direct cause and effect could not be made due to methodological differences between transect counts and BBS counts. The reason for using BBS data was to provide some context to the data, because published data from other reservoir construction projects does not exist. To that end, the use of BBS data provided tangential support of the results of this study regarding the impacts on bird communities from reservoir construction: some bird species declined on the reservoir site during the study period, but showed no such trend in the surrounding area. Given that reservoir development will continue in order to provide for a growing Texas population, further research on existing and future reservoir sites would seem warranted. Existing reservoirs could provide opportunities to investigate long term effects of habitat loss and fragmen¬ tation on abundance, richness, diversity and the degree of species recovery over time; such data would be especially valuable where they exist in proximity to established BBS routes. New reservoir construction projects could offer opportunities to further quantify the immediate post construction impacts on richness, diversity and abundance of avian communities. Acknowledgments We acknowledge the help of T. Harris-Haller, M. Hoy and J. Hinson with data collection during the pre- impoundment, and the first and second post- impoundment studies, respectively. M. Brown developed the Aquilla Lake GIS data. This work was funded by USACOE, Fort Worth District. Additional support was provided by the Department of Wildlife and Fisheries Sciences at Texas A&M University. RANSOM & SLACK 195 Literature Cited Ambuel, B. & S. A. Temple. 1983. Area dependent changes in the bird communities and vegetation of southern Wisconsin forests. Ecol., 64:1057-1068. Andren, H. 1994. Effects of habitat fragmentation on birds and mammals in landscapes with different proportion of suitable habitat: a review. Oikos, 71:355-366. Carothers, S. W. & R. R. Johnson. 1975. Water management practices and their effects on non-game birds in range habitats, pages 210-222, in D. R. Smith, (tech, coord.). Proceedings of the symposium on management of forest and range habitats for nongame birds. USFS Gen. Tech. Rep. WO-1, 343 pp. Coppedge, B. R., D. M. Engle, R. E. Masters & M. S. Gregory. 2001. Avian response to landscape change in fragmented southern great plains grasslands. Ecol. Appl., 11:47-59. Croonquist, M. J. & R. P. Brooks. 1993. Effects of habitat disturbance on bird communities in riparian corridors. J. Soil and Water Cons., 48:65-70. Douglas, D. C., J. T. Ratti, R. A. Black & J. R. Alldredge. 1992. Avian habitat associations in riparian zones of Idaho’s Centennial Mountains. Wilson Bull., 104:485-500. Frye, R. G. & D. A. Curtis. 1990. Texas water and wildlife: an assessment of direct impacts to wildlife habitat from future water development projects. Spec. Admin. Rep., Fed. Aid in Wildl. Rest. Proj. W-107-R, 59 pp. Forman, R. T., A. E. Galli & C. F. Leek. 1976. Forest size and avian diversity in New Jersey woodlots with some land use implications. Oecol., 16:1-8. Galli, A. E., C. F. Leek & R. T. Forman. 1976. Avian distribution patterns in forest islands of different sizes in central New Jersey. Auk., 93:356-65. Gould, F. W. 1975. Texas Plants-a checklist and ecological summary. Texas A&M University, Agr. Exp. Stat., College Station. 112pp. Herkert, J. R. 1994. The effects of habitat fragmentation on mid-western grassland bird communities. Ecol. Appl., 4:461-471. Hill, N. P. & J. M. Hagan, III. 1991. Population trends of some northeastern North American landbirds: a half century of data. Wilson Bull., 103:165-182. James, F. C., D. A. Wiedenfeld & C. E. McColloch. 1992. Trends in breeding population of warblers: declines in the southern highlands and increases in the lowlands. Pp. 43-56, in Hagen III, J. M. and D. W. Johnston, (eds.), Ecology and Conservation of neotropical migrant landbirds). Smith. Inst. Press, Washington, DC, 623 pp. Johnson, D. H. 1999. The insignificance of statistical significance testing. J. Wildl. Manage., 63:763-772. Krebs, C. J. 1989. Ecological Methodology. Harper Collins Publishers, Inc. New York, NY, 654 pp. Naiman, R. J., H. Decamps & M. Pollock. 1993. The role of riparian corridors in maintaining regional diversity. Ecol. Appl., 3:209-211. Robbins, C. S. 1980. Effect of forest fragmentation on bird populations in the Piedmont of the mid-Atlantic region. Amer. Nat., 33:231-36. Sauer, R. J. & S. Droege. 1992. Geographic patterns in population trends in neotropical migrants in North America. Pp. 26-42, in J. M. Hagen III and D. W. Johnston, (eds.), Ecology and Conservation of neotropical migrant landbirds. Smith. Inst. Press, Washington, DC. 623 pp. Sauer, J. R., J. E. Hines & J. Fallon. 2001. The North American Breeding Bird Survey, results and analysis: 1966-2000. Version 2001.2. USGS Patuxent Wildl. Res. Center, Laurel, Maryland, http://www.mbr-pwrc.gs.gov/bbs/bbs2001.html. 196 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Slack, R. D., B. R. Murphy, M. L. Brown, D. Ransom, Jr. & M. F. Cichra. 1996. Aquilla Lake Post-Impoundment Environmental Study (Year 10). Final Rep. to U.S. Army Corps of Engin., Fort Worth Dist. 212 pp. Sedgwick, J. A. & F. L. Knopf. 1987. Breeding bird response to cattle grazing of a cottonwood bottomland. J. Wildl. Manage., 51:230-237. Terbourgh, J. 1989. Where have all the birds gone? Princeton University Press, Princeton, NJ, 207 pp. Texas Water Development Board. 2002. Water for Texans. Capital Station, Austin, TX. Document No. GP-7-1. Whitcomb, B. L., R. F. Whitcomb & D. Bystrak. 1977. Island biogeography and “habitat islands” of eastern Forests. III. Long-term turnover and effects of selective logging on the avifauna of forest fragments. Amer. Birds 31:17-23. Winter, M. & J. Faaborg. 1999. Patterns of area sensitivity in grassland-nesting birds. Cons. Biol., 13:1424-1436. DR at: rdransom@ag.tamu.edu TEXAS J. SCI. 56(3): 197-206 AUGUST, 2004 SEASONAL AND ECOLOGICAL ASSOCIATIONS OF THE AVIFAUNA FROM SIERRA SAN ANTONIO-PENA NEVADA, ZARAGOZA, NUEVO LEON, MEXICO. Irene Ruvalcaba-Ortega, Jose I. Gonzalez-Rojas, Armando J. Contreras-Balderas and Alina Olalla-Kerstupp Laboratorio de Ornitologia, Facultad de Ciencias Bioldgicas Universidad Autonoma de Nuevo Leon Apart ado Postal 25-F, Cd. Universitaria San Nicolas de los Garza, Nuevo Leon, Mexico Abstract. — This study examined the avifauna of three vegetational communities of the Sierra San Antonio-Pena Nevada of northeastern Mexico, from June 2001 to May 2002. A total of 1,084 individuals were recorded, comprising 83 species, 62 genera, 31 families and 9 orders. The ecological associations of the species were as follows: Pine 40; Pine-Oak 48; and Oak 58. The seasonal distribution of the species was: Spring 48; Summer 42; Fall 47; and Winter 40. Based on the Shannon’s Diversity Index, the highest values were obtained for Oak Forest (H’=3.16) and for Spring (H’=3.26). Resumen.— El presente estudio se realizo sobre la avifauna de tres comunidades vegetales de la Sierra San Antonio-Pena Nevada, de junio de 2001 a mayo de 2002. Se registraron 1, 084 individuos, correspondientes a 83 especies, 62 generos, 31 familias y 9 ordenes. La distribucion ecologica de las especies fue la siguiente: Bosque de Pino, 40; Bosque Mixto, 48; y Bosque de Encino, 58. En cuanto a la distribucion estacional, se obtuvo: Primavera, 48 especies; Verano, 42; Otono, 47; e Inviemo, 40. Utilizando el Indice de Diversidad de Shannon se obtuvieron los valores mas altos para el Bosque de Encino (HP =3. 16) y para Primavera (HP =3.26). Many avian studies in coniferous forests of North America have concluded that vegetation coverage or foliage is a factor that positively influences bird species presence, richness and abundance (Tatschl 1967; Baida 1969; Dickson & Segelquist 1979; Beedy 1981; Anderson et al. 1983; Bazakas 1996; Guzman- Velasco 1998; Garcia et al. 1998; Daniel & Flete 1999; Mills et al. 2000; Doherty & Grubb 2000; Latta et al. 2003). Also, in the South American Andes, the distribution of some species is apparently determined by the vegetation type (Terborgh 1971). Avian communities are not static but change seasonally; in fact, bird assemblages in temperate regions are composed by permanent residents and winter and summer visitors that vary throughout the year (e.g., Hilden 1965; Anderson 1972). Several studies in North American forests have found differences in species richness, density and composi¬ tion in different seasons and habitat types (Anderson et al. 1983; Avery & van Riper III 1989; Corcuera & Butterfield 1999; Latta et al. 2003). 198 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 In Mexico, several researchers have established the ecological distribution of the avian communities of some elevated orographic formations, especially with respect to the effects of altitudinal gradients on bird species (Miller 1955; Morales-Perez & Navarro-Siguenza 1991; Navarro 1992; Winker 1992; Garcia et al. 1998). Previous studies on birds diversity in Nuevo Leon are mainly inventories and generally the locality is not mentioned (Friedmann et al. 1950; Miller et al. 1957; Martin-del Campo 1959; Contreras-Balderas et al. 1995; Contreras- Balderas 1997; Howell & Webb 1995). Ecological aspects of the avifauna in the state are almost non-existent, however, Guzman- Velasco’s (1998) study on Cerro El Potosi and Gonzalez-Iglesias’ (2001) research on Sierra Picachos are an exception. This present effort is the first systematic study of the avian community of Sierra San Antonio- Pena Nevada in terms of species richness, abundance, and ecological and seasonal distribution. Study Site The study area (23°52’12" to 23°40’12" N and 99°57,00M to 99° 39’ 36" W) is located in the southeastern region of General Zaragoza municipality of Nuevo Leon. Its total area is approximately 209.57 km2 and its elevation ranges from 2,200 - 3,400 m (INEGI 1986; Arriaga et al. 2000). This mountainous area is also the second highest elevation of Nuevo Leon, exhibiting diverse vegetational communities that vary from chap¬ arral ( Quercus , Dasilyrion , Agave) to fir forests {Abies- Pseudotsuga) , including those specific to this study: Pine Forest ( Pinus ), Pine-Oak Forest {Pinus- Quercus) and Oak Forest {Quercus). It is situated in the Sierra Madre Oriental, but especially in the transition zone between the Neotropical and Neartic biogeographic regions, making this a natural ecotone. The Sierra Pena Nevada is also considered as a Prioritary Terrestrial Region for Conservation (Arriaga et al. 2000) and an Area of Importance for Birds Conservation in Mexico (Arizmendi & Marquez 2000). Materials and Methods The study site was visited monthly from April 1996 to May 2001. Each vegetation type was sampled once each season, using 18 point counts and 18 mist nets (distributed in 9 stations). Point counts followed RUVALCABA-ORTEGA ET AL. 199 Table 1. List of species and their residency status: PR = Permanent resident; SR = Summer Resident; WR = Winter Resident; T=Transient; V = Vagrant; and * = Undetermined. Species Common Name (Spanish) Common Name Residency (English) Coragyps atratus Zopilote comun Black Vulture PR Cathartes aura Zopilote aura Turkey Vulture PR Buteo brachyurus Aguililla cola corta Short-tailed Hawk PR Buteo albonotatus Aguililla aura Zone-tailed Hawk PR Buteo jamaicensis Aguililla cola roja Red-tailed Hawk PR Patagioenas fasciata Paloma de collar Band-tailed Pigeon PR Zenaida macroura Paloma huilota Mourning Dove PR Otus flammeolus Tecolote ojo oscuro Flammulated Owl SR Megascops asio Tecolote oriental Eastern Screech-Owl PR Megascops trichopsis Tecolote ritmico Whiskered Screech-Owl PR Glaucidium gnoma Tecolote serrano Northern Pygmy-Owl PR Micrathene whitneyi Tecolote enano Elf Owl T Caprimulgus vociferus Tapacamino cuerporrin-norteno Whip-poor-will PR Hylocharis leucotis Zafiro oreja blanca White-eared Hummingbird PR Lampormis clemenciae Colibri garganta azul Blue-throated Hummingbird PR Eugenes julgens Colibrf magmfico Magnificent Hummingbird PR Selasphorus platycercus Zumbador cola ancha Broad-tailed Hummingbird SR Trogon mexicanus Trogon mexicano Mountain Trogon PR Melanerpes formicivorus Carpintero bellotero Acorn Woodpecker PR Picoides villosus Carpintero velloso-mayor Hairy Woodpecker PR Colaptes auratus Carpintero de pechera Northern Flicker PR Lepidocolaptes sp. Trepatroncos Woodcreeper * Contopus sp. Pibi Wood-Pewee * Empidonax flaviventris Mosquero vientre amarillo Yellow-bellied Flycatcher T Empidonax hammondii Mosquero de Hammond Hammond’s Flycatcher WR Empidonax wrightii Mosquero gris Gray Flycatcher WR Empidonax occidentalis Mosquero barranqueno Cordilleran Flycatcher PR Empidonax sp. Mosquero Flycatcher * Tyrannus vociferans Tirano griton Cassin’s Kingbird PR Vireo solitarius Vireo anteojillo Blue-headed Vireo WR Vireo huttoni Vireo reyezuelo Hutton’s Vireo PR Aphelocoma ultramarina Chara pecho gris Mexican Jay PR Corvus corax Cuervo comun Common Raven PR Stelgidopteryx serripennis Golondrina ala serrada Northern Rough-winged Swallow PR Poecile sclateri Carbonero mexicano Mexican Chickadee PR Baelophus wollweberi Carbonero embridado Bridled Titmouse PR Psaltriparus minimus Sastrecillo Bushtit PR THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 200 Table 1. Cont. Species Common Name (Spanish) Common Name Residency (English) Sitta carolinensis Sita pecho bianco White-breasted Nuthatch PR Sitta pygmaea Sita enana Pygmy Nuthatch PR Certhia americana Trepador americano Brown Creeper PR Thryomanes bewickii Chivirfn cola oscura Bewick’s Wren PR Troglodytes aedon Chivirin saltapared House Wren WR Regulus calendula Reyezuelo de rojo Ruby-crowned Kinglet WR Polioptila caerulea Perlita azulgris Blue-gray Gnatcatcher PR Polioptila melanura Perlita del desierto Black-tailed Gnatcatcher PR Sialia sialis Azulejo garganta canela Eastern Bluebird WR Myadestes occidentalis Clarfn jilguero Brown-backed Solitaire PR Catharus guttatus Zorzal cola rufa Hermit Thrush WR Turdus migratorius Mirlo Primavera American Robin PR Toxostoma curvirostre Cuitlacoche pico curvo Curve-billed Thrasher PR Melanotis caerulescens Mulato azul Blue Mockingbird V Bombycilla cedrorum Ampelis chinito Cedar Waxwing WR Ptilogonys cinereus Capulinero gris Gray Silky-flycatcher PR Phainopepla nitens Capulinero negro Phainopepla PR Peucedramus taeniatus Ocotero enmascarado Olive Warbler PR Vermivora celata Chipe corona naranja Orange-crowned Warbler WR Vermivora crissalis Chipe crisal Colima Warbler SR Parula superciliosa Parula ceja blanca Crescent-chested Warbler PR Dendroica coronata Chipe Coronado Yellow-rumped Warbler WR Dendroica towns endi Chipe negroamarillo Towsend’s Warbler WR Dendroica occidentalis Chipe cabeza amarilla Hermit Warbler WR Dendroica sp. Chipe Warbler * Mniotilta varia Chipe trepador Black-and-white Warbler WR Wilsonia pusilla Chipe corona negra Wilson’s Warbler WR Myioborus pictus Chipe ala blanca Painted Redstart PR Piranga flava Tangara encinera Hepatic Tanager PR Piranga sp. Tangara Tanager PR Pipilo maculatus Toqui pinto Spotted Towhee PR Pipilo fuscus Toqui pardo Canyon Towhee PR Aimophila cassinii Zacatonero de Cassin Cassin ’s Sparrow PR Spizella passerina Gorrion ceja blanca Chipping Sparrow PR Spizella pallida Gordon palido Clay-colored Sparrow WR Melospiza lincolnii Gorrion de Lincoln Lincoln’s Sparrow WR Melospiza sp. Gorrion Sparrow * Junco phaenotus Junco ojo de lumbre Yellow-eyed Junco PR Pheucticus melanocephalus Picogordo tigrillo Black-headed Grosbeak PR RUVALCABA-ORTEGA ET AL. 201 Table 1. Cont. Species Common Name (Spanish) Common Name (English) Residency Passerina caerulea Picogordo azul Blue Grosbeak PR Passerina cyanea Colorfn azul Indigo Bunting WR Icterus xvagleri Bolsero de Wagler Black-vented Oriole PR Icterus graduacauda Bolsero cabeza negra Audubon’s Oriole PR Icterus parisorum Bolsero tunero Scott’s Oriole PR Euphonia elegantisima Eufonia capucha azul Elegant Euphonia PR Carduelis psaltria Jilguero dominico Lesser Goldfinch PR Ralph (1996) with a fixed radius of 20 m for 10 minutes. Birds captured with mist nets were banded and released. Species were recorded following the systematic nomenclature of the A. O. U. (1998; 2000; Banks et al. 2002; Banks et al. 2003). Their permanency status was determined on the basis of field observations and information provided by Howell & Webb (1995). Guilds were considered following Ehrlich et al. (1988). Shannon’s Diversity Index (1948) and Sorenson’s Index of Similarity (1948) were used to obtain diversity and similarity indices. Results and Discussion Based on records obtained by systematic sampling (point counts or mist nets), 1,080 individuals corresponding to 83 species, 62 genera, 31 families and 9 orders were recorded (Table 1). Seventy percent (54 species) of the species were defined as permanent residents, followed in number by winter residents with 22% (17 species), summer residents with 4% (3 species), transients with 3% (2 species), and vagrants with 1% (1 species). The Oak Forest contained the highest number of species and individu¬ als (58 and 473, respectively), followed by Pine-Oak Forest (48 and 360, respectively), and finally Pine Forest (40 and 251, respectively). The avian community appears distributed into discrete guilds (Table 2) with insectivorous species (42 species, 73%) being the major group in Oak Forest. It is suggested that this is determined by the availability of food (primarly insects) in the Oak Forests, resulting from generally more humid conditions than that of other forest types and the capacity of Quercus bark to support a major richness and abundance of inverte¬ brates. 202 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Table 2. Number of avian species for guilds and type of vegetation. Guilds Total Pine Forest Pine-Oak Forest Oak Forest # Sp. % # Sp. % tt Sp. % # Sp. % Carrion 2 2.4 1 2.5 1 2.1 2 3.4 Prey 4 4.8 3 7.5 2 4.2 1 1.7 Insectivorous 62 75 30 75 36 75 42 72.8 Granivorous 4 4.8 2 5 3 6.2 4 6.8 Nectivorous 4 4.8 2 5 4 8.3 3 5.1 Omnivorous 3 3.4 1 2.5 2 4.2 2 3.4 Frugivorous 4 4.8 1 2.5 0 0 4 6.8 Table 3. Similarity Matrix for vegetational communities (Sorenson’s Index). Pine Pine-Oak Oak Forest Forest Forest Pine Forest 0.465 0.403 Pine-Oak Forest 0.485 Although Shannon diversity values were very similar across vegeta¬ tion types, the highest was the Oak Forest (FF = 3.16), followed by Pine Forest (FT = 2.84), and lowest for Pine-Oak Forest (FT = 2.75). Evenness values were similar across all vegetation types; ranging in value from 0.71 (Pine-Oak Forest) to 0.78 (Oak Forest) to 0.77 (Pine Forest). The Pine-Oak Forest showed the lowest evenness values as a consequence of lower homogeneity in the avian community compared to the other vegetational associations. The least similar avian communities based on Sorenson’s Index were Oak Forest and Pine Forest (Table 3), which shared only 40% of the same species. By contrast, each of these was more similar to Pine-Oak Forest, sharing 48.5% and 46.5% of the species, respectively. The seasonal distribution of species diversity is shown in Figure 1 . The high value for Spring appears to be due to the presence of late winter and early summer migratory species in addition to permanent residents. In Fall, there are occurrences of late summer and early RUVALCABA-ORTEGA ET AL. 203 372 m#Sp. □# Ind. Figure 1 . Seasonal distribution of avian richness and abundance. Numbers indicate the number of species and individuals captured during the study period. §H' DE Figure 2. Shannon’s Index (H’) and evenness (E) values for each season. Table 4. Similarity Matrix for seasons (Sorenson’s Index). Spring Summer Fall Winter Spring 0.489 0.510 0.464 Summer 0.331 0.281 Fall 0.472 winter migrants that result in a greater number of species during this season. Both Spring and Fall appear to be transitional seasons where the replacement of bird species takes place. It is suspected that the high abundance of birds during Winter is due to winter residents and transi¬ ents that migrate in numerically large groups, providing a lower homo¬ geneity in the avian community during this season (Figure 2). The highest similarities among seasons were Spring and Fall (51 %) and the lowest when comparing Summer and Winter (28. 1 %) (Table 4). 204 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 Conclusions The most diverse avian communities were observed in Oak Forests and during the Spring. However, although noticeable differences in richness and abundance of birds exist among the vegetational communi¬ ties and seasons compared, values for diversity and evenness are very similar. This leads the authors to conclude that avian communities in Pine, Pine-Oak, and Oak Forests in the Sierra San Antonio-Pena Nevada system are stable and homogenous throughout the year. Literature Cited American Ornithologists’ Union. 1983. Check-list of North American Birds. 7th ed. American Ornithologists’ Union. Washington, D.C., USA, 829 pp. American Ornithologists’ Union. 2000. Forty-second supplement to the American Ornithologists’ Union Check-list of North American Birds. Auk, 1 17(3):847-858. Anderson, S. H. 1972. Seasonal variation in forest birds of western Oregon. Northwest Sci., 46:194-206. Anderson, B. W., R. D. Ohmart & J. Rice. 1983. Avian and vegetation community structure and their seasonal relationships in the lower Colorado River Valley. Condor, 85:392-405. Arizmendi, M. C. & L. Marquez (editores). 2000. Areas de Importancia para la Conservation de las Aves en Mexico. Comision Nacional para el Conocimiento y Uso de la Biodiversidad. Mexico. 440 pp. Arriaga, L., J. M. Espinoza, C. Aguilar, E. Martinez, L. Gomez & E. Loa (coordinadores). 2000. Regiones Terrestres Prioritarias de Mexico. Comision Nacional para el Conocimiento y Uso de la Biodiversidad. Mexico, 609 pp. Avery, M. L. & C. van Riper III. 1989. Seasonal changes in bird communities of the chaparral and blue-oak woodlands in Central California. The Condor, 91:288-295. Baida, R. P. 1969. Foliage use by birds of the oak-juniper woodland and ponderosa pine forest in Southeastern Arizona. Condor, 7 1(4): 399-41 2. Banks, R. C., C. Cicero, J. L. Dunn, A. W. Kratter, P. C. Rasmussen, J. V. Remsen, Jr, J. D. Rising & D. F. Stotz. 2002. Forty-third supplement to the American Ornithologists’ Union Check-list of North American Birds. Auk, 119:897-906. Banks, R. C., C. Cicero, J. L. Dunn, A. W. Kratter, P. C. Rasmussen, J. V. Remsen, Jr, J. D. Rising & D. F. Stotz. 2003. Forty-fourth supplement to the American Ornithologists’ Union Check-list of North American Birds. Auk, 120(3):923-931 . Bazakas, P. 1996. The effects of vegetative structure on the richness, diversity, and abundance of breeding bird communities in east-central Minnesota. Minnesota Academy of Science Journal, 60(2) :20. Beedy, E. C. 1981. Bird communities and forest structure in Sierra Nevada of California. Condor, 83(2):97-105. Contreras-Balderas, A. J., A. M. Sada-Zambrano, J. A. Garcia-Salas, J. I. Gonzalez-Rojas, A. Guzman- Velasco, J. E. Cisneros-Tello & M. A. Cruz-Nieto. 1995. Capitulo 3. Aves. Pp. 35-54, in Listado Preliminar de la Fauna Silvestre del Estado de Nuevo Leon, Mexico., Contreras-B, S., F. Gonzalez-S, D. Lazcano V. y A. Contreras-A. (Eds.). Consejo Consultivo Estatal para la Preservation y Fomento de la Flora y Fauna Silvestre de Nuevo Leon. Gobierno del Estado de Nuevo Leon, Mexico, 152 pp. Contreras-Balderas, A. J. 1997. Resumen Avifaumstico de Nuevo Leon, Mexico. Pp. RUVALCABA-ORTEGA ET AL. 205 35-44, in The Era of Allan Phillips. A Festschrift. Dickerman, R. W. (compiler). Published by Horizon Communications, New Mexico, 246 pp. Corcuera M. del R., P. & J. E. L. Butterfield. 1999. Bird communities of dry forests and oak woodland of western Mexico. Ibis, 141(2):240-255. Daniel, R. S. & R. R. Fleet. 1999. Bird and small mammal communities of four similar-aged forest types of the Caddo lake area in east Texas. Texas J. Sci., 51(l):65-80. Dickson, J. G. & C. A. Segelquist. 1979. Breeding bird populations in pine and pine-hardwood forest in Texas. Journal Wild. Manage., 43(2):549-555. Doherty, P. E., Jr. & C. T. Grubb, Jr. 2000. Habitat and landscapes correlates of presence, density and species richness of birds wintering in forest fragments in Ohio. Wilson Bulletin, 1 12(3):388-394. Ehrlich, R. P., D. S. Dobkin & D. Wheye. 1988. The birders Handbook. A Fireside Book Published by Simon & Schuster Inc. New York, 785 pp. Garcia, S., D. M. Finch & G. C. Leon. 1998. Patterns of forest use and endemism in resident bird communities of north central Michoacan, Mexico. Forest Ecology and Management, 1 10(1-3): 151-171 . Gonzalez-Iglesias, R. M. 2001. Evaluation de la integridad ecologica en las comunidades de bosque de encino y bosque de pino de la Sierra de Picachos, Nuevo Leon. Unpubl. Master’s thesis. Instituto Tecnologico y de Estudios Superiores de Monterrey. Campus Monterrey, 98 pp. Guzman- Velasco, A. 1998. Distribucion altitudinal de la avifauna del Cerro del Potosi, Galena, Nuevo Leon, Mexico. Unpubl. Master’s thesis. Facultad de Ciencias Biologicas, Universidad Autonoma de Nuevo Leon, 123 pp. Friedmann, H., L. Griscom & R. T. Moore. 1950. Distributional check-list of the Birds of Mexico Part I. Pacific Coast Avifauna, 29: 202 pp. Hilden, O. 1965. Habitat selection in birds: A review. Ann. Zool. Fenn., 2:53-75. Howell, S. N. G. & Webb, S. 1995. A Guide to the Birds of Mexico and Northern Central America. Oxford University Press Inc., New York. U.S.A., 851 pp. INEGI. 1986. Smtesis Geografica del Estado de Nuevo Leon. Instituto Nacional de Estadistica, Geografia e Informatica, 156 pp. Latta, S. C., C. C. Rimmer & K. P. McFarland. 2003. Winter bird communities in four habitats along an elevational gradient on Hispaniola. The Condor, 105(2): 179-197. Martin-del Campo, R. 1959. Contribution al Conocimiento de la Omitologia en Nuevo Leon. Universidad, 16-17:121-180. Universidad Autonoma de Nuevo Leon. Miller, A. 1955. The avifauna of the Sierra del Carmen of Coahuila, Mexico. Condor, 57:154-178. Miller, A., H. L. Friedmann, L. Griscom & R. T. Moore. 1957. Distributional Check-list of the Birds of Mexico. Cooper Ornithological Society Avifauna, II: 1-409. Mills, T. R., M. A. Rumble & L. D. Flake. 2000. Habitat of birds in ponderosa pine and aspen/birch forest in the Black Hills, South Dakota. Journal of Field Ornithology, 71(2): 187-206. Morales-Perez, J. E. & A. G. Navarro-Sigiienza. 1991. Analisis de distribucion de las aves en la Sierra Norte del estado de Guerrero, Mexico. Anales Inst. Biol. Univ. Auton. Mexico, Ser. Zool. 62(3):497-510. Navarro, A. G. 1992. Altitudinal distribution of birds in the Sierra Madre del Sur, Guerrero, Mexico. Condor, 94(l):29-39. Ralph, C. J., G. R. Geupel, P. Pyle, T. E. Martin, D. F. DeSante & B. Mila. 1996. Manual de metodos de campo para el monitoreo de aves terrestres. Gen. Tech. Rep. PSW-GTR-159. Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of Agriculture, 44 pp. 206 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Shannon, C. E. 1948. A mathematical theory of communication. Bell. Syst. Tech. J., 27:379-243. Sorenson, T. 1948. A method of establishing groups of equal amplitude in plant sociology based on similarity of species content and it’s application analyses of the vegetation on Danish Commons. Biol. Skr. (K. Danse Vidensk. Selsk. N.S.), 5:1-4. Tatschl, J. L. 1967. Breeding birds of the Sandia Mountains and their ecological distributions. The Condor, 69:479-490. Terborgh, J. 1971. Distribution on environmental gradients: Theory and preliminary interpretation of distributional patterns in the avifauna of the Cordillera Vilcabamba, Peru. Ecology, 52(l):23-40. Winker, K. 1992. Avian distribution and abundance records for the Sierra de Los Tuxtlas, Veracuz, Mexico. Wilson Bull., 104(4):669-718. AJCB at: arcontre@fcb.uanl.mx TEXAS J. SCI. 56(3):207-214 AUGUST 2004 MATE GUARDING IN NORTHERN MOCKINGBIRDS c M1MUS POLYGLOTTOS) Rebecca Y. Bodily and Diane L. H. Neudorf Department of Biological Sciences, Sam Houston State University Huntsville, Texas 77341 Abstract. — The northern mockingbird, Mimus polyglottos, is a socially monogamous passerine. Behavioral observations during the fertile (nest building and egg laying) and the non-fertile (incubation) stages were used to determine the presence of paternity assurance behaviors. Mockingbird pairs remained close (within 5 m) 76.3% of the time during the fertile period. Median intrapair distance changed significantly from 4.8 m during the fertile period to 11.3 m during the non-fertile period. Males followed females significantly more during the fertile stage than the non-fertile stage. In addition, males sang the most during the fertile period. The male perched higher than the female in all of the breeding stages. Male northern mockingbird behavior was consistent with the mate guarding hypothesis. However, an alternative hypothesis, i.e., that males remain close to females to ensure copulation at the fertile stage, could not be rejected. Ninety percent of bird species are considered monogamous (Lack 1968), however many of these species engage in copulations outside the pair bond (termed extra-pair copulations or EPCs). Extra-pair fertiliza¬ tions (EPFs) result when EPCs are successful. Studies employing modern molecular techniques show that EPFs are common in many bird species with some populations containing 70% extra-pair young (Griffith et al. 2002). In some species females pursue EPCs, which suggests that they benefit from EPC behavior (Kempenares et al. 1992; Neudorf et al. 1997; Double & Cockburn 2000). Potential benefits of EPFs to females include better quality genes for the offspring (Fujioka & Yamagishi 1981; Kempenaers et al. 1992; Burley et al. 1994; Hasselquist et al. 1996), increased genetic variability of the offspring (Birkhead 1993; Petrie et al. 1998) or material benefits such as being allowed to feed on the territory of extra-pair males (Gray 1997). In addition, the extra-pair males may direct aggression toward predators on the territories of their extra-pair females (Gray 1997). In many bird species, mate guarding is a common paternity assurance behavior (reviewed in Birkhead & Moller 1992). Mate guarding is defined as any behavior that functions to reduce the likelihood of encounters between a female and other males during the time when the female is fertile (Hatch 1987). A common form of mate guarding is closely following a mate during her fertile period (Beecher & Beecher 1979; Birkhead et al. 1987; Ritchison et al. 1994). Such behavior may influence a females’ behavior, for example, in pied flycatchers ( Ficedula 208 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 hypoleucia) , the risk of EPCs increases as the distance between pair members increases (Alatalo et al. 1987). Evidence suggests that an intrapair distance greater than 10 m significantly increases the number of EPCs and EPC attempts (Alatalo et al. 1987). Northern mockingbirds (Mimus polyglottos) are socially monogamous but bigamy does occur occasionally (e.g. Laskey 1941). Low EPF frequencies (6.9% of broods, 3.1% of offspring) have been reported for a Texas population of mockingbirds (DeLoach 1997). The low level of EPFs may indicate mockingbirds do not regularly pursue EPCs and thus male paternity guards would not be necessary (Birkhead & Moller 1992). Alternatively, male mate guarding may be effective in prevent¬ ing females from obtaining extra-pair matings (e.g. Chuang- Dobbs et al. 2001, but see Stutchbury and Neudorf 1998). The purpose of this study was to determine if male northern mockingbirds use mate guarding as a paternity assurance strategy. If mate guarding exists, it was predicted that males would maintain a closer proximity to females, a higher perching position than females, and would follow females more during the fertile period than in the non- fertile period. Methods Species and study area— This study was conducted on the campus of Sam Houston State University (SHSU) in Huntsville, Walker County, Texas, during April- August 2000 and 2001 . SHSU is a 85-ha residential campus with an abundance of trees and manicured lawn. Hedge rows, shrubs and trees were common nesting sites of northern mockingbirds on campus. Mockingbirds were trapped using walk-in Potter traps baited with mealworms. Each individual was banded with a U.S. Fish and Wildlife aluminum band and a unique combination of three plastic color bands for visual identification. Sex of individual mockingbirds was determined using behavioral cues (e.g., song) and the presence of a brood patch or cloacal protuberance. Nests were located by following females and males and by checking likely nest sites such as dense shrubs and low dense trees (Joern & Jackson 1983; Means & Goertz 1983). For this study, the female’s fertile period was defined as the period from the initiation of nest building to the laying of the penultimate egg (Birkhead & Moller 1992), which was typically 7-10 days. Behavioral observations.— Over two breeding seasons, 12 different breeding pairs were observed during either the fertile (n = 6 pairs, 1 1 BODILY & NEUDORF 209 h) or nonfertile (n = 6 pairs, 9 h) stages. Ideally, the same female would have been watched during both the fertile and nonfertile stage, however, predation and nest desertion were common on the study site making observations throughout the nesting cycle difficult. No pairs were feeding fledglings from a previous brood at the time of observa¬ tions but some of the nests were the second or third nesting attempt for the season. To determine if and to what extent mate guarding took place, the behavior of individual pairs was sampled during two, 1-h observation periods during the females’ fertile (nest building and egg laying) or nonfertile (incubation) period. Watches were conducted only on pairs whose nest had been located and thus their nest stage was known at the time of the watches. Incubation watches included time females spent on and off the nest. Nest predation and inclement weather prevented two observations from being completed on 4 pairs. Thus, 1 pair at the fertile stage and 3 pairs at the nonfertile stage were watched for 1 h only. Mate guarding behaviors quantified included: (1) Intra-pair distance - distance (m) between a paired male and female every 2 min; (2) Height above mate - recorded which sex was perched higher (m) every 2 min; (3) Movement initiation - determined the frequency that 1 pair member followed the other within 15 sec of a pair member initiating a movement. A movement was defined as flying or walking in a directed manner for at least 1 m from the original position; (4) Song - recorded at 2 min intervals if the male was singing; (5) Fights - noted any observations of fights or intrusions into the focal territory by neighboring individuals or intrusions onto a neighboring territory by focal individuals. Fights were defined as aggression between two individuals that involved contact. Perch height and intra¬ pair distances were estimated visually by the observer. All observations were conducted by RYB. Statistical Analyses— Nonparametric statistics were used due to non- normal data and small sample sizes. Behavior at fertile and non- fertile stages was compared with Mann- Whitney U tests. Wilcoxon signed-rank tests were used to compare male and female behavior. All tests are one-tailed unless indicated otherwise. StatView, V. 5 (SAS Institute, Inc., Cary, NC) was used for all analyses. Results Males remained closer to their mates during the fertile period than during the non- fertile period (Table 1). Males were also within 5 m of the female significantly more during the female’s fertile period (Mann Whitney U test, U = 0.0, P = 0.002), with males within 5 m of 210 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Table 1 . Median (lower, upper interquartile range values) of mate guarding behaviors of northern mockingbirds during the fertile and nonfertile stages. Variable a Fertile n = 6 Nonfertile n = 6 Ub P Intra-pair distance (m) 4.8 (3.7, 5.3) 11.3 (9.7, 15.5) 0.0 0.002 Time male is < 5 m 23.3 (20.0, 26.0) 10.5 (2.0, 12.0) 0.0 0.002 Male follows female 2.8 (1.0, 3.0) 0.5 (0, 2.0) 7.0 0.038 Female follows male 1.5 (1.0, 2.0) 0.3 (0, 1.0) 7.0 0.072 c Male perched above female 14.8 (13.5, 15.5) 15.5 (13.5, 21.0) 14.5 0.285 Female perched above male 3.8 (3.0, 4.0) 5.5 (3.5, 9.0) 11.0 0.26 c Neither perched higher 12.3 (10.5, 13.0) 8.5 (5.0, 12.5) 10.5 0.228 Male song 17.5 (14.5, 18.0) 10.0 (4.0-13.0) 6.5 0.032 Male fighting 0.0 (0, 1.0) .75 (0, 1.0) 14.5 0.271 a Time within 5m, perching and male song are measured as number of 2-min intervals the individuals engaged in behavior. Following and fighting are reported as actual number of times the behaviors occurred. b Fertile and non-fertile stages were compared with a Mann- Whitney test: U values and P values are adjusted for ties. c Indicates two-tailed tests. females 76.3% of the time during the fertile period and 25.8% of the time during the non-fertile period. Males also followed mates more during the fertile period than the non-fertile periods (U = 7.0, P — 0.038, Table 1). Females exhibited a similar tendency, but differences were not significant. During the fertile period, females initiated 64.2% of the pair movements and males initiated 35.8%, and this difference approached significance (Wilcoxon signed-rank test, z = -1.9, P = 0.058, two-tailed). During the fertile period, males more often perched higher than females perched higher (z = -2.2, P = 0.014). However, the number of 2-min intervals during which males were perched higher than females was not significantly different between the fertile and non-fertile periods (U = 14.5, P = 0.29, Table 1). In 40.9% of the time intervals during the fertile stage and 27.1% of the non-fertile time intervals, neither pair member was perched higher than the other and this behavior did not BODILY & NEUDORF 211 differ between nest stages (Table 1). The average percent time males spent singing was 57% during the fertile period, which declined to 33% during incubation. There was a significant difference in song frequency between the fertile and non- fertile stages (U = 6.5, P = 0.032). There was no difference in male fighting behaviors between breeding stages (Table 1). No copulations or copula- tion attempts were observed during observation periods. Discussion These findings support the mate guarding hypothesis. Male northern mockingbirds spent more time within 5 m of mates when they were fertile than when they were non- fertile. This behavior may function to prevent other males from approaching and pursuing EPCs with their mates (Birkhead & Moller 1992). Males also followed females more during the fertile period (Table 1) and this may act to maintain proximi¬ ty (e.g., Beecher & Beecher 1979; Dickinson & Leonard 1996). Male mockingbirds perched higher than females during both the fertile and non- fertile periods. Therefore, this behavior is probably not specific to mate guarding. A higher perching position may permit males to more easily defend their territories, observe neighboring females for extra-pair mating opportunities and be vigilant for predators (Carlson et al. 1985). Hobson and Sealy (1989) found that male yellow warblers ( Dedroica petechia) perched higher than females throughout the nesting cycle and they also suggested multiple benefits to this behavior in addition to a possible mate guarding function. Song output by male mockingbirds was more frequent during the fertile period, which agrees with previous mockingbird studies (Logan 1983). Moller (1991) reported that males may use song in a mate guarding context, however this does not appear to be the case in mockingbirds. Logan (1988) found playbacks of song during the fertile period did not elicit more aggressive responses in male mockingbirds than did playbacks at incubation. If song functioned in mate guarding then males would be expected to respond to playbacks more aggressively while their mates were fertile. Studies of the effectiveness of mate guarding have generated equivocal results (e.g. Alatalo et al. 1987; Moller 1987; Kempenaers et al. 1995). Despite intense mate guarding relatively high EPFs still occur in many passerine species (e.g. Kempenaers et al. 1995; Wagner et al. 1996). The fact that mockingbirds have such low EPFs may indicate they do not regularly pursue EPFs or that males are extremely effective in 212 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 preventing EPFs. The frequency of EPFs in one population of northern mockingbirds is relatively low (6.9% of 130 broods contained extra-pair young, Deloach 1997) compared to many other passerines (see Griffith et al. 2002). Although EPFs can vary between populations of the same species (Bjornstad & Litjeld 1997), there is no reason to expect this population would have a significantly different EPF frequency than that reported by Deloach (1997). The population is located in a similar habitat (urban college campus) and is located only 120 km north of Deloach’s population. Alternative hypotheses may explain male proximity to the female at the fertile stage (Birkhead & Moller 1992; Dickinson & Leonard 1996). The “copulation access hypothesis” states that males remain close to females more often at the fertile stage to increase within-pair copulation opportunities. This hypothesis predicts males should remain close to females during the times when copulations are more likely to occur (Birkhead and Moller 1992). In many species, copulations occur most frequently in the morning (e.g. Birkehead et al. 1987) whereas in others there is no diurnal pattern (e.g. Vernier et al. 1993; Hanski 1994). To the author’s knowledge, the timing of within-pair copulations in mock¬ ingbirds has not been studied. To test the copulation access hypothesis, observation trials would be needed at different times throughout the day. Presumably males should remain closer to their mates in the morning (or the time of day that copulations normally occur) if it increases their opportunities for copulation. Conversely, males maintaining proximity for mate guarding purposes should be vigilant throughout the day as extra-pair copulations can potentially occur at any time of day (Venier et al. 1993). The “predation hypothesis” states that males maintain proximity to females to act as sentinels and warn females when predators are near. This hypothesis predicts that both males and females should equally attempt to remain in close proximity to facilitate male vigilance (Dickinson & Leonard 1996). However, it was found that male mock¬ ingbirds were more likely to follow females than the reverse, which supports the mate guarding hypothesis. In conclusion, male northern mockingbirds exhibited behaviors consistent with paternity assurance strategies. Males remained closer and followed their mates more frequently at the fertile stage. These behaviors have typically been regarded as methods to prevent females from engaging in EPCs. However, one cannot completely rule out alternative explanations for the observed behaviors. Future studies should focus on potential extra-pair mating tactics in mockingbirds to BODILY & NEUDORF 213 determine the extent to which mate guarding behavior may be selected for in males. Acknowledgments We are grateful to C. Logan, E. Morton and B. Stutchbury for hints on capturing mockingbirds. A. Dewees, G. Ritchison, M. Thies and anonymous reviewers provided valuable comments on the manuscript. Financial and logistical support was provided by Sam Houston State University. Literature Cited Alatalo, R. V., K. Gottlander & A. Lundberg. 1987. Extra-pair copulations and mate guarding in the polyterritorial pied flycatcher, Ficedula hypoleucia. Behaviour, 101:139-155. Beecher, M. D. & I. M. Beecher. 1979. Sociobiology of bank swallows: reproductive strategy of the male. Science, 205:1282-1285. Birkhead, T. R. 1993. Avian mating systems: the aquatic warbler is unique. Trends in Ecology and Evolution, 8:390-391. Birkhead, T. R., L. Atkin, & A. P. Moller. 1987. Copulation behaviour in birds. Behaviour, 101:101-138. Birkhead, T. R. & A. P. Moller. 1992. Sperm competition in birds: Evolutionary causes and consequences. Academic Press, San Diego, California, 282 pp. Bjornstad, G. & J. T. Lifjeld. 1997. High frequency of extra-pair paternity in a dense and synchronous population of Willow Warblers Phylloscopus trochilus. Journal of Avian Biology, 28:319-324. Breitwisch, R. & G. Whitesides. 1987. Directionality of singing and non-singing behaviour of mated and unmated northern mockingbirds, Mimus polyglottos. Animal Behaviour, 35:331-339. Burley, N. T., D. A. Enstrom & L. Chitwood. 1994. Extra-pair relations in zebra finches: differential male success results from female tactics. Animal Behaviour, 48:1031-1041. Carlson, A., L. Hillstrom & J. Moreno. 1985. Mate guarding in the wheater Oenanthe oenanthe. Ornis Scandinavica, 16:113-120. Chuang-Dobbs, H. C., M. S. Webster & R. T. Holmes. 2001. The effectiveness of mate guarding by male black-throated blue warblers. Behavioral Ecology, 5:541-546. DeLoach, D. M. 1997. Breeding success, mating success, and mating strategies of the northern mockingbird, Mimus polyglottos. Unpublished Ph.D. dissertation, Rice University, Houston, Texas, 241 pp. Derrickson, K. C. & R. Breitwisch. 1992. Northern Mockingbird. Pp. 1-26, in: (Poole, A., P. Stettenheim, & F. Gill, editors) Birds of North America, No. 7. The Academy of Natural Sciences, Philadelphia; American Ornithologists’ Union, Washington. Dickinson, J. L. & M. L. Leonard. 1996. Mate attendance and copulatory behavior in western bluebirds: evidence of mate guarding. Animal Behaviour, 52:981-992. Double, M. & A. Cockbum. 2000. Pre-dawn infidelity: females control extra-pair mating in superb fairy-wrens. Proceedings of the Royal Society Biological Sciences Series B., 267 (1442): 465-470. Fujioka, M. & S. Yamagishi. 1981. Extramarital and pair copulations in the cattle egret. Auk, 98:134-144. Gray, E. 1997. Female red-winged blackbirds accrue material benefits from copulating with extra-pair males. Animal Behaviour, 53:625-639. 214 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 Griffith, S. C., I. P. F. Owens & K. A. Thuman. 2002. Extra-pair paternity in birds: a review of interspecific variation and adaptive function. Molecular Ecology, 11:2195- 2212. Hanski, I. K. 1994. Timing of copulations and mate guarding in the chaffinch Fringilla coelebs. Omis Fennica, 71:17-25. Hasselquist, D., S. Bensch & T. von Schantz. 1996. Correlation between male song repertoire, extra-pair paternity and offspring survival in the great reed warbler. Nature, 381:229-232. Hatch, S. A. 1987. Copulation and mate guarding in the northern fulmar. Auk, 104:450- 461. Hobson, K. A. & S. G. Sealy. 1989. Mate guarding in the yellow warbler Dendroica petechia. Ornis Scandinavica, 20:241-249. Joern, W. T. & J. F. Jackson. 1983. Homogeneity of vegetational cover around the nest and avoidance of nest predation in mockingbirds. Auk, 100:497-498. Kempenaers, B., G. R. Verheyen, M. Van den Broek, T. Burke, C. van Broeckhoven, & A. A. Dhondt. 1992. Extra-pair paternity results from females preference for high- quality males in the blue tit. Nature, 357:494-496. Kempenaers, B., G. R. Verheyen, & A. A. Dhondt. 1995. Mate guarding and copulation behaviour in monogamous and polygynous blue tits: do males follow a best of a bad job strategy? Behavioral Ecology and Sociobiology, 31:89-96. Lack, D. 1968. Ecological adaptations for breeding in birds. Methuen and Co., London, 409 pp. Laskey, A. 1941. An instance of mockingbird bigamy. The Migrant, 12:65-67. Logan, C. A. 1983. Reproductively dependent song cyclicity in mated male mockingbirds (Mimus polyglottos). Auk, 100:404-413. Logan, C. A. 1988. Breeding context and response to song playback in mockingbirds ( Mimus polyglottos) . Journal of Comparative Psychology, 102:136-145. Means, L. L. & J. W. Goertz. 1983. Nesting activities of northern mockingbirds in northern Louisiana. Southwestern Naturalist, 28(l):61-70. Moller, A. P. 1987. Extent and duration of mate guarding in swallows Hirundo rustica. Ornis Scandinavica, 18:95-100. Moller, A. P. 1991. Why mated songbird sing so much: mate guarding and male announcement of mate fertility status. American Naturalist, 138:994-1014. Neudorf, D. L., B. J. Stutchbury, & W. H. Piper. 1997. Covert extraterritorial behavior of female hooded warblers. Behavioral Ecology, 8:595-600. Petrie, M., C. Doums, & A. P. Moller. 1998. The degree of extra-pair paternity increases with genetic variability. Proceedings of the National Academy of Sciences, 95:9390-9395. Ritchison, G., P. H. Klatt, & D. F. Westneat. 1994. Mate guarding and extra pair paternity in northern cardinals. Condor, 96:1055-1063. Stutchbury, B.J.M. & D. L. Neudorf. 1998. Female control, breeding synchrony, and the evolution of extra-pair mating tactics. Pp. 103-121 , in Avian reproductive tactics: female and male perspectives (P. G. Parker and N. Burley, EDS). Washington, DC: Ornithological Monographs, American Ornithologists’ Union. Venier, L. A., P. O. Dunn, J. T. Lifjeld & R. J. Robertson. 1993. Behavioural patterns of extra-pair copulations in tree swallows. Animal Behaviour, 45:412-415. Wagner, R. H., M. D. Schug & E. S. Morton. 1996. Condition-dependent control of paternity by female purple martins: implications for coloniality. Behavioral Ecology and Sociobiology, 38:379-389. DLHN at: bio_dln@shsu.edu TEXAS J. SCI. 56(3):215-222 AUGUST, 2004 A LATE CRETACEOUS DUROPHAGUS SHARK, PTYCHODUS MARTINI WILLISTON, FROM TEXAS Shawn A. Hamm1 and Kenshu Shimada2 department of Geology, Wichita State University 1845 Fairmount Street, Wichita, Kansas 67260 Environmental Science Program and Department of Biological Sciences DePaul University, 2325 North Clifton Avenue Chicago, Illinois 60614 and Sternberg Museum of Natural History, Fort Hays State University 3000 Sternberg Drive, Hays, Kansas 67601 Abstract.— The Late Cretaceous durophagus shark previously described as Ptychodus connellyi (Family Ptychodontidae) by MacLeod & Slaughter is here diagnosed as a junior synonym of Ptychodus martini Williston. The occurrence of the holotype (SMU-SMP 6903 1 ) in the Roxton Limestone Member (upper Lower Campanian) of the Gober Chalk in Fannin County, Texas is significant both geographically and stratigraphically. Whereas the present fossil record suggests that P. martini is endemic to the Western Interior Sea, this specimen represents the only record of P. martini outside Kansas. If the tooth was not subjected to any significant reworking, the specimen not only represents the youngest occurrence for the species, but also one of the youngest occurrences of the genus and family. Ptychodus is a Cretaceous shark genus occurring in Albian to Early Campanian marine deposits of North and South America, Europe, Africa and Asia (Cappetta 1987). The genus is known primarily by its teeth, which are characterized by a massive crown suited for crushing shelled macroinvertebrates (durophagy: e.g., see Kauffman 1978; Stewart 1988a). Based on articulated specimens (e.g., MacLeod 1982), teeth were arranged in parallel rows in both the upper and lower jaws, forming a pavement-like dentition. Species of Ptychodus are differentiated on the basis of variations in dental morphology (e.g., Cappetta 1987). The tooth crown of Ptychodus is generally square to rectangular when viewed occlusally, and the central portion of the crown surface has several parallel or radial ridges. Surrounding the central portion of the crown is the marginal area, which exhibits various textural patterns (e.g., granular, concentric, radial) formed by numerous small ridges, pits and tubercles. The crown rests on top of a massive tooth root, which may be weakly bilobed. The tooth root is smaller in dimension than the crown and has many forami¬ na located at the crown-root interface. The criteria used to distinguish various species of Ptychodus include crown height, the configuration and number of ridges on the tooth crown, and the ornamentation on the marginal area. 216 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Figure 1. Ptychodus martini Williston 1900a: (A) Occlusal view of SMP-SMU 69031 from Roxton Limestone Member (upper Lower Campanian) of Gober Chalk, Texas, initially described as P. connellyi MacLeod & Slaughter 1980; (B) basal view of SMP-SMU 69031; (C) anterior view of SMP-SMU 69031; (D) occlusal view of FHSM VP-2121 from Smoky Hill Chalk Member of Niobrara Chalk, Kansas; (E) basal view of FHSM VP-2121 ; (F) occlusal view of one of the teeth in holotype of P. martini (KUVP 55277: see Fig. 1G) from Smoky Hill Chalk Member of Niobrara Chalk, Kansas, which resembles SMP-SMU 69031 and FHSM VP-2121; (G) entire view of holotype of P. martini (KUVP 55277: arrow points to tooth shown in Fig. IF). Scale bar = 5 mm. MacLeod & Slaughter (1980) described a new species of Ptychodus , P. connellyi , based on a single tooth (Figs, la-c) recovered from the Roxton Limestone Member (Lower Campanian) of the Upper Cretaceous Gober Chalk (Fig. 2) in northeastern Texas. This specimen (the holo¬ type) remains the only known example of the species (Welton & Farish 1993, p. 58). However, comparisons with other Ptychodus specimens suggest that P. connellyi is conspecific with another species, P. martini (Williston 1900a). Therefore, the purpose of this paper is to reinterpret the holotype as P. martini , and discuss the geographic and stratigraphic significance of the specimen. Specimens in the following institutions are discussed in this paper: Fort Hays State University, Sternberg Museum of Natural History (FHSM), Hays, Kansas; the University of Kansas Vertebrate Paleontology Collection (KUVP), Lawrence, Kansas and the Shuler Museum of Paleontology at Southern Methodist University (SMP-SMU), Dallas, Texas. HAMM & SHIMADA 217 ! CHRONOLOGIC UNIT KANSAS TEXAS Period Stage Group Formation Group Formation c CO 'c to Late Pierre Lake Creek lu o Pecan Gap 03 Weskan >» £ Wolf City | CL s E CO 5 Sharon Springs Ozan O Roxton SMP-SMU 69031 CO UJ Gober C/> 13 O CD c (0 Late O CO •*- • 2> o ‘c O c CO Middle 33 Smoky Hill CO «z c Austin T? €0 UJ k. A o z w 3 < .ate KUVP 55271, FHSM VP-2121 c CO CD o CO ’c o T3 1 o Early Fort Hays Atco Figure 2. Generalized Upper Cretaceous stratigraphy (formations and members) of western Kansas and northeastern Texas (after Kennedy et al. 1997), indicating the stratigraphic horizons of Ptychodus martini specimens. Systematic Paleontology Ptychodus martini Williston 1900a Material.— SMP-SMU 69031 (Figs, la-c), a single tooth initially described as Ptychodus connellyi MacLeod & Slaughter (1980). Occurrence.— Roxton Limestone Member of the Gober Chalk (Fig. 2) exposed along the banks of Brushy Creek, 1.5 miles (2.4 km) southeast of the town of Barkley Woods, Fannin County, Texas (MacLeod & Slaughter 1980: Fig. 3). Description.— SMP-SMU 69031 is rectangular when viewed occulusally and measures 37 mm wide and 21 mm in anteroposterior length. The crown is flat and measures only 5 mm in height. Eight low transverse ridges extend over much of the surface, and the marginal area 218 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Figure 3. Geographic distribution of Ptychodus martini teeth recovered. is very narrow and smooth lacking any ornamentation. However, this lack of ornamentation appears to be due to weathering as the tooth exhi¬ bits signs of abrasion. The tooth root is tabular and porous, and lacks a nutrient groove. The total tooth height (crown -I- root) is 16 mm. Discussion Taxonomic remarks.— Based on SMP-SMU 69031, MacLeod & Slaughter (1980) differentiated Ptychodus connellyi from all other Ptychodus species by the flat occlusal surface (i.e., without an elevated HAMM & SHIMADA 219 cusp). However, observations suggest that the morphology of SMP- SMU 69031 (Fig. la-c) closely resembles teeth from the median row in the holotype of P. martini (KUVP 55277: Figs, lf-g) and FHSM-2121 (Figs, ld-e) recovered from the Upper Cretaceous Smoky Hill Chalk Member of the Niobrara Chalk in western Kansas. Because of their close resemblance, and the fact that no other Ptychodus species possess rectangular teeth with a flat occlusal surface (e.g., see Cappetta 1987; Wei ton & Farish 1993), the authors consider P. connellyi to be con- specific with P. martini. Because P. martini Williston (1900a) was described earlier than P. connellyi MacLeod & Slaughter (1980), P. connellyi is considered a junior synonym of P. martini following the International Code of Zoological Nomenclature (ICZN 1999). Anatomical remarks. —The holotype of Ptychodus martini (Fig. lg) consists of a set of 1 10 teeth. Although they were discovered disassoci¬ ated, the teeth presumably come from an individual shark and were arranged artificially (for naturally arranged, general dental pattern of Ptychodus , see Woodward 1911). The occlusal surfaces of some teeth in the specimen are exceptionally flat and possess low, thin transverse ridges that extend fully to the marginal area. These are interpreted to come from the median tooth row because they are the largest, most symmetrical teeth in the dentition. Other teeth in the dentition, which are interpreted to represent teeth of lateral rows, are less elongate and have a slightly elevated crown with wider marginal areas. The mor¬ phology of SMP-SMU 69031 (Figs, la-c) suggests that the tooth is from the medial tooth row (cf. Fig. If). Geographic remarks.— Reports on Ptychodus martini are scarce. The only previously reported specimens are KUVP 55277 (holotype: Williston 1900a; 1900b; Schultze et al. 1982, p. 13; Fig. lg) and FHSM VP-2121 (isolated tooth: Hamm 2002; Figs, ld-e) from western Kansas. The occurrence of P. martini in Texas is significant because it extends the geographic distribution of the species from the Western Interior to near the Gulf of Mexico (Fig. 3). Nevertheless, the present fossil record suggests that P. martini is endemic to the Western Interior Sea. Stratigraphic remarks.— The genus Ptychodus had a nearly worldwide distribution from Albian to Campanian time (Cappetta 1987; Welton & Farish 1993). The two previously reported P. martini specimens (KUVP 55277 and FHSM VP-2121) occurred in the Smoky Hill Chalk Member of the Niobrara Chalk (Fig. 2). Stewart (1990, p. 24) noted 220 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 that P. martini occurs only in his Proto sphyraena pemicosa biozone (Stewart 1988b). This biozone corresponds to Hattin’s (1982) lithostrati- graphic Marker Units 1, 2 and 3, which are collectively Late Coniacian in age. In northeastern Texas, the Gober Chalk is interpreted to be the upper tongue of the Austin Chalk (Stephenson 1927). The uppermost part (0.3-3 m) of the Gober Chalk, referred to as the Roxton Limestone (Fig. 2), consists of skeletal limestone rich in inoceramid [ Inoceramus balticus (Boehm)] and ammonite remains (Fisher 1965). The occurrence of the ammonites Menabites delawarensis (Morton) and Scaphites hippocrepis (deKay) dates the Roxton Limestone as late Early Campanian in age (Cobban & Kennedy, 1992). The surface of the Ptychodus martini tooth described in this paper (SMP-SMU 69031) shows extensive signs of abrasion (Figs. la-c). The abrasion could have resulted from a combination of pre-burial deposi- tional activities and/or reworking. Because it was recovered from the banks of Brushy Creek (Macleod & Slaughter 1980), the abrasion may also be due to modern fluvial processes. It should be noted that the only Upper Cretaceous rocks in which Brushy Creek cuts through are the Gober Chalk (including the Roxton Limestone) and the overlying Ozan Formation (Fig. 2) where it intersects with the main channel of the North Sulphur River (based on UTBEG 1966; Mark McKenzie pers. comm. 2002). Ptychodus has been reported from the Albian to the Campanian in North America (e.g., Williston 1900a; Applegate 1970; Meyer 1974; Cappetta 1987). Dibley (1911) reported 17 teeth of P. poly gyrus Agassiz from northern France in the zone of Actinocamax quadratus (De Blaiville), which is Early Campanian in age. Schwimmer & Williams (1994) reported the occurrence of P. mortoni in an early Early Campani¬ an deposit in eastern Alabama. If indeed SMP-SMU 69031 occurred in the Roxton Limestone (with no or insignificant reworking) , the specimen is important because it represents the youngest occurrence of P. martini (giving the stratigraphic range of the taxon from Late Coniacian to late Early Campanian). Together with Dibley (1911) and Schwimmer & William’s (1994) data, the specimen also marks one of the youngest occurrences for the genus Ptychodus and family Ptychodontidae (see also Cappetta et al. 1993). HAMM & SHIMADA 221 Acknowledgments We thank the following individuals for allowing us access to specimens in their care: R. J. Zakrzewski (FHSM); D. Maio (KUVP) and K. Newman (SMU). The senior author would also like to thank M. McKenzie (Grapevine, Texas) for discussions on the geology of the Gober Chalk, as well as his wife, Amy Hamm, for her help and support. We would also like to thank David Cicimurri (Bob Cambell Geology Museum, Clemson, South Carolina) and David Schwimmer (Columbus State University, Columbus, Georgia) for their reviews and comments. Literature Cited Applegate, S. P. 1970. The vertebrate fauna of the Selma Formation of Alabama. VIII, The fishes. Fieldiana, Geol. Mem., 3(8) :383-433 . Cappetta, H. 1987. Chondricthyes II. Mesozoic and Cenozoic Elasmobranchii. Pp. 1-193, in Handbook of Paleoicthyology: Volume 3B (H.-P. Schultze, ed.), Gustav Fischer Verlag, Stuttgart, 193 pp. Cappetta, H., C. Duffin & J. Zidek. 1993. Chondrichthyes. Pp. 593-609, in The Fossil Record 2 (M. J. Benton, ed.), Chapman & Hall, London, 845 pp. Cobban, W. A. & W. J. Kennedy. 1992. Campanian ammonites from the Upper Cretaceous Gober Chalk of Lamar County, Texas. J. Paleont., 66(3):440-454. Dibley, G. E. 1911. On the teeth of Ptychodus in the English Chalk. Quart. J. Geol. Soc. London, 67:263-277. Fisher, W. J. 1965. Rock and mineral resources of East Texas. Univ. Texas, Bureau Econ. Geol. Rep. Invest., 54:1-439. Hamm, S. A. 2002. First occurrence of Ptychodus martini (Ptychodontidae) from the Roxton Member of the Gober Chalk. J. Vert. Paleont., 22(Supp. to No. 3):62A. Hattin, D. E. 1982. Stratigraphy and depositional environment of Smoky Hill Chalk Member, Niobrara Chalk (Upper Cretaceous) of the type area, western Kansas. Kansas Geol. Survey Bull., 225:1-108. ICZN (International Commission on Zoological Nomenclature). 1999. International Code of Zoological Nomenclature (fourth edition). International Trust for Zoological Nomenclature, London, 308 pp. Kauffman, E. G. 1978. Ptychodus predation upon a Cretaceous Inoceramus. Palaeont., 15:439-444. Kennedy, W. J., W. A. Cobban & N. H. Landman. 1997. Campanian ammonites from the Tombigbee Sand Member of the Eutaw formation, the Mooreville Formation, and the basal part of the Demopolis Formation in Mississippi and Alabama. American Museum Novitates, 3201:1-44. MacLeod, N. 1982. The first North American occurrence of the Late Cretaceous elasmobranch Ptychodus rugosus Dixon with comments on the functional morphology of the dentition and dermal denticles. J. Paleont., 56:403-409. MacLeod, N. & B. Slaughter. 1980. A new ptychodontid shark from the Upper Cretaceous of northeast Texas. Texas J. Sci., 32(4): 333-335. Meyer, R. L. 1974. Late Cretaceous elasmobranchs from the Mississippi east Texas embayments of the Gulf Coastal Plain. Unpublished Ph.D. dissertation, Southern 222 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Methodist Univ., Dallas, Texas, 419 pp. Schultze, H.-P., J. D. Stewart, A. M. Neuner & R. W. Coldiron. 1982. Type and figured specimens of fossil vertebrates in the collection of the University of Kansas Museum of Natural History. Part I. Fossil fishes. Misc. Pub. Univ. Kansas Mus. Nat. Hist., 73:1-53. Schwimmer, D. & G. D. Williams. 1994. Vertebrate-based Upper Cretaceous biostratigraphy for the Gulf and Atlantic Coastal Plains. J. Vert. Paleont., 14(Supp. to No. 3):56A. Stephenson, L. W. 1927. Notes on the stratigraphy of the Upper Cretaceous formations of Texas and Arkansas. Bull. Am. Asso. Petrol. Geol., 11:1-17. Stewart, J. D. 1988a. Paleoecology and the first North American West Coast record of the shark genus Ptychodus. J. Vert. Paleont., 8(Supp. to No. 3):27A. Stewart, J. D. 1988b. The stratigraphic distribution of Late Cretaceous Protosphyraena in Kansas and Alabama. Fort Hays Studies (Third Ser.), 10:80-94. Stewart, J. D. 1990. Niobrara Formation vertebrate Stratigraphy. Pp. 19-30, in Niobrara Chalk Excursion Guidebook (S. C. Bennett, ed.), Univ. Kansas Mus. Nat. Hist. & Kansas Geol. Survey, Lawrence, 81 pp. UTBEG (University of Texas Bureau of Economic Geology). 1966. Geologic Atlas of Texas, Texarkana Sheet. Univ. Texas Bureau Econ. Geol., scale 1:250,000. Welton, B. & R. F. Farish. 1993. The collectors guide to fossil sharks and rays from the Cretaceous of Texas. Before Times, Lewisville, Texas, 204 pp. Williston, S. W. 1900a. Some fish teeth from the Kansas Cretaceous. Kansas Univ. Quart., 9:27-42. Williston, S. W. 1900b. Cretaceous fishes, selachians and ptychodonts: Univ. Geol. Survey Kansas, 6(2):237-255. Woodward, A. S. 1911. The fishes of the English Chalk. Palaeontogr. Soc., London, 6:185-224. SAH at: sahamm@sbcglobal.net TEXAS J. SCI. 56(3):223-230 AUGUST, 2004 NEW RECORDS OF THE TEXAS HORNSHELL POPENAIAS POPEll (BIVALVIA: UNIONIDAE) FROM TEXAS AND NORTHERN MEXICO Ned E. Strenth, Robert G. Howells and Alfonso Correa-Sandoval Department of Biology, Angelo State University San Angelo, Texas 76909, Texas Parks and Wildlife Department, HOH Fisheries Science Center HC07, Box 62, Ingram, Texas 78025 and Laboratorio de Zoologi'a, Instituto Tecnologico de Cd. Victoria A.P. 175, C.P. 87010, Cd. Victoria, Tamaulipas, Mexico Abstract.— The Texas hornshell (Popenaias popeii) is reported and documented from the South Concho River in west central Texas and the Rio Sabinas of northern Coahuila, both new site records. These records confirm the known distributional range of this species in the Colorado River drainage of central Texas and establishes a new interior state record for Coahuila. Recently collected shell material of P. popeii is also reported from the Devils River above Amistad Reservoir and from the Rio Salado above Falcon Reservoir. Resumen.— El bivalvo texano conocido como concha cuerno ( Popenaias popeii ) es registrado en el Rio Concho Sur en el centro-oeste de Texas y el Rio Sabinas en el norte de Coahuila. Ambos sitios son nuevos registros geograficos. Estos registros confirman el ambito de distribution conocido de la especie en el drenaje del Rio Colorado del centro de Texas y establece un nuevo registro estatal interior para Coahuila. Especimenes de P. popeii tambien son registrados en el Rio Devils arriba de la Presa La Amistad y en el Rio Salado arriba de la Presa Falcon. The freshwater bivalve Popenaias popeii was originally described from the "Devil’s River and Rio Salado, Texas" by Lea (1857) as Unio popeii. Both the designation of the type-locality as well as the scientific name have undergone subsequent revision. While the designation of the Devils River as one of the original collection sites of P. popeii by Lea (1857) is undisputed by subsequent authors, some confusion existed early relative to the exact location of the Rio Salado. Lea (1857) originally placed it in "Texas". Stearns (1891) gave the location as "near Leon, Mexico" and noted additional specimens from the "Rio Salado, New Mexico"; Singley (1893) referred to its location as "New Mexico" and Simpson (1914) cited its location as "New Leon, Mexico" (state of Nuevo Leon). Johnson (1999:21) noted that the lectotype USNM 85895 from the Rio Salado in Nuevo Leon was "inadvertently" selected by Johnson (1974:115) as the "figured holotype." The Texas hornshell historically ranged south in the coastal systems of northeastern Mexico to at least the Rio Cazones of Vera Cruz 224 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 (Johnson 1999). In addition to the Devils River and Rio Sal ado, it has been found upstream in the Pecos River to Ward County in Texas (Singley 1893) and several locations in New Mexico (Cockerell 1902; Metcalf 1982; Lang 2000); upstream in the Rio Grande to sites just downstream of Big Bend, Brewster County, Texas (Howells 1994) as well as several Mexican tributaries of the lower Rio Grande (Johnson 1999). A shell found in the Llano River in 1972 at Castell in Llano County (Ohio State University Museum collection OSUM 1976.365) was reported by Howells et al. (1997). Both Howells (2001a) and Smith et al. (2003) have recently mapped the distribution of P. popeii from the drainage systems of Texas, New Mexico and northern Mexico. As a result of recent field collections, this report documents additional range extensions for Popenaias popeii from west-central Texas and northern Coahuila. Additionally, recent examinations of several pre¬ viously known collection localities were conducted in both Texas and Mexico. Voucher specimens are deposited with the holdings of the Illinois National History Survey (INHS), the Instituto Tecnologico de Ciudad Victoria (ITCV) and the Angelo State University Natural History Collections (ASNHC). The following listing is abbreviated and cites only those synonymies/citations deemed relevant to this study. Popenaias popeii (Lea 1857) Texas Hornshell Unio popeii.— Lea 1857:102; Binney 1863:387; Cockerell 1902:69; Diaz de Leon 1912:136; Simpson 1914:700; Johnson 1974:115. Unio pop ei. — Stearns 1891:104; Singley 1893:322. Elliptio popei . — Ortmann 1912:271; Strecker 1931:17; Murray & Roy 1968:26. Elliptio ( Popenaias ) popei.— Frierson 1927:38. Nephronaias ( Popenaias ) popeii.— Haas 1969:201. Popenaias popei.— Heard & Guckert 1970:339; Burch 1973:16; Neck 1984:11; Neck & Metcalf 1988:262; Howells et al. 1996:93; Johnson 1999:21. Popenaias popeii.— Metcalf 1982:45; Howells 2001 a: 62; Smith et al. 2003:333. STRENTH, HOWELLS & CORREA 225 New Records South Concho River. — A single left valve was collected in 1991 from among flotsam at the low water crossing of the South Concho River and U.S. Highway 277 within the city limits of Christoval, Texas. Heavy flooding had occurred in the area several weeks prior to the collection date. Material examined. — South Concho River in Christoval (N 3 1 ° 1 1’ 15" W 100°29’59"), Tom Green County, Texas, 21 July 1991, a single left valve (INHS 29012). Remarks.— All previous records of Popenaias popei from Texas except the single specimen from the Llano River reported by Howells et al. (1997) have been made from the Rio Grande or its tributaries. This current record is noteworthy in that the South Concho River, like the Llano River, is a tributary of the Colorado River drainage system. The exact nature of the significance of these distributional records of P. popeii from the Colorado River drainage currently remains unknown. Numerous additional collections by Texas Parks and Wildlife Depart¬ ment from 1992 through the present failed to find any other specimens of P. popeii in the Llano or Concho rivers, or elsewhere in the Colorado drainage basin (Howells 2001b). Collected along with the single speci¬ men of P. popeii were several single valves of Cyrtonaias tampicoensis (Tampico pearly mussel). Rio Sabinas.— Specimens of Popenaias popeii were initially collected from the dry river bed of the Rio Sabinas in the Rio Los Sabinitos Park area on Highway 20 (Coahuila) just west of Rio Villa de San Juan Sabinas, Coahuila in August of 2001. A second collection in January of 2002 was made approximately 0.5 km upstream from the original site. Material examined.— Rio Sabinas west of Rio Villa de San Juan Sabinas (N 27°55,23" W 101 ° 1 8’21 "), Coahuila, Mexico, 2 August 2001, three complete sets of valves (INHS 29013); 19 January 2002, three complete sets of valves (ITCV 8002), three complete sets of valves (ASNHC 0049). Remarks. — Although this report represents the first interior record (other than the Rio Grande) of nonfossil material of Popenaias popeii from the state of Coahuila in northern Mexico, it should be noted that the Rio Sabinas is an upstream tributary of the Rio Sal ado. At the time of the collections in August 2001 and January 2002, the Rio Sabinas was completely dry and without any evidence of recent water flow. Workers in the municipality of Sabinas, approximately 20 km downstream from 226 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 the collection site, reported a cessation of water flow in the Rio Sabinas in the spring of 2000. Dead shell material of Cyrtonaias tampicoensis and Utterbackia imbecillis (paper pondshell) was also present in the dry river bed. Additionally, the Rio Sabinas was examined approximately 50 km downstream from the municipality of Sabinas at Juarez just before the river enters the impoundment of the Presa Don Martin (listed on some maps as Presa Venustiano Carranza). At this location (N 27°36’42" W 100°43’29"), the river is accessible beneath the Highway 35 (Coahuila) bridge. Metcalf (1982) earlier reported fossil material of Popenaias popeii from this location. On 27 October 2001 and 3 March 2002 the river exhibited no flow and was characterized by a series of large isolated pools. Numerous intact pairs of valves of dead specimens of Cyrtonaias tampicoensis and Utterbackia imbecillis were common along the bank and shallow soft substrate of the stream bed. No specimens of Popenaias popeii were found. Previously Reported Records In addition to collection efforts in the South Concho River and the Rio Sabinas, both of the originally designated type-localities of the Devils River and Rio Salado as well as the Llano River were revisited in an effort to assess the current existence of specimens of Popenaias popeii at each of these three different locations. Devils River . — Considerable anthropogenic changes have occurred in the area of the lower Devils River since the original collection of Popenaias popeii in the 1800’s. The Amistad Reservoir Dam (Presa La Amistad) was completed on the Rio Grande between Texas and Mexico in 1968. The resulting lake area included the confluence of the Rio Grande with both the Devils River and the Pecos River. Popenaias popeii requires a shallow stream environment and is not currently known from impoundments (Lang 2000); consequently the man-made Amistad Reservoir does not appear to provide suitable habitat for this species. The area of the Devils River immediately above the lake level was examined in July of 2001. Material examined.— 200 m upstream from the confluence of the Devils River and Amistad Reservoir (N 29°39’54" W 100°55’58"), Val Verde County, Texas, 14 July 2001, two complete (but damaged) sets of valves and broken shell material from two additional specimens (ASNHC 0050). All of the P. popeii shell material was old and weathered; no fresh shell material was found at this location. STRENTH, HOWELLS & CORREA 227 Remarks . — Despite changes associated with the construction of the Amistad Reservoir Dam, that section of the Devils River immediately above the current lake level appears to provide a physical habitat capable of sustaining extant populations of Popenaias popeii. While the presence of the above recently collected shell material of P. popeii in July of 2001 would appear to support the above proposal, only addition¬ al and more detailed field studies in this area can determine the current status of this species in the lower Devils River. Collected along with the specimens of P. popeii in 2001 were valves of Cyrtonaias tampicoensis . Rio Salado.— In a fashion similar to that of the Devils River, the Rio Salado has also undergone considerable anthropogenic changes since the original collection of Popenaias popeii in the 1800’s. Falcon Dam (Presa Falcon) was constructed on the Rio Grande between Texas and Mexico in 1953. The resulting Falcon Reservoir included the conflu¬ ence of the Rio Salado with the Rio Grande. As previously mentioned in reference to Amistad Reservoir, the resulting reservoir does not appear to provide suitable habitat for adult specimens of P. popeii. The area of the Rio Salado above the lake level was examined in March of 2002. Material examined. — Rio Salado 100 m downstream from bridge on Highway 2 (Mexico) in northern Tamaulipas (N 26°47’23" W 99°25’ 20"), 2 March 2002, a single heavily worn right valve (ASNHC 0051). Remarks. — At the time of the March 2002 collection, the Rio Salado exhibited no flowing water in the area of the Highway 2 bridge. The river was characterized by a series of large pools, which were separated by narrow bars of exposed substrate. Numerous intact pairs of valves of dead specimens of Cyrtonaias tampicoensis , Utterbackia imbecillis and Quadrula apiculata (Southern mapleleaf) were common in the stream bed. Anahuac.— Rio Salado beneath and downstream of the Highway 1 (Nuevo Leon) bridge within the municipality of Anahuac, Nuevo Leon (N 27° 14’ 1.4" W 100°08’21 .9"), 2 June 2002; three complete sets of valves and four single valves (one of the single valves was very recent) (ASHC 0052). Remarks.— The river at the time of the collection exhibited no detect¬ able flow and was under considerable influence of untreated household waste pollutants. Several specimens of Cyrtonaias tampicoensis and Utterbackia imbecillis were also found at this location. Llano River.— A single specimen of Popenaias popeii collected in 228 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 1972 was reported by Howells et al. (1997) from the Llano River (N 30°42’ 13" W 98°57’32") and the crossing of Highway 2768 at Castell in Llano County. Several recent visits in 1992, 1997, 1999, 2000 and 2001 to the Castell area yielded no additional specimens of P. popeii. Corbicula sp. Shell material of the Asian clam was present at every collection site in both Texas and Mexico examined during the course of this study and is therefore not individually reported as part of the additional faunal listings. Discussion This study extends the known range of Popenaias popeii to include the South Concho River of west central Texas and the Rio Sabinas of northern Coahuila. It also confirms the earlier report of this species by Howells et al. (1997) from the Colorado River drainage system of central Texas. These additional extensions to the known range of this freshwater bivalve would initially appear to represent positive indications to the overall conservation status of this species. It should be noted, however, that current conditions related to reduced water flow, drying of stream beds, or both, in the Rio Sal ado and Rio Sabinas of northern Mexico do not appear capable of supporting significant populations of Popenaias popeii. While isolated or protected areas of both of these rivers or their tributaries may in fact support limited numbers of surviving individuals or populations, the decline in suitable habitat in the area of northern Tamaulipas, Nuevo Leon and Coahuila does not appear favorable to the overall survivability of this species. Even though heavy rains in April of 2004 returned the Rio Sabinas to normal flow, the Devils, Llano and South Concho rivers of Texas cur¬ rently appear to provide a greater range of both available and seemingly suitable habitat for maintaining Popenaias popeii than do most of the rivers of northern Mexico. However, no extant populations are current¬ ly known from these three rivers. These rivers appear to provide both adequate levels of water and the necessary current flow capable of main¬ taining surviving populations of P. popeii. Very little is known about this species in Texas and no living specimens were observed during the course of this study. The extreme rarity of recovered shell material from both the Llano and South Concho would appear indicative of popu¬ lations at or near the extinction level in these two rivers. Indeed, the only known populations of P. popeii are present in a short stretch of the STRENTH, HOWELLS & CORREA 229 Black River, New Mexico (Lang 2000; Howells 2001a) and the Rio Grande, Webb County, Texas (Howells 2003, 2004), with recently dead shells found in the Rio Grande between Big Bend and the mouth of the Pecos River, Texas, suggesting survivors may also persist there as well (Howells 2004). Additional and more detailed study would be required to determine the current status of this species in the rivers of west central Texas. However, this study indicates that P. popeii is at least rare or endangered throughout its range in Texas and New Mexico. Acknowledgments The authors wish to thank David Marsh, James Holm, Jeff Masters, Barbara Strenth, Brad Henry, Lynn McCutchen and Kathryn Perez for assistance in the collection of specimens during the course of this study. Appreciation is extended to Kevin Cummings (Illinois Natural History Survey), Arthur Bogan (North Carolina State Museum of Natural Sciences) and two anonymous reviewers for their comments and suggestions for improving this manuscript. Literature Cited Binney, W. G. 1863. Bibliography of North American Conchology Previous to the Year 1860, Smithsonian Miscellaneous Collections, Part I, 650 pp. Burch, J. B. 1973. Freshwater Unionacean clams (Mollusca: Pelecypoda) of North America. Biota of Freshwater Ecosystems Identification Manual 11. US Environmental Protection Agency, 176pp. Cockerell, T. D. A. 1902. Unio popeii, Lea, in New Mexico. The Nautilus, 16:69-70. Diaz de Leon, J. 1912. Catalogus Molluscarum Maxicanae Reipublicae hucusque descripta. La Naturaleza Tercera Siere, 1 (4):93- 143 . Frierson, L. S. 1927. A classified and annotated check list of the North American Naiades. Baylor University Press, Waco, Texas, 111 pp. Haas, F. 1869. Superfamilia Unionacea. Das Tierreich (Berlin) 88x + 663 pp. Heard, W. H. & R. H. Guckett. 1970. A Re-Evaluation of the Recent Unionacea (Pelecypoda) of North America. Malacologia, 10(2):333-335. Howells, R. G. 1994. Preliminary distributional surveys of freshwater bivalves in Texas: progress report for 1992. Texas Parks and Wildlife Department, Management Data Series 105, Austin, 16 pp. Howells, R. G. 1999. Distributional surveys of freshwater bivalves in Texas: progress report for 1998. Texas Parks and Wildlife Department, Management Data Series 161, Austin, 28 pp. Howells, R. G. 2001a. Status of freshwater mussels of the Rio Grande, with comments on other bivalves. Texas Parks and Wildlife Department, Austin, 81 pp. Howells, R.G. 2001b. Declining status of freshwater mussels of the Rio Grande, with comments on other bivalves. Pages 59-73 in G. P. Garrett and N. L. Allan. Aquatic fauna of the northern Chihuahuan Desert. Museum of Texas Tech University, Special Publication 46, Lubbock, 160 pp. Howells, R. G. 2003. Distributional surveys of freshwater bivalves in Texas: progress report for 2002. Texas Parks and Wildlife Department, Management Data Series 214, Austin, 35 pp. Howells, R. G. 2004. Distributional surveys of freshwater bivalves in Texas: progress 230 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 report for 2003. Texas Parks and Wildlife Department, Management Data Series 222, Austin, 50 pp. Howells, R. G., C. M. Mather & J. A. M. Bergmann. 1997. Conservation status of selected freshwater mussels in Texas. Pages 117-129 in K.S. Cummings, A.C. Buchanan, C.A. Mayer, and T.J. Naimo. Conservation and management of freshwater mussels II: initiatives for the future. Proceedings of a UMRCC symposium, St. Louis, Missouri, October 1995. Upper Mississippi River Conservation Committee, Rock Island, Illinois, 293 pp. Howells, R. G., R. W. Neck & H. D. Murray. 1996. Freshwater Mussels of Texas. Texas Parks and Wildlife Department, Austin, Texas, 218 pp. Johnson, R. I. 1974. Recent and Fossil Taxa of Unionacea and Mutelacea introduced by Issac Lea, Including the Location of all of the Extant Types. Museum of Comparative Zoology Special Occasional Publication, 2:1-159. Johnson, R. I. 1999. Unionidae of the Rio Grande (Rio Bravo del Norte) System of Texas and Mexico. Occasional Papers on Mollusks, 6(77): 1-65. Lang, B. L. 2000. Status of the Texas homshell and native freshwater mussels (Unionoidea) in the Rio Grande and Pecos River of New Mexico and Texas. New Mexico Department of Game and Fish, Performance Report, Santa Fe, 17 pp. Lea, I. 1857. Description of Six New Species of Fresh Water and land Shells of Texas and Tamaulipas, from the Collection of the Smithsonian Institution. Proceedings of the Academy of Natural Sciences of Philadelphia, 9(1857): 101-102. Metcalf, A. L. 1982. Fossil Unionacean Bivalves from Three Tributaries of the Rio Grande. Pp. 43-58, 2 plates, in Proceedings of the Symposium on Recent Benthological Investigations in Texas and Adjacent States (J. R. Davis, Editor). Texas Academy of Science Aquatic Sciences Section, Austin, Texas, 278 pp. Murray, H. D. & E. C. Roy, Jr. 1968. Checklist of Freshwater and Land Mollusks of Texas. Sterkiana, 30:25-42. Neck, R. W. 1984. Restricted and Declining Nonmarine Molluscs of Texas. Texas Parks and Wildlife Department Technical Series No. 34:1-17. Neck, R. W. & A. L. Metcalf. 1988. Freshwater Bivalves of the Lower Rio Grande, Texas. The Texas Journal of Science, 40(3): 259-268. Ortmann, A. E. 1912. Notes on the Families and Genera of the Najades. Annuals of the Carnegie Museum 8(2):222-365 +3 plates (18-20). Simpson, C. T. 1914. A Descriptive Catalogue of the Naiades or Pearly Fresh-water Mussels. Bryant Walker, Detroit, Michigan, 1540 pp. Singley, J. A. 1893. Contributions to the Natural History of Texas. Geological Survey of Texas, Part I, Texas Mollusca, pp. 297-342. Smith, D. G., B. K. Lang & M. E. Gordon. 2003. Gametogenetic Cycle, Reproductive Anatomy, and Larval Morphology of Popenaias popeii (Unionoida) from the Black River, New Mexico. The Southwestern Naturalist, 48(3): 333-340. Sterns, R. E. C. 1891. List of North American land and fresh-water shells received from the U.S. Department of Agriculture, with notes and comments thereon. Proceedings National Museum, Vol. XIV, No. 844:95-106. Strecker, J. K. 1931. The Distribution of the Naiades or Pearly Fresh-Water Mussels of Texas. Baylor University Museum Special Bulletin, 2:1-71. NES at: ned.strenth@angelo.edu TEXAS J. SCI. 56(3):23 1-236 AUGUST, 2004 PARABOLOIDS FOR MAXIMUM SOLAR ENERGY COLLECTION Ali R. Amir-Moez Department of Mathematics Texas Tech University Lubbock, Texas 79409 Abstract.— Paraboloids of revolution have been used for many purposes such as searchlights, radars and other operations concentrating on the broadcasting of waves. This article is a study of some variations of these ideas. 1. Parabolas.- Let F(0,p) by the focus of y = -p the directrix of the parabola x2 — 4py (Fig. 1). It is well-known that the tangent line PT to the parabola at any point P is the bisector of the angle between PF and PH, the perpendicular from P to the directrix. This implies that the normal of P, PN, is the bisector of the corresponding supplement angle (Fig. 2). This idea suggests that some parabolic surfaces are useful in collecting solar energy. A few samples will be given. Fig. 2 232 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 2. Paraboloid of revolution. -Rotating a parabola about its axis, one obtains a paraboloid of revolution (Fig. 3). Since any plane containing the axis of rotation intersects the paraboloid in a parabola of the same size as the original one, the paraboloid has a single focus F. Thus F collects the maximum amount of energy when the rays are parallel to the axis. Indeed, this is quite well known and will not be further elaborated here. z Fig. 3 3. Elliptic Paraboloids. -Consider a concave mirror of elliptic paraboloid shape. The equation of the corresponding surface can be chosen to be where a and b are positive real numbers and we may choose a > b (Fig. 4). Consider a plane containing the z-axis. This plane intersects the xy-plane in a line. Choose an axis Ot on this line. Let (l,m)= (cos a, sin a), 0 0. One can easily see that there is a line of foci whose equations are x=0, z=— . 4a z Indeed a concave mirror of this shape is able to collect enough energy that one can cook a shish kabob or roast hot dogs in the line of the foci. AMIR-MOEZ 235 A parabolic cylinder is the simplest parabolic tube. One may study other tubes which collect more energy. Two interesting ones shall be studied. Consider the parabola y= — ( x2-b 2). 2b Rotating this parabola about the x-axis, one obtains 4 b2 Fig. 6 It is clear that every plane that contains the x-axis intersects this surface in a parabola whose focus is the origin (Fig. 6). Thus a portion of this surface may be used as a concave mirror for collecting energy. Now rotating the parabola about a line perpendicular to its axis which does not pass through the focus, one obtains another tube. Consider the parabola y-a(x2-b 2),a*^—,a > 0 ,b > 0. 2b 236 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Let the focus be at (0,q). Now we rotate the parabola about the x- axis. A portion of this surface may be made into a concave mirror. If q > 0, then we obtain a circular locus of foci which arches downward (Fig. 7), while if q < 0, one obtains a circular arc of foci bending upwards (Fig. 8). In this latter case one can collect the maximum amount of energy from rays along this con¬ cave arc. Acknowledgments Both the author and the Editorial Staff wish to thank David Cecil of TAMU-Kingsville for his review and suggestions relative to the publishing of this manuscript. Literature Cited Lockwood, E. H. 1961 . A Book of Curves. Cambridge at the University Press, Cambridge, U.K. ,199 pp. TEXAS J. SCI. 56(3):237-252 AUGUST, 2004 CHARACTERISTICS OF PERIPHERAL POPULATIONS OF PARTHENOGENETIC CNEM1DOPHORUS LAREDOENSIS A (SQUAMATA: TEIIDAE), IN SOUTHERN TEXAS James M. Walker, James E. Cordes and Mark A. Paulissen Department of Biological Sciences, University of Arkansas , Fayetteville, Arkansas 72701, Division of Sciences, Louisiana State University at Eunice Eunice, Louisiana 70535 and Department of Biological and Environmental Sciences McNeese State University, Lake Charles, Louisiana 70609 Abstract. — From 1984-2004 the distributional ecology of the parthenogenetic Cnemi- dophorus laredoensis ( = Aspidoscelis laredoensis) complex both north and south of the Rio Grande between Amistad Reservoir and the Gulf of Mexico was studied. Although dozens of sites inhabited by clonal complex A of C. laredoensis were discovered within a few km of the river (over a geographic range in parts of Webb, Zapata, Starr and Hidalgo counties, Texas, and Tamaulipas State, Mexico), the species was observed at only three sites in two Texas counties that were widely removed and apparently disjunct from the river-centered zone. In order to better understand what factors limit the distribution of C. laredoensis A, these three most distant sites from the Rio Grande (55.5 to 75.5 km) where this hybrid- derived species is in syntopy with maternal progenitor C. gularis ( = Aspidoscelis gularis) : Catarina, Dimmit County, and Encinal and Artesia Wells, La Salle County, Texas were studied. Each peripheral site was characterized by sandy substrate that is known to be one of the most important requirements for C. laredoensis A. The relative amounts of the original thorn scrub vegetation favorable to C. gularis and chronically disturbed habitat favorable to C. laredoensis A at each site constituted the major determinant of the relative size of populations of the two species. The absence of C. laredoensis A north of these sites in Dimmit and La Salle counties is probably a result of ecological resistance to expansion consisting of unsuitable substrate and vegetation. There was no evidence that a low frequency of hybridization between normally parthenogenetic females of C. laredoensis A and males of C. gularis or periodic collection of C. laredoensis A at Catarina and Artesia Wells measurably destabilized these populations. The hypothesis that diploid parthenogenetic Cnemidophorus laredoen¬ sis (McKinney et al. [1973]; = Aspidoscelis laredoensis sensu Reeder et al. [2002]; Sauria: Teiidae), represents the descendants of one hybrid female between the gonochoristic species C. gularis and C. sexlineatus ( = Aspidoscelis gularis and A. sexlineata respectively, sensu Reeder et al. [2002]) has received support from electrophoretic studies (McKinney et al. 1973; Parker et al. 1989; Dessauer & Cole 1989), mitochondrial DNA analysis (Wright et al. 1983), and skin histocompatibility experi¬ ments (Abuhteba 1990; Abuhteba et al. 2000; 2001). This mode of origin for clonal complex A of C. laredoensis necessitated an improba- 238 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Figure 1. Map of Texas and position (line) of the counties (top Z = Zavala, D = Dimmit, L = La Salle, W = Webb, Z = Zapata, S = Starr and H = Hidalgo), Rio Grande, and state in Mexico (T = Tamaulipas) referenced in this paper. Cnemidophorus gularis occurs in suitable habitats throughout the enlarged area, C. laredoensis A occurs in habitats in the immediate vicinity of the river in Webb, Zapata, Starr and Hidalgo counties and Tamaulipas and at outlying sites in Dimmit (open circle = Catarina), La Salle (upper circle = Artesia Wells and Lower circle = Encinal), and Starr (site marked by circle is not likely disjunct from the distribution of the species in the valley) counties, and C. sexlineatus is limited to parts of Dimmit, Webb (where marginal syntopy occurs with C. laredoensis A) and Starr counties. ble sequence of events involving a single ancestral hybrid female: (1) its growth to adulthood in syntopy with both parental species; (2) initial avoidance of back crossing with C. gularis and C. sexlineatus ; and (3) presence of cytogenetic determinants for production of eggs with parthe- nogenetic potential. Success of C. laredoensis A became possible when the descendants of the original hybrid completed fixation of partheno¬ genesis in successive generations and the incipient species “captured a habitat” (Wright & Lowe 1968). The geographic range of C. laredoensis A is situated between the southern edge of the range of its paternal progenitor C. sexlineatus and the Rio Grande in parts of Webb (only known sites of syntopy are listed by Walker et al. 2001), Dimmit, La Salle, Zapata, Starr and Hidalgo counties, Texas, USA, and the riverine zone bordering Mexico from Nuevo Laredo southeast to Nuevo Progreso, Tamaulipas, Mexico (Fig. 1; see also Walker 1987a; 1987c; Walker et al. 1990; Paulissen & Walker 1998). Remarkably, unlike its largely allopatric relationship to C. sexlineatus , the entire range of C. laredoensis A has developed within a small part of the vast binational distributional area of its WALKER, CORDES & PAULISSEN 239 maternal progenitor C. gularis (Conant & Collins, 1998). Despite extensive searching during over 50 expeditions from 1984- 2004 involving both sides of the Rio Grande between Amistad Reservoir and the Gulf of Mexico, populations of C. laredoensis A have never been located more than about 80 km N or a few km S of the river (Walker 1987a; 1987c; Walker et al. 1990). In fact, all except three of the 51 sites discovered for this parthenogen were either located within 16 km N (n = 35) and 10 km S (n = 1 1) of the river or were apparent¬ ly contiguous with this zone (n = 2 sites in Starr County). The other three are the most distant sites from the Rio Grande known for C. laredoensis A at 55.5 to 75.5 km to the north in Catarina, Dimmit County, and Encinal and Artesia Wells, La Salle County, Texas (Fig. 1). Several collecting trips were made to these peripheral sites inhabited by C. laredoensis A between 1986 and 2000 allowing (1) description of the habitat and substrate characteristics that affect whiptail lizards at each site; (2) estimation of the relative abundance of the parthenogen and C. gularis and characterization of the nature of syntopy between these species; (3) gauging of the impact of interspecific hybridization on both species at Artesia Wells and Catarina; and (4) estimation of the impact of collecting on populations of both species at each site. In this paper, the data obtained on these trips are used to identify the factors which may limit the distribution of C. laredoensis A in areas removed from the Rio Grande. Materials and Methods The capture of a single individual of C. laredoensis A in September 1985 at Encinal, La Salle County, approximately 56 km from the Rio Grande, was the first indication that the species inhabited areas well removed from the river. Subsequently, JMW led a number of sanc¬ tioned collecting trips to explore surrounding areas of La Salle, Dimmit and southern Zavala counties in search of the parthenogen (Walker 1987a). Sites at Catarina, Valley Wells, Asherton, Carrizo Springs and 3.2 km southwest of Carrizo Springs in Dimmit County, sites at Artesia Wells, Cotulla, Gardendale and Millet in La Salle County, and two sites at Crystal City in southern Zavala County were explored (Walker 1987a; 1987c). Each site was systematically searched by three or more collec¬ tors and an attempt was made to collect all lizards seen with air guns; on average, about one in three lizards observed was captured. Collec¬ tions were made between 0900 and 1700 CDT on clear to partly cloudy days in spring and summer during the peak period of whiptail lizard activity; visits were also made to some sites in September and October (Table 1). 240 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 ~0 -52 § S so ='1 II 5-u a. — \ _ - CO r- 00 ON > x _T J£n 00 .00 V J ON ON P ^ S o g n£8 § Sss *-7 % ON P in Tf NO r- ON — — Tj- It? 'n D 1 55 ■£ m O rt CO ^ , N + Q CO N d < o D r-~ .NO P Tf p^ 0 ncT ..NO 73 ^t D »n Sp || ' '-J o «n CQ N fc 9 rr < .P .O O S . . co O in . (N O NO IT) Tt CO CO in 00 (N ^ ^ O P* co co ^ N N ^ Q Q (N <• <£ rf p P >n P P co s •rj »n r- wo 00" 'cr r- «o N Q < P p" O ON >n r- */-> (N »o r- «n 1 + «n P Q O O'® «rj r- Tj- r- co in NO NO o in r- rr co in N N Q Q < < P P PC PC PC <<< P P P . . ^ PC PC § PC < < in < P P P O CO o n N N N N QQQQ I— I I l-H I— 1 1 >-* < < < < PPPP CO W H Q < r* NO 00 00 00 00 00 00 00 ON ON 00 00 ON %} 00 ON VJA ON U X) P W NO 00 ON < Z ON On ON ON ON 00 P < Q S _ , D •E ^-g >. >> (U >-» >1 1 m 2r * £ 0) GO :c CO W » (U < < X P WoW 0 CO ^ 00 CO 00 AT X P~ 00 On 2 •— i On m O 00 GPn ^5 00 ON ^ UO >n (N ^ CO 1 Tj \D - •4- Ov NO 00 CO ON o 5 r- ^ On co O OO ^ ON 00 > co Tt Os rt oo ON o ^ »n ^ << < < < < < < < < H 1 NO u, wOO —> < H P P aJ s X sj cS < H co >< o- < w ^ vo O < NO o O < cs CO Tf NO O NO NO 00 (N H O H < CP O ’■0 2 © cp TO 3 O 12 O CA *8 S.

*o 3. co ON > X3 3. c/a Z o oa o LA r 00 m 3 3 CL CL 3 nearby o 3 CP 3 3 8 3.X) CP 3 ^ 8* cr i *8 3 C/5 3 3 ►o 3 1* 3- a CP 3 CA g CA 3 00 o pr 3 £S. 3 03 3_ cT oa CP -o CP 3 z o z > 3 T3 8. s* CP CA_ o - -+5 a* i § 'C < CP cl »c O ^ ?S. cr o* g. *< ^ H-H o o 3 3 cr CP r P 3s 3 CP a* IB¬ S' era 3 era 1 g Vi % If & V5* CA- 3 NO 00 CA UP 3_ O CO 3 3 g* 3 i/i 0> o* 3 §* l 3 a* 'a* "oj n Is) 245 Table 2. Habitat characteristics of the three peripheral sites north of the Rio Grande known to be inhabited by the allodiploid parthenogenetic species Cnemidophorus laredoensis A (C. gularis was present and C. sexlineatus was absent at all sites). Site Codes follow Walker (1987a). 246 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 individuals of C. laredoensis A collected or observed at Encinal were syntopic with C. gularis either immediately west of main street near a patch of sand spilled from railroad cars that had become interspersed with grasses/weeds on both sides of railroad tracks ( n = 2) or in an adjacent weedy lot (n = 1 + several observed). The small number of C. laredoensis A observed at Encinal during the study were in patches of habitat within an area of about two acres (Table 2). The large number of C. gularis at parts of this site can be attributed to the initial migration of individuals from relatively undisturbed thorn scrub forma¬ tions nearby into the altered grassy habitat in railroad-right-of-ways that typically are not preferred by C. laredoensis A. Characteristics of peripheral sites not inhabited by C. laredoensis A— Many unsuccessful searches for C. laredoensis A have been con¬ ducted in La Salle County north of Artesia Wells along 1-35 (at Cotulla, Gardendale and Millett), in Dimmit County at sites other than Catarina (e.g., Valley Wells, Asherton, Carrizo Springs and SW of Carrizo Springs) and in parts of Zavala County (vicinity of Crystal City) (Fig. 1). The presence of C. gularis and absence of C. laredoensis at the Cotulla, Crystal City, Asherton and Carrizo Springs study sites is cor¬ related with the habitat characteristics of gravelly (not sandy) substrate and relatively undisturbed thorn scrub vegetation. Gardendale, Millett, and 3.2 km SW of Carrizo Springs initially seemed suitable for habita¬ tion by C. laredoensis A based on the presence of deep sandy soil, although only the latter site closely duplicated the chronically disturbed vegetation structures found at Artesia Wells and Catarina. Only C. gularis was recorded on three visits to Gardendale and during two visits to Millett. The site at 3.2 km SW of Carrizo Springs near FM 2644 inhabited by whiptail lizards comprised approximately five acres with sandy soil, large clumps of cacti, scattered mesquites and sparse ground cover of grasses/weeds that had been heavily trampled, trailed and grazed by cattle. Although this habitat type and pattern of chronic disturbance seemed ideal for C. laredoensis A, it was the parthenogen’s paternal progenitor Cnemidophorus sexlineatus that was the most abundant whip- tail lizard at the site (n — 25 + 10 observed); relatively few C. gularis (n = 9 + 5 observed) were present (Table 1). The conclusion that neither C. laredoensis A nor C. sexlineatus were broadly distributed in La Salle and Dimmit counties was also supported WALKER, CORDES & PAULISSEN 247 by information pertaining to the 15, 200 acre Chaparral Wildlife Management Area provided by C. Ruthven (pers. comm.). This area is located 12.8 km west of Artesia Wells on Texas FM 133 in parts of both counties. Ruthven stated that since 1996 the Chaparral WMA staff had been sampling herpetofauna with drift fence arrays (totaling over 3900 drift fence days). They found that Cnemidophorus gularis is very common on the area (1,147 captures), C. sexlineatus is very rare (18 captures), but C. laredoensis A is absent. Role of habitat and substrate characteristics in limiting C. laredoensis A . — The three sites in Dimmit and La Salle counties where C. laredo¬ ensis A has been found away from the Rio Grande are characterized by sandy soil and chronic to catastrophic habitat disturbance (e.g., Fig 2). Most of the other sites in these counties lacked either one of both of these critical habitat characteristics and so it is not surprising that C. laredoensis A did not occur at them. Further north, the substrate becomes generally less sandy; this combined with an more or less unbroken expanse of undisturbed thorn scrub habitat suggests that C. laredoensis A is unlikely to be found much further north than the Dimmit and La Salle county sites documented in this paper. Potential role of interspecific hybridization with C. gularis— Indi¬ viduals of C. laredoensis A and C. gularis were occasionally observed in the same field of vision at Catarina and Artesia Wells and copulation between the two species was observed in the horse pasture at Catarina (Walker et al. 1991). That such copulations can lead to fertilization of the unreduced 2n = 46 eggs of normally parthenogenetic C. laredoensis A by the In = 23 sperm of C. gularis is indicated by the presence of hybrids of both sexes among the lizards obtained at Catarina and Artesia Wells. The seven C. laredoensis x C. gularis hybrids from Catarina were identified as follows: five males based on morphological characters and erythrocyte nuclear diameters (UADZ 1944, snout vent length = SVL 65 mm; 1945, SVL 62 mm; 1946, SVL 66 mm; 2987, SVL 78 mm; 3506, SVL 62 mm); one subadult female based on erythrocyte nuclear diameters (UADZ 2975, SVL 55 mm); and one female based on skin histocompatibility experiments and triploid chromosome complement (UADZ 3541, SVL 74 mm). Confirmed hybrids constituted only 6.4% of all whiptails obtained at Catarina. The hybrid males were readily identifiable based on a dorsal pattern closely resembling C. laredoensis 248 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 A, ventral colors resembling C. gularis, low numbers of granules around midbody, and postantebrachial scales of intermediate size be¬ tween the two parental species. That two females obtained at Catarina initially appeared to be individuals of C. laredoensis A based on color pattern, but were subsequently found to be hybrids based on other techniques, indicates that some female C. laredoensis A x C. gularis hybrids from this and other sites are not identifiable by external mor¬ phology. The most apparent meristic consequence of hybridization between the two species at Catarina was a reduction in the number of granules around midbody (mean 85.6, range 82-87, n = 7) in the confirmed hybrids. Two C. laredoensis x C. gularis hybrids were collected at Artesia Wells, one male (UADZ 1626, SVL 69 mm) and one female (UADZ 2017, SVL 88 mm). The hybrid male was similar in color pattern to the five hybrid males from Catarina. The hybrid female resembled indi¬ viduals of C. laredoensis A from Artesia Wells in dorsal pattern; however, it exceeded the maximum SVL of 80 mm for the species at the site and had a red-pink throat and remarkable purple-blue chest and abdomen resembling adult males of C. gularis. Erythrocyte nuclear diameters confirmed the hybrid status of both specimens. Theoretically, fertilization of normally parthenogenetic females of C. laredoensis A by males of C. gularis could destabilize the parthenogen by reducing successful reproduction at sites such as Artesia Wells and Catarina (Cuellar 1977). To date, this outcome has not been docu¬ mented for any pair of species of Cnemidop horns . At Catarina, seven hybrids were conclusively identified and an additional 10 specimens were putatively identified to C. laredoensis A (SVLs 44-71 mm) with such low numbers of granules around midbody (mean 85.8, range 83- 88) as to arouse suspicion that they might also be hybrids (UADZ 1650 [24 May 1986]; 2733 [8 October 1987]; 2965, 2966, 2969, 2974, 2983, 2986 [13 May 1988]; 3544 [19 May 1989]; 3707 [31 July 1989]). Even if all these individuals are hybrids, the fact that so few have been col¬ lected over the span of four years, combined with the fact that the size of the C. laredoensis A population has shown no sign of declining (see below and Table 3), suggests that interspecific hybridization of C. laredoensis A with C. gularis is not an important factor affecting the population of the parthenogen at Catarina. Presumably the same is true at Artesia Wells. WALKER, CORDES & PAULISSEN 249 Table 3. Summary of collecting success of each species and at the three peripheral sites inhabited by Cnemidophorus laredoensis A, C. gularis and hybrids in Dimmit and La Salle counties, Texas. Site (Visits) C. laredoensis A C. gularis Hybrids Catarina (D-3, 11 visits) Captured/Observed Collected per visit 80/210 (38.0%) 7.3 24/44 (54.5%) 2.2 7/8 (87.5%) 0.6 Artesia Wells (L-2, 6 visits) Captured/Observed Collected per visit 21/39 (53.8%) 3.5 18/22 (81.8%) 2.0 2/2 (100%) 0.3 Encinal (L-3, 3 visits) Captured/observed Collected per visit 4/7 (57.1%) 1.3 6/40 (15.0%) 2.0 None Totals (20 visits) Captured/Observed Collected per visit 105/256 (41.0%) 5.2 48/104 (46.1%) 2.4 9/10 (90.0%) 0.5 Impact of periodic collections. — Collecting trips to Catarina and Artesia Wells made over the span of several years allowed determination if removal of lizards had any effect on abundance of C. laredoensis A (or C. gularis). The fact that the number of lizards captured per trip does not show a decline from the first collecting trip to the last (Table 1) suggests that periodic collecting had no measurable impact on popula¬ tions of either species. Negative impact of collecting on each population was mitigated by the infrequency of removal of individuals between 1985 and 1997 and escape behaviors of the species which reduced the effectiveness of all methods of collection. The yield (% of lizards observed that were collected per site) ranged from 38.0% at Catarina to 57.1% at Encinal for C. laredoensis A and from 15.0% at Encinal to 81.8% at Artesia Wells for C. gularis using air guns, large rubber bands and nooses (Table 3). Each of these methods was ineffective for col¬ lection of hatchlings of C. laredoensis A (Table 1, see 8 October 1987 and 31 July 1989 results). Overall, C. laredoensis A was the most abundant lizard at Catarina, C. gularis was the most abundant species at Encinal, and these two species were roughly equally abundant at Artesia Wells (Tables 1, 3). Conclusions.— Cnemidophorus laredoensis A is one of the most abundant vertebrates at many sites within its restricted range in southern Texas and Tamaulipas. The ancestor of this parthenogenetic species originated at a site inhabited by C. gularis and C. sexlineatus , possibly either in northern Webb County, the only point of syntopy presently 250 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 known for all three species, or in northern Starr County, where the ranges of C. laredoensis A and C. sexlineatus are separated by about 30 km and C. gularis occurs throughout the area. The 10-50 km wide separation of the ranges of C. laredoensis A from C. sexlineatus (extending from northern Webb County through Zapata, Starr and Hidalgo counties) is probably not a result of competitive exclusion of one species by the other. In the absence of major geographic barriers to the northward expansion of C. laredoensis A toward the range of G sexlineatus and the southward expansion of C. sexlineatus toward the Rio Grande, it appears that both are hampered by subtle ecological barriers to expansion (i.e., substrate and/or vegetation structure). It is possible that C laredoensis A has been able to expand more rapidly along both sides of the Rio Grande in a zone of frequent habitat disturbance (that may temporarily displace C. gularis) than northward from the river through more stable habitats (inhabited by C. gularis) toward the southern range limits of C. sexlineatus. Although both C. laredoensis A and C. sexlineatus are sand-loving species, the former is mostly limited to alluvial deposits (Walker 1987a) whereas the latter is mostly limited to broadly distributed eolian deposits where species of the lizard genus Holbrookia and the sandbur genus Cenchrus are ecological indicators (Paulissen et al. 1997). To rephrase the question posed by Paulissen et al. (1992) “Can parthenogenetic Cnemidop horns laredoensis (Teiidae) coexist with its bisexual (progenitors)?” the answer in the case of C. sexlineatus , broadly speaking, is no; the answer in the case of C. gularis is emphatically yes. Broad syntopy between C. laredoensis A and maternal progenitor C. gularis within the range of the former in areas of Texas and Mexico stems from one of two conditions. Syntopic contacts at sites such as Encinal, Artesia Wells and Catarina could involve a temporal dynamic in which one species is eventually excluded from the site by the inter¬ play between interspecific competition and habitat characteristics. A stronger possibility is that syntopy is maintained through mitigation of these effects by a variety of responses (e.g., microhabitat selection, reproductive adaptations, tolerance of diet niche overlap and/or relaxed selection pressure in disturbed habitats: Paulissen et al. 1992; Paulissen 2001). That habitat structure and history of land use are crucial compo¬ nents in the complex syntopic relationship between C. laredoensis A and G gularis at particular sites (Walker 1987a; 1987b; 1987c) is consistent with observations on these species at Encinal, Artesia Wells and Catarina. At each site, the amounts of relatively undisturbed thorn scrub WALKER, CORDES & PAULISSEN 251 vegetation favorable to C. gularis versus disturbed habitats favorable to C. laredoensis A constitute the major determinants in the relative size of populations of the two species (Tables 1, 2, 3). Catastrophic altera¬ tion of any of these sites would be expected to result in the reduction or exclusion of C. gularis and rapid repopulation by C. laredoensis A (Walker 1987b), whereas restoration of the original thorn scrub habitat would likely lead to the reverse of this outcome. Acknowledgments Specimens of Cnemidophorus employed in this study were collected under the terms of yearly permits issued to each of us by Texas Parks and Wildlife. Assistance in the field was provided by Ramadan M. Abuhteba, University of Arkansas, and Stanley E. Trauth, Arkansas State University. C. Ruthven, Assistant Area Manager, Chaparral Wildlife Management Area, La Salle and Dimmit counties, kindly supplied information on Cnemidophorus studies conducted by him and others on this area in Texas. Literature Cited Abuhteba, R. M. 1990. Clonal diversity in the parthenogenetic whiptail lizard, Cnemidophorus ‘laredoensis’ complex (Sauria: Teiidae), as determined by skin transplantation and karyological techniques. Unpubl. Ph. D. Diss., Univ. Arkansas, Fayetteville., 82 pp. Abuhteba, R. M., J. M. Walker & J. E. Cordes. 2000. Genetic homogeneity based on skin histocompatibility and the evolution and systematics of parthenogenetic Cnemidophorus laredoensis (Sauria: Teiidae). Can. J. Zool., 78:895-904. Abuhteba, R. M., J. M. Walker & J. E. Cordes. 2001. Histoincompatibility between clonal complexes A and B of parthenogenetic Cnemidophorus laredoensis : evidence of separate hybrid origins. Copeia, 2001:262-266. Conant, R. & J. T. Collins. 1998. A field guide to reptiles and amphibians of eastern and central North America. 3rd ed., expanded. Houghton Mifflin Co., Boston, Massachusetts, 616 pp. Cuellar, O. 1977. Animal parthenogenesis. Science, 197:837-843. Dessauer, H. C. & C. J. Cole. 1989. Diversity between and within nominal forms of unisexual Teiid lizards. Pp. 49-71, in Evolution and ecology of unisexual vertebrates (R. M. Dawley and J. P. Bogart, eds.), Bull. 466. New York State Museum, Albany, New York, 302 pp. McKinney, C. O., F. R. Kay & R. A. Anderson. 1973. A new all-female species of the genus Cnemidophorus. Herpetologica, 29:361-366. Parker, E. D., Jr., J. M. Walker & M. A. Paulissen. 1989. Clonal diversity in Cnemidophorus : ecological and morphological consequences. Pp. 72-86, in Evolution and ecology of unisexual vertebrates (R. M. Dawley and J. P. Bogart, eds.), Bull. 466. New York State Museum, Albany, New York, 302 pp. Paulissen, M. A. 2001. Ecology and behavior of lizards of the parthenogenetic 252 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 Cnemidophorus laredoensis complex and their gonochoristic relative Cnemidophorus gularis : implications for coexistence. J. Herpetol., 35:282-292. Paulissen, M. A. & J. M. Walker. 1998. Cnemidophorus laredoensis. SSAR Catalogue of American reptiles and amphibians, 673.1-673.5 Paulissen, M. A., J. M. Walker & J. E. Cordes. 1992. Can parthenogenetic Cnemidophorus laredoensis (Teiidae) coexist with its bisexual congeners? J. Herpetol., 26:153-158. Paulissen, M. A., J. M. Walker & J. E. Cordes. 1997. Diet of the Texas yellow-faced racerunner, Cnemidophorus sexlineatus stephensi (Sauria: Teiidae), in southern Texas. Texas J. Sci., 49(2): 143-150. Paulissen, M. A., J. M. Walker & J. E. Cordes. 2001. Status of the parthenogenetic lizards of the Cnemidophorus laredoensis complex in Texas: re-survey after eleven years. Texas J. Sci., 53(2):121-138. Reeder, T. W., C. J. Cole & H. C. Dessauer. 2002. Phylogenetic relationships of whiptail lizards of the genus Cnemidophorus (Squamata: Teiidae): a test of monophyly, reevaluation of karyotypic evolution, and review of hybrid origins. Breviora, 3365: 1-61 . Walker, J. M. 1987a. Distribution and habitat of the parthenogenetic whiptail lizard, Cnemidophorus laredoensis (Sauria: Teiidae). Amer. Midi. Nat., 117:319-332. Walker, J. M. 1987b. Habitat and population destruction and recovery in the parthenogenetic whiptail lizard, Cnemidophorus laredoensis (Sauria: Teiidae), in southern Texas. Texas J. Sci., 39(1 ) :8 1 -88. Walker, J. M. 1987c. Distribution and habitat of a new major clone of parthenogenetic whiptail lizard (genus Cnemidophorus) in Texas and Mexico. Texas. J. Sci., 39(4):3 13-334. Walker, J. M., J. E. Cordes & M. A. Paulissen. 1989. Hybrids of two parthenogenetic clonal complexes and a gonochoristic species of Cnemidophorus , and the relationship of hybridization to habitat characteristics. J. Herpetol., 23:119-130. Walker, J. M., J. E. Cordes, R. M. Abuhteba & M. A. Paulissen. 1990. Additions to the distributional ecology of two parthenogenetic clonal complexes in the Cnemidophorus laredoensis subgroup (Sauria: Teiidae) in Texas and Mexico. Texas J. Sci., 42(2): 129-135. Walker, J. M., R. M. Abuhteba & J. E. Cordes. 1991. Copulation in nature between all-female Cnemidophorus laredoensis and gonochoristic Cnemidophorus gularis (Teiidae). Southwestern Nat., 36(2): 242-244. Walker, J. M., J. E. Cordes, R. M Abuheteba & M. A. Paulissen. 2001. Syntopy between clonal complexes A and B of parthenogenetic Cnemidophorus laredoensis (Sauria: Teiidae) and both of their gonochoristic progenitors. Amer. Midi. Nat., 145:397-401. Wright, J. W. & 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 & W. M. Brown. 1983. The origin of the parthenogenetic lizard Cnemidophorus laredoensis inferred from mitochondrial DNA analysis. Herpetologica, 39:410-416. MAP at: mpauliss@mail.mcneese.edu TEXAS J. SCI. 56(3):253-262 AUGUST, 2004 COMPARISON OF BRANCH ELONGATION AMONG FOUR ACACIA SPECIES AND TEXAS EBONY IN THE LOWER RIO GRANDE VALLEY OF TEXAS Melissa R. Eddy and Frank W. Judd Department of Biology, University of Texas-Pan American Edinburg, Texas 78541-2999 Abstract.— Branch elongation was compared among four Acacia species ( Acacia berlandieri , A.farnesiana, A. rigidula , A. schaffneri ) and Texas ebony ( Chloroleucon ebano) at three sites in Hidalgo and Starr counties, Texas. Most of the branch elongation occurred in fall and early winter in A. berlandieri , A. farnesiana and A. rigidula, but in A. schaffneri most of the growth occurred in late winter and spring. Branch elongation in Texas ebony was not concentrated in a given season. Acacia berlandieri, A. farnesiana and A. rigidula had significant positive correlations between branch elongation and rainfall, but A. schaffneri and Texas ebony did not. Variation in branch elongation among Acacia species is as great as that which occurs between the Acacia species and Texas ebony. Phenological studies are important because they provide descriptive information essential to the elucidation of reproductive and growth patterns. Such studies are a crucial prelude to formulation of hypotheses in experimental investigations (Bullock & Solis-Magallanes 1990; Eddy & Judd 2003). There have been only two studies (Vora 1990; Eddy & Judd 2003) of the phenology of woody plants in the Lower Rio Grande Valley of Texas (LRGV). Vora (1990) reported on flowering, fruiting, leaf growth and leaf drop of 19 native species (most were woody) occurring primarily at Santa Ana National Wildlife Refuge, 12.1 km south of Alamo, Hidalgo County, Texas. He did not quantitatively analyze comparisons among species in the characteristics he examined, and he did not quantify the relationships between climatic factors and the reproductive and vegetative responses of the species studied. Eddy & Judd (2003) described and quantified the flowering and fruiting phenolo¬ gy of Acacia berlandieri , A. minuata (= A. farnesiana) , A . rigidula , A. schaffneri and Chloroleucon ebano at two sites in Hidalgo County and one site in Starr County. The objectives of this study were to: (1) describe and quantify patterns of branch elongation among four Acacia species (A. berlandieri , A. farnesiana, A. rigidula and A. schaffneri) and Chloroleucon ebano ; (2) quantitatively examine the relationships between climatic factors and branch elongation of the species studied; and (3) determine if the magnitude of differences in branch elongation between members of the 254 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 genus Acacia are as great as those between any of the Acacia species and Chloroleucon ebano. The null hypotheses tested were: (1) there are no significant differences in the patterns of branch elongation of the Acacia species studied; (2) variation in branch elongation among the Acacia species is less than the variation between any of the Acacia species and Chloroleucon ebano ; and (3) there are no significant correlations between climatic factors and branch elongation. Materials and Methods Study area. — The LRGV comprises the southernmost four counties of Texas (Cameron, Hidalgo, Starr and Willacy counties). This study was conducted in Hidalgo and Starr counties. The climate is semi-arid and subtropical. Summers are long and hot and winters are short and mild (Lonard et al. 1991; Eddy & Judd 2003). The mean length of the frost- free period is 330 days, but winters often pass without a freezing temperature. Mean monthly temperature is greater than 16°C in all months throughout the LRGV. In summer, a temperature of 32.5 °C or greater occurs for 116 or more days. Mean annual rainfall ranges from a high of 71.5 cm at Harlingen, Cameron County to a low of 54.9 cm at Rio Grande City, Starr County. From 28 to 33 % of the annual rainfall occurs in September and October and 65 to 73% of the annual rainfall occurs from May through October. Most of the precipitation results from thunderstorms. Vegetation of the study sites is brush grassland and thorn woodland (Lonard et al. 1991; Eddy & Judd 2003). Study sites were the Castilla Ranch (CR) 11.9 km north of Rio Grande City, Starr County, Yturria Brush Tract (YBT) 7. 1 km west of La Joya, Hidalgo County and Santa Ana National Wildlife Refuge (SAN) 12. 1 km south of Alamo, Hidalgo County. Description of species .—Acacia berlandieri (guajillo) is a semi¬ evergreen shrub ranging in height from 1.0 to 4.0 m (Lonard et al. 1991; Everitt et al. 2002; Richardson 1995). It is found on a variety of soils, but is especially abundant in the LRGV on caliche soils in western Hidalgo and Starr counties. The leaves are fern-like, bipinnately compound, alternate and have 30 to 50 pairs of leaflets per pinna (Lonard et al. 1991; Eddy & Judd 2003). The flowers are white, and the legumes are 10.2 to 15.2 cm long with 5 to 10 dark brown seeds (Taylor et al. 1999; Eddy & Judd 2003). EDDY & JUDD 255 Acacia famesiana (huisache) is a small, spiny tree or shrub ranging from 2.0 to 4.0 m tall (Lonard et al. 1991; Everitt et al. 2002; Eddy & Judd 2003). It occurs on a variety of soil types (Lonard et al. 1991). The leaves are bipinnately compound, alternate, with 2 to 8 pairs of pinnae and 10 to 25 pairs of leaflets per pinna (Lonard et al. 1991). The flowers are yellow to gold, and the fruit can be reddish brown, purple, or black (Everitt et al. 2002). The legumes are 5.1 to 7.6 cm long and the seeds are in 2 rows within them (Everitt et al.2002; Taylor et al. 1999). Acacia rigidula (black brush) is a white-spined, multiple-stemmed shrub that grows to a maximum height of 3.0 m (Lonard et al. 1991; Eddy & Judd 2003). It is often found with guajillo. Black brush is found on clay or gravelly soils in the LRGV (Richardson 1995). The leaves are alternate, bipinnately compound with 1 or 2 pairs of pinnae and 2 to 4 leaflets per pinna (Lonard et al. 1991). The flowers are yellowish or white. The legume is black to reddish black, 5. 1 to 8.9 cm long, and constricted between the seeds (Richardson 1995; Taylor et al. 1999). Acacia schaffneri (huisachillo) is a spiny, rounded shrub that grows to a maximum height of 2.0 m (Lonard et al. 1991). It occurs on sandy and clay soils in the LRGV (Richardson 1995). Leaves are alternate, bipinnately compound with 2 to 5 pairs of pinnae and 10 to 15 pairs of leaflets per pinna (Lonard et al. 1991). Flowers are yellow. The fruit is a linear, black, pubescent legume from 4.0 to 13.0 cm long and constricted between the seeds (Correll & Johnston 1979; Lonard et al. 1991; Everitt et al. 2002; Richardson 1995). Chloroleucon ebano (Texas ebony) is a tree with a maximum height of 15 m (Richardson 1995), but usually it is less than 10 m tall (Lonard et al. 1991). It has zig-zag branches with stout stipular spines. The leaves are alternate or fascicled and bipinnately compound with 3 to 6 pairs of leaflets per pinna. Texas ebony occurs on sandy loam soils in the LRGV (Lonard et al. 1991). The flowers are white, and the fruit is a thick- walled woody legume. Field and statistical methods.— Only black brush was present at all three study sites (Table 1). Each of the other four species was present at two sites. SAN and YBT each had four of the five species and CR had three species present. Ten individuals from each of the species 256 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Table 1. Species present at study sites in Hidalgo and Starr counties. NWR = National Wildlife Refuge. Species Castilla Ranch Santa Ana NWR Yturria Brush Tract Acacia berlandieri X X Acacia farnesiana X X Acacia rigidula X X X Acacia schaffneri X X Chloroleucon ebano X X present at a site were marked for study. Shrubs (guajillo, huisachillo and black brush) were 1.5 m in height or taller. Huisache and Texas ebony were 3 m or taller. Shrubs and trees of these heights were known to be capable of possessing fruit. Distance between marked individuals ranged from 8 m to 2,320 m. All plants selected were healthy. Plants were marked with colored flagging and two aluminum tags bearing a unique identification number. Branch elongation was monitored by applying a ring of paint just below the terminal bud on three randomly selected branches on each individual. The distance from the paint mark to the tip of the branch was measured to the nearest mm at monthly intervals from October 1998 through August 1999. The mean elongation of the three branches was recorded as the shoot elongation for the individual for a given month. Daily air temperatures, precipitation and photoperiod were obtained from the National Climatic Data Center for McAllen, Texas. Long-term precipitation and temperature data were obtained from the Office of the Texas State Climatologist. Results Mean monthly photoperiod at McAllen, Texas ranged from 10 h and 32 min in December 1998 to 13 h and 45 min in June 1999 (Table 2). The study sites varied from McAllen by less than 15 min latitude, so there was little variation between photoperiod at McAllen and any of the three study sites. Likewise, there was little variation in photoperiod among the study sites. Because of the distance between the study sites and the distance between them and McAllen, it was possible that rain might have occurred at McAllen and not at any of the study sites. Likewise, it was EDDY & JUDD 257 Table 2. Climatic data for McAllen, Texas. Month Rain (cm) 1998-99 Rain (cm) 1958-98 Mean Temp. (°C) 1998-99 Mean Temp. (°C) 1958-98 Mean Daylight (min) 1998-99 Sept. 24.09 11.11 28.7 29.1 738 Oct. 7.23 7.93 24.8 25.2 692 Nov. 2.61 2.82 21.6 20.6 652 Dec. 0.71 2.81 16.7 16.6 632 Jan. 0.08 3.74 18.2 15.0 643 Feb. 0.03 3.63 21.9 17.3 677 Mar. 5.74 2.02 23.0 21.1 721 Apr. 0.10 3.65 26.9 25.0 767 May 3.17 6.70 28.9 27.6 806 Jun. 1.27 7.06 30.9 29.9 825 Jul. 0.41 3.71 29.8 30.6 816 Aug. 7.82 5.51 31.2 30.9 782 possible that rain occurred at a study site and not at McAllen or that rain occurred at one study site and not at the other two sites. Using local observer reports it was previously shown (Eddy & Judd 2003) that there was less than 1 .0 cm difference in monthly rainfall total of the SAN and YBT sites in all months of this study. The CR site generally was within 1.5 cm in monthly rainfall of the other two sites, but in October 1998, CR received 2.6 cm more rain than the other sites and in August 1999, CR received 3.0 cm less rain than the other two sites. Branches were first marked for monitoring growth in length in October 1998. Consequently, November 1998 is the first month that data on branch elongation was reported. Rainfall in October and November 1998 was close to the long-term average for these months (Table 2). However, rainfall in December, January and February was 92% lower than the long-term average. And, rainfall from April through July, 1999 was 77% lower than the long-term average. Air temperature from January through June, 1999 was markedly higher than the 40-year average (Table 2). Mean monthly branch elongation is shown among species, months and sites in Table 3. Analysis of variance (ANOVA) showed significant variation in branch elongation among months in all species (Table 4), but there was no significant variation in branch elongation among months in black brush at the CR site or in Texas ebony at the SAN site. In guajillo, 63.1% of the increase in branch length occurred in November, December and January at the YBT site and 69.1% of the growth occurred in these same three months at the SAN site. Much of 258 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Table 3. Comparison of mean branch elongation (cm) per month among months, species, and study sites. N = 10 for each mean. Numbers in parenthesis equal one standard error of the mean. Sp = species, A. b. — Acacia berlandieri, A. f = Acacia fames iana, A. r. = Acacia rigidula, A. s. = Acacia schajfneri, C. e. = Chloroleucon ebano, CR = Castilla Ranch, SAN = Santa Ana National Wildlife Refuge and YBT = Yturria Brush Tract. Sp & Site Nov 98 Dec 98 Jan 99 Feb 99 Mar 99 Apr 99 May 99 Jun 99 Jul 99 Aug 99 A. b. 3.67 1.91 1.38 0.02 0.06 0.73 0.93 1.07 0.44 0.79 YBT (0.72) (0.70) (0.52) (0.02) (0.03) (0.27) (0.51) (0.70) (0.19) (0.35) A. b. 4.44 3.01 0.84 0.88 0.92 0.50 1.05 0.01 0.34 0.00 SAN (0.58) (1-41) (0.52) (0.46) (0.55) (0.26) (0.62) (0.01) (0.34) (0.00) A.f 7.82 0.13 0.10 0.33 0.22 2.40 2.75 0.41 0.15 0.29 CR (1.01) (0.11) (0.07) (0.19) (0.16) (1.16) (1.32) (0.28) (0.15) (0.28) A.f 9.04 1.82 0.83 0.00 0.08 3.87 0.34 0.49 0.89 1.41 SAN (1.69) (1.59) (0.45) (0.00) (0.08) (1.74) (0.16) (0.40) (0.71) (1.01) A. r. 4.45 1.73 0.12 0.64 0.49 2.14 3.92 1.19 4.83 3.20 CR (1.07) (1.69) (0.07) (0.40) (0.33) (0.98) (1.28) (0.62) (1.73) (1.19) A. r. 5.31 1.05 0.19 0.01 0.23 0.35 0.12 0.04 0.02 1.23 YBT (0.91) (0.42) (0.12) (0.01) (0.20) (0.19) (0.12) (0.04) (0.01) (0.88) A. r. 4.52 1.13 0.31 0.18 0.18 2.21 0.45 1.24 0.46 1.15 SAN (1.11) (0.50) (0.21) (0.15) (0.16) (0.80) (0.21) (0.59) (0.38) (0.60) A. s. 1.10 1.07 0.41 2.87 2.47 2.27 5.10 0.24 0.10 2.35 CR (0.51) (0.59) (0.25) (1.25) (0.69) (0.97) (1.47) (0.20) (0.06) (1-25) A. s. 1.34 0.00 0.34 0.86 6.95 2.78 2.93 0.15 0.06 0.06 YBT (1.09) (0.00) (0.35) (0.70) (1.90) (1.70) (1.41) (0.07) (0.04) (0.05) C. e. 2.89 0.14 0.45 0.04 0.13 2.36 1.95 0.11 0.02 3.61 YBT (0.86) (0.09) (0.42) (0.04) (0.13) (1.01) (0.73) (0.09) (0.01) (1.61) C. e. 0.45 1.67 0.01 0.10 0.00 1.69 0.29 0.54 0.41 0.39 SAN (0.45) (0.77) (0.01) (0.07) (0.00) (1.12) (0.19) (0.33) (0.35) (0.33) the growth in branch length took place in November alone in huisache (53.6% at the CR site and 48.2% at the SAN site). Increase in branch length was concentrated in November and December in black brush at two of the three sites (74.4% at the YBT site, 47.8% at the SAN site). Branch elongation at the CR site was distributed more evenly among months, but was low in January, February and March. Branch elongation in huisachillo showed a very different seasonal pattern than the other three Acacia species. At both the CR site (70.7%) and the SAN site (87.4%) most of the growth occurred in late winter and spring, i.e., February, March, April and May. In Texas ebony, branch elongation was distributed at peaks throughout the ten months. At the YBT site growth was concentrated in November, April, May and August, while at the SAN site growth was greatest in December and April. EDDY & JUDD 259 Table 4. Analysis of Variance of mean monthly branch elongation among species and sites. CR = Castilla Ranch, YBT = Yturria Brush Tract and SAN = Santa Ana National Wildlife Refuge. DF = degrees of freedom, SS = Sums of Squares, MS = Mean Squares, F = AN OVA value. NS = Not Significant (P > .05), * = P < .01, ** = P < .001. Species and Site Source DF SS MS F Acacia berlandieri Among months 9 99.368 11.041 39.573** YBT Within months 90 25.122 0.279 Acacia berlandieri Among months 9 193.913 21.546 6.069** SAN Within months 90 319.517 3.550 Acacia farnesiana Among months 9 536.198 59.575 13.680** CR Within months 90 391.198 4.355 Acacia farnesiana Among months 9 686.108 76.234 7.371** SAN Within months 90 930.789 10.342 Acacia rigidula Among months 9 70.136 7.793 0.562 NS CR Within months 90 1,247.210 13.858 Acacia rigidula Among months 9 237.293 26.367 14.017** YBT Within months 90 169.315 1.881 Acacia rigidula Among months 9 160.316 17.813 5.754** SAN Within months 90 278.645 3.096 Acacia schajfneri Among months 9 212.858 23.651 3.192* CR Within months 90 666.922 7.410 Acacia schajfneri Among months 9 433.642 48.182 4.683** YBT Within months 90 925.927 10.288 Chloroleucon ebano Among months 9 173.204 19.245 3.779** YBT Within months 90 458.286 5.092 Chloroleucon ebano Among months 9 34.732 3.859 1.582 NS SAN Within months 90 219.555 2.440 Correlation between mean monthly branch elongation and the previous month’s rainfall is compared between species and sites in Table 5. This correlation allows time for growth after rainfall occurs. Guajillo, huisache and black brush showed significant positive correlations at one or two sites. Huisachillo and Texas ebony did not have significant correlations with the previous month’s rainfall. There were no signifi¬ cant correlations in any species between mean monthly branch elonga¬ tion and mean monthly temperature. Only guajillo at the SAN site showed a significant correlation between mean monthly branch elonga¬ tion and mean monthly photoperiod (r = -0.682, 8 df,P< 0.05). Mean branch elongation over the 10 months of study was used to compare growth between sites within species. Guajillo, huisache and huisachillo did not exhibit significant variation between sites. Conversely, black brush had significant variation among the three sites where it was studied (F = 11.897, 2 & 27 df9 P < 0.001). The SAN 260 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 Table 5. Correlation coefficients for mean monthly branch elongation versus the previous month’s rainfall. N = 10 for all species and locations. CR = Castilla Ranch, YBT = Yturria Brush Tract and SAN = Santa Ana National Wildlife Refuge. NS = not significant ( P > .05). * = P < .05, ** = P < .01. Species CR YBT SAN Acacia berlandieri 0.715 * 0.563 NS Acacia farnesiana 0.739 * 0.884 ** Acacia rigidula 0.304 NS 0.688 * 0.923 ** Acacia schaffneri - 0.324 NS - 0.116 NS Chloroleucon ebano 0.385 NS 0.554 NS site had greater mean branch elongation than the YBT site (t = 2.166, 18 df, P < 0.05) and the CR site had a greater mean than either the SAN site (t = 2.996, 18 df \ P < 0.01), or the YBT site (t = 4.050, 18 df ’ P < 0.001). Texas ebony also showed significant variation in branch elongation between sites ( t = 2.287, 18 df, P < 0.05). Discussion Hypothesis 1 that there are no significant differences in the patterns of branch elongation of the Acacia species studied was falsified. Branch elongation occurred primarily in fall and early winter in guajillo, huisache and black brush but in huisachillo, branch elongation prin¬ cipally took place in late winter and spring. Eddy & Judd (2003) also found significant differences in the flowering and fruiting phenologies of these Acacia species. Hypothesis 2 that variation in branch elongation among the Acacia species was less than the variation between any of the Acacia species and Texas ebony also was falsified. Huisachillo differed from the other species of Acacia in the timing of branch elongation (as explained above) and unlike the other Acacia species, huisachillo did not show a significant correlation with rainfall. It was similar to Texas ebony in this respect. Eddy & Judd (2003) found that the flowering and fruiting of these Acacia species were more similar to each other than to Texas ebony. Thus, the data on branch elongation are very different from that on flowering and fruiting. Hypothesis 3 that there are no significant correlations between climatic factors and branch elongation also was falsified. Guajillo, huisache and black brush showed significant positive correlations with rainfall. Additionally, guajillo at the SAN site had a significant inverse correlation with mean monthly photoperiod. Thus, these findings EDDY & JUDD 261 support the conclusion of New (1984) that growth in Acacia species is often correlated with moisture. Vora (1990) stated that plant growth and reproduction were keyed to rainfall and soil moisture for most of the 19 species he studied at Santa Ana National Wildlife Refuge. Also, Nilsen & Muller (1981) found that branch elongation in the legume Lotus scoparius in California was primarily influenced by soil moisture and they suggested that this is a common response in chaparral plants. These data were obtained during a drought. Rainfall from November 1998 through August 1999 in the LRGV was only about half (47.3%) of the long-term average for this time period. Clearly, the drought may have influenced the phenol ogical responses of the species studied. Furthermore, it is possible that data for September and October, which are lacking here, might have produced different conclusions about the seasonal patterns of branch elongation since these are the two months with the greatest rainfall in the LRGV. However, this seems unlikely because there was no correlation between rainfall and branch elongation in huisachillo. Additional information, especially from wet years, is needed to elucidate the full range of growth responses for these and other species of Acacia in the LRGV. This study points to the need for experiments on the effects of soil moisture on growth to help explain the differences observed between huisachillo and the other three Acacia species. Among sites variation is not often assessed in phenological studies. It was shown that this was an important factor in two of the five species studied. In arid environments, variation in soil moisture is common both within and between sites (Beatley 1974) and it may be the proxi¬ mate cause of variation in phenological responses in this study. Acknowledgments This paper is part of a master’s thesis by M. Eddy submitted to the Department of Biology at the University of Texas- Pan American. Thanks go to D. Howell and C. Best of the United States Fish and Wildlife Service for permits to study phenology at Santa Ana National Wildlife Refuge. Special thanks go to D. R. Rios, Sr., D. R. Rios, Jr., J. Rios and C. Eddy for field assistance. Literature Cited Beatley, J. C. 1974. Phenological events and their environmental triggers in Mojave Desert ecosystems. Ecology, 55(4): 856-863. 262 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 Bullock, S. H. & J. A. Solis-Magallanes. 1990. Phenology of canopy trees of a tropical deciduous forest in Mexico. Biotropica, 22(l):22-35. Correll, D. S. & M. C. Johnston. 1979. Manual of the vascular plants of Texas. Texas Research Foundation, Renner, Texas, 1881 pp. Eddy, M. R. & F. W. Judd. 2003. Phenology of Acacia berlandieri , A. minuata, A. rigidula, A. schqffneri, and Chloroleucon ebano in the Lower Rio Grande Valley of Texas during a drought. Southwest. Nat., 48(3) :321 -332. Everitt, J. H., D. L. Drawe & R. I. Lonard. 2002. Trees, shrubs, and cacti of South Texas. Texas Tech University Press, Lubbock, 249 pp. Lonard, R. I., J. H. Everitt & F. W. Judd. 1991. Woody plants of the Lower Rio Grande Valley, Texas. Misc. Publications, No. 7. Texas Memorial Museum. Univ. of Texas at Austin, 179 pp. New, T. R. 1984. A biology of Acacias. Oxford Univ. Press, Melbourne, Australia, 153 pp. Nilsen, E. T. & W. H. Muller. 1981. Phenology of the drought-deciduous shrub Lotus scoparius : climatic controls and adaptive significance. Ecological Monographs, 51(3): 323-341. Richardson, A. 1995. Plants of the Rio Grande Delta. Univ. Texas at Austin Press. 322 pp. + 94 color plates. Taylor, R. B., J. Rutledge & J. G. Herrera. 1999. A field guide to common South Texas shrubs. Texas Parks & Wildlife Press, Austin, 106 pp. Vora, R. 1990. Plant phenology in the Lower Rio Grande Valley of Texas. Texas J.Sci., 42(2): 137-142. FWJ at: Qudd@panam.edu TEXAS J. SCI. 56(3), AUGUST, 2004 263 GENERAL NOTES SYSTEMATIC AND ECOLOGICAL NOTES ON TUBIF1COIDES HETEROCHAETUS (OLIGOCHAET A : TUBIFICIDAE) FROM THE NECHES RIVER ESTUARY, TEXAS Richard C. Harrel Department of Biology, Lamar University Beaumont, Texas 77710 Tubificoides heterochaetus (Michaelsen 1926) is an estuarine oligo- chaete in the Family Tubificidae that has been reported in Europe and North America. North American records include Virginia, North Carolina, Florida, Louisiana and the Sabine-Neches estuary in Texas (Wern 1980; Shirley & Loden 1982; Harrel & Hall 1991; Milligan 1996; Harrel & Smith 2002). All of the publications concerning this species, except Shirley & Loden (1982), are taxonomic, and no informa¬ tion is given concerning its water quality tolerance. The taxonomic status of this species was in a state of confusion until recently. It was originally described by Michaelsen (1926) and placed in the genus Limnodrilus and later transferred to the genus Peloscolex (Lastockin 1937; Cekanovskaya 1962; Brinkhurst & Jamison 1971). Holmquist (1978) established the genus Tubificoides and in 1979 Brinkhurst & Baker transferred the marine and estuarine Peloscolex to the genus Tubificoides . Descriptions of T. heterochaetus in the literature vary from one author to another and most were based on specific lectotypes and did not con¬ sider all of the morphological variation that occur in the species. Tubificoides heterochaetus was originally described by Michaelsen (1926) as possessing a cuticular penis sheath. Brinkhurst & Jamison (1971) and Brinkhurst & Baker (1979) described it as lacking a penis sheath. Baker (1981) redescribed the species to correct this. Milligan (1996) contains the only taxonomic key, known by this author, that can be used by an applied biologist for proper identification of T. hetero¬ chaetus. However, numbers of setae per bundle, lengths of setae, and width and length of the penis sheath vary more than the scattered litera¬ ture states. Thus, an updated description of the species is given based on the literature and examination of 302 specimens collected from the Neches River estuary in Texas. The diagnostic characteristics of the 264 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 genus are based on histological genitalia structures and these are not often visible in specimens collected and prepared for ecological pur¬ poses. Thus, the description below is based on structures visible without special handling or dissection. All specimens examined were killed and preserved in formalin containing rose bengal stain, stored in 70 percent ethanol and mounted in CMC- 10 media on microscope slides. Complete specimens 5 to 9 mm long and ranged from 46 to 66 segments, but most were incomplete. Maximum width ranged from about 375 to 500 /x,m at segment X or XI. Anterior segments (I-XII) are non-papillate and distinctly wider than posterior papillate segments which are 70 to 160 /xm wide (Figure 1). The posterior papillate segments are elongate and often constricted at their base. The prostomium is conical and shorter or equal to its base at the peristomium. Anterior segments II through XII become progressively longer. Segments II through IX have secondary annulations and have 3 to 8 (mostly 5 or 6) 38 to 50 /xm long ventral and dorsal bifid setae per bundle with equal length teeth. Segment IX may have one, two or no setae. Clitellar segments X, XI and XII lack setae. A short thimble-shaped penis ranging from 36 to 37 /xm wide at the base and 37 to 46 /xm long with a thin cuticular sheath may be present in or just outside of segment XI. Only eight of 302 specimens examined had a visible penis sheath; two collected in February, two in May, one in August and three in November. Segment XIII decreases in width from anterior to posterior and scattered papille first appear. Segments behind XIII are covered with oblong papillae, but the posterior segments of complete specimens had very few or lacked papillae. Some post-clitellar segments possess 1 , 2 or occasional¬ ly 3 apparently simple pointed setae per bundle 54 to 67 /xm long. Some posterior setae are actually bifid and the upper tooth is longer and thicker than the shorter, thinner lower tooth, which is not visible unless turned just right. The posterior setae are often broken, difficult to see or absent in some segments. If all of the papillate segments of a speci¬ men are missing it could easily be misidentified as Limnodrilus. Harrel et al. (1976), Harrel & Hall (1991) and Harrel & Smith (2002) conducted three year-long surveys, with seasonal sampling of macro¬ benthos at the same seven collection stations in the highly industrialized, tidal, lower Neches River. A 1971-72 study (Harrel et al. 1976) was conducted before implementation of the Clean Water Act (CWA) when this section of the river was listed as the second most polluted waterway in the state with a permitted BOD (biochemical oxygen demand) waste load of 123,125 kg/day. Oxygen depletion (concentrations <2 mg/L) TEXAS J. SCI. 56(3), AUGUST, 2004 265 Figure 1. Tubificoides heterochaetus : (a) body, (b) tip of anterior and dorsal seta, and (c) tip of posterior weakly bifid seta. occurred at all stations and toxic pollutants were present in the water and the substrate. No T. heterochaetus were collected during this survey and they may have been excluded by the heavy load of organic and toxic pollutants in the river. During a 1984-85 study (Harrel & Hall 1991), after implementation of the first two phases of the CWA and a 93 percent reduction in the permitted BOD pollution load in the river to 8,717 kg/day, a total of 525 specimens of T. heterchaetus were collected from six of the seven sampling stations. Density at individual collection stations ranged from zero to 1196/m2 and maximum density occurred during February. Salinity ranged from <0.5 ppt to 8.5 ppt at the stations and depths where it occurred. During a 1999 study (Harrel & Smith 2002), after implementation of phase 3 of the CWA, but a 19 percent increase in the permitted BOD waste load in the river, 302 specimens of T. heterochaetus were col¬ lected at five of the seven collecting stations. Density at individual collecting stations ranged from zero to 991/m2 and maximum density occurred in November. Salinity ranged from <0.5 ppt to 13.2 ppt. During 1978 and 1979 Wern (1980) conducted monthly collections of macrobenthos from 12 stations in the Keith Lake system of marsh lakes located between the Sabine-Neches navigation channel, the Gulf of Mexico and the Intracoastal Waterway. She collected 1254 specimens 266 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 of T. heterochaetus and some specimens were collected at all 12 stations at some time during the study. Density ranged from zero to 3075/m2 and highest densities occurred during July and August, which was attributed to a reproductive event. Salinity ranged from < 0.5 to 20 ppt. Mean station bottom water dissolved oxygen concentrations ranged from 6.0 to 7. 1 mg/L. No permitted effluents were released directly into this system, but some contaminants (e.g., metals, oil and grease) were present in the sediments and were probably transported in by tidal action from the Intracoastal Waterway and the Sabine- Neches Navigation channel or from oil field activity in the area. These occurred in higher concentrations at some stations than at others, but no differences in macrobenthos distribution, abundance or diversity could be attributed to pollution. Shirley & Loden (1982) reported T. heterochaetus from the Calcasieu River estuary in Louisiana, which is located about 80 km east of the Neches River and Keith Lake estuaries. Specimens were collected from 10 of 27 stations sampled during 1974 to 1976. No specimens were collected at stations where oxygen depletion occurred and environmental parameters where they were collected included: (1) salinity - 2.3 to 14. 1 ppt, (2) oxygen percent saturation - 68 to 1 12%, (3) depth - 1.0 to 5 m, and (4) substrate - clay and silt. Density rarely exceeded 100/m2 and average density was 46.2/m2. Other Oligochaetes that occurred with T. heterochaetus in the Neches River and Keith Lake estuaries include Limnodrilus hoffineisteri , L. udekmianus , Ilyodrilus templetoni , Aulodrilus piguetti, A. pluriseta , Dero nivae, D. Jurcata, Slavinia appendiculata , Nais variabilis and Paranais grandis. All of these are considered freshwater species, except P. grandis which has been reported only from coastal Louisiana and Texas. Polychates that were common where T. heterochaetus occurred were Hobsoni grayi , Parandalia americana, Neanthes succine , Laeonereis culveri, Poly dor a socialis , Streblospio benedicti and Mediomastus calif omiensis . Tubificoides heterochaetus is a oligohaline to mesohaline estuarine species restricted to habitats where the salinity varies from <0.5 to 20 ppt, but was uncommon where salinity was <2 ppt or > 14 ppt. It occurred in sand, silt and clay substrates and at depths to at least five meters. It is tolerant to moderate pollution and cannot tolerate oxygen depletion or severe pollution. It was not collected in the Neches River estuary until after pollution abatement occurred resulting in improved water quality when it became a common component of the benthic community. TEXAS J. SCI. 56(3), AUGUST, 2004 267 Literature Cited Baker, H. R. 1981. A redescription of Tubificoides heterochaetus (Michaelsen) (Oligochaeta: Tubificidae). Proc. Biol. Soc. Wash., 94:564-568. Brinkhurst, R. O. & B. G. M. Jamieson. 1971. Aquatic Oligochaeta of the world. Univ. of Toronto Press, 860 pp. Brinkhurst, R. O. & H. R. Baker. 1979. A review of the marine Tubificidae (Oligochaeta) of North America. Can. J. Zool., 67:1553-1569. Chekanovskaya, O. V. 1962. Aquatic Oligochaeta of the USSR. Translated from Russian in 1981 by Amerind Publ. Co. Ltd., New Delhi, 513 pp. Harrel, R. C., J. Ashcraft, R. Howard & L. Patterson. 1976. Stress and community structure of macrobenthos in a Gulf Coast riverine estuary. Cont. Mar. Sci., 20:69-81. Harrel, R. C. & M. A. Hall. 1991. Macrobenthic community structure before and after pollution abatement in the Neches River estuary (Texas). Hydrobiologia, 211:241-252. Harrel, R. C. & S. T. Smith. 2002. Macrobenthic community structure before, during, and after implementation of the Clean Water Act in the Neches River estuary (Texas). Hydrobiologia, 474:213-222. Holmquist, C. 1978. Revision of the genus Peloscolex (Oligochaeta, Tubificidae). 1. Morphological and anatomical scrutiny; with discussion on the generic level. Zool. Scr. , 7:187-208. Lastockin, D. A. 1937. New species of Oligochaeta limicola in the European part of the USSR. Dokl. Akad. Nauk. SRR, 17:233-235. Michaelsen, W. 1926. Oligochaeten aus dem. Ryck bei Greifswald und von benachbarten Meeresgebieten. Mitt. Hamb. Zool. Mus. Inst., 42:21-29 Milligan, M. R. 1996. Identification manual for the aquatic Oligochaeta of Florida, Volume II Estuarine and nearshore marine oligochaetes. Bureau of Water Resources Protection, Florida Dept, of Environmental Protection, Tallahassee, 239 pp. Shirley, T. C. & M. S. Loden. 1982. The Tubificidae (Annelia, Oligochaeta) of a Louisiana estuary: ecology and systematics, with the description of a new species. Estuaries, 5:47-56. Wern, J. O. 1980. A study of the macrobenthos of the brackish lakes in Sea Rim State Park, Texas and contiguous Keith Lake. Unpublished M.S. thesis, Texas A & M Univ., College Station, 215 pp. RCH at: biology@hal.lamar.edu * * * REPRODUCTION IN THE WESTERN HOGNOSE SNAKE, HETERODON NASICUS (SERPENTES: COLUBRIDAE) FROM THE SOUTHWESTERN PART OF ITS RANGE Stephen R. Goldberg Department of Biology, Whittier College Whittier, California 90608 The western hognose snake, Heterodon nasicus ranges from southern Canada to San Luis Potosi, Mexico and southeastern Arizona to central Illinois where it frequents prairies, open woodlands and floodplains of 268 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 rivers; in the extreme western part of its range it occurs in semidesert habitats (Stebbins 2003). Most of the information on reproduction in this species was reported by Platt (1969) who studied a Kansas popula¬ tion of H. nasicus. Anecdotal information on reproduction is in: Marr (1944); Werler (1951); Moore (1953); Wright & Wright (1957); Fitch (1970); Pendlebury (1976); Tennant (1984); Lowe etal. (1986); Taggart (1992); Iverson (1995); Degenhardt et al. (1996) and Stebbins (2003). Ernst & Ernst (2003) summarized information on reproduction in H. nasicus. Information on the biology of this species is in Walley & Eckerman (1999). The purpose of this paper is to present the first reproductive data on H. nasicus from the southwestern part of its range based on a histological examination of reproductive tissues from museum specimens. Studying the reproductive cycle in different parts of a snake’s range allows one to see the extent of geographic variation in reproduction within a species. Also presented is the first histological evidence that H. nasicus females initiate yolk deposition (= secondary vitellogenesis sensu Aldridge 1979) during late summer in follicles that will be ovulated the following year. A sample of 37 specimens of H. nasicus (19 females, mean snout- vent length, SVL = 480.4 ± 71.3 SD, range: 361-613; 18 males, SVL = 324.3 mm + 34.9 SD, range: 290-390 mm) from Arizona, New Mexico and Mexico was examined from the herpetology collections of Arizona State University (ASU), the Natural History Museum of Los Angeles County, Los Angeles (LACM) and the University of Arizona, Tucson (UAZ). Most snakes (33/37) 89% were from Arizona. Snakes were collected 1949-1999. Counts were made of enlarged follicles > 8 mm length or oviductal eggs. The left testis, vas deferens and a portion of the kidney were removed from males; the left ovary was removed from females for histological examination (except for females with enlarged follicles or oviductal eggs which were counted). Tissues were embedded in paraffin and sectioned at 5 [im . Slides with tissue sections were stained with Harris’ hematoxylin followed by eosin counterstain. Testes slides were examined to determine the stage of the male cycle; ovary slides were examined for the presence of yolk deposition (secondary yolk deposition sensu Aldridge 1979). Some snakes were road-kills so not all tissues were available for examination. Number of specimens histologically examined by reproductive tissue were: testis = 18, vas deferens = 16, kidney = 18, ovary = 16. Male and female mean body sizes were compared using an unpaired Mest. TEXAS J. SCI. 56(3), AUGUST, 2004 269 Table 1 . Monthly distribution of conditions in seasonal testicular cycle of Heterodon nasicus from examination of museum specimens. Values shown are the numbers of males exhibiting each of the three conditions. Month n Regressed Recrudescence Spermiogenesis May 1 1 0 0 June 1 1 0 0 July 1 0 1 0 August 2 0 0 2 September 6 0 0 6 October 7 0 0 7 Material examined.— The following specimens of Heterodon nasicus were examined: ARIZONA: COCHISE COUNTY, (ASU 22859, LACM 109514, 115794, 145667, UAZ 9365, 24934, 24935, 24937, 24938, 24941, 24942, 35159, 39611, 39612, 39617, 39618, 40146, 41146, 41147, 41152, 43892, 46321, 46322, 46833, 48011, 50017, 51822) GRAHAM COUNTY, (ASU 7029, 22461) SANTA CRUZ COUNTY, (UAZ 40778, 43756, 43799, 50066). NEW MEXICO: HIDALGO COUNTY, (ASU 31499) LUNA COUNTY, (LACM 109527). MEXICO: CHIHUAHUA, (UAZ 39198, 39199). Testicular histology was similar to that of the two colubrid snakes, Masticophis taeniatus and Pituophis catenifer ( = P. melanoleucus ) as reported by Goldberg & Parker (1975). In the regressed testes, semi¬ niferous tubules contained spermatogonia and Sertoli cells. In recrudes¬ cence (recovery) there was renewal of spermatogenic cells characterized by spermatogonial divisions; primary and secondary spermatocytes were typically present. In spermiogenesis, metamorphosing spermatids and mature sperm were present. Testes undergoing spermiogenesis were found August-October (Table 1). Testes from the two spring males (one from May and one from June) were regressed. The smallest reproduc- tively active male (spermiogenesis in progress) measured 290 mm SVL (UAZ 46322). Platt (1969) found motile spermatozoa in cloacal smears of 3/7 (43%) H. nasicus < 300 mm SVL from Kansas. As was the case for Kansas (Platt 1969), H. nasicus from the southwestern extreme of its range undergoes a postnuptial spermatogenesis = aestival sperma¬ togenesis ( sensu Saint Girons 1982) which is completed before winter with sperm stored over winter in the vas deferens. All vasa deferentia (n = 16) contained sperm: May (1), August (2), September (6), October (7). Tubules of all kidney sexual segments, except for the one July male 17/18 (94%), were enlarged and contained secretory granules: May (1), June (1), August (2), September (6), October (7), a condition that 270 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Table 2. Monthly distribution of conditions in seasonal ovarian cycle of Heterodon nasicus from examination of museum specimens. Values shown are the numbers of females exhibiting each of the four conditions; *squashed oviductal eggs, clutch could not be counted. Month n Inactive Early yolk deposition Enlarged follicles > 12 mm length Oviductal eggs May 2 I 0 0 1 June 2 0 1 1 0 July 5 0 4 0 1* August 2 1 1 0 0 September 6 2 4 0 0 October 2 2 0 0 0 typically coincides with breeding (Saint Girons 1982). According toPlatt (1969), the principal H. nasicus mating period is in the spring, although some mating may also occur in autumn. Females were significantly larger than males (t = 8.4, df = 35, P < 0.0001). One female H. nasicus from Cochise County, Arizona with five oviductal eggs (UAZ 24941) was collected 28 May. Another, (UAZ 24938) from Cochise County, with six enlarged follicles (> 8 mm length) was collected 6 June. A third female from Cochise County collected in July (ASU 22859) contained squashed oviductal eggs that could not be counted. Females with early yolk deposition (secondary yolk deposition sensu Aldridge (1979) were found June-September (Table 2). This yolk deposition was in the form of a small band of discrete yolk granules. Because the yolk occupied only a limited area of the follicles it would have been unlikely for yolk deposition to have been completed during the current reproductive season. However, since Platt (1969) reported H. nasicus deposits eggs in August (locality not given), one must consider the possibility that in some females, yolk deposition might have been completed during the current year. How¬ ever, it appears that in at least some cases H. nasicus females initiate yolk deposition (vitellogenesis) the summer prior to completing it. For example (Fig. 1), early yolk deposition is present in UAZ 43892, a road-kill from 31 July in which the two largest follicles had lengths of 2 mm. It is doubtful that these follicles would have completed yolk deposition in the current year. These findings agree with Ernst & Ernst (2003) who reported a complement of small follicles in H. nasicus females which represent ova to be matured the following year. There was a report of an August Hypsiglena torquata female with yolk deposition in Goldberg (2001). Whether starting yolk deposition in the TEXAS J. SCI. 56(3), AUGUST, 2004 271 Figure 1. Yolk deposition in ovarian follicle of Heterodon nasicus (UAZ 43892) collected 31 July 1980. Bar represents 15 /xm. summer prior to ovulation is common in North American colubrid snakes needs to be investigated. The two clutch sizes reported herein (5, 6) are near the lower end of the ranges for H. nasicus 4-23 clutch sizes reported by Platt (1969) and 4-25 reported by Stebbins (2003). The smallest reproductively active female (yolk deposition in progress, UAZ 39611) measured 361 mm SVL. This was close to the smallest gravid H. nasicus female (SVL = 366 mm) from Kansas (Platt 1969). Eight female H . nasicus from Harvey County, Kansas deposited egg clutches from 2-23 July (Platt 1969). The presence of one Arizona female H. nasicus (UAZ 24941) with oviductal eggs on 28 May and a female from Valencia County, New Mexico that deposited eggs on 12 June (Degenhardt et al. 1996) may suggest that females from the southern portion of the range produce eggs earlier in the year than females from the northern part. Small sample sizes prevent an analysis of geographic variation in clutch sizes, although Fitch (1985) found no evidence of geographic change in clutch sizes between northern and southern populations of H. nasicus. In conclusion, there does not appear to be differences in the timing of the seasonal testicular cycle of H. nasicus between Kansas and the south- 272 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 western part of its range in that sperm produced in autumn are stored through winter in the vasa deferentia in both areas. There is a sug¬ gestion that eggs may be produced earlier in the season in the south. Additional females need to be examined to determine if this occurs. It appears that some H . nasicus females initiate yolk deposition (vitello¬ genesis) in follicles the summer before eggs are produced. Acknowledgments I thank Andrew T. Holycross (Arizona State University), George L. Bradley (University of Arizona) and David A. Kizirian (Natural History Museum of Los Angeles County) for permission to examine H. nasicus Literature Cited Aldridge, R. D. 1979. Female reproductive cycles of the snakes Arizona elegans and Crotalus viridis. Herpetologica, 35(3) :256-261 . Degenhardt, W. G., C. W. Painter & A. H. Price. 1996. Amphibians and reptiles of New Mexico, University of New Mexico Press, Albuquerque, xix + 431 pp. Ernst, C. H., & E. M. Ernst. 2003. Snakes of the United States and Canada. Smithsonian Books, Washington, ix + 668 pp. Fitch, H. S. 1970. Reproductive cycles of lizards and snakes. Misc. Publ. Mus. Nat. Hist., Univ. Kansas, 52:1-247. Fitch, H. S. 1985. Variation in clutch and litter size in New World reptiles. Misc. Publ. Mus. Nat. Hist., Univ. Kansas, 76:1-76. Goldberg, S. R. 2001. Reproduction in the night snake, Hypsiglena torquata (Serpentes: Colubridae), from Arizona. Texas J. Sci., 53(2): 107-1 14. Goldberg, S. R. & W. S. Parker. 1975. Seasonal testicular histology of the colubrid snakes, Masticophis taeniatus and Pituophis melanoleucus. Herpetologica, 3 1 (3):3 1 7-322. I verson, J. B. 1995. Heterodon nasicus (Western Hognose Snake). Reproduction. Herpetol. Rev., 26(4):206. Lowe, C. H., C. R. Schwalbe & T. B. Johnson. 1986. The venomous reptiles of Arizona. Arizona Game & Fish Department, Phoenix, ix + 115 pp. Marr, J. C. 1944. Notes on amphibians and reptiles from the central United States. Am. Midi. Nat., 32(2):478-490. Moore, J. E. 1953. The Hog-nosed snake in Alberta. Herpetologica, 9(4): 173. Pendlebury, G. B. 1976. The western hognose snake, Heterodon nasicus nasicus, in Alberta. Can. Field Natur., 90(4) :4 16-422. Platt, D. R. 1969. Natural history of the hognose snakes Heterodon platyrhinos and Heterodon nasicus. Univ. Kansas Publ., Mus. Nat. Hist., 18(4):253-420. Saint Girons, H. 1982. Reproductive cycles of male snakes and their relationships with climate and female reproductive cycles. Herpetologica, 38( 1 ) :5- 16. Stebbins, R. C. 2003. A field guide to western reptiles and amphibians, 3rd ed. Houghton Mifflin Company, Boston, Massachusetts, xiii + 533 pp. Taggart, T. W. 1992. Observations on Kansas amphibians and reptiles. Kansas Herpetol. Soc. Newsletter, 88:13-15. Tennant, A. 1984. The snakes of Texas. Texas Monthly Press, Inc., Austin, 561 pp. Walley, H. D., & C. M. Eckerman. 1999. Heterodon nasicus Baird and Girard. Western TEXAS J. SCI. 56(3), AUGUST, 2004 273 hognose snake. Cat. Amer. Amphib. Rept., 698.1-698.10. Werler, J. E. 1951. Miscellaneous notes on the eggs and young of Texan and Mexican reptiles. Zoologica, 36(l):37-48. Wright, A. H. & A. A. Wright. 1957. Handbook of snakes of the United States and Canada. Vol. I., Comstock Publ. Assoc., Ithaca, New York, xviii -I- 564 pp. SRG at: sgoldberg@whittier.edu * * * ENDOPARASITES OF THE SEQUOYAH SLIMY SALAMANDER, PLETHODON SEQUOYAH (CAUDATA: PLETHODONTID AE) , FROM MCCURTAIN COUNTY, OKLAHOMA Chris T. McAllister and Charles R. Bursey Department of Biology, Texas A&M University -Texarkana Texarkana, Texas 75505 and Department of Biology, Pennsylvania State University -Shenango Valley Campus Sharon, Pennsylvania 16146 The Sequoyah slimy salamander, Plethodon sequoyah , is a medium¬ sized plethodontid that is restricted to McCurtain County, Oklahoma (Conant & Collins 1998) and perhaps adjacent Sevier County, Arkansas (Trauth et al. 2004). This salamander occurs in upland forests where it inhabits seeps and springs hiding beneath rocks, clumps of moss, or under decaying logs. This species, one of several belonging to the P. glutinosus group 10 complex, was described by Highton (1989) as hav¬ ing a unique Mdh-2 allele that distinguishes it from 15 other species of the P. glutinosus group. In addition, this evolutionary lineage has also been recognized by Powell et al. (1998) and Duellman & Sweet (1999), and most recently was included on a list of standard and common current scientific names (Collins & Taggart 2002). Although information is available on parasites of other species within the P. glutinosus complex (Baker 1987; McAllister et al. 1993; 2002), nothing, to the authors’ knowledge, has been published on protozoan or helminth parasites of P. sequoyah. This study provides the first report of endoparasites from this host. Twenty-five juvenile and adult salamanders (mean + 1 SD snout- vent length [SVL] = 46.5 ± 14.3, range 24-74 mm) were collected by hand, two on 13 September 2002, six on 3 June 2003, and 17 on 15 April 2004 from Beaver’s Bend State Park, McCurtain County, Oklahoma (33° 7.7’N, 94° 41.9’W, elev. 153.6 m). Specimens were placed in 274 THE TEXAS JOURNAL OF SCIENCE- VOL. 56, NO. 3, 2004 damp collecting bags on ice and returned to the laboratory within 24h for processing. Specimens were killed by prolonged immersion in a dilute Chloretone® solution. For necropsy, a midventral incision was made and the entire gastrointestinal tract, liver, gallbladder, spleen and gonads were examined for helminths. Blood smears were taken from the exposed heart and stained with DifQuick. Feces from the colon and rectum were collected and placed in individual vials containing tap water supplemented with antibiotic (100 I. U./mL penicillin-G 100 pgt mL streptomycin) and examined directly without sucrose flotation by microscopy for coccidia. The integument was examined closely for intradermal mites (Hannemania) . Tapeworms were relaxed in cold tap water, fixed in 70% ethanol, stained with Semichon’s acetocarmine and mounted entire with Permount®. Nematodes were placed in a drop of glycerol on microscopic slides and identifications were made from these temporary mounts. Flelminth voucher specimens were deposited in the United States National Parasite Collection (USNPC), Beltsville, Maryland, USA, and the Harold W. Manter Laboratory of Parasitology, Lincoln, Nebraska, USA: Cepedietta michiganensis (HWML 45996), Cylindrotaenia idahoensis (USNPC 94810, 95245), Mesocestoides sp. (USNPC 94811), Batracholandros magnavulvaris (USNPC 94812, 95246), Cosmocercoides variabilis (USNPC 94813). Host voucher specimens were deposited in the Arkansas State University Museum of Zoology (ASUMZ 27250, 27920-27924) and University of Oklahoma Museum of Natural History (OMNH 39181). Eighteen of 25 (72%) of the P. sequoyah were infected with one of five parasite species, including one (4%) with Cepedietta michiganensis in the small intestine, seven (28%) with Cylindrotaenia idahoensis (mean intensity 6.3, range 1-19) in the small intestine, two (8%) with Meso¬ cestoides sp. in the mesenteries and peritoneal cavity, three with Cosmocercoides variabilis (mean intensity 9.7, range 5-17, 13 females, 16 males) in the rectum, and eight (32%) each with a single female of Batracholandros magnavulvaris in the rectum; six salamanders (24%) harbored multiple infections. Blood smears were negative for hemato- zoa, the feces did not contain coccidia, and none of the salamanders were infested with Hannemania . The astomatous ciliate, C. michiganensis has been reported previously from various salamanders and frogs (Joy & Tucker 2001; McAllister & Bursey 2004), including the Fourche Mountain salamander, P. fourchen- sis , western slimy salamander, P. albagula, and Rich Mountain sala- TEXAS J. SCI. 56(3), AUGUST, 2004 275 mander, P. ouachitae from Arkansas (Winter et al. 1986; McAllister et al. 1993; 2002), and the southern redback salamander, P. serratus from Oklahoma (McAllister et al. 2002). This study represents the first report of this protist in P. sequoyah. The cyclophyllidean tapeworm, C. idahoensis was originally described from the Coeur d’Alene salamander, P. idahoensis from Idaho (Waitz & Mehra 1961). Since then, this cestode has been reported in Jordan’s redcheek salamander, P. jordani from North Carolina (Dyer 1983; Jones 1987), the western redback salamander, P. vehiculum from Oregon (Panitz 1969), and the Caddo Mountain salamander, P. caddoensis, P. ouachitae and P. serratus from Arkansas and Oklahoma (McAllister et al. 2002). This study documents a new host record for the parasite in P. sequoyah. The cestode, Mesocestoides sp. is an enigmatic tapeworm whose complete life cycle is unknown. The initial report in salamanders of the world was by McAllister et al. (1995) who reported this parasite in eight of 41 (20%) Ouachita dusky salamanders, Desmognathus brimleyorum from Arkansas. This study reports a second salamander host for this tapeworm. This cestode has also been previously reported from various anurans (McAllister & Conn 1990). The ascarid nematode, C. variabilis has been commonly reported from both amphibians and reptiles in the United States and Canada (summarized by McAllister & Bursey 2004). This parasite (as Oxy- somatium sp.) has also been previously reported from Oklahoma in bullfrogs, Rana catesbeiana (Trowbridge & Hefley 1934); however, this study reports a new host for this roundworm. The nematode, B . magnavulvaris is a pinworm with a direct life cycle that exhibits little host specificity. It has been previously reported in P. caddoensis , P. fourchensis, P. ouachitae , P. serratus and D. brimleyorum in Arkansas and Oklahoma (Winter et al. 1986; McAllister et al. 1995; 2002). In addition, this parasite has a wide geographic range as it has been reported in salamanders of the genera Aneides, Desmognathus , Eurycea, Leurognathus , Notopthalmus from California, Illinois, Michigan, New Hampshire, North Carolina, Pennsylvania, Tennessee, Virginia and West Virginia (see Joy & Tucker 2001 for summation). Plethodon sequoyah represents a new host for this parasite. 276 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 In summary, although no new geographic records are documented, this study provides the first report of endoparasites from P. sequoyah. Several parasite species reported herein are shared with other Plethodon sp., and as in previous surveys on salamanders, this limited data supports Aho’s (1990) suggestion that the parasite community structure is depauperate and noninteractive. Acknowledgments The senior author thanks Joel Johnson (Univ. Oklahoma), Boy Scout Troop 1, Indian Nations Council, Tulsa, and the Spring 2004 TAMU-T Herpetology class (especially Zach Ramsey) for assistance in collecting, and the Oklahoma Department of Wildlife Conservation for Scientific Collecting Permit Nos. 3172 and 3376. We also thank Drs. Dan Brooks and Bruce Conn for examining the Mesocestoides sp. Literature Cited Aho, J. M. 1990. Helminth communities in amphibians and reptiles: comparative ap¬ proaches to understanding patterns and processes. Pp. 157-195, in Parasite Communities: Patterns and Processes (G. W. Esch, A. O. Bush, and J. M. Aho, eds), Chapman and Hall, New York, 304 pp. Baker, M. R. 1987. Synopsis of the Nematoda parasitic in amphibians and reptiles. Mem. Univ. Newfoundland Occas. Pap. Biol., 11:1-325. Collins, J. T. & T. W. Taggart. 2002. Standard common and current scientific names for North American amphibians, turtles, reptiles & crocodilians. 5th Edition. Center for North American Herpetology, Lawrence, Kansas, 44 pp. Conant, R. & J. T. Collins. 1998. A field guide to reptiles and amphibians of eastern and central North America. 3rd Edition, expanded. Houghton Mifflin, Boston, Massachusetts, 616 pp. Duellman, W. E. & S. Sweet. 1999. Pp. 31-109, in Patterns of Distribution of Amphibi¬ ans: A Global Perspective (W. E. Duellman, W. E. ed), Johns Hopkins University Press, Baltimore, viii + 633 pp. Dyer, W. G. 1983. A comparison of the helminth fauna of two Plethodon jordani popu¬ lations from different altitudes in North Carolina. Proc. Helm. Soc. , Washington 50:257- 260. Highton, R. 1989. Part 1. Geographic protein variation. Pp. 1-78, in Biochemical Evolu¬ tion in the Slimy Salamanders of the Plethodon glutinosus Complex in the Eastern United States (R. Highton, G. C. Maha, and L. R. Maxson, eds). Illinois Biol. Monogr., 57:1- 153. Jones, M. K. 1987. A taxonomic revision of the Nematotaeniidae Liihe, 1910 (Cestoda: Cyclophyllidea). Syst. Parasitol., 10:165-245. Joy, J. E. & R. B. Tucker. 2001. Cepedietta michiganensis (Protozoa) and Batracholandros magnavulvaris (Nematoda) from plethodontid salamanders in West Virginia, U.S.A. Comp. Parasitol., 68:185-189. McAllister, C. T. & C. R. Bursey. 2004. Endoparasites of the dark-sided salamander, Eurycea longicauda melanopleura, and the cave salamander, Eurycea lucifuga (Caudata: Plethodontidae), from two caves in Arkansas, U.S.A. Comp. Parasitol., 71:61-66. TEXAS J. SCI. 56(3), AUGUST, 2004 277 McAllister, C. T. & D. B. Conn. 1990. Occurrence of Mesocestoides sp. tetrathyridia (Cestoidea: Cyclophyllidea) in North American anurans (Amphibia). J. Wild. Dis., 540- 543. McAllister, C. T., S. J. Upton & S. E. Trauth. 1993. Endoparasites of western slimy salamanders, Plethodon albagula (Caudata: Plethodontidae), from Arkansas. J. Helm. Soc. Washington, 60:124-126. McAllister, C. T., C. R. Bursey & S. E. Trauth. 2002. Parasites of four species of endemic Plethodon from Arkansas and Oklahoma. J. Arkansas Acad. Sci., 56:239-242. McAllister, C. T., C. R. Bursey, S. J. Upton, S. E. Trauth & D. B. Conn. 1995. Parasites of Desmognathus brimleyorum (Caudata: Plethodontidae) from the Ouachita Mountains of Arkansas and Oklahoma. J. Helm. Soc. Washington, 62:150-156. Panitz, E. 1969. Helminth parasites of salamanders of the genus Plethodon in western Oregon. Canadian J. Zool., 47:157-158. Powell, R., J. T. Collins & E. D. Hooper, Jr. 1998. A key to the amphibians and reptiles of the continental United States and Canada. University Press of Kansas, Lawrence, Kansas, vi + 131 pp. Trauth, S. E., H. W. Robison & M. V. Plummer. 2004. The amphibians and reptiles of Arkansas. University of Arkansas Press, Fayetteville, Arkansas, xv + 421 pp. Trowbridge, A. H. & H. M. Hefley. 1934. Preliminary studies of the parasite fauna of Oklahoma anurans. Proc. Oklahoma Acad. Sci., 14:16-19. Waitz, J. A. & K. N. Mehra. 1961. Baerietta idahoensis n. sp. a nematotaeniid cestode from the intestine of Plethodon vandykei idahoensis from northern Idaho. J. Parasitol., 47:806-808. Winter, D. A., W. M. Zawada & A. A. Johnson. 1986. Comparison of the symbiotic fauna of the family Plethodontidae in the Ouachita Mountains of western Arkansas. Proc. Arkansas Acad. Sci., 40:82-85. CTM at: chris.mcallister@tamut.edu 278 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 3, 2004 Plan Now for the 108th Annual Meeting of the Texas Academy of Science March 3 - 5, 2005 University of Texas-Pan American Program Chair Damon Waitt Lady Bird Johnson Wildflower Center 4801 LaCrosse Ave. Austin, Texas 78739 Phone: 512.292.4200 E-mail: dwaitt@wildflower.org Local Host Hudson DeYoe Dept, of Biology and Center for Subtropical Studies University of Texas-Pan American 1201 West University Dr. Edinburg, Texas 78541 Phone: 956.381.3538 FAX: 956.381.3657 E-mail: hdeyoe@panam.edu For additional information relative to the Annual Meeting, please access the Academy homepage at: www. texasacademyofscience. org Future Academy Meetings 2006 - Lamar University 2007 - Baylor University THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 3, 2004 279 IN RECOGNITION OF THEIR ADDITIONAL SUPPORT OF THE TEXAS ACADEMY OF SCIENCE DURING 2004 Patron Members Ali R. Amir-Moez Deborah D. Hettinger Don W. Killebrew David S. Marsh John Sieben Ned E. Strenth Sustaining Members James Collins Dovalee Dorsett Stephen R. Goldberg Donald E. Harper, Jr. Norman V. Horner Michael Looney Sammy M. Ray Fred Stevens William F. Thomann Milton W. Weller Supporting Members David A. Brock Frances Bryant Edens Michael Halbouty Patrick Donegan Hollis George D. McClung Jimmy T. Mills Jim Neal Gary Powell Judith Schiebout Lynn Simpson THE TEXAS ACADEMY OF SCIENCE www . texasacademy ofscience . org Membership Information and Application MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. (Please print or type) Name Last First Middle MailingAddress City State Zip Ph FAX E-mail: Regular Member $30.00 Supporting Member $60.00 Sustaining Member $100.00 Patron Member $150.00 Student-Undergraduate $15.00 Student- Graduate $15.00 Joint $35.00 Emeritus $10.00 Corporate Member $150.00 Affiliate $5.00 (list name of organization) Contribution AMOUNT REMITTED SECTIONAL INTEREST AREAS: A. Biological Science B. Botany C. Chemistry D. Computer Science E. Conservation & Management F. Environmental Science G. Freshwater & Marine Sciences H. Geography I. Geology K. Mathematics L. Physics M. Science Education O. Systematics & Evolutionary Biology P. Terrestrial Ecology Q. Anthropology R. Threatened or Endangered Species Please indicate your Sectional interest(s) below: 1. _ 2. _ 3. _ Send Application Form and Check or Money Order to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 Please photocopy this Application Form THE TEXAS ACADEMY OF SCIENCE, 2004-2005 OFFICERS John A. Ward, Brook Army Medical Center Damon E. Waitt, Lady Bird Johnson Wildflower Center David S. Marsh, Angelo State University John T. Sieben, Texas Lutheran University Fred Stevens, Schreiner University Deborah D. Hettinger, Texas Lutheran University Ned E. Strenth, Angelo State University Robert J. Edwards, University of Texas-Pan American James W. Westgate, Lamar University Sandra S. West, Southwest Texas State University DIRECTORS President : President Elect : Vice-President : Immediate Past President: Executive Secretary : Corresponding Secretary: Managing Editor: Manuscript Editor: Treasurer: AAAS Council Representative: 2002 Sushma Krishnamurthy, Texas A&M International University Raymond D. Mathews, Jr., Texas Water Development Board 2003 Hudson R. DeYoe, University of Texas-Pan American Cynthia Contreras, Texas Parks and Wildlife Department 2004 Benjamin A. Pierce, Baylor University Donald L. Koehler, Balcones Canyonlands Preserve Program SECTIONAL CHAIRPERSONS Anthropology: Roy B. Brown, Instituto Nacional de Antropologia y Historia Biological Science: Francis R. Horne, Texas State University Botany: Herbert D. Grover, Harden-Simmons University Chemistry: Mary Kipecki-Fjetland, St. Edward’s University Computer Science: Laura J. Baker, St. Edwards University Conservation and Management: Felipe Chavez-Ramirez, Platte River Whooping Crane Trust Environmental Science: William Thomann, University of the Incarnate Word Freshwater and Marine Science: Sharon Conry, Baylor University Geology and Geography: Carol Thompson, Tarleton State University Mathematics: Hueytzen J. Wu, Texas A&M University-Kingsville Physics: David Bixler, Angelo State University Science Education: Jimmy Hand, Austin, Texas Systematics and Evolutionary Biology: Kathryn Perez, University of Alabama Terrestrial Ecology: Jerry Cook, Sam Houston State University Threatened or Endangered Species: Alice L. Hempel, Texas A&M University-Kingsville COUNSELORS Collegiate Academy: William J. Quinn, St. Edward’s University Junior Academy: Vince Schielack, Texas A&M University Nancy Magnussen, Texas A&M University PERIODICAL POSTAGE PAID AT LUBBOCK TEXAS 79402 RETURN SERVICE REQUESTED THE TEXAS JOURNAL OF SCIENCE PrinTech, Box 43151 Lubbock, Texas 79409-3151 0343 u) THE JOURNAL OF SCIENCE GENERAL INFORMATION MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. For more information regarding membership, student awards, section chairs and vice-chairs, the annual March meeting and author instructions, please access the Academy’s homepage at: www . texasacademy ofscience . org Dues for regular members are $30.00 annually; supporting members, $60.00; sustaining members, $100.00; patron members, $150.00; associate (student) members, $15.00; family members, $35.00; affiliate members, $5.00; emeritus members, $10.00; corporate members, $250.00 annually. Library subscription rate is $50.00 annually. The Texas Journal ofScience is a quarterly publication of The Texas Academy of Science and is sent to most members and all subscribers. Payment of dues, changes of address and inquiries regarding missing or back issues should be sent to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 E-mail : FStevens@schreiner . edu The Texas Journal of Science (ISSN 0040-4403) is published quarterly at Lubbock, Texas, U.S.A. Periodicals postage paid at San Angelo, Texas and additional mailing offices. POSTMASTER: Send address changes and returned copies to The Texas Journal of Science, Dr. Fred Stevens, CMB 5980, Schreiner University, Kerrville, Texas 78028-5697, U.S.A. The known office of publication for The Texas Journal of Science is the Department of Biology, Angelo State University, San Angelo, Texas 76909; Dr. Ned E. Strenth, Managing Editor. COPYRIGHT POLICY All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, recording or otherwise, without the prior permission of the Managing Editor of the Texas Journal ofScience. THE TEXAS JOURNAL OF SCIENCE Volume 56, No. 4 November, 2004 CONTENTS Potential Causes of a Decline in American Beech (Fag us grandifolia Ehrh.) in Wier Woods, Texas. By S. Jha, P. A. Harcombe, M. R. Fulton and 1. S. Elsik . 285 Comparative Analysis of Growth and Mortality Among Saplings in a Dry Oak-Pine Forest in Southeast Texas. By Jie Lin, Paul A. Harcombe, Mark R. Fulton and Rosine W. Hall . 299 Structural Changes after Prescribed Fire in Woody Plant Communities of Southeastern Texas. By Changxiang Liu, Paul A. Harcombe and Robert G. Knox . 319 Growth of Chinese Tallow Tree ( Sapium sebiferum ) and Four Native Trees under Varying Water Regimes. By Bradley J. Butterfield, William E. Rogers and Evan Siemann . . . 335 Effects of Temperature and Mulch Depth on Chinese Tallow Tree (Sapium sebiferum ) Seed Germination. By Candice Donahue, William E. Rogers and Evan Siemann . 347 The Effect of Mycorrhizal Inoculum on the Growth of Five Native Tree Species and the Invasive Chinese Tallow Tree (Sapium sebiferum). By Somereet Nijjer, William E. Rogers and Evan Siemann . 357 Characterization of Arthropod Assemblage Supported by the Chinese Tallow Tree (Sapium sebiferum) in Southeast Texas. By Maria K. Hartley, Saara DeWalt, William E. Rogers and Evan Siemann . 369 Diel Activity Patterns of the Louisiana Pine Snake (Pituophis ruthveni) in Eastern Texas. By Marc J. Ealy, Robert R. Fleet and D. Craig Rudolph . 383 Arboreal Behavior in the Timber Rattlesnake, Crotalus horridus, in Eastern Texas. By D. Craig Rudolph, R. R. Schaefer, D. Saenz and R. N. Conner . 395 Nesting Habitat of Eastern Wild Turkeys (Meleagris gallopavo sylvestris) in East Texas. By Bobby G. Eichler and R. Montague Whiting, Jr . 405 The Red-Cockaded Woodpecker: Interactions with Fire, Snags, Fungi, Rat Snakes and Pileated Woodpeckers. By Richard N. Conner, Daniel Saenz and D. Craig Rudolph . 415 Feeding Habits of Songbirds in East Texas Clearcuts During Winter. By Donald W. Worthington, R. Montague Whiting, Jr. and James G. Dickson . . . 427 Index to Volume 56 (Subject, Authors & Reviewers) . 441 Recognition of Member Support . 453 Membership Application . 454 Postal Notice . 455 THE TEXAS JOURNAL OF SCIENCE EDITORIAL STAFF Managing Editor: Ned E. Strenth, Angelo State University Manuscript Editor: Robert J. Edwards, University of Texas- Pan American Associate Editors for this Issue: Paul Harcombe, William Marsh Rice University Craig Rudolph, U.S. Forest Service Evan Siemann, William Marsh Rice University Manuscripts intended for publication in the Journal should be submitted in TRIPLICATE to: Dr. Robert J. Edwards TJS Manuscript Editor Department of Biology University of Texas-Pan American Edinburg, Texas 78541 r edwards@panam . edu Scholarly papers reporting original research results 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 Manuscript Editor at the above address. They are also available on the Academy’s homepage at: www . texasacademy ofscience . org AFFILIATED ORGANIZATIONS American Association for the Advancement of Science, Texas Council of Elementary Science Texas Section, American Association of Physics Teachers Texas Section, Mathematical Association of America Texas Section, National Association of Geology Teachers Texas Society of Mammalogists The 3rd Big Thicket Science Conference, "Biodiversity and Ecology of the West Gulf Coastal Plain Landscape", was held October 9-11, 2003 in Beaumont, Texas. The Big Thicket is a biologically rich area within the West Gulf Coastal Plain where the influences of southeastern swamps, eastern deciduous forests, central plains, pine savannas and xeric sandhills meet and intermingle. The region provides habitat for many rare species and favors unusual combinations of plants and animals. The purpose of the Big Thicket Science Conference is to highlight the results of recent ecological research and conservation efforts to understand, manage and restore the unique biological diversity of the Big Thicket and surrounding West Gulf Coastal Plain. The event brought together a diverse group of individuals representing government, academia, conservation organizations, private industry and local residents. It took the efforts of many people to produce this document. Numerous people reviewed the manuscripts included in this volume. We appreciate their input that greatly improved the quality of the manuscripts and their willingness to review manuscripts in a short period of time. We thank the contributing authors for their patience with the editorial process. We thank the Texas Academy of Science for their support of this project. We are particularly grateful to Ned Strenth for his assistance as managing editor. We hope this publication increases our understanding of the biological resources of this region. The 4th Big Thicket Science Conference is scheduled for Fall 2007. Information regarding this event will be forwarded to registered participants of the 3rd conference. Other interested parties may contact: Chief of Resources Management, Big Thicket National Preserve, 3785 Milam, Beaumont, Texas 77701 (phone: 409.839.2689). Big Thicket Science Conference Publication Committee Paul Harcombe, William Marsh Rice University Maxine Johnston, Big Thicket Association Wendy Ledbetter, The Nature Conservancy Ricky Maxey, Texas Parks and Wildlife Craig Rudolph, U.S. Forest Service Evan Siemann, William Marsh Rice University Program Committee Judy Aronow, Lamar University Center for the Study of the Big Thicket James Barker, Big Thicket National Preserve Carroll Cordes, U.S. Geological Survey Deanna Fusco, Big Thicket National Preserve Cathy Guivas, Big Thicket National Preserve Paul Harcombe, William Marsh Rice University Chuck Hunt, Big Thicket National Preserve Fulton Jeansonne, Big Thicket National Preserve Maxine Johnston, Big Thicket Association Wendy J. Ledbetter, The Nature Conservancy Ricky Maxey, Texas Parks and Wildlife Department Kim McMurray, Entergy, Inc. Jim Neal, U.S. Fish and Wildlife Service Jeff Pittman, Lamar University Craig Rudolph. USD A Forest Service Evan Siemann, William Marsh Rice University Underwritten by: Entergy, Inc. Sponsored By: Beaumont Convention & Visitors Bureau Big Thicket Association ExxonMobil Corporation Lamar University’s Center for the Study of the Big Thicket National Park Service (Big Thicket National Preserve) The Nature Conservancy Texas Parks & Wildlife Department USD A Southern Research Station USGS National Wetlands Research Center U.S. Fish & Wildlife Service William Marsh Rice University TEXAS J. SCI. 56(4):285-298 NOVEMBER, 2004 POTENTIAL CAUSES OF A DECLINE IN AMERICAN BEECH ( FAGUS GRANDIFOLIA EHRH.) IN WIER WOODS, TEXAS S. Jha, P. A. Harcombe, M. R. Fulton and I. S. Elsik Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77005 Abstract.— In a mature southern mixed hardwood stand in Hardin County, Texas, American beech ( Fagus grandifolia) declined in basal area by 38% between 1985 and 2001, and 59% of the largest trees (>45 cm dbh) died (4.10%/yr). The mortality rate was nearly triple that of understory trees (4.5-14cm dbh) (1.13%/yr). Mortality increased in 1987 following a hurricane, and remained high for the 15-year duration of the study. Dead trees were aggregated in space, causing the population to change in distribution from regular to random. Evidence for pathogen damage was mostly circumstantial. Night-time tempera¬ tures, to which beech is susceptible, have been increasing over the last 20 years. No single factor (increasing temperatures, moderate hurricane damage, or pathogens) alone appears sufficient to explain the decline of large American beech trees in this forest over the past 20 years. Instead, a combination of factors seems most likely. Southern mixed forests contain an unusually high diversity of woody plant species (Marks & Harcombe 1975). In general, they exhibit a successional trend towards dominance by beech and magnolia (Gano 1917; Kurz 1944; Glitzenstein et al. 1986), though the mixed species nature of these southern hardwood forests is hypothesized to result from complex disturbance regimes (Glitzenstein et al. 1986; Platt & Schwartz 1990). In addition to its shade tolerance and longevity, American beech may also be resistant to exogenous damage caused by tropical storms (Batista et al. 1998; Batista & Platt 2003). Consequently, beech-domi¬ nated forests might be expected to be relatively stable. However, a beech population in southeast Texas showed substantial decline between 1987 and 1999 (Harcombe et al. 2002). In this paper, the decline is analyzed and several hypotheses are tested to explain it. One possible explanation could be a hurricane which hit the site in 1986. Within southern mixed hardwood forests, hurricanes can slow the replacement of shade-intolerant species by shade tolerant species (Glitzenstein et al. 1986; Cain & Shelton 1995; Arevalo et al. 2000). Peters & Poulson (1994) suggested that hurricanes may limit beech dominance in beech forests around the world. However, there is also contrary evidence; hurricanes did not strongly reduce American beech growth rates in northern Florida (Batista et al. 1998) or in east Texas (Bill 1995). 286 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 Another possible explanation involves climate change. Recent model¬ ing research indicates that American beech distribution within the United States is governed by temperature, precipitation, soil, and elevation- related variables (Iverson & Prasad 1998). Box et al. (1993) define the climatic space corresponding to the geographic species range as a "climate envelope." The envelope for American beech involves, among other measures, a relatively moist climate and maximum daily tempera¬ tures between 17°C and 29°C. Davis & Zabinski (1992) modeled the distribution of American beech with respect to temperature and predicted that if temperature increased, species at their southern range limits, including American beech, would exhibit immediate declines in seedling density and an eventual decline of canopy trees after a few decades of warming. Other studies in North America have also suggested that increasing summer temperature significantly reduces American Beech growth (Fritts 1958; Tubbs & Houston 1990, Tardiff et al. 2001). Finally, a recent dendroecological study in Texas showed high sensitivity of American beech to temperature and precipitation between the summer months of May and July (Cook et al. 2001). Particularly in east Texas, where American beech reaches its southwestern range limit, increasing summer temperatures may exceed heat tolerance limits of the species, affecting growth and mortality. American beech is also vulnerable to sucking insects, decay fungi, and pathogens (Tubbs & Houston 1990). The most notorious example is Beech Bark Disease, which has affected American beech trees in the northeastern United States (Ehrlich 1934; Houston et al. 1979). In the southern United States, the bark canker fungus Hy poxy Lon atropunctatum has been documented on American beech trees (Thompson 1963; Pase 2002). Hy poxy Ion first affects the cambium; it is thought to be triggered by low moisture in the xylem and can take three to four years to kill a tree (Pase 2002). Aphid infestation can also damage beech; it has recently been documented in east Texas (Hemmingsen 2002; Siemann & Rogers 2003). A variety of abiotic and biotic factors clearly influence American beech populations. Declines of woody species could also be related to species population structure and natural population dynamics (reviewed in Mueller- Dombois 1992). Furthermore, population dynamics may be strongly influenced by the series of stresses each individual in a popula¬ tion experiences. Manion (1981) classified stresses into two categories: "predisposing factors," which are long term stresses, and "inciting factors, " which are short-duration stresses. Pederson (1998) showed that trees with a negative response to a prior stress were more likely to have JHA ET AL. 287 a negative response to a subsequent stress. Thus, tree mortality can be the result of a variety of factors that act over a lifetime, and growth and mortality may be synchronized in a population that has a history of stresses. The hurricane, changing climate conditions, and pathogens could be acting together or separately, along with stress history or population structure, to cause the decline of American beech in Wier Woods. In this paper, these hypotheses are examined by analyzing 22 years of data on spatial and temporal variation in beech growth, recruitment, and mortality. Study Site The study site is a 4 ha plot in Wier Woods Preserve (The Nature Conservancy), located about 16 km north of Beaumont, Hardin County (30° 16’ N, 94° 12’ W), Texas (Figure 1). Wier Woods is located 140 km east of the western range limit of the species (McLeod 1975), just 5 km north of the southern range limit for American beech (Little 1971). The site is part of the Big Thicket (Marks & Harcombe 1981), a 2500 km2 forested region located 50-100 km inland from the Gulf of Mexico. The soil is a siliceous, thermic, Susquehanna fine sandy loam (Deshotels 1978). Average annual temperature is 20.4°C, with a long growing season (approx. 240 days) from March to November (Harcombe et al. 2002). Species composition in the Wier Woods is typical of southern mesic forests (Quarterman & Keever 1962; Blair & Brunett 1976; Glitzenstein, et al. 1986). The important species in Wier Woods include loblolly pine (Pinus taeda), water oak ( Quercus nigra), American beech (Fagus grandifolia) , southern magnolia (Magnolia grandiflora) , and white oak (Quercus alba) (Harcombe et. al. 1998). Glitzenstein et al. (1986) found that disturbance at Wier Woods may accelerate early successional stands of pine and oak towards beech and magnolia domi¬ nance and also re-initiate new regeneration of pine and oak in areas currently dominated by beech and magnolia. Harcombe et al. (2002) noted the rapid decline in basal area of beech, in spite of increases in most other species and an overall increase in stand basal area. On June 26, 1986, Hurricane Bonnie, with winds estimated at 120 km/hr, passed over the site (Doyle & Girod 1997; NOAA 1986). Methods Data for this research are from a permanent sample plot of approxi¬ mately 4 ha. An irregular polygon was divided into 101 contiguous 20 by 20 m cells, and stems with DBH > 4.5 cm were tagged and mapped; species identity and DBH was measured for each stem. All 288 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 4, 2004 Fig. 1. American beech distribution in southeastern U.S.A. (Little 1971). tagged trees were measured in May or June of 1980, 1982, 1985, 1987, 1989, 1992, 1995, 1998, and 2001. All stems that had reached 4.5 cm DBH since the last measurement (ingrowth) were tagged and mapped. Trees with missing or anomalous DBH values were assigned an interpolated value. Trees missed in earlier surveys were assigned DBH values by back projection, based on calculated mean growth rates for the appropriate time period, species, size, and class (Bill 1995). All of the interpolated values were used in calculations of basal area and density, but not in calculations of growth rates, even though they had only a small effect on growth values. Data were analyzed using SAS (SAS Institute) or Microsoft Excel (ver. MS2000). Average annual growth rates were calculated by dividing change in DBH by the number of years between measurements. Mortality surveys were conducted annually, and percent mortality was calculated as the number of individuals found dead in a single year divided by the number of individuals in the living population in the previous year. The possible existence of a temporal pattern in mortality (as opposed to a random fluctuation) was evaluated by comparing two models of large beech mortality using the Akaike Information Criteriron or AIC (Burnham & Anderson 2002). The AIC incorporates both the likelihood of the data given in the model, and the number of free parameters in the model; the model with the lowest AIC is considered to be the best supported by the data. The first model assumed a constant probability of mortality, with the average mortality rate as the one free parameter. The second model approximated mortality by a step JHA ET AL. 289 function, with one probability before and one probability after the step. For this model there were three free parameters: the first and second mortality probabilities and the time of the step between the two. Mortality as a function of DBH was also predicted with logistic re¬ gression using the Weibull distribution (Antle & Wain 1988). Models were fitted for the intervals six years before Hurricane Bonnie (198 1 - 1986), the year immediately after the hurricane (1987), six years after the hurricane (1987-92), and the 14 years after the hurricane (1987- 2001). The Clark- Evans Nearest Neighbor Test (Clark & Evans 1954) was used to test for aggregation of the American beech population. Because distributions at any time are highly influenced by the prior population distribution, a randomization test was also performed to determine whether mortality was aggregated given the initial spatial distribution. The test calculates the mean nearest neighbor distances for dead and live trees where the null hypothesis takes the initial spatial distribution of the population as a given. The null distribution is created by shuffling the identity of living and dead trees 1000 times. Meteorological data (NOAA 2002) were obtained for Liberty, Texas, 56 km west of the study site. This station provided the longest temporal record within a reasonable distance. Less-extensive records from the Beaumont Research Station, 16 km from the study site, were also examined; they indicated similar weather patterns. Mean temperature of the warmest month was calculated by averaging the daily minimum and the daily maximum for the summer months and then averaging the daily averages to get monthly means. The century average for the August mean temperature was also calculated. To calculate the mean temperature of the coldest month, this same procedure was repeated for the month of January. Night-time temperature of the coldest month was approximated by averaging the daily minimum temperatures in the month of January. An annual moisture index (annual precipitation/ potential evapotranspiration; Box et al. 1993) was calculated. Potential evapotranspiration for Wier Woods was obtained from Caird (1996). Average precipitation of the driest month was calculated by summing daily precipitation per month, calculating the century average for each month, and then choosing the driest month. For the age distributions of American Beech individuals at Wier Woods, data were used from 136 tree cores extracted from a random sample of the tree population as reported by Glitzenstein (1984; Glitzenstein et al. 1986). 290 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 160 4.5-14cm □ individuals in 1980 m individuals in 2001 ■ 15-29cm 30-45cm > 45cm DBH class Fig. 2. Number of individuals in 1980 and 2001 by DBH class. Results Population trends.— The original population of American beech trees tagged in 1980 consisted of 153 individuals > 4.5 cm DBH. Basal area of beech was 4.0 m2/ha, 11.5% of stand basal area. By 2001, basal area of beech had declined to 2.7 m2/ha in spite of an overall increase in stand basal area from 34.7 to 36.4 m2/ha. The decline was strongly concentrated in the largest trees; the number of individuals > 30 cm dropped and the number of smaller individuals rose between 1980 and 2001 (Fig. 2). The > 45 DBH class experienced mortality at a rate of 4.10%/year, more than double the rate for the smaller size classes (Table 1). Mortality of largest trees was consistently higher than that of the whole population across the 20-year study period (Fig. 3). The step- function mortality model was significantly better in predicting large tree mortality than was the model assuming a constant probability of mortality over time (AIC of 257 vs. 267); the best step function was the one in which the increase occurred in 1986. Average tree growth rates were variable. Dying large trees did not show significantly lower growth before mortality than surviving large trees (FU66 = .50, />=0.48). Hurricane Beech mortality was high in 1987, the survey after the storm. Nevertheless, this was not the highest yearly mortality rate JHA ET AL. 291 Table 1. Percent mortality of American beech by size class six years before Hurricane Bonnie (1981-1986), one year after the hurricane (1987), six years after the hurricane (1987-92), the average across the study period (1981-2001), and the maximum annual rate. Size Class (cm DBH) Six Years Before One Year After Six Years After Average Highest 4.5 - 14 0.90 0.00 0.97 1.13 4.08 (1996) 15 - 30 0.00 0.00 1.85 1.55 10.00 (1992) 30 - 45 0.52 6.90 1.67 1.70 14.81 (1995) >45cm 1.16 7.80 5.92 4.10 13.89 (1990) total 0.77 3.52 2.50 2.03 5.11 (1990) □ % mortality total population years Fig. 3. Percent mortality of total beech population (dark bars) and large individuals of beech (light bars). across the long-term study (Fig. 3, Table 1). Rather, Wier Woods lost most of its large trees gradually between 1987 and 2001. Mortality in the storm interval itself was not significantly different from the six years before the storm (x2 test; df— 1, P=. 282); however, mortality was significantly greater after the storm than before for both the six-yr interval (x2 test, df= 1, P< .001) and the entire post-hurricane interval (x2 test, df— 1 , P— .031). Logistic regressions of mortality versus DBH show the same pattern, i.e., that there was a significantly higher probability of mortality after the storm interval than before (for both the six-year and 15-year intervals) but not between the pre-storm interval and the storm interval itself (Fig. 4). Climate.— After 1972, minimum temperature in August for Liberty, Texas rose steadily through 2001. In fact, 21 of last 23 summers ex¬ ceeded the century average for minimum summer temperature, while precipitation showed no trends (Fig. 5). Wier Woods was above the 292 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 4, 2004 co •e o E <4— o >» j5 CO .a o CL •O Q) O T3 £ CL 0 20 40 60 80 Diameter at breast height (cm) Fig. 4. Logistic regression curves of probability of mortality of American beech as a function of DBH for (a) six years before Hurricane Bonnie (1981-1986) (b) the census year including the hurricane (1987), (c) the six years after the hurricane (1987-92), and (d) the 15 years after the hurricane (1987-2001). Dashed lines represent 95% confidence intervals. climate envelope of American beech for mean temperature of the warm¬ est month (34° vs 29 °C) but did not go below the bottom of the enve¬ lope for mean minimum temperature in the coldest month (15° vs 9°C). The annual moisture index (precipitation/potential evapotranspiration) , was slightly below the minimum threshold (1.0 vs 1.1). Mean precipi¬ tation of the driest month was above the minimum threshold (88 vs 40 mm). Pathogens /pests .—Dying trees had thin canopies and exhibited sub¬ stantial leaf yellowing. Otherwise, none of the dying beech trees exhibited physical characteristics that might suggest death was caused by pathogens or parasites. Spatial aggregation analysis showed that in 1980 living American beech trees of all sizes were uniformly distributed (R= 1.11, P< .01) JHA ET AL. 293 Year Fig. 5. Meteorological data from Liberty, Texas, (a) Minimum August temperatures. Dashed lines represent average minima for 1906-1980 (21.9°C) and for 1981-2002 (23.2°C). (b) Summer (May, June, July) precipitation. The solid line represents average summer precipitation for 1905-2002. according to the Clark Evans nearest neighbor test. Large individuals were also uniformly distributed (R— 1.27, P< .01). Large trees dying between 1980 and 2001 were significantly aggregated (R= .65, P< .01), causing large living trees in 2001 to be randomly distributed. The randomization test confirmed that mortality was significantly aggregated (P< .05), even considering the initial distribution of the population. Synchronous death. —The age distribution (Fig. 6) shows that beech trees have been germinating steadily since 1850, except for peaks in 294 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 1820-1870 and 1910-1950. Minima in ring widths in the 1920s and 1950s suggest that the population experienced two important periods of marked environmental change, the pine logging of 1917 and the drought of 1950 (Glitzenstein, et al. 1986). The drought of the 1950s was indeed the longest and worst in the state’s climate history registering "severe" on the Palmer Drought Severity Index (NOAA 2003). Discussion Beech mortality was clearly not randomly distributed in time or space, nor was the population even-aged, and so the decline is neither a conse¬ quence of a random fluctuation in large-tree mortality nor a result of synchronous death of an even-aged population; there was a significant decline beginning in 1987 in an all-aged population. To explain this decline, predisposing stress, pest/pathogen, and hurricane disturbance are considered independently, and then a combination of these causes is proposed. Stress. — Stress due to drought is one of the most common factors that predispose populations to respond negatively to environmental stresses in the future (Pederson 1998). However, the 1950s drought occurred many years ago and so it is hard to imagine it had a major effect (but see Pederson 1998). A more immediate stress is the summer tempera¬ tures that are outside the climate envelope of beech, as defined by Box et al. (1993), recalling the high sensitivity of beech ring widths to August temperature (Cook et al. 2001). Large trees may be especially prone to temperature-related stress because of their greater exposure to sunlight and higher respiration. Further support for climate stress is provided by recent range limit studies for American beech. For example, Iverson & Prasad (1998) predicted current distribution of American beech to be north of its actual distribution range, and Davis & Zabinski (1992) predicted that the American beech population would shift north if temperatures increased. The proximity of Wier Woods to the southwestern range limit of beech (Fig. 1) is relevant in this context since the influence of changing temperature would logically be expected to appear here first. Pathogens /pests. —Two influences, aphid infestation (Siemann & Rogers 2003) and Hypoxylon (Pase 2002), have been documented on American beech in east Texas. However, neither aphids nor patches of their ‘honey dew’ secretions, were noted to be particularly abundant in Wier Woods during annual mortality surveys. Also, aphid feeding has minimal impact on large mature trees, causing its greatest damage and 295 JHA ET AL. Approximate Decade of Germination Fig. 6. The age distribution of a random sample of the beech population at Wier Woods based on ring counts from increment cores gathered by Glitzenstein (1986). dieback in small trees less than 3 meters tall (Hemmingsen 2002) . Hypoxylon- infected water oaks are common, but Hypoxylon fungal cankers were observed on only a single beech tree in Wier Woods in a special inspection conducted in May 2003. Furthermore, patterns in growth rates at Wier Woods do not support the hypothesis that a fungal pathogen is causing beech decline. At Wier Woods, growth rates of dying large trees were not lower than those for living large trees, as might have been expected (see Houston 1979). However, it should be noted that the sample size for growth rates was small, and growth rate trends may be unclear given a small sample size and the inherently low growth rates of large, old trees. Although there is little direct evidence for an effect of Hypoxylon , the aggregated mortality of large beech trees is consistent with the influence of a pathogen (but see below). A pathogen might also explain the extended duration of high mortality at Wier Woods, since pathogens may take months or years to affect their host (Hepting 1971), rather than causing mortality in one short time period. Given the high susceptibility of American beech to pathogens, the wide variety of pathogens known to affect beech, and the difficulty in documenting pathogen influences, this possible cause cannot be completely ruled out. Hurricane .—In a mesic forest in northern Florida, after a hurricane more severe than Hurricane Bonnie, Batista et al. (1998) found that large American beech trees experienced moderate direct hurricane mortality (8.2%) and low overall post-hurricane mortality (Batista & Platt 2003). Assuming that a more intense storm would cause greater immediate damage to large American beech trees (Batista et al. 1998, Batista & Platt 2003), the lower immediate mortality and higher post- 296 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 hurricane mortality in Wier Woods suggests that Hurricane Bonnie was not intense enough to account for the decline of the beech population. Furthermore, although the hurricane could have resulted in delayed mortality (cf., Putz & Brokaw 1989), mortality has remained high for more than 15 years after the hurricane, which suggests that other factors are influencing population mortality. Synthesis. — Tree population declines occasionally may be the result of a single environmental factor, but they most often have multiple causes (Manion 1981). Some of these are predisposing factors, occur¬ ring months or even years, before tree mortality, and others are inciting factors precipitating an episode of mortality (Houston 1987). The hurricane could have been such an inciting factor. By damaging trees, it might have triggered an increase in mortality (Putz & Brokaw 1989) in a population already weakened by a predisposing factor such as the consistently increasing summer temperatures in the 1980s and 1990s. Pathogens often appear on host species after periods of climate stress, and trees weakened by climate stresses (Houston 1987) or hurricane injury (Putz & Brokaw 1989) can be especially susceptible to attacks of insects or fungi. Hurricane damage to the crowns of large beech trees could also increase heat loading on remaining nearby trees and could therefore explain the spatial aggregation of mortality. Thus stress due to high summer temperatures, in conjunction with hurricane disturbance and possible pathogen influence, provides the most consistent hypothesis to explain the observed decline in American beech at Wier Woods. Further empirical observation of this beech population, as well as surveys of other beech populations in southeast Texas, will be required to fully evaluate this hypothesis. Acknowledgments We thank Saara DeWalt and Jie Lin for their advice on statistics and SAS; the Wier family and the Nature Conservancy for permission to work in Wier Woods; and Peter Marks (Cornell University) David Appel (Texas A&M University) and Elgene Box (University of Georgia) who shared their thoughts and research results with us. Literature Cited Antle, C. & L. Wain. 1988. Weibull distribution. Pp 549-556, in Encyclopedia of statistical science: Volume 9 (S. Kotz & N. Johnson, eds.), Wiley, NY, 432 pp. Arevalo, J. R., J. K. DeCoster, S. D., McAlister & M. W. Palmer. 2000. Changes in two Minnesota forests during 14 years following catastrophic windthrow. J. Veg. Sci., 11:833-840. Batista, W. B. & W. J. Platt. 2003. Tree population responses to hurricane disturbance: syndromes in a south-eastern USA old-growth forest. J. Ecol., 91:197-212. JHA ET AL. 297 Batista, W. B., W. J. Platt & R. E. Macchiavelli. 1998. Demography of a shade-tolerant tree in a hurricane disturbed forest. Ecology, 79:38-53. Bill, C. J. 1995. Hurricane effects on a Fagus/ Magnolia forest in southeast Texas, USA, in the context of long term forest monitoring. Master’s Thesis, Rice University, 91 pp. Blair, R. M. & L. E. Brunett. 1976. Phytosociological changes after harvest in southern pine ecosystems. Ecology, 57:18-32. Box, E. O., D. W. Crumpacker & E. D. Hardin. 1993. A climatic model for the location of plant species in Florida, U.S.A. J. Biogeogr., 20:629-644. Burnham, K. P. & D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Springer- Verlag, New York, NY, 488 pp. Cain, M. D. & M. D. Shelton. 1995. Thirty-eight years of autogenic, woody understory dynamics in a mature temperate pine-oak forest. Can. J. For. Res., 25:1997-2007. Caird, P. S. 1996. Soil moisture modeling, characteristics, and measurement in the Big Thicket. Master’s Thesis, Rice University, 155 pp. Clark P. J. & F. C. Evans. 1954. Distances to nearest neighbor as a measure of spatial relationships in populations. Ecology, 35:445-453. Cook, E. R. J. S. Glitzenstein, F. J. Krusic & P. A. Harcombe. 2001. Identifying functional groups of trees in West Gulf Coast forests (USA): A tree ring approach. Ecol. Appl., 11:883-903. Davis, M. B. & C. Zabinski. 1992. Changes in geographical range resulting from greenhouse warming effects on biodiversity in forests. Pp. 298-308, in Global warming and biological diversity (R.L. Peters & T.L. Lovejoy, eds.), Yale University Press, New Haven, CT, 386 PP- Deshotels, J. D. 1978. Soil survey for Big Thicket National Preserve, Texas. Texas Agricultural Experiment Station, College Station, Texas, 233 pp. Doyle, T. W. & G. F. Girod. 1997. The frequency and intensity of Atlantic hurricanes and their influence on the structure of South Florida mangrove communities. Pp. 109-120, in Hurricanes: Climate and Socioeconomic Impacts (H.F. Diaz & R.S. Pul warty, eds.), Springer-Verlag, New York, NY., 292 pp. Ehrlich, J. 1934. The beech bark disease. A Nectria disease of Fagus, following Cryptococcus fagi (Baer). Can. J. For. Res., 10:593-692. Fritts, H. C. 1958. An analysis of radial growth of beech in a central Ohio forest during 1954-1955. Ecology, 39:705-720. Gano, L. 1917. A study in the physiographic ecology in northern Florida. Bot. Gaz., 63:337-372. Glitzenstein, J. S. 1984. Succession in a beech-magnolia forest in East Texas. Ph.D. Dissertation, Rice University, 86 pp. Glitzenstein, J. S., P. A. Harcombe & D. R. Streng. 1986. Disturbance, succession, and maintenance of species diversity in an East Texas forest. Ecol. Monogr., 56:243-258. Harcombe, P. A., R. B. W. Hall, J. S .Glitzenstein, E. S. Cook, P. Krusic, M. R. Fulton & D. R. Streng. 1998. Sensitivity of Gulf Coast Forests to Climate Change: In: Vulnerability of coastal wetlands in the southeastern United States: climate change research results. Biological Science Report USGS/BRD/BSR-1998-0002. Edited by G. Guntenspergen & B.A Varain. Harcombe, P. A., C. J. Bill, J. S. Glitzenstein, M. R. Fulton, P. L. Marks & I. S. Elsik. 2002. Stand dynamics over 18 years in a southern mixed hardwood forest, Texas, USA. J. Ecol., 90:947-957. Hemmingsen, K. 2002. Big aphids advancing to Northeast Texas. Country World News- East Texas Edition, Feb 7th. Echo Publishing Company, Sulphur Springs, Texas, 2 pp. Hepting, George H. 1971. Diseases of forest and shade trees of the United States. U.S. Dept, of Agr., Agric. Handb. 386, U.S. Gov. Printing Office, Washington, DC, 658 pp. Houston, D. R., E. J. Parker & D. Lonsdale. 1979. Beech bark disease. Annu. Rev. Phytopathol., 32:75-87. Houston, D. R. 1987. Forest tree declines of past and present: current understanding. Can. 298 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 J. Plant Pathol., 9:349-360. Iverson, L. R. & A. M. Prasad. 1998. Predicting abundance of 80 tree species following climate change in the eastern United States. Ecol. Monogr., 68:465-486. Kurz, H. 1944. Secondary forest succession in the Tallahassee Red Hills. Proc. Florida Acad. Sci., 7:59-100. Little, E. L. 1971. Atlas of United States Trees, Volume 1. U.S. Gov. Printing Office, Washington, D.C., 8 pp., 200 maps. Manion, P. 1981. Tree disease concepts. Prentice Hall Inc., N.J., 389 pp. Marks, P. L. & P. A. Harcombe. 1975. Community diversity of coastal plain vegetation in southern east Texas. Ecology, 56:1004-1008. Marks, P. L. & P. A. Harcombe. 1981. Forest vegetation of the Big Thicket, southeast Texas. Ecological Monographs, 51:287-305. McLeod, C. A. 1975. Southwestern Range Limit of Fagus grandifolia Ehrh. Tex. J. Sci., 26(1&2): 179-184. Mueller-Dombois, D. 1992. A natural dieback theory, cohort senescence, as an alternative to the decline disease theory. Pp. 26-37, in Forest Decline Concepts (P. D. Manion & D. Lachance, eds.), Am. Phytopathol. Soc. Press, 349 pp. NOAA. 1986. Hurricane Bonnie Climate Data. National Oceanic and Atmospheric Administration. Environmental Data and Information Service, National Climate Center, Ashville, North Carolina. NOAA. 2002. Monthly normals of temperature and precipitation, 1903-2002, Liberty, Texas. Climatography of the United States, National Oceanic and Atmospheric Administration. Environmental Data and Information Service, National Climate Center, Ashville, North Carolina. NOAA. 2003. Texas Weather Extremes. Dallas/Fortworth. Climatography of the United States, National Oceanic and Atmospheric Administration. Environmental Data and Information Service, National Climate Center, Ashville, North Carolina. Pase, J. 2002. Hypoxylon Canker. Tex. For. Serv. Feb. 17th Pederson, B. S. 1998. The role of stress in the mortality of Midwestern oaks as indicated by growth prior to death. Ecology, 79:79-93. Peters, R. & Poulson, T. L. 1994. Stem growth and canopy dynamics in a world-wide range of Fagus forests. J. Veg. Sci., 5:421-432. Platt, W. J. & M. W. Schwartz. 1990. Temperate hardwood forests. Pp. 194-229, in Ecosystems of Florida (R.L. Myers and J.J. Ewel, eds.), Univ. of Central Florida Press, Orlando, FL, USA., 765 pp. Putz, F. E. & N. V. L. Brokaw. 1989. Sprouting of broken trees on Barro Colorado Island, Panama. Ecology, 70:508-512. Quarterman, E. & C. Keever. 1962. Southern mixed hardwood forest: climax in the southeastern Coastal Plain, U.S. A. Ecological Monographs, 32:167-185. SAS. 1990b. SAS Language: Reference, Version 6, First Edition. SAS Institute, Cary, NC, USA. Siemann, E. & W. E. Rogers. 2003. Herbivory, disease, recruitment limitation and the success of an alien tree species. Ecology, 84:1489-1505. Tardiff, J., J. Brisson & Y. Bergeron. 2001. Dendroclimatic analysis of Acer saccharum, Fagus grandifolia, and Tsuga canadensis from an old-growth forest, southwestern Quebec. Can. J. For. Res., 31:1491-1501. Thompson, G. E. 1963. The decay of oaks caused by Hypoxylon atropunctatum. U.S. Dept, of Agric. Plant Dis. Rep., 47:202-205. Tubbs, C. H. & D. R. Houston. 1990. Fagus grandifolia Ehrh. Pp 325-332, in. Silvics of North America. Volume 2, Hardwoods (R. M. Burns and B. H. Honkala, eds.), U.S. Dept. Agric. Handb. 654, 877 pp. PAH at: harcomb@rice.edu TEXAS J. SCI. 56(4): 299-3 18 NOVEMBER, 2004 COMPARATIVE ANALYSIS OF GROWTH AND MORTALITY AMONG SAPLINGS IN A DRY OAK-PINE FOREST IN SOUTHEAST TEXAS Jie Lin, Paul A. Harcombe, Mark R. Fulton1 and Rosine W. Hall2 Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77251-1892 Present address: department of Biology, Bemidji State University 1500 Birchmont Dr. NE, Bemidji, Minnesota 56601 department of Biology, Auburn University at Montgomery 7300 University Drive, Montgomery Alabama 36117 Abstract.— The role of shade tolerance in the dynamics of a sandy upland pine-oak forest in Big Thicket National Preserve, southeast Texas was investigated. Using a forest dynamics modeling framework, radial growth of saplings as a function of light availability and mortality as a function of recent growth history for species with a range of shade tolerance levels was investigated. In low light, shade-tolerant species grew faster than shade-intolerant species. However, in high light, shade-intolerant species did not grow faster than shade- tolerant species possibly because some of them are adapted for drought resistance. They did not survive better, either, perhaps because of recent increases in canopy shading. Mesic, shade-tolerant species had better performance at the dry site than at the mesic site, possibly because of a difference in the competitive environment of the two sites. An implication of invasion and higher growth and survival of the mesic species is that these species may have been limited to a larger extent by fire than by site conditions on this site in the past. Broad patterns in species dominance across the landscape are well known for the southeastern United States (Christensen 1988; Ware et al. 1993), and these are consistent with general understanding of physio¬ logical tolerances of the major tree species. In southeast Texas, interspecific differences in response to light are consistent with trends in species dominance at a mesic site (Lin et al 2001; 2002), and thereby help provide mechanistic underpinning for observed species dominance on mesic sites. At a wet site, light was important in helping to explain species dominance, but only if response to flooding was considered, as well (Hall 1993; Hall & Harcombe 1998; 2001; Lin et al. 2004). In the study reported here, analysis of the light response to a dry site is extended, partly to further investigate the effects of site differences on light responses, and partly also to determine whether differences in light response among species help explain changes in species dominance. The approach is based on the general understanding that light, soil moisture and nutrients are important factors that determine species 300 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 composition of many terrestrial plant communities (e.g. Huston & Smith 1987; Smith & Huston 1989; Pacala et al. 1994; Knox et al. 1995; Sipe & Bazzaz 1995; Grubb et al. 1996; Catvosky & Bazzaz 2000). Mortali¬ ty-growth-light relationships based on the forest dynamics model, SORTIE (Pacala et al. 1993; 1994; 1996; Kobe et al. 1995) are used. The model assumes resource competition among coexisting species, as do most forest dynamics models (e.g. Botkin et al. 1972; Shugart 1984; Smith & Huston 1989; Pacala et al. 1996). Through repeated iterations of the model, light competition results in shifting dominance from shade- intolerant species to shade- tolerant species over the course of stand development. Extending SORTIE by incorporating soil moisture into the mortality-growth model, Caspersen & Kobe (2001) found that species ranks in mortality-growth relationship shifted substantially across soil moisture gradient, resulting in shifting dominance. Although competition for soil moisture provides a possible process- level explanation for the broad pattern of species segregation across the landscape in southeast Texas (Marks & Harcombe 1981; Harcornbe et al. 1993) and across the southeastern United States (Christensen 1988; Ware et al. 1993), fire also plays a role (Harcombe et al. 1993; 1998). Under the fire scenario, sites with longleaf pine ( Pinus palustris ), a species highly tolerant to fire, would not support mature hardwood forests. One way to investigate the question of the relative importance of soil moisture and fire is to compare growth-mortality relationships of species under different moisture regimes. In essence, this is asking whether consistency can be found between process (growth/mortality) and pattern, and tie it to a mechanism (competition for light and/or mois-ture). If growth and mortality for species present at different sites are lower at the dry site, the inference that soil controls vegetation pattern cannot be ruled out. If, on the other hand, growth and mortality are higher at the dry site under the current fire suppression scenario, then fire may have been the major limiting factor at the dry site in the past. In this study, light competition in a mixed pine-oak stand in the Turkey Creek Unit of the Big Thicket National Preserve, southeast Texas was investigated. In addition, growth and mortality of species common to both this dry site and a nearby mesic site were compared. Compared with the mesic site, the dry site is characterized by coarser soils and lower soil moisture availability (Caird 1996). Widespread presence of charcoal on stumps and the prevalence of longleaf pine LIN ET AL. 301 indicates that the dry site probably burned relatively frequently (Harcombe et al. 1993). Under the current fire suppression scenario, the site is being invaded by mesic species (Harcombe et al. 1998). The invasion of mesic species suggests that they may have been limited by fire in the past, and not by low soil moisture. The following questions are addressed: Do differences in mortality-growth-light relationship among species within and between sites explain differences in dominance between the dry site and the mesic site? Can species responses to site conditions explain differences in species composition or must historical disturbances (e.g., fire) be invoked? Study Sites and Species The dry study site is located on a low, sandy ridge in the Turkey Creek Unit of the Big Thicket National Preserve about 10 km southeast of Warren, Tyler County, Texas (30°35’N, 94°24’W). The climate of the area is humid subtropical with an annual rainfall around 1475 mm. The soil is a sandy loam of Landman series, loamy, siliceous thermic Grossarenic Paleudalf (Caird 1996). Light measurements obtained from hemispherical photos taken at plot centers (100 plots in total) indicated a light range in the understory from 1.7% full sun to 33.5% full sun with a mean of 12.8%. The vegetation is dominated by oaks and pines. Ranked in decreasing order of relative abundance, post oak ( Quercus stellata Wang.), southern red oak (Quercus falcata Michx.), black hickory (Cary a texana Buckl.), longleaf pine (Pinus palustris Mill.), loblolly pine (Pinus Taeda L.) and shortleaf pine (Pinus echinata Mill.) form a relatively open canopy 15-20 m tall. Basal area increased from 21m2/ha in 1982 to 28 m2/ha by 1999. Red maple (Acer rub rum L.) and sweetgum (Liquidambar styraciflua L.) are minor canopy components. The understory is a moderately dense mixture of tree saplings and shrubs; flowering dog¬ wood (Comus florida L.), yaupon (Ilex vomitoria Ait.) are abundant. Saplings of mesic species, such as Southern magnolia (Magnolia grandi- flora L.) and American holly (Ilex opaca Ait.) have become more abun¬ dant since 1980 (Harcombe et al. 1998). American holly and flowering dogwood are very shade-tolerant; sweetgum and most dry-site species are shade- intolerant. The above shade tolerance categories are based on conventional wisdom regarding shade tolerance as summarized by Burns & Honkala (1990). These shade tolerance classifications are based largely on field observations regarding the relative abundance of differ¬ ent species in the forest understory. 302 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Table 1. Latin names, common names, name codes and shade tolerance of major species. Species are arranged in ascending order of shade tolerance according to Burns & Honkala (1990). Latin Name Common Name Species Code Shade Tolerance Site Affiliation Quercus stellata Post oak QUST Intolerant Dry Cary a texana Black hickory CATE Intolerant Dry Pinus palustris Longleaf pine PIPA Intolerant Dry Pinus echinata Shortleaf pine PIEC Intolerant Dry Pinus Taeda Loblolly pine PITA Intolerant Mesic, dry Liquidambar styraciflua Sweetgum LIST Intolerant Mesic, dry Quercus falcata Southern red oak QUFA Intermediate Dry Acer rubrum Red maple ACRU Tolerant Mesic, dry Magnolia grandiflora Southern magnolia MAGR Tolerant Mesic, dry Ilex opaca American holly ILOP Very tolerant Mesic dry Comus florida Flowering dogwood COFL Very tolerant Mesic, dry The dry site was logged in 1930 but the stand is not strongly even- aged (Harcombe et al. 1993; Kaiser 1995); apparently many old hardwoods and older pines were left in the site. Exactly how long ago fire occurred on this site is unknown. The presence of charcoal on stumps implies relatively frequent fire prior to 1930 and relatively infrequently after that until 1974. Fire has been absent since 1974 (Kaiser 1995; P. Harcombe, personal communication). A nearby mesic site was chosen for comparison. The mesic site is located in Hardin County, Texas (30°16’N, 94°12’W) approximately 14 km away from the dry site. Species composition of this site represents many typical mesic sites throughout the Coastal Plain area of the south¬ eastern U.S. (Marks & Harcombe 1981). The site is dominated by loblolly pine ( Pinus taeda L.), water oak ( Quercus nigra L.), white oak (Quercus alba L.), American beech ( Fagus grandifolia Ehrh.) and southern magnolia ( Magnolia grandiflora L.). Red maple (Acer rubrum L.), blackgum (Nyssa sylvatica Marsh.) and sweetgum (Liquidambar styraciflua L.) are abundant as small to medium stems but are infrequent as large trees. Important understory trees include American holly (Ilex opaca Ait.) and flowering dogwood (Comus florida L.). Basal area has varied between 33.7 m2/ha (after hurricane) and 35.1 m2/ha over the last 20 years. More detailed description can be found in Glizenstein et al. (1986) and Lin et al. (2001; 2002). See Table 1 for shade tolerances and affiliations of species with sites. LIN ET AL. 303 Data Collection and Analyses Sapling growth. — The dry study site is 4 ha divided into 100 con¬ tiguous tree plots. Each plot is 20m by 20m. Tree surveys were performed in 1980, 1982, 1985, 1989, 1994, 1997 and 2000. During tree surveys, stems with a Diameter at Breath Height (DBH) > 2 cm are measured with a diameter tape. A subset of 16 plots was chosen randomly for annual measurement of saplings (height > 140 cm and DBH <4.5 cm), in which DBH of all saplings was measured to the nearest 0. 1 cm from 1980-2000. All trees and saplings are tagged with an identification number. For each sapling (height > 140 cm and DBH <4.5 cm), annual radial growth rate over three years was calculated as the difference in radius between year 1999 and year 1996 divided by 3. The average over 3 years was used to reduce measurement variation. Calculations of growth were made for all species with more than 15 individuals in the sample. As approximations of high-light growth and low-light growth, top quartile growth rate (TQGR) and bottom quartile growth rate (BQGR) were calculated. Approximations were chosen because it was not possible to model mortality-growth-light relationships owing to small sample sizes and/or insufficient range of light conditions, TQGR is a reasonable approximation of high- light growth because saplings that have high growth rates are unlikely to be growing in low light. Comparison of TQGR and the actual high-light growth in the mesic site where both measures are available showed a good agreement between the two (data not shown). It is important to note that bottom quartile growth rate is only a rough approximation of low-light growth because low growth could result from many reasons other than low light. Top quartile growth rate was computed as follows: First, the radial growth rate over the first 3 years after the sapling first entered the survey was calculated. After calculating growth rates of all first-year saplings, growth rates were sorted in descending order. Then saplings with growth rates in the top 25% were chosed and their growth rates were averaged. To see whether TQGR of first-year sapling obtained this way might underestimate maximum growth, it was compared with TQGR for all saplings present in one period (1996-1999); it did not (results not shown) . The bottom quartile growth rates were obtained by taking the bottom 25% growth rates and computing the average. Light measurement.— A subset of live saplings was selected from the 304 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 database for light measurements. In keeping with the protocols of previous studies, the goal was to find at least 50 saplings per species for light measurement. The final sample size ranged from 45 to 59 saplings per species. The six species are: red maple, sweetgum, loblolly pine, post oak, Southern magnolia and American holly. Saplings were selected in a stratified random fashion by plot to obtain a broad range of light conditions. Fish-eye photographs were taken at the top of each sapling (following Rich 1989; Pacala et al. 1994) in mid summer (late June to mid July), 1999. To increase contrast, all photos were taken early in the morning before sunrise and late in the afternoon after sunset when skylight is evenly distributed. Moreover, all photos were taken on Kodak TMAX ASA 400 (black and white) film and the film was under¬ exposed by 1 f-stop to further enhance contrast. The images were scanned, digitized and analyzed using CANOPY (Rich 1989). Thres¬ hold values were set individually to minimize the “halo effects” (Anderson 1964). The global site factor (GSF) was estimated from each photo. GSF is an estimation of the fraction of total radiation (both diffuse and direct) a sapling experienced during the growing season. The GSF value was converted to percent of full sun by multiplying GSF by 100. Since no major canopy disturbances occurred during the 1996- 1999 period, the light level captured in 1999 was considered to be a reasonable representation of average light environment over the three- year period at a given location. Sapling mortality .—In addition to periodic measurement, each sapling was checked annually to see whether it was dead or alive. Survival time was calculated as the length of time a sapling was followed during the course of the study. If a sapling died, then its survival time would be the difference between the year of death and the year it entered the study. If a sapling was alive at the end of the study (Year 1999), its survival time was the difference between the ending year and the year it entered the study. Saplings that were alive at the end of the study were flagged as right censored (Cox & Oakes 1984; Lee 1992). All saplings (dead or alive) that had been recorded since the beginning of the long-term study (Year 1980) were included. To model mortality as a function of recent growth, pre-mortality growth rate was calculated for dead saplings as the difference in radius over the last 3 years prior to death divided by 3. Growth-light analysis.— The goal of this analysis is to model growth response from light availability using a Michaelis-Menten function, as LIN ET AL. 305 in previous studies (cf. Pacala et al. 1994; Wright et al. 1998). How¬ ever, because of sampling limitations, the asymptote parameter was replaced by TQGR, which is treated as a constant instead of a para¬ meter, because of inadequate range of conditions and small sample sizes for some species. The one-parameter model takes the following form: aL a/S + L (1) Where yu. is the mean growth response given light availability; a is the TQGR; S is the slope at low light; L is the light availability (% of full sun). The maximum likelihood methods to estimate parameter S was used. The final likelihood function is: n nr f=l ^27iC[aL/(a/S + L)}' - exp(- [Gi aLKa/S + L)]1 (2) 2 C[aL/(a/S + L)]‘ where Gj is the radial growth rate of sapling i (3-year average); C, D are two parameters that account for heteroscedasticity. Confidence intervals of S were obtained by bootstrapping. Both model fitting and bootstrapping were done using Splus 6.0 on Unix (Mathsoft, Inc. 2000). A more detailed description of the maximum likelihood estimation method can be found in Lin et al. (2002). Mortality risk (annual death rate) as a function of growth.— Survival analysis was used to model mortality risk as a function of growth. The likelihood function for censored and non-censored saplings is (Lee 1992): n*e i = 1 AT n~r -Ati Y\e i = 1 (3) where r is the number of saplings that died during the study and n-r is the number of saplings that are right-censored, f and are lifetimes of a non-censored and right-censored sapling i, respectively; X is the parameter of mortality risk (annual mortality risk). A negative exponential function was used to estimate X from predictor variables A = e-0°-P'Xl-0iXlx0 (4) 306 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 where Xj is the radial growth rate (mm/yr); X2 is the initial size (radius in mm). The parameters to be estimated are the fts. 6 is the error term. Estimates of parameters Bi and ft2 were found by maximizing the likelihood function (3). Maximum likelihood estimation of annual death rate— To further explore how mortality might be different among species with different shade tolerance, annual death rate was also compared. The maximum likelihood estimator of annual death rate is (Lee 1992): D (5) 2 = ZD rj-, yf—~\N-D Ti+y t, ;=1 Z— (,= 1 Where D is the number of deaths during the time interval The 95% confidence interval of X is: ~ 2 x 1.96 (6) Results Growth response to light and interspecific tradeoff. —Growth in¬ creased with light for all species (Figure 1). Except for sweetgum, which showed higher growth than red maple, the pattern of low-light growth was consistent with the expectation that shade-tolerant species grow faster in low light than shade-intolerant species (Figure 1). The low-light growth index, slope at low light, was highest for American holly, followed by southern magnolia (Table 2). Two shade-intolerant species, loblolly pine and post oak, ranked low in slope (Table 2). The correspondence between low-light growth and shade tolerance ranks was further supported by the comparison of bottom quartile growth rates among species (Figure 2a): Shade-tolerant species ranked higher than most shade- intolerant species in bottom quartile growth rates, though bottom quartile growth rate of sweetgum and loblolly pine were higher than expected based on standard shade tolerance ranks. In contrast, for high-light growth, the order of TQGR did not cor¬ respond to shade tolerance expectation: First, shade- intolerant post oak and loblolly pine showed low TQGR; second, shade-tolerant southern magnolia and American holly grew more rapidly than expected (Figure 1, Table 2). Top quartile growth rates of xeric dominants (e.g., post LIN ET AL. 307 Percent of full sun (%) Fig. 1. Fitted growth-light regression curves for different species using equation (1). The horizontal axis represents percent of full sun (log scale); the vertical axis represents annual radial growth. Table 2. Top quartile growth rates (TQGR, a in equation 2) and estimated slope at low light (S in equation 2) with 95% confidence intervals (Cl). N is the sample size. NA stands for not available. Species Shade tolerance N TQGR Cl of TQGR S Cl of S Post oak intolerant 53 0.905 0.736-1.074 0.026 0.014-0.046 Black hickory intolerant 78 0.718 0.641-0.795 NA NA Loblolly pine Intolerant 59 1.720 1.643-1.798 0.058 0.033-0.099 Sweetgum Intolerant 58 2.006 1.912-2.099 0.654 0.357-1.100 Southern red oak Intolerant 16 1.263 1.155-1.370 NA NA Red maple Tolerant 45 1.728 1.599-1.857 0.347 0.232-0.530 Southern magnolia Tolerant 52 2.363 2.205-2.516 0.755 0.545-1.123 American holly Very tolerant 47 1.847 0.901-1.282 2.911 1.650-5.144 Flowering dogwood Very tolerant 33 1.944 1.831-2.057 NA NA 308 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Bottom quartile growth rates of first year saplings (N=25)(N=33)(N=25) (N=23) (N=1 6) (N=63) (N=78)(N=31 )(N=121 ) Top Quartile growth rate of first year saplings I LOP COFL MAGR ACRU QUFA LIST CATE QU ST PITA (N=25) ( N=33) (N=25) (N=23XN= 1 6) ( N=63) ( N=78) ( N=3 1 ) (N= 1 21 ) Fig. 2. Bottom quartile growth rates for different species (a) and top quartile growth rates for different species (b). Values not sharing the same letter are significantly different (ANOVA followed by Tukey’s multiple comparison adjustment, P < 0.05). N is the number of saplings. Species are arranged in descending order of shade tolerance from left to right. See Table 1 for key to species codes. oak, black hickory, southern red oak) were significantly lower (P < 0.05; ANOVA followed by Tukey’s multiple comparison adjustment) than mesic invaders (e.g., American holly, Southern magnolia, sweet- gum). Even within the six mesic species, top quartile growth rates did not conform to expectation: shade-tolerant southern magnolia grew significantly faster than shade- intolerant sweetgum and loblolly pine (Figure 2b). LIN ET AL. 309 Mortality risk as a function of growth.— Mortality risk as a function of growth was used to characterize shade tolerance in previous studies (e.g., Kobe et al. 1995; Lin et al. 2001). In this study, the low number of dead saplings of American holly, southern magnolia and red maple made survival analysis on these species unreliable (e.g., there was only one dead American holly sapling and two dead southern magnolia saplings found in the long-term study data base). Thus, at this site, the only shade-tolerant species included in survival analysis was flowering dogwood. In contrast to results of a previous study performed at the mesic site (Lin et al. 2001), both growth and size were significant predictors of mortality risk in the dry site. Overall, mortality risk decreased as growth increased and decreased with increasing size (Table 3). The mortality-growth relationship was not consistent with the expectation that shade-intolerant species have higher mortality risk at zero growth and steeper slope than shade- tolerant species (Table 3). Annual death rare.— Interpretation of the above mortality-growth responses in terms of shade tolerance expectation was limited by the fact that only one shade-tolerant species (dogwood) was involved in the analysis. Therefore, annual death rates among species were also compared (Figure 3). Mesic species such as American holly, southern magnolia, red maple exhibited extremely low annual death rate (Figure 3), which is consistent with the previous finding that they have become more abundant and species typical of dry sites have experienced dramatic decline (Harcombe et al. 1998). Death rates of dry site dominants (longleaf pine, post oak, southern red oak) were consistently higher than mesic site species. Cross-site comparisons Growth-light curves of southern magnolia and American holly were significantly higher at the dry site than at the mesic site over the light range (Figure 4a and b): For red maple, growth rates were significantly higher only above 60% full sun (Figure 4c). For sweetgum, there was no significant difference between sites (confidence interval overlapped, not shown) (Figure 4d). Annual death rates were significantly higher at the mesic site than at the dry site for all species common to the two sites except flowering dogwood (Figure 5). Discussion Growth, mortality and tolerance.— Results show that growth responses to low light are roughly consistent with one of the expectations regarding shade tolerance: in low light, shade- tolerant species grow faster than Table 3. Parameter estimates of the mortality-growth model (equation 4) with 95% confidence intervals (Cl) for different species. N is the total number of saplings (both dead and live); fis are parameters in equation 4. X is the mortality risk at zero growth at size class 0.5 mm. 310 THE TEXAS JOURNAL OF SCIENCE- VOL. 56(4), 2004 tj- m r- o oo Tfr o n r- in <4- — 1 •n O s ri — H ri 1) a g G G G G 4) O H-H M >— 1 i— ■* g 5 cm DBH (large trees) and were measured for DBH, identified by species, and tagged. Stems 2-5 cm DBH (small trees) were counted in three categories (2-3 cm, 3-4 cm and 4-5 cm) by species. Stems were considered alive if they had living tissues above breast height. (2) Saplings. Large (DBH < 2 cm but taller than 1.4 m; shrub species were included in this category) and small (between 0.5 m and 1.4 m in height, tree species only) stems were counted in a 2 by 10 m strip centered on the central line in each plot parallel with the transect. (3) Seedlings ( < 50 cm in height, tree species only) were tallied in a 1 by 10 m strip within the sapling plot by species. Densities of seedlings and small saplings were combined in analysis. (4) Shrubs. For clumps of these characteristically multi-stemmed woody species < 1.4 m tall, cover was measured along an intercept line with a length of 10 m (in RL) or 20 m (in BTNP) in each plot parallel with the transect if 1.4 m tall). Larger shrub stems (height > 1.4 m) were tallied with the large sapling class. When two plots were so close that the shrub cover measurement would overlap with that of another plot, the central line was extended accordingly in the opposite direction along the transect to avoid measurement overlap. (5) Fuels: fine fuel (1-hour fuel) was collected in a 50 by 50 cm quadrat at one of the four corners outside the plot and sorted into duff, needles, leaves, twigs, cones, barks, and live materials. Sorted samples were dried at 70 °C for 72 hours and weighed. Fuel depth was measured at 1 m, 3 m, 5 m, 7 m, and 10 m along a 10-m central line. (6) Fire temperature: fire-sensitive tablets (Tempil of Big Three Industries, Inc., New Jersey, USA) were placed in the center of each plot to obtain a fire temperature estimate. The tablets were wrapped in aluminum foil and placed 20 cm above ground. Tablets had following discrete melting points: 52 °C, 101°C, 153°C,204°C, 262°C, 305 °C, 343°C, 399°C, 454°C, 500°C, and 545°C. For statistical analysis, a nested-factorial analysis was used to compare fire effects within types by differencing (i.e. , by comparing the magnitude of change in the burn plots with the magnitude of change in the controls). About two hundred plots representing 10 sites and five vegetation types were used in this comparison (Table 2). For burn plots 324 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Table 2. Study plots selected for within-type and cross-type comparisons. Unit Site Vegetation Type i Total WS SH UP US MS B2 C2 B C B C B C B C Big Sandy RC3 5 5 5 5 20 Big Sandy 06 3 5 5 10 10 30 Big Sandy 15 5 5 5 5 20 Big Sandy 15 3 5 5 5 5 20 Lance Rosier 53 5 5 Lance Rosier 54 10 10 20 Turkey Creek IS3 10 10 5 5 5 5 40 Turkey Creek 36 3 5 5 10 10 30 Roy Larsen BF3 5 5 4 4 18 Roy Larsen HL 5 4 5 5 19 Total 20 14 20 20 15 15 44 44 15 15 222 1 Vegetation type: WS— wetland pine savanna; SH— sandhill; UP— upland pine; US— upper- slope; MS— midslope 2 Treatment: B— Burn; C— Control 3 Sites used for cross-type comparison at two sites, there was a delay of >2 yr between measurement and burning, and so preburn values were adjusted to account for the succes- sional change that would have taken place (based on measured changes in control plots). Changes in the control plots were generally small for large saplings and trees, but there were large fluctuations in small individuals (seedlings and saplings) and in fuel components from year to year. Because differencing increases variances of adjusted changes the small individuals were not as useful in addressing fire effects. The before-after fire comparisons within types involved more than one site for each type; 4-5 plots were nested within each study site. In the statistical model, site and plot were treated as factors. Because there was no replication across sites within burn blocks, the error term was not retrievable. The following model was used: Yj = mean + site + plot (site) + time + plot (site) x time Where Yj — response variable (dependent); LIU, HARCOMBE AND KNOX 325 mean — overall mean response; site — site effect; time — before vs. after; plot (site) — plot effect nested within site; plot (site) x time — interaction between time and site; Here the main focus is the time effect, i.e. is there a significant difference between post-fire and pre-fire measurements? Because plots were chosen randomly, effects of plot(site) and plot(site) x time were treated as random. Because of the unbalanced experimental design (there are sets of four plots instead of five plots at some sites), a nested- factorial analysis was preferred to a repeated measurement analysis (SAS Institute; 1992a; 1992b). The two analyses produce identical results. To test whether fire had different effects on different vegetation types, the sites that had more than one vegetation type and were burned on a single day were chosen. Six sites and four types were appropriate for such an analysis (Table 2). The four vegetation types that could be compared were sandhill, upland pine, upperslope, and midslope. Suc- cessional change unrelated to fire was adjusted for using changes in control plots as described above for the within- types comparison. How¬ ever, pre-fire differences still existed for the burn plots of different vegetation types after the adjustment, and so these differences were adjusted for, as well. The reasoning was that post- fire change measured in absolute terms might not reflect the fire effects but a combination of pre-fire difference and fire effect. For instance, a reduction of 50 out of 100 small trees in an upland pine type by fire is not the same as a loss of 50 of 500 small trees in a midslope type at the same site. The former would have a 50% reduction compared to only 10% in the latter. To overcome this problem, percentage change was calculated with re¬ spect to pre-fire measurement for each site. Thus, the differences in magnitude of change between types reflected differential effects of fire on the types. The GLM procedure (SAS Institute 1992a; 1992b) was used with vegetation type as the only independent variable. Basal area, shrub cover, density of large saplings, and density of seedling and small saplings departed somewhat from a normal distribution so a logarithmic 326 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 transformation to the base e was applied after adding one to every measurement. All transformed data appeared approximately normally distributed. Because the response variables were from the same plots and were possibly correlated, testing the hypotheses for each variable involves multiple comparisons. Therefore, the error rates (type I error) were adjusted according to the Dunn-Sidak method (Day & Quinn 1989). Significant levels were determined by the adjusted error rate according to Dunn-Sidak method (k = 8). P< =0.0064 (overall error rate of 0.05) was considered highly significant (**); P< =0.0131 (overall error rate of 0.10) was considered significant (*); P> 0.0131 (overall error rate more than 0.10) was not significant (ns) (Figure 1). Results Fire reduced fine fuel load in three of the five types, but not in midslope or savanna (Figure 1). Absence of significant effects in the midslope type was probably a result of cool and patchy fires (see below). In the savanna, grasses and forbs recovered quickly after fire and replaced much of the fuel consumed by fire. Fire reduced fuel depth (fine fuel only) significantly in all types, though the magnitude of reduction appeared greater in sandhill and upland pine (Figure 1). Because heavy needle drop was quite common after hot fires when the canopy was scorched, fuel consumption in the sandhill and upland types was probably greater than the data indicated. Shrub cover was reduced significantly by fire only in the savanna; other types showed no significant differences between pre- and post- fire measurements. The rate of post-fire recovery of shrubs by resprouting was sufficient to return shrub cover to values near pre-fire values, except in the savanna type, which typically has a sparse understory. The seedling-small sapling class also showed no significant differences between pre-fire and post-fire densities, probably because of rapid resprouting. Large saplings decreased significantly in density in all types except midslope after fire. The magnitude of the response was highest in upland pine and lowest in sandhill. The large saplings consisted mostly of post- fire survivors; few hardwood species can resprout rapidly and grow to large sapling sizes (0-1 cm DBH) in one or two years. LIU, HARCOMBE AND KNOX (IQ 5*3 ^ g £ a g ^ § <2 <’ >3 W T3 1—4 a&gg' M&g'S’ * p X 6 2 .1 ^ . 0> •< c/i •-c ^3 p cj* cr^ o Cl» g* * & r« © !-V «-*■ 3 O 3 X 8.3 (S “ S’ 2 8 S-5 SS- __ s-3 r: 8-s|®| srlSf ” 1 8 §1 1 Jill S. s g a g 5/3 2 “ -5 pj a ^ i ® o s O' 3 3 2- ^ 5 i CL ^ o- o sr ’-*■ o § "3 ® 8!Eb - « ^ o. a. an cr *q o 3 , a> s* a> p Ell ,- — v or -og-ff 8~ ttflt 8.*o ” F® <“♦ a. ►— X* CP M Ilf si®* I O 2. 3“ ^ Cl c« cd ' a> CL p _ B3 » 3 3 o p -+> ° ol a CL O g. O- » 3 : % ar g. g> <. C3 o o >-» fuel deplh (cm) 4 6 8 fine fuel (t/ha) 0 5 10 15 20 0 5 10 15 20 15 20 327 sandhill upland pine upperslope midslope savanna 328 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Small tree density was significantly reduced in all types except savanna (Figure 2). Since it takes several years for saplings or seedlings to grow to small tree size, the change can be attributed exclusively to the direct impact of fire. Large tree density declined significantly in the sandhill and upland types (Figure 2). Although tree density declined, the largest trees had high survival, so tree basal area was not signifi¬ cantly affected by the prescription fires. To compare changes among the types within burn, six blocks were selected (Table 2) because each of these blocks was burned on a single day, so the vegetation types being compared were burned under the same conditions. Since the post-fire values for fuel load, seedling, and saplings were a combination of fire-related death and recovery after fire rather than direct fire impact (as described above), the small tree and large tree strata were emphasized in this cross-type or between-type comparison. One block (BS10— MS and LS types) was excluded because the attempt to burn this unit failed. For four of the six blocks, percentage changes in small tree densities due to fire differed significantly between the types ( P < 0.005; Figure 3); two of the four also showed significant differences for percent change in large tree density, as well. The significant differences all involved comparisons between sandhill or upland pine and other types. The two blocks (BS15 and BSRC) which showed no significant differ¬ ences in any of the test variables involved comparison between upper- slope and midslope. The results for all six blocks suggest that fire affected two dry types (sandhill and upland pine) more strongly than it did other types. The greater impact of fire on sandhill and upland pine corresponded to higher fire intensity. For example, the temperature readings from upland plots at BS06 were all 152°C to 399°C., whereas fire tablets melted only in two of the ten upperslope plots (152°C and 204°C). The differences in fire intensity among types in other sites were similar to BS06. In the upperslope - midslope comparisons (BS15 and BSRC), the lack of significant differences in fire response could be a consequence of cool fires in both. At BSRC, temperatures ranged from < 52 °C to 253 °C; At BS15, the fire only partially burned the upperslope plots, and missed three of the five midslope plots completely. LIU, HARCOMBE AND KNOX 329 (rq o c o 5 3 °> T3 K> a: • S' w O o 3 £ 3 2 w g cp — • — cp‘ o "I -+> 05 Vi 3 3 cp £L X — Ust § S 05 ws o’c' 3 s ° ^P ■-+5 <-i O*^ O &5 X 3 T3 CP, 5* tree density (number/ha) 1000 2000 3000 small tree (number/tia)x100 0 20 40 60 80 1000 2000 3000 c TJ P 3 a ■o 5* CD 1000 2000 3000 20 40 80 IE . H . i before uppers! t -m . i Y E . H -1 1 o ® ■§ t EH . 1 | 3 cr CL CP o o5 S. cp 3 g O CL £ -tj S 3 §1 § s CP o 1000 2000 3000 20 40 60 60 Q. 2L O T3 CD < CP (K) I 1000 2000 3000 20 40 80 5- S n . ] before sav [•»] III 1 p 3 m . 1 3 Si P | ED . ] 1 1 § sandhill upland pine upperslope midslope savanna TCIS RLBF BS06 TC36 330 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 ^ r-j rt . c/5 . - g C >£ « • - oo G- g * G ~ O o c PQ .2 o • ^ ^ & S' S § 5 bo *£ boxplots and significant levels. LIU, HARCOMBE AND KNOX 331 Discussion The effects described here on structural characteristics of stands (reduced fuel load, reduced sapling and small-tree density, low mortality of large trees) are consistent with previous finding based on species composition change (Liu et al. 1997) that the vegetation types in Big Thicket studied were only moderately sensitive to the prescription fires. Field observations suggested that an important contributor to this modest sensitivity was rapid response of small hardwoods by regenerating and sprouting, a finding corroborated by many other studies (e.g. , Abraham- son 1984a; 1984b; Westman & O deary 1986; Malanson & Trabaud 1987; Waldrop et al. 1992). This study focused on short-term effects of prescribed fire. How this might translate into long-term impact will depend on how the species respond to repeated fires, whether repeated fire causes a shift in species composition, and how long the short-term impacts last relative to the frequency of prescribed burning. In the slope types, fire effects may disappear in a few years because few large stems are killed by fire. In the upland types, changes will persist for many years because many small trees or even large trees were killed by fire. Whether these types will undergo conversion to longleaf pine forest with continued burning depends on changes in species composition in newly recruited seedlings and saplings, particularly the successful establishment of longleaf pine seedlings. In the current landscape, wherever the longleaf pine is still dominant, it is not difficult to change the structure and appearance of that particular vegetation type. However, at sites where the longleaf pine once was present but is now rare, such as some sandhills and upper slopes, conversion to longleaf pine forest by means of prescribed burn¬ ing will be more difficult. For the midslope and lowerslope types, the intact canopy and low flammability may portend little change in the understory and future regeneration; the lack of response to fire is consistent with the idea that mixed pine-hardwood occurred on such sites in the presettlement landscape (Marks & Harcombe 1981; Harcombe et al. 1993). In this study, modification of vegetation by fire was limited to the dry end of the topographic-moisture gradient, and so the hypothesis of differential fire effects is supported. The effect of fire on current vegetation is conditioned by that vegetation, which is influenced by site characteristics. This is consistent with a growing body of literature (e.g., 332 THE TEXAS JOURNAL OF SCIENCE- VOL. 56(4), 2004 Platt et al. 1989; Gibson et al. 1990; Glitzenstein et al. 1995; 2003; Breininger et al. 2002; Drewa et al. 2002). The results are also consis¬ tent with evidence that present patterns and trends in natural vegetation in the Big Thicket area are strongly influenced by soil factors, site history, and fire (Marks & Harcombe 1981; Streng & Harcombe 1982; Liu 1992; Harcombe et al. 1993; Lin et al. 2004). This work supports an approach to prescribed fire that recognizes natural patterns and natural variation in fire intensity, and thereby promotes the natural diversity of communities and the complexity of the vegetation for which the region is famous. Acknowledgments We are grateful to Chet Cain, Gary Cox, Linda C. Kaiser, Julie Swindell, and Rebecca McCulley for their assistance in the field, and David McHugh of the Big Thicket National Preserve and Ike McWhorter of The Nature Conservancy for directing prescribed burn¬ ing. Special thanks are due to Dr. Katherine B. Ensor and Dr. Joe Ensor of Rice University for their assistance in statistical analysis, and to Jeff Glitzenstein and Donna Streng for their assistance in developing the rationale for this study and in reviewing the manuscript. This research was sponsored by the National Park Service and by The Nature Conservancy. This manuscript is derived from part of the senior author’s Ph.D. dissertation at Rice University. Literature Cited Abrahamson, W. G. 1984a. Post-fire recovery of Florida Lake Wales Ridge vegetation. Am. J. Bot., 71:9-21. Abrahamson, W. G. 1984b. Species responses to fire on the Florida Lake Wales Ridge. Am. J. Bot., 71:35-43. Breininger, D. R., B. W. Duncan & N. J. Dominy. 2002. Relationships between fire frequency and vegetation type in pine flatwoods of east-central Florida, USA. Natural Areas Journal, 22:186-193. Christensen, N. L. 1981. Fire regimes in southeastern ecosystems. Pp. 112-136, in Fire regimes and ecosystem properties. Edited by H. L. Mooney. USDA For. Serv. Gen. Tech. Rep. WO-26, 594 pp. Christensen, N. L. 1988. Vegetation of the southeastern Coastal Plain. Pp. 317-364 in North America terrestrial vegetation. Edited by M. G. Barbour and W. D. Billings. Cambridge University Press, England, 434 pp. Correll, D. S. & M. C. Johnston. 1979. Manual of the vascular plants of Texas. The University of Texas at Dallas, Dallas, Texas, 1881 pp. Day, R. W. & G. P. Quinn. 1989. Comparisons of treatments after an analysis of variance in ecology. Ecol. Monogr., 59: 433-463. Drewa, P. B., W. J. Platt & E. B. Moser. 2002. Fire effects on resprouting of shrubs in LIU, HARCOMBE AND KNOX 333 headwaters of southeastern longleaf pine savannas. Ecology, 83:755-767. Gibson, D. J., D. C. Hartnett & G. L. S. Merrill. 1990. Fire temperature heterogeneity in contrasting fire prone habitats: Kansas tallgrass prairie and Florida sandhill. Bull. Torrey Bot. Club, 117:349-356. Glitzenstein, J, W. J. Platt & D. R. Streng. 1995. Effects of fire regime and habitat on tree dynamics in north Florida longleaf pine savannas. Ecol. Monogr, 65:441-476. Glitzenstein, J. S., D. R. Streng & D. D. Wade. 2003. Fire frequency effects on longleaf pine (Pinus palustris P. Miller) vegetation in south Carolina and northeast Florida. Natural Areas Journal, 23:22-37. Harcombe, P. A., J. S. Glitzenstein, R. G. Knox, S. L. Orzell & E. L. Bridges. 1993. Vegetation of the longleaf pine region of the west Gulf Coastal Plain. Pp. 83-104, in Proceedings of the Tall Timbers Fire Ecology Conference No 18: the longleaf pine ecosystem: ecology, restoration, and management, 30 May— 2Jun. 1991, Tallahassee, FL. Edited by S. M. Hermann. Tall Timbers Research Station, Tallahassee, FL., 418 pp. Lertzman, K. P. & J. Fall. 1998. From forest stands to landscapes: spatial scales and the roles of disturbances. Pp. 339-367, in Peterson, D.L. Parker, V.T., (Eds.) Ecological scale: theory and applications. Columbia University Press, NY, USA, 615 pp. Lin, J., P. A. Harcombe, M. R. Fulton & R. W. Hall. 2004. Comparative analysis of growth and mortality among saplings in a dry oak-pine forest in southeast Texas. Texas Journal of Science 56(4):299-318. Liu, C., P. A. Harcombe & R. G. Knox. 1997. Effects of prescribed fire on the composition of woody plant communities in southeastern Texas. J. Veg. Sci., 8:495-504. Liu, C. 1992. Vegetation patterns in fire-prone habitats, Southeast Texas, U. S. A. Master of Science thesis, Rice University, Houston, Texas, 108 pp. Malanson, G. P. & L. Trabaud. 1987. Ordination analysis of components of resilience of Quercus coccifera Garrigue. Ecology, 68:463-472. Marks, P. L. & P. A. Harcombe. 1981. Forest vegetation of the Big Thicket, southeast Texas. Ecol. Monogr., 51:287-305. Platt, W. J., J. S. Glitzenstein & D. R. Streng. 1989. Evaluating pyrogenicity and its effects on vegetation in longleaf pine savannas. Pp. 143-161, in Proceedings of Tall Timbers fire ecology conference No 17: High-intensity fire in wildlands— management challenges and options. Tallahassee, FL. Edited by S. M. Hermann, Tall Timbers Research Station, Tallahassee, FL., 418 pp. Quarterman, E. & C. Keever. 1962. Southern mixed hardwood forest: climax in the southeastern coastal plain, U.S.A. Ecol. Monogr., 32:167-185. Renkin, R. A. & D. G. Despain. 1992. Fuel moisture, forest type, and lightning-caused fire in Yellowstone National Park. Can. J. For. Res., 22:37-45. Hoshmand, A. R. 1994. Experimental research design and analysis. CRC Press, Inc., Boca Raton, FL. pp. 46-48. Romme, W. H. & D. H. Knight. 1981. Fire frequency and subalpine forest succession along a topographic gradient in Wyoming. Ecology, 62:319-326. SAS Institute. 1992a. SAS/STAT user’s guide, Vol. 1, ver. 6, 4 ed. SAS Institute Inc., Cary, NC. SAS Institute. 1992b. SAS/STAT user’s guide, Vol. 2, ver. 6, 4 ed. SAS Institute Inc., Cary, NC. Schwartz, M. W. 1994. Natural distribution and abundance of forest species and communities in northern Florida. Ecology, 75:687-705. Streng, D. R. & P. A. Harcombe. 1982. Why don’t east Texas savannas grow up to forest? Am. Midi. Nat., 108:278-294. Waldrop, T. A., D. L. White & S. M. Jones. 1992. Fire regimes for pine-grassland 334 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 communities in the southeastern United States. For. Ecol. Manage., 47:195-210. Ware, S., C. Frost & P. D. Doerr. 1993. Southern mixed hardwood forest: the former longleaf pine forest. Pp. 447-493, in Biodiversity of the southeastern United States. Edited by W. H. Martin S. G. Boyce & A. C. Esternacht. John Wiley and Sons, Inc., NY, 373 pp. Wahlenberg, W. G. 1946. Longleaf pine: its use, ecology, regeneration protection, growth, and management. Charles Lathrop Pack Forestry Foundation, Washington, D.C., 429 pp. Westman, W. E. & J. F. O’Leary. 1986. Measures of resilience: the response of coastal sage scrub to fire. Vegetatio, 65:179-189. PAH at: harcomb@rice.edu TEXAS J. SCI. 56(4):335-346 NOVEMBER, 2004 GROWTH OF CHINESE TALLOW TREE {SAPIUM SEBIFERUM) AND FOUR NATIVE TREES UNDER VARYING WATER REGIMES Bradley J. Butterfield*, William E. Rogers and Evan Siemann Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77005 * Current address: School of Life Sciences, Arizona State University Tempe, Arizona 85287 Abstract.— Abiotic stress tolerance may play a role in the invasion and spread of Chinese tallow tree ( Sapium sebiferum). A greenhouse experiment was conducted to determine the effects of water stress on the growth of Sapium and four tree species native to the south¬ eastern United States. Species identity, water treatment, and their interaction significantly influenced growth rate and mass of seedlings. No native species had as high an average growth rate as Sapium. Indeed, Sapium had a higher growth rate than every native species in every water treatment with the exception of a single native species ( Liquidambar styraciflua L.) in the drier treatments (pulse drought, well watered). Sapium exhibits the potential to thrive at any point along the water gradient present in southeastern floodplain forests . Plant species distributions often reflect abiotic conditions. Species composition may shift along a resource gradient based on efficiency of resource use at different concentrations (Tilman 1982; 1985; Huston & Smith 1987). Species distributions in some landscapes are based pri¬ marily on one resource, and in such cases analysis of the performance of species along a gradient of that resource can be useful in predicting community composition (Tilman 1987). Similarly, comparisons of the performance of an invasive species and native species along a gradient of the most limiting abiotic factor in an ecosystem may be a good predictor of the conditions in which the invasive will displace natives (Alpert et al. 2000; Sakai et al. 2001; Daehler 2003). Invasive species often have very different ecological attributes from species in their introduced range (Bruce et al. 1997; Busch & Smith 1995). Comparisons between native and exotic congeners (Schierenbeck et al. 1994; Mack 1996; Gerlach & Rice 2003) and between ecologically similar native and exotic species (Nijjer et al. 2002; Rogers & Siemann 2002; Daehler 2003; Siemann & Rogers 2003a) have produced informa¬ tive results. Studies analyzing plant growth along a resource gradient can be useful for identifying traits that may lead to the competitive dominance of invasive species, as well as for predicting potential range expansions. 336 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 In southeastern floodplain forests, water is a major determinant of the distribution of tree species (e.g. Hall & Harcombe 1998; Wall & Darwin 1999; Denslow & Battaglia 2002; Ernst & Brooks 2003). The elevation- al heterogeneity of these systems positions different plant communities within close proximity to each other (Christensen 2000), which likely results in distribution of propagules into a wide range of moisture conditions, making seedling establishment and growth important aspects of population dynamics. Sloughs and depressions are often flooded year round, while other areas of bottomland forests experience seasonal flooding. Upland areas may never flood, and often experience seasonal droughts (Christensen 2000). Tree species in these forests can be expected to follow different growth strategies depending on their distribution along a water gradient. Stress tolerance is important at extreme elevations where abiotic factors limit seedling growth and survival, while competitive ability is more important in less stressful environments. Stress tolerant species are expected to have relatively restricted phenotypic responses to external stimuli since survival depends on highly conservative growth strategies (Grime 1974; 1977; Campbell & Grime 1992). This is often reflected in slow growth rates and negligible increases in mass and growth rate in less stressful conditions (Grime 1974; 1977; Pigliucci 2001). Tree species adapted to more favorable conditions can be expected to maxi¬ mize resource assimilation and grow rapidly, since biotic competition is often more important than in stressful environments (Grime et al. 1986). A greenhouse experiment was conducted to determine the growth and performance of Sapium sebiferum (L.) Roxb. (Chinese tallow tree) and four native tree species under a range of water conditions representative of natural conditions. Sapium has invaded a variety of ecosystems in the southeastern United States. Even though it thrives in early successional conditions and has extremely high growth rates (Siemann & Rogers 2003a), seedlings are also shade tolerant (Jones & McLeod 1989; Rogers & Siemann 2002; 2003; but see Lin et al. 2004) and flood tolerant (Jones & Sharitz 1990; Conner et al. 1997, 2001). It was predicted that the range of soil moisture conditions in which native tree species sustain high growth rates and mass production would be restricted. Adaptations to particular habitats were expected to cause tradeoffs between stress tolerance and other traits such that native species with the greatest growth rates in optimal conditions should be BUTTERFIELD, ROGERS & SIEMANN 337 more sensitive to extreme conditions. Because of its widespread distri¬ bution in floodplain forests and invasive nature, Sapium was expected to have a higher growth rate and produce more mass than all native species under all water conditions. Methods The experiment was conducted in a climate controlled greenhouse in Houston, Texas between March and August 2003. The roof and walls of the greenhouse were clear glass, and humidity was approximately 100%. Pinus taeda L. (loblolly pine), Liquidambar styraciflua L. (sweetgum), Nyssa aquatica L. (water tupelo), and N. sylvatica Marsh, var. sylvatica (blackgum) seeds were acquired commercially (Louisiana Forest Seed Co. Lecompte, LA). Sapium sebiferum seeds were collect¬ ed in Texas and Georgia. In Texas, seeds were collected from many different trees at the Armand Bayou Nature Center, approximately 35 km southeast of Houston. In Georgia, seeds were collected from numer¬ ous trees on Sapelo Island, a barrier island approximately 55 km south of Savannah. Seeds of all species were germinated in topsoil in early March and transplanted into individual 11 liter plastic pots in April. Potted seedlings of each native species plus one of Texas and Georgia Sapium were assigned to a random position in each of twenty-four 160 liter plastic tubs (a split-plot design). Seedlings were watered daily for two weeks before initiation of the treatments. Each tub was randomly assigned one of four watering treatments, with 6 tubs per water treatment. The treatments were: (1) Control - Pots were watered daily until water flowed out of the bottom of the pot; (2) Flooded - Pots were permanently submerged in water (1-3 cm above soil surface) for the duration of the 16- week experiment. Evaporative losses were replaced with de-ionized water to avoid salt accumulations; (3) Pulsed flood - Pots received the control water treatment for two weeks followed by flood treatment for the following two weeks. This four-week cycle was completed four times during the course of the experiment; (4) Pulsed drought - Pots received the control water treat¬ ment for the first two weeks of each four- week cycle, but received no water for the latter two weeks. Initial stem heights, basal diameters, and leaf counts were recorded for each plant on 9 April. Stem height and number of leaves per seedling were measured weekly during the experiment. After 16 weeks, all of the plants were harvested. Roots, stems, and leaves were separ¬ ated and dried at 60 °C for 96 hours before dry mass was measured. 338 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Figure 1 . Dependence of the growth rate of each tree species on water treatment (mean + 1 SE). Letters indicate significantly different means (P<0.05) within (lowercase) and among (uppercase) species. The mean growth rate across all species and treatments is provided as a reference. All statistical analyses were conducted in SAS Version 8 (SAS Institute 1999). ANOVAs were performed using PROC MIXED to analyze the effects of species identity (split-plot factor), water treatment (whole-plot factor) and their interaction (split-plot factor) on growth rate, total biomass, and mass allocation. Stem growth rate was measured as In (final height/initial height). Total mass was log transformed for analyses. Proportion of total mass allocated to root, stem, and leaf tissues were measured as organ mass/total mass. Fisher’s Least Signifi¬ cant Difference (LSD) was used for means contrasts among treatments. Results Stem growth rate depended on species (F5 >100 = 124.5; P< 0.0001), water treatment (F3 20 — 99.1; P< 0.0001), and their interaction (F3j100 = 18.3; P< 0.0001; Fig. 1). Georgia Sapium grew most rapidly, fol¬ lowed by Liquidambar, Texas Sapium , N. aquatica, N. sylvatica , and Pinus (Fig. 1). Georgia Sapium varied the least in growth across water treatments (1.34- fold difference between treatment in which it grew fastest and the one in which it grew slowest), followed closely by N. aquatica (1.49-fold) and Texas Sapium (1.65-fold), then Pinus (2.17- fold), Liquidambar (4. 34- fold), and N. sylvatica (9. 00- fold). BUTTERFIELD, ROGERS & SIEMANN 339 50 m s «s O H Species Figure 2. Dependence of total mass of each species on water treatment (mean +1 SE). Letters indicate significantly different means (P < 0.05) within (lowercase) and among (uppercase) species. The mean total mass across all species and treatments is provided as a reference. Total mass depended on species (F5A00 = 893.8; P<0.0001; Fig. 2), water treatments (F3 20 = 1 17.7; PcO.0001), and their interaction (F3 100 = 24.2; P< 0.0001; Fig. 2). Nyssa aquatica and Texas Sapium had the highest total mass (Fig. 2), but N. aquatica seedlings were on average between two and four times as tall as the other species at the beginning of the experiment, which likely contributed to the high final mass (Fig. 3). In a split-plot design these differences in starting sizes are difficult to account for with covariates. Texas Sapium had a slightly larger final mass than Georgia Sapium , but this can also be reconciled by initial heights (Fig. 3). Liquidambar and N. sylvatica were both significantly lower than Georgia Sapium but were similar with respect to each other. All species but Pinus exhibited significant reductions in total mass in response to permanent flooding (Fig. 2). Proportion of total mass allocated to roots depended on species identity (F5l00 = 81.4; PcO.OOOl) but not on water treatment (F3 20 = 1.2; P = 0.35) or their interaction (F3 100 = 1.8; P = 0.10). Propor¬ tional leaf mass depended significantly on both species identity (Fs wo = 198.6; PCO.0001) and water treatment (F320 = 5.54; PcO.Ol) but not on their interaction (F3 100 = 0.46; P = 0.94). Stem mass proportion 340 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 O) CD O (/) (/) CD 03 O h- 1 - Sap TX Pinus. N.a 0 Sapium GA A Sapium TX □ Liquidambar V N. aquatica o N. sylvatica 0 Pinus ~r— 10 ~ i — 15 — r- 20 Initial Height (cm) Figure 3. Initial height at planting versus log (total mass) for each species. — i 25 was significantly affected by species identity (F5 100 = 349.2; P< 0.0001) and the interaction between species identity and water treatment (F3 100 = 15.1; P< 0.0001), but not water treatment alone (F3 20 = 0.7; P = 0.54). Sapium seedlings allocated approximately 30% of mass to leaves, 30% to stems and about 40% to roots. Nyssa aquatica had a root to shoot ratio similar to Sapium , but allocated markedly less mass to leaves. Liquidambar and N. sylvatica were similar to each other in their stem versus leaf allocation ratios, but N. sylvatica had the highest root to shoot ratio of any species, while Liquidambar allocated a relatively low amount of mass belowground. Pinus had the lowest root to shoot and stem to leaf ratios (Table 1). Discussion The results of this experiment suggest that Sapium has characteristics of both stress tolerant and rapidly growing species without experiencing the same magnitudes of tradeoffs between these characteristics as are evident for the native tree species in this study. Sapium had high growth rates across all water treatments and experienced only modest BUTTERFIELD, ROGERS & SIEMANN 341 Table 1. Proportion of total mass allocated to root, stem, and leaf parts by species. Species % Total Mass Root Stem Leaf Sapium GA 39 31 30 Sapium TX 42 28 30 Liquidambar 31 27 42 Nyssa aquatica 37 45 18 Nyssa sylvatica 44 21 35 Pinus 26 16 58 reductions in growth in response to water stress (Figs. 1,2). Sapium' s stress tolerance appears to extend across the entire experimental water gradient. Within this range of tolerance, Sapium' s growth rate was always high relative to most native species. The only species that grew faster than Sapium was Liquidambar in drier treatments, and it was a very poor performer in the flood treatment (Fig. 1). While Sapium may not be able to out perform N. aquatica in perma¬ nently flooded conditions if differences in initial seedling sizes observed here are typical of field conditions (Fig. 2), Sapium seedlings may still survive to reproductive maturity due to relatively low competition in such stressful environments (Ernst & Brooks 2002). The high leaf-to- stem mass ratio of Sapium relative to N. aquatica also indicates that Sapium may be able to survive in very wet areas with dense canopies in which N. aquatica may not be able to capture enough light to grow well (Jones & Sharitz 1990). Sapium should also be able to exist in the middle-to-high moisture range of Liquidambar and N. sylvatica. In areas that are highly favorable for either of the natives, Sapium' s shade tolerance (Rogers & Siemann 2002; 2003) and ability to reproduce as a sub-canopy species may favor its presence. The performance of Sapium in areas with drier moisture regimes was not tested in this study, but it has been shown to be much less successful in dry uplands that support Liquidambar and N. sylvatica (Hall & Harcombe 1998; Harcombe et al. 2002; Lin et al. 2004). Sapium also exhibited positive traits similar to Liquidambar and N. sylvatica. High growth rates in non- flood treatments (Fig. 1) and high leaf-to-stem ratios (Table 1) of these two natives are indicative of seedlings adapted to relatively nutrient-rich, disturbed areas (Grime 1974; 1977). Nyssa sylvatica had high root : shoot ratios (Table 1) and relatively greater mass production in flood treatments (Fig. 2) indicating that seedlings of this species may survive periods of flooding and grow 342 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 rapidly when floodwaters subside (Grime et al. 1986). Liquidambar performed as a more typical gap species, allocating more resources to stem growth rate in a relatively narrow range of dry to moist soils (Fig. 1, Table 1). Sapium exhibited growth traits that were characteristic of these two native species including high root to shoot ratios, intermediate leaf to stem ratios, and high growth rates (Table 1). The potential gradient distributions of native seedlings in this experi¬ ment corresponded relatively well with observed distributions of mature trees. Nyssa aquatica was clearly the most tolerant of both flood treat¬ ments. Mature N. aquatica trees often coexist with Taxodium distichum (L.) Rich, as the dominant species in anoxic bottomlands (Marks & Harcombe 1981; Visser & Sasser 1995). Nyssa sylvatica seedlings can likely survive periodic flooding while taking advantage of intermittent dry periods, as well as thrive in moist areas. Distribution of mature individuals of this species also covers a wide range of moisture condi¬ tions, including areas with seasonal flooding and drought (Keeland et al. 1997). Liquidambar performed best in moist to dry conditions, which does appear to deviate slightly from the observed distribution of mature trees. Liquidambar is primarily a floodplain species (Marks & Harcombe 1981; Denslow & Battaglia 2002; Ernst & Brooks 2003), but the drought treatment in this experiment was not severe enough to simulate upland conditions. Therefore, dry conditions in this experiment are similar to more elevated areas within a floodplain. Light may also play an important role in the distributions of Liquidambar and N. sylvatica. Their strategy of maximizing shoot growth in this study is an adaptation consistent with these species being shade intolerant (Hall & Harcombe 1998; Lin et al. 2004). The high variability of total mass and mass allocation under varying water regimes also indicates that these species maximize growth under relatively specific, favorable conditions. Pinus was more flood tolerant in this study than was expected (Kozlowski 1997) and was relatively incongruous with respect to distri¬ bution of mature trees. Light availability is another important predictor of Pinus distribution in nature, which may explain this discrepancy (Harcombe et al. 2002). The apparent flood tolerance may also be a reflection of the fine-grained soils used in this study, which may have stunted the growth of seedlings in all water treatments. It is not clear what mechanism would contribute to the superior performance of the invasive species observed in this study. One possi¬ bility is that Sapium possesses novel physiological or biochemical traits as a result of taxonomic novelty or an evolutionary history in a different BUTTERFIELD, ROGERS & SIEMANN 343 biotic province or under different selection pressures (Tilman 1999). This possibility cannot be discounted. Sapium is unusual in that it is the only tree in the southeastern U.S. that is a member of the Euphorbia- ceae. In addition, Sapium is the only plant from Asia in this study, and it is possible that in general Asian trees would outperform North American trees in this type of experiment. Finally, Sapium has a long history of being cultivated in Asia for its oil rich seeds, and was origin¬ ally introduced to the U.S. as an agricultural crop (Bruce et al. 1997). The traits observed here could be the result of artificial selection prior to introduction to North America. There are, however, proximate eco¬ logical factors that contribute to the success of invasive plants that may have relevance to the results of this experiment. Low herbivore loads in the introduced range is one of the factors that is widely believed to contribute to the greater vigor of exotic plants (Keane & Crawley 2002), and has been shown to contribute to Sapium' s success (Rogers & Siemann 2002; Siemann & Rogers 2003a). One way in which plants may benefit from low herbivore loads is by a plastic phenotypic response to low losses to herbivores in which additional resources are used for growth (Elton 1958). In this greenhouse study, however, there was negligible damage to any plants, either natives or Sapium , so this is unlikely to be the cause of Sapium' s unusual combina¬ tion of high growth rates and high flood tolerance observed here. In fact, Liquidambar , the only species that was able to outperform Sapium in this study, sometimes suffers extremely high herbivore damage in natural settings (Siemann & Rogers 2003a) which would only strengthen the conclusion that Sapium has an unusual combination of growth and tolerance to stress. Release from herbivory may also affect plant performance by direc¬ tional selection on plant defense and growth (Blossey & Notzold 1995). Sapium' s high level of vigor in a wide range of conditions may be due to genetic responses to low herbivory resulting in reallocation of re¬ sources from defense to faster growth (Siemann & Rogers 2001 , 2003b) and perhaps also to phenotypic plasticity (Bazzaz et al. 1987, Alpert et al. 2000). If this is true, the tradeoff between growth rates and stress tolerance examined in this study may be applicable to plant responses under varying conditions of other resources and other forms of stress. Comparisons of the results of a greenhouse study such as this and natural distributions may give insights into the role of other factors, such as herbivory, in determining plant distributions. 344 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 This study adds further support to the importance of stress tolerance in the invasion of southeastern floodplain forests by Sapium. The two primary determinants of species distribution in these forests are light and water (Hall & Harcombe 1998). Other studies have demonstrated Sapium' s ability to grow in a variety of light levels (Jones & McLeod 1989; Rogers & Siemann 2002; 2003; Siemann & Rogers 2003c). In accordance with other studies on soil moisture regimes (Jones & Sharitz 1990; Barrilleaux & Grace 2000; Conner et al. 2001), this experiment confirms that Sapium can perform well under a wide range of water conditions. Regardless of the mechanism, Sapium is able to exhibit traits of both rapidly growing and stress tolerant species, which may allow it to spread into bottomlands with anoxic soils as well as into seasonally dry areas of floodplain forests. Perhaps more importantly, this study demonstrates the ability of an introduced species to minimize tradeoffs that substantially affect the performance and growth strategies of native species. Acknowledgments We would like to thank Saara DeWalt for statistical assistance and comments on the manuscript; Paul Harcombe and two anonymous reviewers for comments on the manuscript; Candice Donahue, Maria Hartley, and Somereet Nijjer for comments; Philemon Chow, Zac McLemore, Rachel Tardif, and Terris White for assistance; Armand Bayou Nature Center, University of Georgia Marine Institute and Georgia Department of Natural Resources for permission to collect seeds on their properties; the National Science Foundation (DEB-9981654) and EPA (R82-8903) for support. Literature Cited Alpert, P., E. Bone & C. Holzapfel. 2000. Invasiveness, invasibility and the role of environmental stress in the spread of non-native plants. Perspectives in Plant Ecology, Evolution and Systematics, 3:52-66. Barrilleaux, T. C. & J. B. Grace. 2000. Growth and invasive potential of Sapium sebiferum (Euphorbiaceae) within the coastal prairie region: the effects of soil and moisture regime. American Journal of Botany, 87:1099-1106. Bazzaz, F. A., N. R. Chiariello, P. D. Coley & L. F. Pitelka. 1987. Allocating resources to reproduction and defense. BioScience, 37:58-67. Blossey, B. & R. Notzold. 1995. Evolution of increased competitive ability in invasive nonindigenous plants: a hypothesis. Journal of Ecology, 83:887-889. Bruce, K. A., G. N. Cameron, P. A. Harcombe & G. Jubinsky. 1997. Introduction, impact on native habitats, and management of a woody invader, the Chinese tallow tree, Sapium sebiferum (L.) Roxb. Natural Areas Journal, 17:255-260. Busch, D. E. & S. D. Smith. 1995. Mechanisms associated with decline of woody species BUTTERFIELD, ROGERS & SIEMANN 345 in riparian ecosystems of the southwestern U.S. Ecological Monographs, 65:347-370. Campbell, B .D. & J. P. Grime. 1992. An experimental test of plant strategy theory. Ecology, 73:15-29. Christensen, N. L. 2000. Vegetation of the southeast coastal plain. Pp. 357-396, in North American Terrestrial Vegetation (M. G. Barbour. & W. D. Billings, eds.). Cambridge University Press, New York, NY, 708 pp. Conner, W. H., K. W. McLeod & J. K. McCarron. 1997. Flooding and salinity effects on growth and survival of four common forested wetland species. Wetlands Ecology and Management, 5:99-109. Conner, W. H., L. W. Inabinette & C. A. Lucas. 2001. Effects of flooding on early growth and competitive ability of two native wetland tree species and an exotic. Castanea, 66:237-244. Daehler, C. C. 2003. Performance comparisons of co-occurring native and alien invasive plants: implications for conservation and restoration. Annual Review of Ecology and Systematics, 34:183-211. Denslow, J. S. & L. L Battaglia. 2002. Stand composition and structure across a changing hydrological gradient: Jean Lafitte National Park, Louisiana, USA. Wetlands, 22:738-752. Elton, C. S. 1958. The ecology of invasions by plants and animals. Chapman and Hall, London, UK, 181 pp. Ernst, K. A. & J. R. Brooks. 2003. Prolonged flooding decreased stem density, tree size and shifted composition towards clonal species in a central Florida hardwood swamp. Forest Ecology and Management, 173:261-279. Gerlach, J. D. & K. J. Rice. 2003. Testing life history correlates of invasiveness using congeneric plant species. Ecological Applications, 13:167-179. Grime, J. P. 1974. Vegetation classification by reference to strategies. Nature, 250:344-347. Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist, 111:1 169-1 194. Grime, J. P., J. C. Crick & J. E. Rincon. 1986. The ecological significance of plasticity. Pp. 5-30, in Plasticity in Plants (D. H. Jennings & A. J. Trewavas, eds.). The Company of Biologists Limited, Cambridge, 372 pp. Hall, R. B. W. & P. A. Harcombe. 1998. Flooding alters apparent position of floodplain saplings on a light gradient. Ecology, 79:847-855. Harcombe, P. A., C. J. Bill, M. Fulton, J. S. Glitzenstein, P. L. Marks & I. S. Elsik. 2002. Stand dynamics over 18 years in a southern mixed hardwood forest, Texas, USA. Journal of Ecology, 90:947-957. Huston, M.. & T. Smith. 1987. Plant succession: life history and competition. American Naturalist, 130:168-198. Jones, R. H. & K. McLeod. 1989. Shade tolerance in seedlings of Chinese tallow tree, American sycamore, and cherrybark oak. Bulletin of the Torrey Botanical Club, 116:371-377. Jones, R. H. & R. R. Sharitz. 1990. Effects of root competition and flooding on growth of Chinese tallow tree seedlings. Canadian Journal of Forest Research, 20:573-578. Keeland, B. D., W. H. Conner & R. R Sharitz. 1997. A comparison of wetland tree growth response to hydrologic regime in Louisiana and South Carolina. Forest Ecology and Management, 90:237-250. Keene, R. M. & M. J. Crawley. 2002. Exotic plant invasions and the enemy release hypothesis. Trends in Ecology and Evolution, 17:164-170. Kozlowski, T. T. 1997. Responses of woody plants to flooding and salinity. Tree Physiology Monograph No. 1. Heron Publishing, Victoria, Canada, 29pp. 346 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Lin, J., P. A. Harcombe, M. R. Fulton & R. W. Hall. 2004. Sapling growth and survivorship as affected by light and flooding in a river floodplain forest of southeast Texas. Oecologia, 139:399-407. Mack, R. N. 1996. Predicting the identity and fate of plant invaders: emergent and emerging approaches. Biological Conservation, 78:107-121. Marks, P. L. & P. A. Harcombe. 1981. Forest vegetation of the Big Thicket, southwest Texas. Ecological Monographs, 51 :287-305. Nijjer, S., R. A. Lankau, W. E. Rogers & E. Siemann. 2002. Effects of temperature and light on Chinese Tallow {Sapium sebiferum) and Texas Sugarberry ( Celtis laevigata) seed germination. Texas Journal of Science, 54(l):63-68. Pigliucci, M. 2001. Phenotypic plasticity: beyond nature and nurture. Johns Hopkins, Baltimore, MD, 328 pp. Rogers, W. E. & E. Siemann. 2002. Effects of simulated herbivory and resource availability on native and invasive exotic tree seedlings. Basic and Applied Ecology, 3:297-307. Rogers, W. E. & E. Siemann. 2003. Effects of simulated herbivory and resources on Chinese tallow tree {Sapium sebiferum , Euphorbiaceae) invasion of native coastal prairie. American Journal of Botany, 90:241-247. Sakai, A. K., F. W. Allendorf, J.S. Holt, D. M. Lodge, J. Molofsky, K. A. With, S. Baughman, R. J. Cabin, J. E. Cohen, N. C. Ellstrand, D. E. McCauley, P. O’Neil, I. M. Parker, J. N. Thompson & S. G. Weller. 2001. The population biology of invasive species. Annual Review of Ecology and Systematics, 32:305-332. SAS Institute. 1999. SAS version 8. SAS Institute, Cary, North Carolina, USA. Schierenbeck, K. A., R. N. Mack & R. R. Sharitz. 1994. Effects of herbivory on growth and biomass allocation in native and introduced species of Lonicera. Ecology, 75:1661-1672. Siemann, E. & W. E. Rogers. 2001. Genetic differences in growth of an invasive tree species. Ecology Letters, 4:514-518. Siemann, E. & W. E. Rogers. 2003a. Herbivory, disease, recruitment limitation and the success of an alien tree species. Ecology, 84:1489-1505. Siemann, E. & W. E. Rogers. 2003b. Reduced resistance of invasive varieties of the alien tree Sapium sebiferum to a generalist herbivore. Oecologia, 135:451-457. Siemann, E. & W. E. Rogers. 2003c. Changes in light and nitrogen availability under pioneer trees may indirectly facilitate tree invasions of grasslands. Journal of Ecology, 91:923-931. Tilman, D. 1982. Resource Competition and Community Structure. Princeton University Press, Princeton, NJ, 296 pp. Tilman, D. 1985. The resource-ratio hypothesis of plant succession. American Naturalist, 125:827-852. Tilman, D. 1987. Secondary succession and the pattern of plant dominance along experimental nitrogen gradients. Ecological Monographs, 57:189-214. Tilman, D. 1999. The ecological consequences of changes in biodiversity: a search for general principles. Ecology, 80:1455-1474. Visser, J. M. & C. E. Sasser. 1995. Changes in tree species composition, structure and growth in a bald cypress water tupelo swamp forest, 1980-1990. Forest Ecology and Management, 72:119-129. Wall, D. P. & S. P. Darwin. 1999. Vegetation and elevational gradients within a bottomland hardwood forest of southeastern Louisiana. American Midland Naturalist, 142:17-30. BJB at: Bradley.J.Butterfield@asu.edu TEXAS J. SCI. 56(4): 347-356 NOVEMBER, 2004 EFFECTS OF TEMPERATURE AND MULCH DEPTH ON CHINESE TALLOW TREE (SAP1UM SEBIFERUM) SEED GERMINATION Candice Donahue*, William E. Rogers and Evan Siemann Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77005 * Current address: Armand Bayou Nature Center PO Box 58828 Houston, Texas 77258 Abstract.— Shredding mowers can be used in prairie and savannah restoration to quickly eliminate trees, such as the invasive Chinese tallow tree ( Sapium sebiferum), and leave a layer of mulch on the ground. Sapium has shown highest germination rates in fluctuating daily temperatures, and mulch has been shown to damp those fluctuations in the field. A lab study was conducted to separate direct effects of mulch depth and indirect effects from changes in soil temperatures on Sapium seed germination. Sapium seeds were exposed to different combinations of mulch depth and temperature oscillations. Sapium seeds showed highest germination in large temperature oscillation treatments regardless of the depth of the mulch. Seedlings were able to emerge through mulch up to 10 cm deep, the maximum used in this study. While herbicide use appears to be necessary because of resprouting from stumps, this study indicates that mulching Sapium trees shows promise as a restoration tool by removing existing trees as well as by reducing Sapium regeneration from seed through the indirect effects of mulch on seed germination. The lower subsequent seedling numbers might reduce the frequency and intensity of future herbicide treatments. The invasive Chinese tallow tree ( Sapium sebiferum (L.) Roxb.), Euphorbiaceae, was introduced to the United States from Asia in 1772 and has spread across the southeastern states (Barrilleaux & Grace 2000; Bruce et al. 1997). Grasslands have always been subject to woody encroachment, but the great seed output, bird dispersal, rapid growth, and adaptation to wide environmental conditions of Sapium (Renne & Gauthreaux 2000; Rogers et al. 2000; Siemann & Rogers 2003a) have allowed it to become the most serious threat to endangered prairies along the upper coast of the Gulf of Mexico (Grace 1998). Once Sapium becomes established, it shades out the native herbaceous vegetation and forms a monospecific forest (Bruce et al. 1997; Siemann & Rogers 2003b). This also displaces native animal species, such as several federally endangered grassland birds (Herkert et al. 2003; Perkins et al. 2003). The loss of prairie bunchgrasses and rapid decomposition of Sapium litter (Cameron & Spencer 1989) leave the soil bare beneath the trees; such a condition may reduce bioremediation of anthropogenic 348 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 pollutants and speed the flow of water and sediments to rivers (Fajardo et al. 2001; Harbor et al. 1995; Liaghat & Prasher 1996). Sapium invasion is not limited to prairies. A 20-yr forest dynamic study (Harcombe et al. 1999) revealed that Sapium had increased dramatically in the Neches Bottom Unit of the Big Thicket National Preserve between 1981 and 1995. Among small saplings, Sapium growth was three times the median of all species studied during that period, and among large saplings, Sapium growth significantly exceeded that of all other species. In another study of the area, Hall & Harcombe (1998) documented an interaction of shade tolerance and flood tolerance among the species present. For example, species often were found in higher light conditions than would be expected from their known tolerance for shade, apparently having to make environmental trade-offs to survive both stresses of shade and flooding. Since Sapium is known to perform well in shade (Jones & McLeod 1989; Rogers 2002) and withstand flooding (Conner 1994; Grace 1998), it may become a serious threat to native tree species in the Big Thicket. Effective control for Sapium has been elusive, and a great percentage of coastal prairie has been displaced by this exotic species. A promising new technique for prairie restoration uses shredding mowers to mulch stands of Sapium. This method employs a large shredding mower to chip entire trees at ground level. Herbicide is manually applied to the cut surface of the stumps to reduce resprouting. For restoration to be successful, Sapium regeneration needs to be controlled while simul¬ taneously promoting native prairie plant regeneration. Mulch from Sapium trees may contribute to successful prairie restoration by limiting Sapium regeneration from seed. However, mulch depths necessary for suppression of Sapium seed germination and the mechanisms that contri¬ bute to suppression are not known. Armand Bayou Nature Center, located 44 km southeast of Houston, Texas, has twice mulched Sapium trees on invaded prairie with a shredding mower, once in summer of 2000 and again in fall 2002/spring 2003. In the 2000 restoration, the stand was more mature and resulting mulch depths ranged up to 15 cm. In the younger stand mulched in 2002/2003, average mulch depths were approximately 5 cm. The subsequent emergence of Sapium seedlings in the area mulched in 2000 appeared lower than in the area where Sapium trees were killed with herbicide and left standing. DONAHUE, ROGERS & SIEMANN 349 The mulch layer might have reduced germination by limiting day / night variation in surface soil temperatures. Experimental studies have shown highly variable germination rates for Sapium , depending on the geographic source of the seeds (Cameron et al. 2000) and the germina¬ tion protocols. Conway et al. (2000) only achieved 0-10% gemination on filter paper in petri dishes under an oscillating light and temperature regime, but Cameron et al. (2000) and Renne et al. (2001) achieved 26% and 22.5% gemination rates, respectively, for seeds planted in soil in greenhouses under natural temperatures and light. Seeds under these conditions would be expected to experience natural daily fluctuations in soil temperatures. In another study, highest germination rates were obtained for seeds planted in soil under experimentally controlled fluctuating daily temperatures (Nijjer et al. 2002). The objective of this lab study was to separate direct effects of mulch and indirect effects by changes in soil temperatures on Sapium seed germination by maintaining constant temperature regimes under varying mulch depths. If direct effects of mulch on seed germination are the primary cause of lower germination rates, then germination should decrease as mulch depth increases for all temperature treatments. How¬ ever, if indirect effects via changes in soil temperatures are more important, germination should be greatest in high oscillating tempera¬ tures regardless of the mulch depth. Materials and Methods Seeds of Sapium were collected from trees at the University of Houston Coastal Center in Galveston County, Texas, from August to September, 2002 and stored at room temperature. On 16 July 2003, 50 seeds were planted in each of 48 plastic bins (16 by 30 by 10 cm deep) on a 2.5 cm layer of commercially available topsoil and covered with another 2.5 cm layer of topsoil. Bins were randomly assigned to a temperature treatment (high oscillation, low oscillation, warm, and cool) and a mulch treatment (bare soil, 5 cm Sapium mulch, and 10 cm Sapium mulch) in a full-factorial design. Temperature treatments were chosen based on field soil temperatures measured during spring 2003 in the field that was mulched in late 2002 (Fig. 1). Bins were in a temperature controlled room (21 °C) without windows or artificial light for the duration of the experiment. Sapium germination is independent of light conditions (Nijjer et al. 2002). 350 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Figure 1. Sample of field soil temperatures recorded every 30 minutes, by mulch depth, in a field that had Sapium trees removed with a shredding mower in late 2002. Vertical bars indicate midnight on successive days. Electric roof de-icing cables (EASYHEAT, New Carlisle, IN) laid in the bottoms of the bins raised the soil temperatures. Cables passed once through low-oscillation bins and twice through high-oscillation and warm bins. Oscillation treatments were warmed for 16 hours and allowed to return to room temperature over eight hours. The high oscillation temperature maximum was 33 °C, and the low oscillation temperature maximum was 27 °C. The warm treatment was a constant 33 °C, and the cool treatment was constant room temperature (21 °C). Fresh Sapium mulch was collected from a recently mulched Sapium restoration area at Armand Bayou Nature Center. Mulch was spread evenly across the soil in the 5 cm and 10 cm mulch treatment bins. Plastic baffles were used to support the mulch layer at the edges of the 10 cm treatment bins. Because the 0 cm and 5 cm mulch treatments lost more heat to the air than the 10 cm mulch treatment and did not main¬ tain the desired soil temperatures, heavy-duty plastic sheeting was cut slightly larger than each bin and laid over the tops of the bins for these two treatments. The plastic was neither sealed to the bins nor in contact with the soil or mulch layers. DONAHUE, ROGERS & SIEMANN 351 Table 1 . Dependence of Sapium germination on experimental temperature and mulch depth treatments in an AN OVA. Factor df SS F- value P-value Temperature 3 112.2 123.5 <0.0001 Mulch Depth 2 2.0 3.4 <0.05 Temperature*Mulch 6 3.5 1.9 0.11 Error 36 10.9 All treatments were thoroughly watered three times each week until water drained from the bins, and newly germinated seeds were counted and removed from the bins during these periods. The experiment was conducted for 125 days, but no seeds germinated after 110 days. ANOVA was used to compare the different experimental treatments and Fisher’s LSD tests were used for post-hoc means contrasts (Statview 5.0, SAS Institute, 1998, Cary, North Carolina). Data were checked for normality and square root transformed to meet the assumptions of ANOVA. Data were back- transformed for presentation. Results Temperature treatment and mulch depth treatment, but not their interaction, had significant effects on seed germination; however, temperature alone explained 87% of the variation in germination (Table 1). All pairwise comparisons among temperature treatments were significantly different (P ranging from <0.0001 to 0.0152) with the greatest germination in the high oscillation (217 germinants from 600 seeds total) followed by low oscillation (34 germinants), warm (18 germinants) and cool (1 germinant) treatments (Fig. 2). The only significant difference among mulch treatments was the lower germination rate under 5 cm of mulch compared to bare soil (Fig. 2). Discussion Germination success for Sapium clearly depends on daily fluctuations in temperature, and the amplitude of the fluctuation is critical, as evidenced by the magnitude of the difference between germinants in the high-oscillation treatment and the low-oscillation treatment (Fig. 2). Pioneer species and wetland species commonly use diurnal temperature fluctuations as an indicator of canopy gaps (Fenner 1985; Baskin & Baskin 1989), proximity to the soil surface (Thompson & Grime 1983; 352 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Hi-Osc Lo-Osc Warm Cool Temperature Treatment Figure 2. Number of Sapium seeds germinating in each bin (means + 1 SE) for each combination of temperature treatment (Hi-Osc = 21-33°C, Lo-Osc = 21-27°C, Warm = constant 33 °C, Cool = constant ambient 21 °C) and mulch depth (0 cm, 5 cm, 10 cm). Ghersa et al. 1992), or recession of standing water (Fenner 1985). These environmental conditions are often critical to subsequent seedling success (Thompson & Grime 1983; Fenner 1985; Vleeshouwers et al. 1995). Several studies of invasive species have shown dependence on temperature fluctuations for successful germination (Ghersa et al. 1992; Lonsdale 1993; Young & Clements 2001). Also, several threatening invasives are woody invaders of wetland areas, including Sapium (Davis et al. 1946; Bruce et al. 1997), Schinus terebenthifolius Raddi, or Brazilian peppertree (Wheeler et al. 2001; Hight et al. 2003), and Melaleuca quinquenervia (Cav.) Blake, or punktree (Costello et al. 2003; Johnston et al. 2003). Mulching might be an effective control method for other invasive woody species as well. Germination and emergence from under 10 cm of mulch was not significantly different from that from bare soil (P = 0.6575), and there was no consistent trend in germination rates as mulch depth increased. This supports a conclusion that the indirect effect of mulch on soil DONAHUE, ROGERS & SIEMANN 353 temperature oscillations is more important than mulch depth alone for Sapium seed germination. It is encouraging for the potential success of this restoration method that only 5 cm of mulch in the field was required to damp the soil temperature oscillations sufficiently (Fig. 1) to achieve the germination suppression evidenced by the low oscillation treatment in Figure 2. The cotyledons of the seedlings in 10 cm of mulch were on long attenuated stems. The large Sapium seed (0.16 g/seed, Bonner 1989) apparently provides adequate resources for the seedling to emerge through deep mulch before reaching light where it can begin to photo- synthesize. Several studies in different environments have shown a positive correlation between seed mass and ability for seedlings to become established (Dzwonko & Gawronski 2002; Christie & Armesto 2003). When they modeled the emergence response of weed seeds to burial depth, Grundy et al. (2003) also found that some species had adequate reserves to emerge from a wider range of depths than might be expected in the field, as Sapium demonstrated in the present study. This may contribute to Sapium' s ability to invade and exploit many different environmental conditions. To be useful, the mulching treatment should have minimal effects on native prairie species. Foster & Gross (1998) found that prairie forbs and the prairie grass, Andropogon gerardi , were able to establish a significant number of seedlings in intact plant litter, even though the densities in litter were significantly lower than where litter was removed. In multiple-site studies, Foster & Gross (1997) and Foster (1999) found that accumulated litter affected Andropogon gerardi seedling establish¬ ment in some sites but not in others. Also, when examining tallgrass prairie recolonization mechanisms after soil disturbance by pocket gophers, Rogers & Hartnett (2001) found that vegetative regrowth after burial under soil was the dominant recolonization mechanism. There¬ fore, possible mulch-induced seed germination suppression could be expected to have little impact on native vegetation. Finally, the high flotation rubber tires of the mulching equipment limit damage to the root structure of existing perennial vegetation. Techniques for control of invasive vegetation include biological, herbicidal, mechanical, or some combination of these. While herbicide use appears to be necessary because of resprouting from stumps (Jubinsky & Anderson 1996), this study indicates that mulching live 354 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 trees can be an effective initial mechanical treatment that reduces subsequent seedling numbers, and thereby reduces the frequency and intensity of herbicide treatments. Acknowledgments The authors would like to thank Armand Bayou Nature Center for mulch, the University of Houston Coastal Center for permission to collect seeds, Brad Butterfield, Summer Nijjer, and Rachel Tardiff for assistance in the lab, and Wray-Todd Fellowship, US EPA (R82-8903), and US NSF (DEB-9981654) for financial support. Literature Cited Barrilleaux, T. C. & J. B. Grace. 2000. Growth and invasive potential of Sapium sebiferum (Euphorbiaceae) within the coastal prairie region: the effects of soil and moisture regime. Am. J. Bot., 87:1099-1106. Baskin, J. M. & C. C. Baskin. 1989. Physiology of dormancy and germination in relation to seed bank ecology. Pp. 53-66, in Ecology of soil seed banks (M. A. Leek, V. T. Parker & R. L. Simpson, eds.). Academic Press, Inc., San Diego, CA, 462 pp. Bonner, F. T. 1989. Sapium sebiferum (L.) Roxb. Chinese tallow tree. Pp. 760, in Seeds of woody plants in the United States (C. S. Schopmeyer, ed.). USDA, Forest Service, Washington, DC, 883 pp. Bruce, K. A., G. N. Cameron, P. A. Harcombe & G. Jubinsky. 1997. Introduction, impact on native habitats, and management of a woody invader, the Chinese tallow tree, Sapium sebiferum (L.) Roxb. Nat. Areas J., 17:255-260. Cameron, G. N. & S. R. Spencer. 1989. Rapid leaf decay and nutrient release in a Chinese tallow forest. Oecologia, 80:222-228. Cameron, G. N., E. G. Glumac & B. D. Eshelman. 2000. Germination and dormancy in seeds of Sapium sebiferum (Chinese tallow tree). J. of Coastal Research, 16:391-395. Christie, D. A. & J. J. Armesto. 2003. Regeneration microsites and tree species coexistence in temperate rain forests of Chiloe Island, Chile. J. Ecol., 91:776-784. Conner, W. H. 1994. The effect of salinity and waterlogging on growth and survival of baldcypress and Chinese tallow seedlings. J. of Coastal Research, 10:1045-1049. Conway, W. C., L. M. Smith & J. F. Bergan. 2000. Evaluating germination protocols for Chinese tallow {Sapium sebiferum) seeds. Tex. J. Sci., 52(3):267-270. Costello, S. L., P. D. Pratt, M. B. Rayamajhi & T. D. Center. 2003. Arthropods associated with above-ground portions of the invasive tree, Melaleuca quinquenervia, in South Florida, USA. Fla. Entomol., 86:300-322. Davis, W. S., R. Stanger, W. L. Nash, J. Kucera & A. Surface. 1946. District Program: Brazoria-Galveston Soil Conservation District No. 318, Texas. US Department of Agriculture, Soil Conservation Service, Washington, DC. Dzwonko, Z. & S. Gawronski. 2002. Influence of litter and weather on seedling recruitment in a mixed oak-pine woodland. Ann. Bot., 90:245-251. Fajardo, J. J., J. W. Bauder & S. D. Cash. 2001. Managing nitrate and bacteria in runoff from livestock confinement areas with vegetative filter strips. J. Soil Water Conserv., 56:185-191. Fenner, M. 1985. Seed Ecology. Chapman & Hall Ltd., London, 151 pp. DONAHUE, ROGERS & SIEMANN 355 Foster, B. L. 1999. Establishment, competition and the distribution of native grasses among Michigan old-fields. J. Ecol., 87:476-489. Foster, B. L. & K. L. Gross. 1997. Partitioning the effects of plant biomass and litter on Andropogon gerardi in old-field vegetation. Ecology, 78:2091-2104. Foster, B. L. & K. L. Gross. 1998. Species richness in a successional grassland: effects of nitrogen enrichment and plant litter. Ecology, 79:2593-2602. Ghersa, C. M., R. L. Benech & M. A. Martinez. 1992. The role of fluctuating temperatures in germination and establishment of Sorghum halepense. Regulation of germination at increasing depths. Funct. Ecol., 6:460-468. Grace, J. B. 1998. Can prescribed fire save the endangered coastal prairie ecosystem from Chinese tallow invasion? Endangered Species Update, 15:70-76. Grundy, A. C., A. Mead & S. Burston. 2003. Modelling the emergence response of weed seeds to burial depth: interactions with seed density, weight and shape. J. Appl. Ecol., 40:757-770. Harcombe, P. A., R. B. W. Hall, J. S. Glitzenstein, E. S. Cook, P. Krusic, M. Fulton. 1999. Sensitivity of Gulf Coast forests to climate change. Vulnerability of coastal wetlands in the Southeastern United States: climate change research results. Pp. 45-66 in Biology Science Report USGS/BRD/BSR -1998-0002 (G. Gunterspergen & B. A. Varain, eds.). United States Geological Survey, Washington, DC, 101 pp. Hall, R. B. W. & P. A. Harcombe. 1998. Flooding alters apparent position of floodplain saplings on a light gradient. Ecology, 79:847-855. Harbor, J. M., J. Synder & J. Storer. 1995. Reducing nonpoint source pollution from construction sites using rapid seeding and mulching. Phys. Geogr., 16:371-388. Herkert, J. R., D. L. Reinking, D. A. Wiedenfeld, M. Winter, J. L. Zimmerman, W. E. Jensen, E. J. Finck, R. R. Koford, D. H. Wolfe, S. K. Sherrod, M. A. Jenkins, J. Faaborg & S. K. Robinson. 2003. Effects of prairie fragmentation on the nest success of breeding birds in the midcontinental United States. Conserv. Biol., 17:587-594. Hight, S. D., I. Horiuchi, M. D. Vitorino, C. Wikler & J. H. Pedrosa-Macedo . 2003. Biology, host specificity tests, and risk assessment of the sawfly Heteroperreyia hubrichi, a potential biological control agent of Schinus terebinthifolius in Hawaii. Biocontrol, 48:461-476. Johnston, S. G., P. G. Slavich & P. Hirst. 2003. Alteration of groundwater and sediment geochemistry in a sulfidic backswamp due to Melaleuca quinquenervia encroachment. Aust. J. Soil Res., 41:1343-1367. Jones, R. H. & K. W. McLeod. 1989. Shade tolerance in seedlings of Chinese tallow tree, American sycamore, and cherrybark oak. Bull. Torrey Bot. Club, 116:371-377. Jubinsky, G. & L. C. Anderson. 1996. The invasive potential of Chinese tallow-tree {Sapium sebiferum Roxb.) in the southeast. Castanea, 61:226-231. Liaghat, A. & S. O. Prasher. 1996. A lysimeter study of grass cover and water table depth effects on pesticide residues in drainage water. Trans. ASAE, 39:1731-1738. Lonsdale, W. M. 1993. Losses from the seed bank of Mimosa pigra: soil micro-organisms vs. temperature fluctuations. J. Appl. Ecol., 30:654-660. Nijjer, S., R. A. Lankau, W. E. Rogers & E. Siemann. 2002. Effects of temperature and light on Chinese tallow {Sapium sebiferum ) and Texas sugarberry {Celtis laevigata ) seed germination. Tex. J. Sci., 54(l):63-68. Perkins, D. W., P. D. Vickery & W. G. Shriver. 2003. Spatial dynamics of source-sink habitats: effects on rare grassland birds. J. Wildl. Manage., 67:588-599. Renne, I. J. & S. A. Gauthreaux Jr. 2000. Seed dispersal of the Chinese tallow tree {Sapium sebiferum (L.) Roxb.) by birds in coastal South Carolina. Am. Midi. Nat., 144:202-215. Renne, I. J., T. P. Spira & W. C. Bridges, Jr. 2001. Effects of habitat, burial, age and 356 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 passage through birds on germination and establishment of Chinese tallow tree in coastal South Carolina. J. Torrey Bot. Soc., 128:109-119. Rogers, W. E., S. Nijjer, C. L. Smith & E. Siemann. 2000. Effects of resources and herbivory on leaf morphology and physiology of Chinese tallow ( Sapium sebiferum ) tree seedlings. Tex. J. Sci., 52(4)Supplement:43-56. Rogers, W. E. & D. C. Hartnett. 2001. Temporal vegetation dynamics and recolonization mechanisms on different-sized soil disturbances in tallgrass prairie. Am. J. Bot., 88:1634-1642. Rogers W. E. & E. Siemann. 2002. Effects of simulated herbivory and resource availability on native and invasive exotic tree seedlings. Basic Appl. Ecol., 3:297-307. Siemann, E. & W. E. Rogers. 2003a. Herbivory, disease, recruitment limitation and success of alien and native tree species. Ecology, 84:1489-1505. Siemann, E. & W. E. Rogers. 2003b. Changes in light and nitrogen under pioneer trees may facilitate tree invasions of grasslands. J. Ecol., 91:923-931. Thompson, K. & J. P. Grime. 1983. A comparative study of germination responses to diurnally fluctuating temperatures. J. Appl. Ecol., 20:141-156. Vleeshouwers, L. M., H. J. Bouwmeesterl & C. M. Karssen. 1995. Redefining seed dormancy: an attempt to integrate physiology and ecology. J. Ecol. 83:1031-1037. Wheeler, G. S., L. M. Massey & M. Endries. 2001. The Brazilian peppertree drupe feeder Megastigmus transvaalensis (Hymenoptera : Torymidae): Florida distribution and impact. Biol. Control, 22:139-148. Young, J. A. & C. D. Clements. 2001. Purple loosestrife ( Lythrum salicario ) seed germination. Weed Technol., 15:337-342. CD at: candy@abnc.org TEXAS J. SCI. 56(4):357-368 NOVEMBER, 2004 THE EFFECT OF MYCORRHIZAL INOCULUM ON THE GROWTH OF FIVE NATIVE TREE SPECIES AND THE INVASIVE CHINESE TALLOW TREE ( SAPIUM SEBIFERUM) Somereet Nijjer, William E. Rogers and Evan Siemann Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77005 Abstract. — Mycorrhizal fungi may play an important role in plant invasions, but few studies have tested this possibility. Chinese Tallow ( Sapium sebiferum ) is an invasive tree in the southeastern United States. An experiment was conducted to examine the effects of mycorrhizal inoculation, fungicide application, and fertilization on the growth of Sapium and five native tree species (Liquidambar styracif.ua , Nyssa sylvatica, Pinus taeda, Quercus alba, and Q. nigra) that co-occur in forests in the Big Thicket National Preserve in east Texas. Seedlings were grown in a greenhouse for twenty weeks under full factorial combinations of mycorrhizal inoculum, fungicide, and fertilizer. Mycorrhizal inoculation increased Sapium growth but caused zero to negative growth changes of the five native species. This suggests that Sapium may gain unusual benefits from mycorrhizal associations. Liquidambar styraciflua benefited from mycorrhizal inoculation only in fertilized conditions which indicates that the potential advantage Sapium might gain from mycorrhizal associations may vary with native species and soil fertility. Mycorrhizal fungi form close associations with roots of plants in which in exchange for fixed carbon, the fungi provide essential nutrients to the plant (N, P) and may protect the plant from pathogens, support helpful bacteria, enhance soil aggregation, assist in water transport and gain, and stimulate plant growth through auxin production; these asso¬ ciations can vary from mutualistic to parasitic depending on soil fertility levels (Harley 1968; Allen 1991; Johnson et al. 1997; Smith & Read 1997; Van der Heijden & Sanders 2002). It is possible that mycorrhizae play a key role in temperate forest dynamics and community responses by changing the outcome of competition and by influencing plant fitness (Johnson et al. 1997; Van der Heijden & Sanders 2002). Little attention has focused on how the existing mycorrhizal network of the introduced range may facilitate the invasion of exotic plant species. Sapium sebiferum (L.) Roxb, a native to central China, was intro¬ duced to Georgia in the late 18th century (Bruce et al. 1997). Although present in Texas in the early 1900’s, Sapium did not become invasive until the middle of the century and has only rapidly increased abundance in the past two decades in mesic and hydric forests in the Big Thicket National Preserve (BTNP) in east Texas (Harcombe et al. 1999). Re- 358 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 cent studies have shown that Sapium benefits from low herbivore loads (Siemann & Rogers 2001; 2003ab; Rogers & Siemann 2002; 2003), but Sapium appears to have unusually high growth rates even after account¬ ing for differences in aboveground herbivore impacts. Although a release from belowground pathogens could explain the high growth rates of Sapium , unusually large benefits from mycorrhizal associations are also a factor that may contribute to Sapium* & invasive success. Generalist mycorrhizae with low host specificity may be able to form associations with invasive plants (Richardson et al. 2000). This associa¬ tion by itself would not create unusually high benefits, and thus could not be itself responsible for invasive success, unless the invader could utilize the mycorrhizae in a novel fashion (Richardson et al. 2000). The combination of potentially novel mycorrhizal utilization and the short co-evolutionary history exotic plants have with native mycorrhizal mutualists suggests that these plants could receive unusually high bene¬ fits or extremely high costs their introduced ranges (Richardson et al. 2000) . Another way that exotic invaders could obtain benefits would be to usurp native species’ existing mycorrhizal network connections, or utilize neighbors’ nutrient pools with their own extraradical (soil exploring) hyphae, thus parasitizing neighboring competitors through enhanced nutrient uptake (Marler et al. 1999; Zabinski et al. 2002). Only limited work to date has been done to examine how the existing mycorrhizal network of the introduced range may influence the competi¬ tive ability of exotic invaders (Bray et al. 2003). Understanding how Sapium utilizes mycorrhizal associations in its introduced range may help explain the mechanisms underlying its invasion in the BTNP and in¬ crease general knowledge of the role of mycorrhizae in affecting plant community dynamics. A greenhouse experiment was conducted to test the effects of mycor¬ rhizal inoculation, fungicide application, and fertilization on the growth of Sapium and five tree species native to the BTNP. If mycorrhizae contribute to Sapium invasion, then the performance advantage of Sapium compared to natives should be greater with mycorrhizal inocula¬ tion than without. To potentially decrease the performance advantage of Sapium if mycorrhizal inoculation facilitates invasion, Rovral fungi¬ cide was applied (Gange et al. 1990; Ganade & Brown 1997). Fertiliza¬ tion is predicted to highlight plant alterations in mycorrhizal dependen¬ cies and mimic potential changes in field conditions. Fertilization is predicted to decrease the effect of mycorrhizae on plant performance NIJJER, ROGERS & SIEMANN 359 because carbon costs are not offset by benefits of nutrient gathering in high fertility (Menge et al. 1978; Buwalda & Goh 1981; Hetrick et al. 1988; Hetrick 1991; Johnson 1993; Peng et al. 1993) and additionally because the benefits plants receive from mycorrhizae may be less valuable in higher fertility conditions (Koide 1991; Johnson 1993; Johnson et al. 1997). In Flatland Hardwood Pine Forests of the Lance Rosier Unit in the Big Thicket, which are equivalent to Lower Slope Hardwood Forests found elsewhere, phosphorus tends to be in limited supply (Marks & Harcombe 1981; Knox et al. 1995; BTNP 2003) be¬ cause of its difficulty to acquire at low levels and strong adsorption to soil particles (Nye & Tinker 1977; Read 1991). However, nitrogen deficiencies may limit growth of plants with non-mycorrhizal affiliations because they can only absorb simple forms of N (Chalot & Brun 1998). Together these predictions will begin to answer how mycorrhizae may promote or hinder Sapium’s invasibility and ultimately alter the sur¬ rounding native community. Methods Seeds of five native tree species that are common in mesic and hydric forests in the BTNP and may potentially be outcompeted by Sapium sebiferum ( Liquidambar styraciflua L. [sweetgum], Nyssa sylvatica Marsh [blackgum], Pinus taeda L. [loblolly pine], Quercus alba L. [white oak], and Q. nigra L. [water oak]) were purchased (Louisiana Forest Seed Company, Lecompte, LA) to ensure that seeds were from uniformly healthy trees. Sapium sebiferum seeds were collected at Armand Bayou Nature Preserve (Houston, TX). Stratification took place in a 21 °C cold-room in January- February 2003. Germination of non- surface sterilized seeds occurred in an unheated greenhouse on the Rice University campus during March-May 2003. Germinated seeds were planted in 66 mL Conetainers (Stuewe & Sons, Inc., Corvallis, OR) filled with potting soil. Forty-eight similarly sized seedlings of each species were selected approximately two weeks after germination. All of the plants within each species were randomly assigned to one of eight treatments in a full-factorial experimental design with inoculation (yes or no) , fungicide (yes or no) and fertilizer (yes or no) for a total of six replicates per treatment. Roots were gently brushed free of soil and the soil was retained. Roots were then dipped in either “Silva Dip” (Reforestation Technologies International, Salinas, CA) which contained a total of eight general endo- and ectomycorrhizal species ( Glomus intraradices , Glomus 360 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 aggregation , Glomus mossae , Pisolithus tinctorius , and four species of Rhizopogon sp.) or distilled water. Excluding Rhizopogon sp., which is primarily found in the northwestern United States, at least one of the remainder of the endo- and ectomycorrhizal species listed would be encountered by the focal tree species of this study in the field (Keeley 1980; Black et al. 1981; McIntosh et al. 1985; Weber & Smith 1985; Walker & McLaughlin 1991; Metzler & Metzler 1992; Lewis & Strain 1996; Constable et al. 2001). After dipping, roots were covered with the retained soil and transplanted into 3.8 liter Treepots™ (Stuewe & Sons, Inc.) filled with a mixture of 2/3 potting soil and 1/3 perlite. Pots were placed within blocks grouped by species on plastic pallets on the greenhouse floor because of differences in germination times. Pots were watered as needed and periodically rotated within species blocks to minimize shading and location effects. Fertilizer was applied four times in the course of the 20- week experi¬ ment in weeks 3, 7, 12, and 17. Application rates were equivalent to 4 g/m2 each of N, P and K per application. This mimics field regulation standard rates. Nutrients were added as ammonium nitrate (N), super¬ phosphate (P), and potash (K) dissolved in 40 mL of distilled water. Distilled water was added to non- fertilized controls. Rovral® 4 Flowable Fungicide (Aventis CS, Bridgewater, NJ) was applied three times in the course of the 20-week experiment in weeks 4, 10, and 16. Rovral, active ingredient iprodione, has been shown to reduce mycorrhizal infection in plant roots and is a contact pesticide with no known systemic action (Gange et al. 1990; Ganade & Brown 1997). Application rates followed recommendations for controlling pathogenic root fungi (Aventis 2001). Initial height of each seedling was measured. Initial heights were taken before seedlings were dipped into either inoculum or a distilled water control and as such did not require sterilization of equipment to pre-empt transfer of inoculum between sources. At the end of 20 weeks, roots, leaves, and stems were harvested and dried at 60 °C for at least 72 hours before weighing. An ANCOVA with starting height as a covariate was used to test whether final mass (log transformed to achieve normality) depended on experimental treatments in a model with all possible interactions among experimental treatments (SAS 8.2, SAS Institute, Cary, NC). Mass data were back transformed for graphical presentation. Single species ANOVAs were used to investigate significant interaction terms in the full NIJJER, ROGERS & SIEMANN 361 analysis and Fisher’s Least Significant Difference Test was used to test for differences between treatment means (Stat View 5.0, SAS Institute, Cary, North Carolina). Results The percent of root mass was independent of all factors other than species (F5 238= 1 19.80, PC .0001). It was lowest for Pinus (29%) followed by Liquidambar (40%), Q. nigra (46%), Sapium (47%), Nyssa (53 %) and Q. alba (73 %). The contrasts among species were significant at a =0.05 for all pairs of species except Q. nigra vs. Sapium . Because allocation patterns are independent of treatments (modeled as a percent¬ age of belowground root biomass) and species is the only significant factor explaining the allocation pattern variance, the remainder of the analyses utilized total mass as the dependent variable. Total mass varied among species (Table 1, Fig. 1) and the contrasts among species were significant at a =0.05 for all pairs of species except Q. alba vs. Nyssa and Liquidambar vs. Q. nigra. No other main effect significantly affected mass in the ANCOVA (Table 1). Total mass depended on starting height in the ANCOVA (Table 1). The species which had significant correlations between starting height and log (final mass) in z- tests were Q. alba (r=0.59, P< 0.0001) and Nyssa (r=0.41, P< 0.001). Variation in mass depended on several interactions: species/ noculation, species/fertilization, species/inoculation/fertilization, and species/inoculation/fungicide. Since each interaction term had species as one of the factors, individual species ANOVAs were used to help identify the main factors influencing the interactions. The significant effect of species/inoculation in the full model indicat¬ ed that species differed in the direction or magnitude of their responses to inoculation. All five native species tended to have lower mass when inoculated but this difference was significant only for Nyssa (P<0.01) in single species ANOVAs. Sapium had significantly higher mass when inoculated (P< 0.01). In a separate analysis with a two-level predictor that indicated whether a species was native vs. Sapium , the interaction of this term and inoculation was significant (P<0.05). The significant effect of species/fertilization in the full model indicated that species differed in their responses to fertilization. In single species ANOVAs, Pinus fP<0.01) and Liquidambar (P< 0.05), 362 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Table 1 . The dependence of log(fmal mass) on experimental treatments in an ANCOVA with starting height as a covariate. Significant terms are noted with an asterisk (*). Factor df SS F-Value P- Value Species* 5 48.2 157.7 <0.0001 Fertilizer 1 0.1 1.6 0.20 Fungicide 1 0.0 0.2 0.70 Inoculum 1 0.1 0.9 0.35 Species/Fertilizer* 5 0.8 2.6 <0.05 Species/Fungicide 5 0.6 1.9 0.10 Species/Inoculum* 5 1.3 4.4 <0.001 Fertilizer/Fungicide 1 0.0 0.3 0.58 Fertilizer/Inoculum 1 0.1 0.9 0.34 Fungicide/Inoculum 1 0.0 0.5 0.48 Species/Fertilizer/Fungicide 5 0.3 0.9 0.50 Species/Fertilizer/Inoculum* 5 0.8 2.7 <0.05 Species/Fungicide/Inoculum* 5 0.8 2.7 <0.05 Fertilizer/Fungicide/Inoculum 1 0.1 1.3 0.26 Species/Fertilizer/Fungicide/Inoculum 5 0.3 0.9 0.45 Starting height* 1 1.2 19.6 <0.0001 Error 238 14.5 but no other species, were significantly larger when fertilized (Fig. 1). Pinus had larger mass in fertilized controls and maintained this increase when inoculated. However, Liquidambar1 s growth had significant mass increases with inoculation in the fertilized treatments only. Single species ANOVAs show that the significant interaction of species/fertilization/inoculation in the full model was related to the idiosyncratic effect of these treatments on Liquidambar mass CP <0.01, Fig. 1). Inoculation reduced Liquidambar mass in low fertility condi¬ tions but increased it in high fertility conditions. The significant effect of species/inoculation/fungicide largely reflect¬ ed the distinct responses of Sapium to fertilizer and fungicide since the interaction of these treatments was only significant for Sapium (P< 0.01) in single species ANOVAs. Submodels showed fungicide-non-inoculated plants to be significantly different from fungicide-inoculated plants and contol (non-fungicided, non- inoculated) plants to be significantly differ¬ ent from fungicided- inoculated plants by Fisher’s Least Significant Difference Test, respectively (P< 0.01, /><0.05). Specifically, Sapium mass was lowest in the fungicide only treatment (average = 7.8 g) followed by control (non-inoculated and non-fungicided), (15.1 g), inoculation only (15.1 g), and finally the combination of inoculation and fungicide (20.9 g). NIJJER, ROGERS & SIEMANN 363 Liquidambar Nyssa Pinus Q. alba Q. nigra Sapium Figure 1. The dependence of mass (g) of Liquidambar styraciflua, Nyssa sylvatica, Pinus taeda, Quercus alba , Quercus nigra , and Sapium sebiferum seedlings on fertilization (con = no fertilizer, fert = fertilized) and mycorrhizal inoculation after 20 weeks. Fungicide treatments are not shown. See Table 1 for statistical results. Discussion Sapium' s striking positive growth response to mycorrhizal inoculation (65% increase) differed markedly from the neutral to negative responses of native tree species (Fig. 1). The magnitude of reductions in growth of the five native tree species in response to inoculation ranged from negligible ( Q . alba = 1% reduction, Q. nigra = 6%) or minor ( Pinus = 17%, Liquidambar = 24%) to large and significant ( Nyssa = 46%) but the direction of the response to inoculation was always negative. Sapium was clearly able to gain large benefit from mycorrhizal associa¬ tions with a generalist mycorrhizal inoculum in conditions where natives could not. It appears that natives were unable to benefit from the generalist inoculum in this study suggesting that mycorrhizal specificity is important (Bever 2002; Klironomos 2003). The strains used in this study may not be beneficial in these conditions and may create an un¬ necessary obligate symbiosis with direct translations to decreases in growth (Hetrick et al. 1988; Hetrick 1991). This supports the hypothe¬ sis that unusual relationships between the exotic Sapium and North American mycorrhizae species, such as those in the inoculum, may contribute to Sapium' s success as an invader. 364 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 There are a number of explanations for why lack of specialist mycor- rhizae (Bever 2002; Klironomos 2003), which was predicted to be bene¬ ficial, appeared to be especially detrimental in fertilized treatments for Nyssa, Pinus, Q. nigra and for the native species Liquidambar and Nyssa in unfertilized treatments in this experiment. First, carbon drain on host plants, which is well documented (Buwalda & Goh 1981; Hetrick 1991; Johnson 1993; Peng et al. 1993; Graham et al. 1996) may have exceeded the benefits of increased nutrients and/or water in these relatively fertile, well-watered greenhouse conditions. Second, mycor- rhizae in this experiment may have used carbon from plants largely for respiration rather than increasing extraradical hyphae surface area and increasing nutrient absorption (Peng et al. 1993; Graham et al. 1996). Increases in maintenance respiration, as well as higher root construction costs due to high lipid vesicle allocation, has been shown in P addition experiments for Citrus volkameriana (Peng et al. 1993, Graham et al. 1996) and has been attributed to decreases in carbohydrate root exudates from plants in highly fertilized soils (Johnson et al. 1997). The unexpected results for fungicide and inoculation combinations, in particular the effects on Sapium mass, were inconsistent with the expec¬ tation that seedlings in the two treatments, non- inoculated fungicide only and inoculation plus fungicide, would be identical in size. This suggests that fungicide applications were not an effective method of fungal control. One possible explanation is that non-spore ingredients in the mycorrhizal inoculum had phytotoxic effects on seedling growth in the presence of fungicide. The reduction of Sapium mass by fungicide application (without inoculum) might indicate that beneficial microbes (phosphate-solubilizing microbes and plant growth-promoting bacteria) were present in the potting soil which were killed by the fungicide (Allen 1992). Alternatively, it might indicate direct toxic effects of fungicide on Sapium. The recovery of Sapium growth with inoculation in fungicide treatments suggests that the mycorrhizal inoculum was not effectively suppressed and that mycorrhizae may be acting synergistically with microbes in the fungicided soil that were not effective or prevalent in the non- fungicided soil. One goal of this greenhouse experiment was to develop methods that could be applied in field experiments. Further work with direct assays of mycorrhizal and non-mycorrhizal fungi in experiments with Sapium is needed to complete the identification of reliable field methods and identify the cause of the seemingly anomalous inoculation and fungicide result. NIJJER, ROGERS & SIEMANN 365 The prediction of decreased response of all species to mycorrhizal inoculation in high fertility environments was based on the assumption that mycorrhizal carbon costs are not offset by the benefits of nutrient gathering in conditions in which nutrients are abundant (Menge et al. 1978; Buwalda & Goh 1981; Hetrick et al. 1988; Hetrick 1991; Johnson 1993; Peng et al. 1993). The positive response of Liquidambar to mycorrhizal inoculation only in fertilized conditions was opposite the prediction that the benefit of mycorrhizal associations would be lower in more fertile conditions (Fig. 1). Indeed, the reverse pattern observed here suggests that there may be potential for strong competition for nutrients between mycorrhizae or other soil microbes and plants in low fertility environments that may counteract the potential benefit of mycor¬ rhizal associations in these conditions (Bardgett et al. 2003). The strong benefit of mycorrhizal inoculation for Liquidambar in some conditions (Figure 1) indicates that the competitive advantage Sapium might gain from mycorrhizal associations may vary with native species and soil fertility (Marler et al. 1999). One theory explaining the success of invaders in their introduced range is the Enemy Release Hypothesis. It predicts that invasives experience a release from the pressures of the natural enemies in their native range and can therefore allocate additional resources to growth and reproduction (Alpert et al. 2000; Maron & Vila 2001; Keane & Crawley 2002; Mitchell & Power 2003). However, little attention has been given to belowground enemies. This experiment raises the possi¬ bility that the large size of Sapium in all conditions, although doing better with inoculum than natives, (Figure 1) reflects presence of below¬ ground pathogenic fungi that more readily attack native tree species. The results reported here would be more compelling with confirma¬ tion of mycorrhizal colonization and dependence by direct examination. Further, it is imperative that these results be verified in field trials as well as in experiments including competitive interactions between species. Such experiments are currently underway to rigorously test the preliminary conclusion presented here that interactions with soil mi¬ crobes play a role in Sapium invasions in east Texas forests. Acknowledgments We would like to thank: the National Science Foundation (DEB- 9981546) and a Wray-Todd Fellowship for financial support; Bradley 366 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Butterfield, Philemon Chow, Saara DeWalt, Candice Donahue, Maria Hartley, Catherine LaMaur, Rick Lankau, Zack McLemore, Jay Nijjer, Rachel Tardiff, Emily Wheeler, and Terris White for assistance and support; and Armand Bayou Nature Center for permission to collect seeds, Paul Harcombe and two anonymous reviewers for their com¬ ments. Literature Cited Allen, M. 1991. Ecology of mycorrhizae. Cambridge University Press: Great Britain, 184 pp. Allen, M. 1992. Mycorrhizal functioning: an integrative plant-fungal process. London: Chapman & Hall, 534 pp. Alpert, P., E. Bone & C. Holzapfel. 2000. Invasiveness, invasibility and the role of environmental stress in the spread of non-native plants. Perspect. Plant Ecol. Evol. Syst., 3:52-66. Aventis Supplemental Label: Rovral brand 4 Flowable Fungicide. Aventis CropScience USA LP. 2001. Bardgett, R. D., T. A. Streeter & R. Bol. 2003. Soil microbes compete effectively with plants for organic-nitrogen inputs to temperate grasslands. Ecology, 84:1277-1287. Barrilleaux, T. C. & J. B. Grace. 2000. Growth and invasive potential of Sapium sebiferum (Euphorbiaceae) within the coastal prairie region: the effects of soil and moisture regime. Am. J. Bot., 87:1099-1106. Bever, J. 2002. Host specificity of AM fungal population growth rates can generate feedback on plant growth. Plant Soil, 244:281-290. Big Thicket National Preserve: plant associations: appreciating the nuances of the Big Thicket. 2003. http://www.nps.gov/bith/plant_associations.htm. Black, H. D., L. B. Coons & S. J. Warrington. 1981 . Entry of Pisolithus tinctorius hyphae into Pinus taeda roots. Can. J. Bot., 59:2135-2139. Bray, S., K. Kitajima & D. Sylvia. 2003. Differential response of an exotic invasive shrub to mycorrhizal fungi: growth, physiology, and competitive interaction. Ecol Appl., 13:565-574. Bruce, K. A., G. N. Cameron, P. A. Harcombe & G. Jubinsky. 1997. Introduction, impact on native habitats, and management of a woody invader, the Chinese tallow tree, Sapium sebiferum (L.) Roxb. Nat. Areas J., 17:255-260. Buwalda, J. & K. Goh. 1981. Host-fungus competition for carbon as a cause of growth depressions in vesicular-arbuscular mycorrhizal ryegrass. Soil Biol. Biochem., 14:103- 106. Chalot, M. & A. Brun. 1998. Physiology of organic nitrogen acquisition by ectomycorrhizal fungi and ectomycorrhizas. FEMS Microbiology Reviews, 22:21-44. Constable, J. V. H., H. Bassirirad, J. Lussenhop & A. Zerihun. 2001. Influence of elevated C02 and mycorrhizae on nitrogen acquisition: contrasting responses in Pinus taeda and Liquidambar styraciflua. Tree Physiol., 21:83-91. Ganade, G. & V. K. Brown. 1997. Effects of below-ground insects, mycorrhizal fungi and soil fertility on the establishment of Vicia in grassland communities. Oecologia, 109:374-381. Gange, A. C., V. K. Brown & L. M. Farmer. 1990. A Test of Mycorrhizal Benefit in an Early Successional Plant Community. New Phytol., 115:85-91. Graham, J., D. Drouillard & N. C. Hodge. 1996. Carbon economy of sour orange in NIJJER, ROGERS & SIEMANN 367 response to different Glomus spp. Tree Physiol., 16:1023-1029. Harcombe, P. A., R. B. W. Hall, J. S. Glitzenstein, E. S. Cook, P. Krusic, M. Fulton & D. R. Streng. 1999. Sensitivity of Gulf Coast forests to climate change. Pp. 47-67, in Vulnerability of coastal wetlands in the southeastern United States: climate change research results. Biological Science Report USGS/BRD/BSR- 1998-0002. (G. Gunterspergen & B. A. Varain, eds.), 114 pp. Harley, J. L. 1968. Presidential address: fungal symbiosis. Trans. Br. Mycol. Soc., 51:1-11. Hetrick, B., J. F. Leslie, G. Thompson Wilson & D. Gerschefske Kitt. 1988. Physical and topological assessment of effects of vesicular-arbuscular mycorrhizal fungus on root architecture of big bluestem. New Phytol., 110:85-96. Hetrick, B. 1991. Mycorrhizas and root architecture. Experientia (Basel), 47:355-362. Johnson, N. 1993. Can fertilization of soil select less mutualistic mycorrhizae? Ecol Appl., 3:749-757. Johnson, N., J. H. Graham, F. A. Smith. 1997. Functioning of mycorrhizal associations along the mutualism-parasitism continuum. New Phytol., 135:575-586. Keane, R. M. & M. J. Crawley. 2002. Exotic plant invasions and the enemy release hypothesis. Trends Ecol. Evol., 17:164-170. Keeley, J. E. 1980. Endomycorrhizae influence growth of blackgum seedlings in flooded soils. Am. J. Bot., 67:6-9. Klironomos, J. 2003. Variation in plant response to native and exotic arbuscular mycorrhizal fungi. Ecology, 84:2292-2301. Knox, R. G., P. A. Harcombe & I. S. Elsik. 1995. Contrasting patterns of resource limitation in tree seedlings across a gradient in soil texture. Can. J. For. Res., 25:1583-1594. Koide, R. T. 1991. Tansley review no. 29. Nutrient supply, nutrient demand and plant response to mycorrhizal infection. New Phytol., 117:365-386. Lewis, J. D. & B. R. Strain. 1996. The role of mycorrhizas in the response of Pinus taeda seedlings to elevated C02. New Phytol., 133:431-443. Marks, P. L. & P. A. Harcombe. 1981. Forest vegetation of the Big Thicket, southeast Texas. Ecol. Monogr., 51:287-305. Marler, M. J., C. A. Zabinski & R. M. Callaway. 1999. Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology. , 80: 1 180-1186. Maron, J. L. & M. Vila. 2001. When do herbivores affect plant invasion? Evidence for the natural enemies and biotic resistance hypotheses. Oikos, 95:361-373. McIntosh, M. S., P. R. Beckjord & J. H. Melhuish. 1985. Effects of nitrogen and phosphorus fertilization on growth and formation of ectomycorrhizae of Quercus alba and Q. rubra seedlings by Pisolithus tinctorius and Scleroderma auranteum. Can. J. Bot., 63:1677-1680. Menge, J. A., D. Steirle, D. J. Bagyaraj, E. L. V. Johnson & R. T. Leonard. 1978. Phosphorous concentrations in plants responsible for inhibition of mycorrhizal infection. New Phytol., 80:575-578. Metzler, S. & V. Metzler. 1992. Texas mushrooms: a field guide. University of Texas Press: Austin, 350 pp. Mitchell, C. & A. Power. 2003. Release of invasive plants from fungal and viral pathogens. Nature, 421:625-627. Nye, P. H. & P. B. Tinker. 1977. Solute movement in the soil-root system. Studies in Ecology Vol. 4, Blackwell Scientific, Oxford, UK, 342 pp. Peng, S., D. Eissenstat, J. H. Graham, K. Williams & N. Hodge. 1993. Growth depression in mycorrhizal citrus at high-phosphorous supply: analysis of carbon costs. Plant Physiol., 101:1063-1071. 368 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Read, D. J. 1991. Mycorrhizas in ecosystems. Experientia, 47:376-391. Richardson, D. M., N. Allsopp, C. M. D’ Antonio, S. J. Milton & M. Rejmanek. 2000. Plant invasions - the role of mutualisms. Biol Rev., 75:65-95. Rogers, W. E. & E. Siemann. 2002. Effects of simulated herbivory and resource availability on native and invasive exotic tree seedlings. Basic Appl. Ecol., 4:297-307. Rogers, W. E. & E. Siemann. 2003. Effects of simulated herbivory and resources on Chinese tallow tree (Sapium sebiferum, Euphorbiaceae) invasion of native coastal prairie. Am. J. Bot., 90:243-249. Siemann, E. & W. Rogers. 2001. Genetic differences in growth of an invasive tree species. Ecol Lett., 4:514-518. Siemann, E. & W. E. Rogers. 2003a. Herbivory, disease, recruitment limitation and success of alien and native tree species. Ecology, 84:1489-1505. Siemann, E. & W. E. Rogers. 2003b. Changes in resources under pioneer trees may facilitate tree invasions of grasslands. J. Ecol., 91:923-931. Smith, S. E. & D. J. Read. 1997. Mycorrhizal symbiosis 2nd edition. Academic Press: New York, 605 pp. Van der Heijden, M. & I. Sanders. 2002. Mycorrhizal ecology. Ecological Studies (157) Springer, Germany, 469 pp. Walker R. F. & S. B. McLaughlin. 1991. Growth and root system development of white oak and loblolly pine as affected by simulated acidic precipitation and ectomycorrhizal inoculation. For. Ecol. Manage., 46:123-133. Weber, N. S. & A. H. Smith. 1985. A field guide to southern mushrooms. The University of Michigan Press: Ann Arbor, 280 pp. Zabinski, C. A., L. Quinn & R. M. Callaway. 2002. Phosphorous uptake, not carbon transfer, explains arbuscular mycorrhizal enhancement of Centaurea maculosa in the presence of native grassland species. Funct Ecol., 16:758-765. SN at: somereet@rice.edu TEXAS J. SCI. 56(4): 369-3 82 NOVEMBER, 2004 CHARACTERIZATION OF ARTHROPOD ASSEMBLAGE SUPPORTED BY THE CHINESE TALLOW TREE (SAPIUM SEBIFERUM) IN SOUTHEAST TEXAS Maria K. Hartley, Saara DeWalt, William E. Rogers and Evan Siemann Department of Ecology and Evolutionary Biology Rice University, Houston, Texas 77005 Abstract.— Arthropod abundance, species richness and trophic structure were measured on the introduced species Chinese Tallow tree ( Sapium sebiferum (L.) Roxb.) in southeast Texas. Samples were collected using sweep nets between June and October of 2001 . A total of 811 individuals and 160 arthropod species were caught. Orders Diptera, Acari, and Araneida were abundant on Sapium , while orders such as Thysanoptera, Neuroptera, Orthoptera were present in much lower relative abundances. The order Hemiptera was markedly low in abundance and species richness. Compared to available data on native ecosystems, predators and detritivores were relatively abundant while herbivores and total arthropod diversity were relatively low on Sapium. These results suggest that Sapium has not yet acquired an insect fauna comparable to native plants in Texas. Arthropods represent a significant proportion of faunal community diversity and have vital roles in ecosystem functioning (Wilson 1992; Price 1997). A number of ecosystem services are performed by arthropods, such as nutrient recycling, seed dispersal, herbivory, and pollination (Proctor & Yeo 1972; Petrusewicz & Grodzinski 1975; Davidson & Morton 1981; Jones et al. 1994). Introduced plant species have been shown to alter ecosystem functioning, reduce native diversity, and promote extinction of native species (Vitousek 1986; Liebhold et al. 1995; Mack et al. 2000), and through changes in vegetation structure, composition and host quality, they may affect arthropod assemblages. Insect diversity is frequently correlated with the diversity of plants (Schowalter 1995; Siemann 1998) and architectural complexity of a habitat (Strong et al. 1984). When previously diverse habitats are converted to monospecific stands of non- native plants, insect species richness will often be lower. Factors that influence arthropod colonization rates on introduced plant species may affect subsequent community composition and structure. Strong et al. (1984) suggested that taxonomic, phenological , biochemi¬ cal, and morphological similarities between introduced and native plants, as well as geographic range, may influence how quickly introduced plants are colonized by native arthropods. However, arthropod host 370 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 choice is typically driven by physiological and behavioral adaptations in response to host plant quality (Price 1997; Schowalter 2000). Host plants considered low quality for arthropod growth and development, are typically highly defended and/or nutritionally poor (Price et al. 1980). Host choice usually divides herbivorous insects into two categories, gen¬ eralists and specialists (Feeny 1976). Generalists capitalize on the most abundant and obvious resource, whereas specialists possess increased efficiency but reduced resource choice (Feeny 1976; Brown 1984). Therefore, generalist arthropods are thought to be more commonly found on introduced plant species than native species (Strong et al. 1984; Lankau et al. 2004), but little empirical evidence supports this assertion. Sapium sebiferum (L.) Roxb. (Euphorbiaceae) also known as Triadica sebifera, invades coastal tallgrass prairie, disturbed areas, and intact floodplain forests in east Texas (Bruce et al. 1997). The enemy release hypothesis has been used to explain the success of some introduced species including Sapium (Elton 1958; Keane & Crawley 2002; Siemann & Rogers 2003a). It asserts that alien species are introduced without their co-evolved specialist herbivores and pathogens. This release from natural enemies may give alien species a competitive advantage over native plants (Elton 1958; Groves 1989; Lodge 1993; Tilman 1999). Indeed, there is evidence that herbivore loads are lower on introduced plant species than native species (South wood et al. 1982; Strong et al. 1984; Yela & Lawton 1997). Furthermore, biological control agents can sometimes control alien plant populations (Goeden & Louda 1976; Groves 1989). If the enemies release hypothesis is valid, insects may play an important role in the invasion of Sapium. The objective of this study was to characterize the arthropod community by quantifying arthropod taxonomic richness and abundance on a monospecific stand of Sapium , growing on a former coastal prairie in southeast Texas, and comparing to data from native habitats in southeastern Texas (Birch 1975; McFadden 1978; Cameron & Byrant 1999). It was predicted that: (1) fewer herbivore species would be found on Sapium than in native communities if Sapium is avoided by North American herbivores, and (2) the arthropod community structure on Sapium would be different from that found in native habitats, as Sapium has been present for a shorter time and is therefore less likely to have acquired a full insect fauna. HARTLEY ET AL. 371 Materials and Methods Focal study species. —Originally from Asia, Chinese tallow tree ( Sapium sebiferum ) was introduced to Georgia in the late eighteenth century and subsequently into Texas in the early 1900’s (Bruce et al. 1997). Sapium is a dominant invasive species in the southeastern United States (Flack & Furlow 1996; Bruce et al 1997). Once established it can form dense monospecific stands with little under story vegetation (Bruce et al. 1997). It experiences low levels of herbivory in Texas (Siemann & Rogers 2001; 2003a; 2003b) but the diversity and composition of associated arthropods in Texas is not known. Study site. — The study was conducted at the University of Houston Coastal Center (henceforth known as UHCC), a 374 ha research area, located 50 km SE of Houston, Texas. Most of the research site consists of Sapium stands in areas that originally would have been tallgrass prairie. This study was conducted in a monospecific Sapium stand that was estimated to be 30 years old. Sampling protocol. — This Sapium stand was sampled 16 times between 8 June and 24 October 2001. The sampling frequency was devised for taxa that emerge for only short periods and, or have short life spans. On each sampling occasion, four samples were collected randomly from Sapium. Each sample was collected along a 16 m transect. Transects were selected for minimal undergrowth to minimize the influence of other plant species on the focal arthropod community. Each transect was sampled for arthropods using 30 swings of a sweep net (15 inches diameter) that reached 5.8m into the canopy (see Siemann 1998 for comparisons of sampling methods affecting relative abundance and species richness). Sampling was conducted at approximately the same time of day and under similar weather conditions (dry and warm). Arthropod identification.— Arthropod specimens were sorted under magnification and identified to either species or morphospecies within family or genus, and abundance, and trophic group was recorded by taxon. Individuals from the order Araneida (spiders) were often not identified beyond order due to their taxonomic complexity and lack of a local reference collection. Morphospecies have been shown to correlate with arthropods identified by entomologists (Oliver & Beattie 1996), and this technique is often effectively utilized in the characteriza¬ tion of communities (Ingham & Samways 1996; Siemann 1998; Symstad et al. 2000). 372 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Determination of trophic level.— For each species or morphospecies, a trophic group was determined for the developmental stage at which the individual was caught by referring to relevant literature (Arnett 1960 1993; Borror & White 1970; McAlpine et al. 1981; 1987; Schuh & Slater 1995). The functional groups were the following: herbivore, detritivore, predator, parasite, omnivore, non- feeding, and unknown. Herbivores included any arthropod feeding primarily on living plant material. Omnivores were defined as individuals feeding on plants and animals. The group ‘unknown’ was assigned for those whose trophic grouping could not be determined through lack of available knowledge or insufficient taxonomic determination. Little is known about feeding habits for some taxonomic groups, especially those without agricultural or medical importance. There are some arthropods that only feed in their larval stage; therefore, a non- feeding group was included. Data from previous studies.— The native sites and habitats sampled by Cameron & Byrant (1999) were located near Sealy, Texas, approximate¬ ly 1 10 km NW of the Sapium study site (UHCC). They sampled using a beating net for woody areas and a sweep net in herbaceous vegetation. The beating nets usually have heavier canvas fabric that collects smaller individuals than a sweep net. The habitats included: riparian woodland with ungrazed pasture and savanna woodland (RW1), dense riparian woodland with less open grassland (RW2), bottomland woodland with dense herbaceous understory (BW3), fluvial woodland with open under¬ story with periodic flooding and bordered by pasture (FWP4), dense drier woodland with woody understory (DW5), grazed pasture with a few woody species (GP6), abandoned pasture with patches of riparian woodland (PW7), and coastal prairie with no woody vegetation, surrounded by agriculture and grazing (CP8). Cameron & Byrant (1999) did not include non-insect arthropods in their study so these groups were excluded from the UHCC data (including Sapium data) for comparative analyses. Two studies from UHCC on arthropod communities were also included in this study (Birch 1975; McFadden 1978). Arthropod data from high (HDB) and low densities (LDB) of Baccharis halimifolia L. were utilized from an earlier study by Birch (1975). Like Sapium , Baccharis is both common and woody, yet Baccharis is native to the area. Birch (1975) sampled the stands on four occasions in 1975, using a D-vac. Siemann (1998) found that relative richness and abundance values for D-vac and sweep net samples were strongly correlated. McFadden (1978) collected arthropod data in the coastal prairie at HARTLEY ET AL. 373 UHCC (UHCP) every two months, a total of seven times in the year, using a sweep net. Sampling effort was standardized for McFadden (1978), Birch (1975), and Cameron & Byrant (1999) by using relative rather than absolute values. Birch (1975) and McFadden (1978) are the only available studies on arthropod communities at the UHCC. Data analyses. — To assess the differences in the Sapium insect com¬ munity from those in native Texas habitats, a non-metric multidimen¬ sional scaling (NMS) ordination was conducted using relative abundance of seven insect orders from Sapium, high and low densities of Baccharis (Birch 1975), coastal prairie (McFadden 1978), and eight native sites studied by Cameron & Bryant (1999). Araneida and Acari were ex¬ cluded. NMS is a non-parametric, iterative technique based on ranked distances among sites (McCune & Grace 2002). The number of dimen¬ sions was determined by a minimal stress (departure from monotonicity) . The distance matrix of sites used for ordination was 1-DS, in which Ds is Sorensen’s similarity index. Using the distance matrix output by PC-ORD Version 4, the distance ordination was conducted in SAS V.8 (SAS Institute 2000) with routine PROC NMS. Results A total of 811 individuals and 160 species in 15 orders of arthropods were caught in a total of 1920 sweeps. Some orders were abundant on Sapium , such as Acari (mites), Araneida (spiders), and Diptera (flies), which accounted for 78% of the individuals in the community (Table 1). The most diverse orders were Diptera (36% species richness) and Acari (13% species richness). Coleoptera (beetles), Homoptera (leafhoppers) , Hymenoptera (wasps and ants) and Psocoptera (barklice) were less abundant on Sapium. Eight orders were rarely encountered (Collembola (springtails), Dictyoptera (mantids and cockroaches), Ephemeroptera (mayflies), Hemiptera (true bugs), Lepidoptera (moths and butterflies), Neuroptera (lacewings), Orthoptera (grasshoppers and crickets), and Thysanoptera (thrips). Twenty immature individuals were caught, of which 13 were Orthoptera, and the remainder were Coleoptera, Homop¬ tera, and Thysanoptera. A species accumulation curve was constructed to determine the number of species collected versus sampling effort for the data on Sapium. Three saturating equations were fitted to the curve (Tablecurve 2D, Systat, Point Richmond CA). They indicated that the total number of species in the community was 189 (first order intermediate kinetic 374 THE TEXAS JOURNAL OF SCIENCE- VOL. 56(4), 2004 Table 1 . Abundance and species richness of arthropods by taxonomic order summed over all samples. Order Abundance Species Richness Acari 165 20 Araneida 248 — Coleoptera 25 14 Collembola 1 1 Dictyoptera 2 2 Diptera 222 57 Ephemeroptera 1 1 Hemiptera 2 2 Homoptera 36 16 Hymenoptera 39 16 Lepidoptera 4 4 Neuroptera 14 7 Orthoptera 13 6 Psocoptera 36 11 Thysanoptera 3 3 TOTAL 811 160 function), 191 (simple equilibrium, net rate and equilibrium concentra¬ tion function), or 208 (first order intermediate kinetic function with equilibrium) which suggests the sampling effort on Sapium caught 85%, 84%, or 77% of the species respectively. A species-sweep curve con¬ structed by McFadden (1978) showed that 1000 sweeps would contain 85% of the diversity. Cameron & Byrant (1999) also estimated they collected 85% of the diversity (based on McFadden 1978). Birch (1975) did not create a sampling curve. The most abundant family encountered was Oripodidae (beetle or armored mites), which accounted for 14% of total arthropod community abundance (Table 2). Chironomidae (non-biting midges), Lauxaniidae (Lauxaniid flies), and Dolichopodidae (long legged flies) were also relatively common (Table 2). The most diverse (species rich) among these were Dolichopodidae and Chironomidae. Other common families were Psocidae (common barkl ice), Sciaridae (dark winged fungus gnats), Formicidae (ants) and Coccidae (scales) (Table 2). Only two families were encountered that might be considered as specialist herbivores. These were Coccidae (scales) and Cicadellidae (leafhoppers) both in the order Homoptera. Homoptera are often known to stay on host plants where their eggs are laid. Predators (326 individuals) and detritivores (241 individuals) together represented 70% of the arthropod assemblage supported by Sapium. Herbivores were considerably less abundant and composed only 7 % of HARTLEY ET AL. 375 Table 2. Fifteen most abundant families sampled from Sapium. The families listed account for 53% of total arthropod community abundance and 55% of total species richness. Order Family Abundance Species Richness Acari Oripodidae 115 5 Diptera Chironomidae 61 12 Diptera Lauxamidae 60 9 Diptera Dolichopodidae 53 13 Psocoptera Psocidae 25 6 Diptera Sciaridae 18 8 Hymenoptera Formicidae 16 4 Homoptera Coccidae 14 5 Homoptera Cicadellidae 11 5 Hymenoptera Braconidae 11 3 Diptera Chloropidae 10 6 Orthoptera Gryllidae 9 4 Neuroptera Chrysopidae 8 3 Psocoptera Pseudocaeciliidae 8 3 Coleoptera Coccinellidae 7 2 all Sapium community arthropods (58 individuals). Insect relative abundance for the additional trophic categories were 3 % for omnivores and parasites, 10% unknown, and 8% non- feeding on Sapium. How¬ ever, species richness was more evenly proportioned among the trophic categories. Detritivores were the most species rich (43 species or morphospecies) but only represented 27% of the community diversity. Both herbivores and predators had similar levels of diversity, represent¬ ing 20% and 17% respectively. The arthropod community on Sapium differed from the communities found on native sites sampled by Birch (1975), McFadden (1978), and Cameron & Byrant (1999) (Table 3). After Acari and Araneida data were removed, relative species richness and abundance were recalculated to make all the data sets comparable. The relative richness of herbi¬ vores (29%) was approximately 50% less on Sapium than on native vegetation (native herbivore range 49-67%). In contrast, both predator and detritivore relative richness was higher on Sapium (24% and 38% respectively) than the native site averages (12% and 16% respectively). The average relative species richness for predators from native sites was 12% (range 6-19%), and the average for detritivores (native sites) was 16% (range 7-24%). Parasites on Sapium were similar in their relative species richness (9%) compared to the native sites (range 8-21%). Cameron & Byrant (1999) did not present results on the trophic distribu¬ tion of arthropod abundance. 376 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Table 3. Arthropod relative species richness by trophic group for Sapium samples in this study (“Sapium") and habitats sampled by Birch (1975), McFadden (1978), and Cameron & By rant (1999). Refer to methods for description of sites. Sites Herbivore % Predator % Parasite % Detritivore % RW1 57 9 15 20 RW2 57 7 11 24 BW3 55 12 16 17 FWP4 61 11 12 17 DW5 59 11 9 21 GP6 58 6 14 22 PW7 67 8 10 15 CP8 54 13 16 16 UHCC prairie 67 19 8 7 HD Baccharis 56 10 21 12 LD Baccharis 49 14 21 16 Sapium 29 24 9 38 The comparison of community composition of Sapium and native sites sampled by Birch (1975), McFadden (1978), and Cameron & By rant (1999) showed both differences and similarities in the relative abundance of orders (Table 4). Arthropod relative abundance on Sapium was comparable within the range of relative abundance at native sites for Homoptera, Coleoptera, Orthoptera, Hymenoptera, and Lepidoptera (Table 4). However the relative abundance found on Sapium was higher for Diptera and ‘others’, and lower for Hemiptera (Table 4). The NMS ordination of relative abundance of orders indicated that the insect community on Sapium differed substantially from that of native sites (Figure 1). A 3-dimensional solution was found. However, a two dimensional graph is presented, for ease of interpretation (Figure 1). A total of 38 iterations were run for the final solution, and the final stress was 0.08196. A final stress value between 0.1 and 0.05 is generally interpreted as a good ordination with negligible risk of inferring false conclusions (McCune & Grace 2002) . The UHCC sites were distinctly separated from Cameron & Byrant’s (1999) sites along dimension 1 (Figure 1). The Sapium site was located at the extremes of both axes (Figure 1). The grazed pasture site (GP6) was the most similar native site to Sapium in insect community (Sorenson’s similarity index (SSI) = 0.75), followed by UHCC coastal prairie (UHCP) (SSI = 0.56), while abandoned pasture with patches of riparian woodland (PW7) was the most different (SSI = 0.31). HARTLEY ET AL. 377 Table 4. Relative abundance of insects (Acari and Araneida excluded) by order from the native habitats sampled by Birch (1975), McFadden (1978), Cameron & Byrant (1999), and for Sapium samples in this study. These values are the percentage of each order within each site. ‘Others’ include all other orders not already listed. Refer to methods for site abbreviations. HOM = Homoptera, HEM = Hemiptera, COL = Coleoptera, ORT = Orthoptera, DIP=Diptera, HYM = Hymenoptera, LEP = Lepidoptera. Sites HOM HEM COL ORT DIP HYM LEP Others RW1 13.2 39.0 33.1 3.6 6.9 3.8 0.3 0.4 RW2 29.8 11.6 14.6 26.0 11.8 5.1 1.0 0.0 BW3 23.0 8.7 47.3 7.5 8.3 4.3 0.6 0.3 FWP4 12.0 2.7 66.3 7.6 7.0 4.1 0.1 0.5 DW5 20.0 18.8 21.0 10.0 25.0 4.0 1.0 1.0 GP6 12.1 16.4 6.9 11.9 49.5 3.2 0.0 0.0 PW7 28.1 1.8 23.5 37.6 5.7 2.9 0.3 0.1 CP8 4.0 38.7 6.2 23.2 21.7 6.2 0.3 0.1 UHCC prairie 15.7 14.1 20.1 4.4 20.0 20.7 1.6 3.2 HD Baccharis 34.0 19.3 3.9 0.3 9.7 31.8 0.5 0.5 LD Baccharis 17.8 39.7 1.9 0.0 11.6 23.4 1.0 4.7 Sapium 9.0 0.5 6.3 3.3 55.8 9.8 1.0 14.3 Discussion Consistent with the enemies release hypothesis, Sapium woodlands in southeastern Texas supported communities depauperate in herbivores and specialists, and were instead composed primarily of predators and detritivores (Table 1, Table 3). These data support earlier predictions of fewer herbivores and a differing arthropod community structure on Sapium compared to native habitats. The differences in arthropod abundance and species richness between Sapium woodlands and native habitats were substantial (Figure 1 , Table 3). Nevertheless, Sapium may be in the early stages of acquiring a more typical insect assemblage. Other work has shown that introduced plants may take up to 300 years to support an insect fauna indistinguishable from native plants (Strong 1974; Strong et al. 1984). Therefore the difference in the fauna documented on Sapium might be consistent with only 100 years of colonization time in Texas. A large proportion of the species or morphospecies were infrequently encountered on Sapium , suggesting either a high number of transient individuals or rare individuals. This is considered typical in arthropod communities (Siemann et al. 1999). The differences in arthropod communities between Sapium woodlands and native habitats might reflect unusual taxonomic, phenological, 378 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Dimension 1 Figure 1 . Non-metric multi dimensional scaling ordination of the relative abundance of the seven major insect orders (see Table 4) sampled from Sapium, UHCC coastal prairie (McFadden 1978), high and low densities of Baccharis (Birch 1975), and the eight native habitats from Cameron & Byrant (1999). Refer to methods for site abbreviations. biochemical, and morphological properties of the exotic species (Strong et al. 1984). Taxonomically, there are no other native tree species belonging to the Euphorbiaceace family, although there are a number of herbs such as Euphorbia bicolor (snow-on-the-prairie) and Croton capitatus (woolly croton). However, phenologically and morphological¬ ly it is similar to the native mid-sized, broad-leaved deciduous trees, such as Celtis laevigata (Bush & Van Auken 1986; Bruce et al. 1997), suggesting that Sapium is not unusual in this regard. Sapium' s ability to form dense monospecific stands and reduce habitat complexity in the HARTLEY ET AL. 379 understory is unprecedented in this region, thus simple plant architecture and or low local plant diversity might account for reduced arthropod diversity and abundance. Of all the native habitats examined, the grazed pasture site was most similar in arthropod composition to Sapium woodlands (Figure 1, Table 4). Both Sapium and grazed pasture are unnatural types of habitat. Originally the Sapium sampling location would have been coastal tail- grass prairie approximately 100 years ago, although 90 hectares of coastal prairie has now been restored. The UHCC coastal prairie site (McFadden 1978) was the second most similar native site, while the coastal prairie (Cameron & Byrant 1999) was the fourth most similar. Sapium woodlands may have recruited some arthropods from adjacent prairie habitat, and this may account for some degree of similarity between the arthropod community composition of Sapium and native coastal prairie sites sampled by McFadden (1978) and Cameron & Bryant (1999). Comparisons to Birch (1975), McFadden (1978), and Cameron & Byrant (1999) are informative. However, there are differences between the approaches that should be noted (also see methods). First, sampling was conducted at different times and years. Birch (1975), McFadden (1978), and Cameron & Byrant (1999) all sampled in the mid to late 1970’s, although there have been no significant, sudden, or large scale changes (such as land use change) in the UHCC vicinity. Furthermore Cameron & Byrant (1999) only sampled in the spring. Generally, insect communities increase in abundance at the beginning of the growing season and decrease at the end of the growing season, yet many popula¬ tions display substantial fluctuations. Sapium arthropod data (total abundance and species richness) exhibited no significant pattern of variation among the sampling periods. Secondly, sampling efforts could differ, but are difficult to quantify or compare. Thirdly, Birch (1975) also used a D-vac in addition to a sweep net (see Siemann 1998). Although there are differences in approaches, the overall relative results should not be greatly influenced by them, especially considering that both McFadden (1978) and Cameron & Byrant (1999) state they col¬ lected 85% of the diversity, which is comparable with the Sapium data (77-84%). This would suggest that their results are representative of the communities they sampled. Finally, the authors determined trophic data for McFadden (1978) from an appendix of the most common 95 species and morphospecies (from a total of 535). It was assumed that the 380 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 complete data would have been driven by the most abundant species and morphospecies and so the trophic data determined would reflect this. These factors may have influenced the contrast between the insects found on Sapium and in native habitats, but the data indicate a paucity of herbivores found on Sapium. In conclusion, Sapium woodlands seem to presently support an atypical arthropod fauna, with Diptera (flies), Acari (mites) and Araneida (spiders) as the dominant orders. Sapium' s fauna is mostly composed of predators and detritivores with very few herbivores. The apparent relative lack of a herbivorous food chain supports the predic¬ tion and may have important implications in ecosystem functioning. Although Sapium woodlands in southeastern Texas appear to have acquired few herbivores in the 100 years it has been present, it is expected that arthropod diversity and possibly abundance will continue to increase on Sapium and the composition of associated arthropod fauna will change to be more similar to native communities over time. Per¬ haps the accumulation of a more robust herbivore fauna will limit Sapium' s success as an invader in the future. Acknowledgments The authors would like to thank: University of Houston for access, EPA (R82-8903), NSF (DEB-9981654) and a Wray Todd fellowship (for MKH) for support; John Jackman, John Oswald, Ed Riley, and Jim Woolley from the Texas A & M University Entomology department and William Mackay at the University of Texas at El Paso, for help with identifications; Glenn Aumann, Ann Awantang, Tim Becker, Brad Butterfield, Candy Donahue, Will Gordon, James Hammer, Paul Harcombe, Stephanie Hsia, June Keay, Viki Keener, Rick Lankau, Mary Mackay, Daniel Mee, Summer Nijjer, Rachel Tardif, and Liz Urban for assistance; Paul Harcombe, Guy Cameron and an anonymous reviewer for helpful comments. Literature Cited Arnett, R. H. Jr. 1960. The Beetles of the United States (A manual for Identification). The Catholic University of America Press, Washington, D.C, 1112 pp. Arnett, R. H. Jr. 1993. American insects: A handbook of the insects of America north of Mexico. Sandhill Crane Press, Florida, 1003 pp. Birch, J. B. 1975. Faunal effects of shrub concentration. Unpublished Ph.D. dissertation. University of Houston, 226 pp. Borror, D. J. & R. E. White. 1970. Peterson Field Guides. Insects. Houghton Mifflin Company, Boston, 404 pp. HARTLEY ET AL. 381 Brown, V. K. 1984. Secondary succession: Insect-plant relationships. BioScience, 34:710-716. Bruce, K. A., G. N. Cameron, P. A. Harcombe & G. Jubinsky. 1997. Introduction, impact on native habitats, and management of a woody invader, the Chinese tallow tree ( Sapium sebiferum (L.) Roxb). Nat. Areas J., 17:255-260. Bush, J. K. & O. W. Van Auken. 1986. Light requirements of Acacia smallii and Celtis laevigata in relation to secondary succession on floodplains of South Texas. Am. Mid. Nat., 115:118-122. Cameron, G. N. & E. H. Bryant. 1999. Species composition and trophic structure of insect communities in Texas prairies. Prairie Nat., 31(4):221-242. Davidson, D. W. & S. R. Morton. 1981. Competition for dispersal in ant-dispersed plants. Science, 213:1259-1261. Elton, C. S. 1958. The ecology of invasion by plants and animals. Chapman and Hall, London, 181 pp. Feeny, P. 1976. Plant apparency and chemical defense: Rec. Adv. Phytochem., 10:1-40. Flack, S. & E. Furlow. 1996. America’s least wanted: the dirty dozen: Nat. Cons., 46(6): 17-23. Groves, R. H. 1989. Ecological control of invasive terrestrial plants. Pp. 437-462, in Biological invasions: A global perspective (J. A. Drake, H. A. Mooney, F. di Castri, R. H. Groves, F. J. Kruger, M. Rejmanek & M. W. Williamson, eds.), John Wiley and Sons, Chichester, 525 pp. Goeden, R. D. & S. M. Louda. 1976. Biotic interference with insects imported for weed control. Ann. Rev. Ento., 21:325-342. Ingham, D. S. & M. J. Samways. 1996. Application of fragmentation and variegation models to epigaeic invertebrates in South Africa. Cons. Biol., 10(5): 1353-1358. Jones, C. G., J. H. Lawton & M. Shachak. 1994. Organisms as ecosystem engineers. Oikos, 69:373-386. Keane, R. M. & M. J. Crawley. 2002. Exotic plant invasions and the enemy release hypothesis. Trends Ecol. Evol., 17:164-170. Lankau, R .L., W. E. Rogers & E. Siemann. 2004. Constraints on the utilization of the invasive Chinese Tallow Tree {Sapium sebiferum ) by generalist native herbivores in coastal prairies. Ecol. Ento., 29:66-75. Leibhold, A. M., W. L. MacDonald, D. Bergdahl & V. C. Mastro. 1995. Invasion by exotic forest pests: a threat to forest ecosystems: For. Sci. Monog., 30:1-49. Lodge, D. M. 1993. Biological invasions: lessons for ecology. Trends Ecol. Evol., 8:133-137. Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, M. Clout & F. A. Bazzaz. 2000. Biotic invasions: causes, epidemiology, global consequences, and control. Ecol. Appl.. 10:689-710. McAlpine, J. F., B. V. Peterson, G. E. Shewed, H. J. Teskey, J. R. Vockeroth & D. M. Wood. 1981. Manual of Neartic Diptera: vol.l, vi + 674 pp. Research Branch, Agriculture Canada (Monograph No. 27). McAlpine, J. F., B. V. Peterson, G. E. Shewed, H. J. Teskey, J. R. Vockeroth & D. M. Wood. 1987. Manual of Neartic Diptera: vol.2, vi + 1332 pp. Research Branch, Agriculture Canada (Monograph No. 28). McCune, B. & J. Grace. 2002. Analysis of Ecological Communities. MjM Software Design, Gleneden Beach, OR, USA. McFadden, R. T. 1978. Structure and seasonality in an arthropod community. Unpublished Ph.D. dissertation. University of Houston, 61 pp. Oliver, I. & A. J. Beattie. 1996. Invertebrate morphospecies as surrogates for species: a case study. Cons. Biol., 10(1):99-109. 382 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Petrusewicz, K. & W. L. Grodzinski. 1975. The role of herbivore consumers in various ecosystems. Pp. 64-70, in Productivity of world ecosystems (D. E. Reichle, J. F. Franklin, & D. W. Goodall, eds.), National Academy of Sciences, Washington, D.C., 166 pp. Price, P. W., C. E. Bouton, P. Gross, B. A. McPheron, J. N. Thompson & A. E. Weis. 1980. Interactions among three trophic levels: Influence of plants on interactions between insect herbivores and natural enemies. Annu. Rev. Ecol. Syst., 11:41-65. Price, P. W. 1997. Insect Ecology. 3rd Ed. John Wiley & Sons. New York, 874 pp. Proctor, M. & P. Yeo. 1972. The Pollination of Flowers. Taplinger Publishing Company, NY. SAS Institute. 2000. SAS/STAT User’s Guide, Version 8, Volumes 1, 2, & 3, Cary, North Carolina, USA. Schowalter, T. D. 1995. Canopy invertebrate community response to disturbance and consequences of herbivory in temperate and tropical forests. Selbyana, 16(l):41-48. Schowalter, T. D. 2000. Insect Ecology: An Ecosystem Approach. New York: Academic Press, 483 pp. Schuh, R. T. & J. A. Slater. 1995. True Bugs of the World (Hemiptera: Heteroptera) . Classification and Natural History. Cornell University Press, Ithaca, New York, 336 pp. Siemann, E. 1998. Experimental tests of the effects of plant productivity and plant diversity on grassland arthropod diversity. Ecology, 79:2057-2070. Siemann, E. , D. Tilman & J. Haarstad. 1999. Abundance, diversity and body size: patterns from a grassland arthropod community. J. Anim. Ecol., 68:824-835. Siemann, E. & W. E. Rogers. 2001. Genetic differences in growth of an invasive tree species. Ecol. Lett., 4:514-518. Siemann, E. & W. E. Rogers. 2003a. Herbivory, disease, recruitment limitation and the success of an alien tree species. Ecology, 84:1489-1505. Siemann, E. & W. E. Rogers. 2003b. Reduced resistance of invasive varieties of the alien tree Sapium sebiferum to a generalist herbivore. Oecologia, 135:451-457. Southwood, T. R. E., V. C. Moran & C. E. J. Kennedy. 1982. The richness, abundance and biomass of the arthropod communities on trees. J. Anim. Ecol., 51:635-649. Strong, D. R. 1974. Rapid asymptotic species accumulation in phytophagous insect communities: the pests of cacao. Science, 185:1064-1066. Strong, D. R., J. H. Lawton & Sir R. Southwood. 1984. Insects on plants: community patterns and mechanisms. Blackwell Scientific publications, 313 pp. Symstad, A. J., E. Siemann & J. Haarstad. 2000. An experimental test of the effect of plant functional group diversity on arthropod diversity. Oikos, 89:243-253. Tilman, D. 1999. The ecological consequences of changes in biodiversity: a search for general principles. Ecology, 80:1455-1474. Vitousek, P. M. 1986. Biological invasions and ecosystem properties: can species make a difference? Pp. 163-176, in Ecology of biological invasions of North America and Hawaii: Ecological studies (H. A. Mooney, & J. A. Drake, eds.), 58. Springer- Verlag, NY Inc., 321 pp. Wilson, E. O. 1992. The Diversity of Life. Cambridge, MA: Harvard Univ. Press, 424 pp. Yela, J. L. & J. H. Lawton. 1997. Insect herbivore loads on native and introduced plants; a preliminary study. Entomol. Exp. Appl., 85:275-279. MKH at: mariak@rice.edu TEXAS J. SCI. 56(4):383-394 NOVEMBER, 2004 DIEL ACTIVITY PATTERNS OF THE LOUISIANA PINE SNAKE {PITUOPHIS RUTHVENI) IN EASTERN TEXAS Marc J. Ealy, Robert R. Fleet and D. Craig Rudolph Texas Parks and Wildlife Department, 1700 7th St., Rm. 101 Bay City, Texas 77414; Department of Mathematics and Statistics, Stephen F. Austin State University Nacogdoches , Texas 75962 and USD A Forest Service, Southern Research Station, 506 Hay ter St. Nacogdoches, Texas 75965 Abstract.— This study examined the diel activity patterns of six Louisiana pine snakes in eastern Texas using radio-telemetry. Snakes were monitored for 44 days on two study areas from May to October 1996. Louisiana pine snakes were primarily diurnal with moderate crepuscular activity, spending the night within pocket gopher burrows or inactive on the surface. During daylight hours, snakes spent approximately 59% of their time underground within gopher burrows, burned out/ rotten stumps, or nine-banded armadillo (. Dasypus novemcinctus ) burrows. Remaining time was spent on the surface either close to subter¬ ranean refuge, or in long distance movements that generally terminated at another pocket gopher burrow system. Long distance movements occurred on 45% of the days snakes were monitored and averaged 163 m/movement. When snakes were active, movements related to ambientair temperature; 82% of these movements occurred between 1000 and 1800 hours. These results confirm that Louisiana pine snakes are diurnal and closely associated with Baird’s pocket gophers and their burrow systems, and have provided new insight on the ecology of this rare snake. The Louisiana pine snake ( Pituophis ruthveni), first described by Stull (1929), is a large-bodied constrictor of the family Colubridae and until recently was considered one of 15 subspecies of Pituophis melanoleucus (see Sweet & Parker 1990; Collins 1991; Crother et al. 2003). The Louisiana pine snake is allopatric to other Pituophis and its distribution is primarily restricted to the longleaf pine ( Pinus palustris ) ecosystem of west-central Louisiana and eastern Texas (Conant 1956; Reichling 1995). The longleaf pine ecosystem is perpetuated by frequent fire (Platt et al. 1988; Frost 1993). Louisiana pine snakes are semi-fossorial and are closely associated with Baird’s pocket gopher ( Geomys breviceps ) burrow systems (Rudolph & Burgdorf 1997). Baird’s pocket gophers are the predominant prey of Louisiana pine snakes and their burrow systems are used for foraging, shelter, escape from frequent fires, and hibernation (Rudolph et al. 1998; 2003). Many have reported on the apparent rarity of P. ruthveni ; this can be 384 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 partly attributed to its semi-fossorial habits and secretive nature (Conant 1956; Young & Vandeventer 1988; Rudolph & Burgdorf 1997). Only 57 records of P. ruthveni were available through 1990 (Conant 1956; Jennings & Fritts 1983; Young & Vandeventer 1988; Reichling 1989). As a result, this species is considered to be one of the rarest snakes in North America (Thomas et al. 1991). Extreme rarity has prevented researchers from collecting substantial ecological and natural history data on the species and accounts for the paucity of available literature. In 1993, the USD A Forest Service Southern Research Station initiated a long term study of home range and habitat use of free ranging Louisiana pine snakes in eastern Texas and west-central Louisiana through the use of radio- telemetry. This portion of the study was con¬ ducted from May through October 1996 to elucidate diel activity patterns of this snake in eastern Texas. Study Areas Two areas were used to monitor Louisiana pine snakes in eastern Texas. Foxhunter’s Hill is a 500 ha longleaf pine savanna located on the Sabine National Forest approximately 25.5 km south of Hemphill, Texas, in Sabine County. The second area, Scrappin’ Valley, owned by Temple-Inland Forest Products Corporation, is approximately 29 km south of Hemphill, Texas, in Newton County. The portion of Scrappin’ Valley used as the study area is a 450 ha longleaf pine savanna. Characteristics common to both sites are: soils with high sand content; diverse herbaceous flora dominated by little bluestem ( Schizachyrium scoparium) and bracken fern ( Pteridium aquilinum ); over story domi¬ nated by longleaf pine ( Pinus palustris ), sparsely distributed blackjack oak ( Quercus marilandica) and blue jack oak ( Quercus incana ); and areas of encroachment by sweet gum ( Liquidambar styraciflua ), sassa¬ fras ( Sassafras albidum ), and yaupon (Ilex vomitoria) as a result of past fire suppression. Foxhunter’s Hill possesses moderate topographic relief, average basal area of 9 m2/ha, and heavy leaf litter accumulation and was burned by prescription in late winter of 1993. Scrappin’ Valley has lower topographic relief than Foxhunter’s Hill, average basal area of 6 m2/ha, moderate leaf litter accumulation, and was burned in late winter of 1995. Generally, Scrappin’ Valley was burned annually while Foxhunter’s Hill was burned every 3-5 years, resulting in differential leaf litter accumulation in the two areas. EALY, FLEET & RUDOLPH 385 Materials and Methods Transmitter implantation.— Louisiana pine snakes were captured on the study areas by hand or in drift fence/funnel traps. Temperature sensitive transmitters (Holohil Systems Ltd., SI-2T) 29mm long and 10 mm in diameter with 28 cm whip antennae were implanted subcutane¬ ously following the general procedure of Weatherhead & Anderka (1984). Transmitter life-span was approximately 18 months and maxi¬ mum transmission range was approximately 1200 m. Radio-telemetry /data collection.— Snakes were located early in the morning before they became active and emerged from subterranean shelter. A Trimble GPS Professional unit and data logger was used to record each snake’s location. Air temperature at the snake’s location was measured with a mercury thermometer 0.5 m above the ground in the shade. Substrate temperature was recorded in one of two ways: if the snake was aboveground, the thermometer was placed on the substrate as close as possible to the snake without disturbing it; if below ground, the thermometer was inserted approximately 5 cm into the soil. Snake body temperature was determined by comparison of transmitter pulse rate with a calibration curve for each transmitter. Throughout the day until sunset, transmitter pulse counts and air temperatures were recorded at 30-45 minute intervals. When the pulse count of a transmitter changed by becoming much slower or faster, indicating a temperature change of the implanted transmitter, the snake was relocated to determine if snake activity had occurred. Six snakes, three on Foxhunter’s Hill, and three on Scrappin’ Valley were monitored from dawn to dusk for a total of 44 snake days. Movements were recorded and calculated only if an individual moved more than 10 m from its previous location on a given day (Slip & Shine 1988). Move¬ ments on six additional days were recorded during the course of other data collection and were also available. Movement distances were calculated through the use of Trimble GPS Pathfinder Office software (Trimble Mapping and GIS Systems Division, Sunnyvale, CA). Periodic night checks were conducted by locating snakes at sunset and again at midnight and before sunrise to determine if the snakes were active nocturnally. Additional data regarding movement and choice of underground refugia were collected from these and other snakes in addition to the 44 snake monitoring days. Habitat measurements were taken at each snake relocation point as required for various aspects of research on P. ruthveni. Additional 386 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 habitat measurements were taken at 100 stratified random points deter¬ mined by overlaying a grid on the overall study site and using the inter¬ sections of the grid lines as the random points. The only habitat measurement relevant to this study was the number of burrows counted within an 11.2 m radius (0.04 ha) of each habitat point. Geomys breviceps “burrows” were counted as the number of visible push-up mounds and all other burrows were enumerated by the number of actual openings at or near the soil surface. Data analysis— Distance moved per snake each day was tested by a Mann-Whitney U-test. Chi-square contingency tests were used to evaluate the time each snake utilized above ground and below ground environments, movement frequency, and refuge/shelter types used. Frequency of movements during 12 two-hour time periods were evalu¬ ated by Chi-square contingency tests and all statistical analyses were performed at an alpha level of 0.05. Results Six P. ruthveni (5 F, 1 M) were monitored during all or most of a total of 44 snake days between July and October, 1996. During the 44 snake days of monitoring, individual snakes were located at the surface between sunrise and sunset for 145 hrs of a total of 354 hrs (41 %). The remainder of their time was spent underground in G. breviceps burrows, nine-banded armadillo burrows, and decayed or burned stump holes and associated root channels. In order to determine nocturnal behavior, the six P. ruthveni were monitored at approximately sunset, midnight, and sunrise for a total of 20 snake days during July and August. With one exception, all snakes were located below ground in G. breviceps burrows each night ( n = 17). The exception, a female, was located on the surface beneath dense herbaceous vegetation at sunset on three separate days and remained in that location until the next morning. One of these instances was during pre-ecdysis. For the 44 snake days when extensive monitoring oc¬ curred, snakes were assumed to have spent the previous night in G . breviceps burrows, based on early morning detections, a total of 29 times. These same snakes were assumed to have spent the succeeding night in subterranean retreats in 38 instances (35 in G. breviceps burrows, three in D. novemcintus burrows) based on detections at dusk. Data are not available for the remaining 21 nights. EALY, FLEET & RUDOLPH 387 Time of Day Figure 1. Body temperature (open circles), air temperature (open squares), and substrate temperature (open triangles) for a Louisiana pine snake ( Pituophis ruthveni ) spending daylight hours underground in a Baird’s pocket gopher ( Geomys breviceps ) burrow. Adult female 143 on 14 July 1996. Pituophis ruthveni monitored for daily activity during this study evinced three general daily activity patterns. In 17 cases, snakes re¬ mained in G. breviceps burrow systems for the entire daily tracking period (Fig. 1). All six snakes except one female from Scrappin’ Valley spent at least one entire day in a G. breviceps burrow. Conversely, three individuals spent an entire day on the surface. Two of these individuals moved significant distances (225 m and 59 m), and the third was in pre-ecdysis condition with clouded eyes. In 24 cases various combinations of time were spent on the surface and below ground. These cases were usually associated with substantial surface movement (19 of 24), usually culminating with entrance into another underground refuge (22 of 24) (Fig. 2). Of these 24 snake days, 12 involved snakes that were on the surface when first located in the morning and 12 were in G. breviceps burrow systems from which they subsequently emerged. It is unclear if the snakes initially located on the surface had emerged from underground refugia early or had spent the night on the surface, although sampling for nocturnal activity sug¬ gests the former in most instances. 388 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Figure 2. Body temperature (open circles), air temperature (open squares), and substrate temperature (open triangles) for a Louisiana pine snake ( Pituophis ruthveni ) spending portions of a day underground in a Baird’s pocket gopher ( Geomys breviceps) burrow and portions above ground. Adult female 118 on 03 August 1996. On the 27 snake days in which at least a portion of the day was spent on the surface plus six additional snake days for which movement distances are available, seven snakes remained in the same location, exhibiting only minor movements of < 10 m throughout the day. One individual moved 72 m from its initial location, but returned to its initial location by dusk. In 25 instances snakes moved substantial distances (> 10 m) during the day and were located an average of 163 m (range 1 1-625 m) from their initial location. Movements occurred from shortly after sunrise until dusk with the majority (82%) between 10:00 and 18:00 hours (Fig. 3). Overall, snakes moved a substantial distance on 20 of 44 days monitored (45.5%). There was a significant difference in frequency of movement between Scrappin’ Valley and Foxhunter’s Hill snakes (%2 = 9.99, df — 1, P < 0.005) with the Scrappin’ Valley snakes moving more frequently (Table 1). Daily movement distances were calculated by summing straight line measurements between con¬ secutive locations and should be interpreted as an underestimation since snakes rarely travel in a straight line (Secor 1994). On days when movement occurred, snakes at Scrappin’ Valley (Table 1) moved greater distances, (jc = 189 m, n = 19) than did those on Foxhunter’s Hill ( x EALY, FLEET & RUDOLPH 389 Time of Day Figure 3. Frequency distribution (%) of movements by six Louisiana pine snakes ( Pituophis ruthveni) relative to time of day. Data for 12 May - 27 October 1996. = 91 m, n = 7); this difference was significant (U = 40.5, df = 26, P < 0.05). Pine snake use of underground refugia was recorded on 44 days during which daily activity patterns were monitored and on other days when snakes were located for home range computation. Snakes used G. breviceps burrows (80.9%), decayed or burned stumps (15.4%), or D. novemcintus burrows (3.7%) as underground refugia. Based on habitat data collected at random points (Table 2), Scrappin’ Valley had signifi¬ cantly higher densities of G. breviceps burrows (x2 = 193.9, df = 1, P < 0.005) and other types of retreats (x2 = 10.2, df = 1, P < 0.005) than Foxhunter’s Hill. Compared to snakes at Foxhunter’s Hill, snakes at Scrappin’ Valley used underground retreats other than pocket gopher burrows more frequently (x2 = 29.31, df = 1, P < 0.001). The percent of time an individual utilized underground environments on days snakes were monitored was determined through visual observa¬ tions and making inferences from temperature relationships based on the snakes’ body temperature compared to air and substrate temperatures. Snakes at Scrappin’ Valley (Table 1) spent a significantly lower propor¬ tion of daylight hours underground (45%) compared to snakes at Fox¬ hunter’s Hill (74%) (x2 =19.96, df= 1, P<0.05). 390 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Table 1. Distance moved, movement frequency, and time spent below ground {% time sunrise to sunset) for six Louisiana pine snakes ( Pituophis ruthveni) at Scrappin’ Valley and Foxhunter’s Hill in eastern Texas. Study Area Range of movement (m) Mean distance moved per day (m) Movement frequency (%) % Time below ground Scrappin Valley 12-625 189 + 35 68 45 Foxhunter’s Hill 11-184 91+22 24 74 Combined 11-625 163+32 46 59 Table 2. Indices of burrow abundance at snake relocation points and random points (0.04 ha plot) (Scrappin’ Valley and Foxhunter’s Hill in eastern Texas). Study Area No. of gopher No. of gopher No. of burrows No. of burrows burrows at burrows at at snake at random snake relocation points random points relocation points points Scrappin Valley 7.74 2.52 1.28 0.70 Foxhunter’s Hill 8.08 0.64 0.62 0.37 During the May through October period when P. ruthveni tempera¬ tures were monitored, subterranean retreats, primarily G. breviceps burrows, provided a refuge from extreme temperatures. Pituophis ruthveni emerged from subterranean retreats at body temperatures ranging from 19 to 29 °C. The lower temperatures were recorded in May and October, and the higher temperatures were presumably associ¬ ated with snakes that were re-emerging within a day or had undergone a period of basking at the burrow entrance prior to actual emergence. Body temperatures of snakes in subterranean retreats were generally within 2°C of soil temperatures at a depth of 5 cm which ranged be¬ tween 20.75 and 32.5 °C. Body temperatures of snakes present on the surface ranged from 20 to 36.75 °C. However, snakes frequently maintained body temperatures between 25.5 and 34.5 °C by basking, even when air temperatures were as low as 22 °C. Air temperatures never exceeded 35.5 °C during moni¬ toring periods, but P. ruthveni frequently moved into subterranean re¬ treats as air temperatures approached 35 °C. Discussion Surface activity of P. ruthveni was determined to be essentially diurnal. Individuals were typically located in subterranean retreats, EALY, FLEET & RUDOLPH 391 generally those of G. breviceps, at night. Snakes located above ground at night were inactive and sheltered under low vegetation. Diurnally, P. ruthveni were located above ground 41% of the time, and all recorded movements occurred during daytime. Diurnal activity is typical of Pituophis sp. with the exception of populations located in desert environ¬ ments where diurnal activity is severely limited by high temperatures (Gibbons & Semlitsch 1987). Pituophis ruthveni also spent a substantial portion of daylight hours underground (59%), generally in burrows of G. breviceps. The close association of P. ruthveni with G. breviceps burrows provides substantial opportunity to avoid extreme air tempera¬ tures. The close association with the burrows of G. breviceps is consistent with other observations of the ecology of P. ruthveni. Geornys breviceps is the primary prey of P. ruthveni (Rudolph et al. 2003), and decline or loss of G. breviceps populations, generally resulting from alteration of the fire regime, is hypothesized to be an important cause of population declines (Rudolph & Burgdorf 1997). In addition, G. breviceps burrows are the only documented hibernaculum sites, and are used for escape from predators and fire (Rudolph et al. 1998). Pituophis ruthveni were relatively immobile (i.e., moved < 10 m) on 54.5% of days monitored. This is consistent with a figure of 43% for northern pine snakes, P. melanoleucus melanoleucus , in New Jersey (Burger & Zappalorti 1989). Relative inactivity has been hypothesized to be a critical component of the thermal ecology of reptiles (Gans & Dawson 1976). This may be the case with P. ruthveni because remain¬ ing immobile near a subterranean retreat provides immediate access to two divergent thermal regimes. Huey (1982) also suggested that inactivity conserves energy and reduces the risk of predation. In a generally more mobile and active species, Coluber constrictor , Plummer & Congdon (1994) found that 90% of inactivity was associated with ecdysis. In P. ruthveni , only 13% of inactive days were associated with ecdysis, suggesting that the previously mentioned factors may be involved in the relative inactivity of this species. Pituophis ruthveni moved an average of 1 63 m/d on those days when substantial movements were undertaken. This is similar to the findings of Fitch & Shirer (1971) for P. catenifer in Kansas (142 m/d) and considerably greater than Parker & Brown (1980) found for P. catenifer deserticola in Utah (71 m/d). Long-distance movements in P. ruthveni 392 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 generally involved movement from one G. breviceps burrow system to another and consequently reflect the dispersed distribution of these burrow systems. Pituophis ruthveni , during this and associated studies were found to move very little while underground in G. breviceps burrows, typically remaining near the point of entrance in the relatively shallow foraging tunnels. This suggests that P. ruthveni behave as sit- and- wait predators when hunting pocket gophers, rather than actively searching within the burrow system. Geomys breviceps maintain an intricate burrow complex that can reach 180 m in length (Schmidly 1983), and they can rapidly construct an earthen plug effectively limiting movement by P. ruthveni (Rudolph et al. 2003). These observations suggest that a sit- and- wait strategy combined with a brief pursuit may be the most effective strategy to capture G. breviceps. Pituophis ruthveni behavior differed significantly, based on three criteria, between the Scrappin’ Valley and Foxhunter’s Hill study sites. Snakes at Scrappin’ Valley moved more frequently, moved greater distances, and spent less time underground compared to snakes at Foxhunter’s Hill. The Scrappin’ Valley site was also characterized by a greater density of both G. breviceps burrows and other types of retreats compared to the Foxhunter’s Hill site. It is possible that the greater availability of subterranean retreats at Scrappin’ Valley resulted in fewer restrictions on above ground activity by P. ruthveni. The greater availability of G. breviceps burrows and other subterranean retreats (primarily burned stump and root channels) is presumably related to the more frequent prescribed fire regime at the Scrappin’ Valley site. The use of subterranean retreats during the active period of the year provided P . ruthveni with predictable escape from excessively high air temperatures. Conversely, snakes also had direct access to basking opportunities on the surface that allowed the snakes to maintain a higher body temperature during substantial periods. This general pattern is similar to the results of Himes et al. (2002) for this species in northern Louisiana. The diel activity budget of P. ruthveni reveals a species that is diurnal and semifossorial as is generally typical of other members of the genus in the United States (Fitch & Shirer 1971 ; Parker & Brown 1980; Sweet & Parker 1990). The importance of burrows of Baird’s pocket gophers when combined with previous data and observations (Rudolph & EALY, FLEET & RUDOLPH 393 Burgdorf 1997; Rudolph et al. 1998; 2003) supports the hypothesis that P. ruthveni is dependent on G. breviceps and ultimately on a frequent fire regime that maintains the herbaceous vegetation that supports G . breviceps populations. Acknowledgments B. Autrey, S. J. Burgdorf, R. R. Schaefer, R. N. Conner, R. Maxey, and C. M. Duran provided assistance in collection of field data and other aspects of this research. Temple-Inland Forest Products Corp. provided access to the Scrappin’ Valley study site. The U.S. Fish and Wildlife Service and Texas Parks and Wildlife Department provided partial funding under Section 6 of the U. S. Endangered Species Act and Texas Parks and Wildlife Department issued the required permits. Literature Cited Burger, J. & R. T. Zappalorti. 1988. Habitat use by pine snakes (Pituophis melanoleucus) in the New Jersey Pine Barrens; individual and sexual variation. J. Herpetol., 23(l):68-73. Collins, J. T. 1991. Viewpoint: A new taxonomic arrangement for some North American amphibians and reptiles. Herpetol. Rev., 22(2):42-43. Conant, R. 1956. A review of two rare pine snakes from the Gulf coastal plain. Amer. Mus. Novitates, (1781): 1-31 . Crother, B. I., J. Boundy, J. A. Campbell, K De Quieroz, D. Frost, D. M. Green, R. Highton, J. B. Iverson, R. W. McDiarmid, P. A. Meylan, T. W. Reeder, M. E. Seidel, J. W. Sites, Jr., S. G. Tilley & D. B. Wake. 2003. Scientific and standard English names of amphibians and reptiles of North America north of Mexico: update. Herp. Rev., 34(3): 196-203. Davis, W. B. & D. J. Schmidly. 1994. The Mammals of Texas. Texas Parks and Wildlife Press, Austin, 338pp. Fitch, H. S. & H. W. Shirer. 1971. A radio telemetric study of spatial relationships in some common snakes. Copeia, 1971(1): 1 18-128. Frost, C. C. 1993. Four centuries of changing landscape patterns in the longleaf pine ecosystem. Proc. Tall Timbers Fire Ecol. Conf. , 18:17-43. Gans, C. & W. R. Dawson. 1976. Reptilian physiology: An overview. Pp. 1-17, in Biology of the Reptilia. C. Gans and W. R. Dawson, (eds). Academic Press, London and New York, 556 pp. Gibbons, J. W. & R. D. Semlitsch. 1987. Activity patterns. Pp. 396-421, in R. A. Siegel, J. T. Collins, and S. S. Novak (eds.), Snakes: Ecology and Evolutionary Biology. McGraw Hill, New York, 529 pp. Himes, J. G., L. M. Hardy, D. C. Rudolph & S. J. Burgdorf. 2002. Body temperature variations of the Louisiana pine snake ( Pituophis ruthveni) in a longleaf pine ecosystem. Herpetol. Nat. History 9(2):1 17-126. Huey, R. B. 1982. Temperature, physiology, and the ecology of reptiles. Pp. 25-91, in Biology of the Reptilia. C. Gans, ed. Academic Press, London and New York, 502 pp. Parker, W. S. & W. S. Brown. 1980. Comparative ecology of two colubrid snakes, Masticophis taeniatus taeniatus and Pituophis melanoleucus deserticola, in northern Utah. 394 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Publ. Biol. Geol. No. 7, Milwaukee Publ. Mus., 104p. Platt, W. J., G. W. Evans & S.L. Rathbun. 1988. The population dynamics of a long-lived conifer {Pinus palustris). Amer. Nat. 131 (4) :49 1-525. Plummer, M. V. & J. D. Congdon. 1994. Radio telemetric study of activity and movement of racers ( Coluber constrictor ) associated with a Carolina Bay in South Carolina. Copeia, 1994(7): 20-26. Reichling, S. B. 1995. The taxonomic status of the Louisiana pine snake ( Pituophis melanolecus ruthveni) and its relevance to the evolutionary species concept. J. HerpetoL, 29(2): 186-198. Rudolph, D. C. & S. J. Burgdorf. 1997. Timber rattlesnakes and Louisiana pine snakes of the west Gulf coastal plain: hypotheses of decline. Texas J. Sci., 49(3) Supplement: 111-122. Rudolph, D. C., S. J. Burgdorf, J. C. Tull, M. Ealy, R. N. Conner, R. R. Schaefer & R. R. Fleet. 1998. Avoidance of fire by Louisiana pine snakes, Pituophis melanoleucus ruthveni. Herpetol . Rev . , 29(3) : 1 46- 1 48 . Rudolph, D. C., S. J. Burgdorf, R. N. Conner, C. S. Collins, D. Saenz, R. R. Schaefer, T. Trees, C. M. Duran, M. Ealy & J. G. Himes. 2003. Prey handling and diet of Louisiana pine snakes {Pituophis ruthveni) and black pine snakes (P. melanoleucus lodingi) with comparisons to other selected colubrid taxa. Herpetol. Nat. History, 9(l):57-62. Schmidly, D. J. 1983. Texas mammals east of the Balcones fault zone. Texas A&M Univ. Press, College Station, 400p. Secor, S. M. 1994. Ecological significance of movements and activity range for the sidewinder, Crotalus cerastes. Copeia, 1994(3): 63 1-645. Slip, D. J. & R. Shine. 1988. Habitat use, movements, and activity patterns of free-ranging diamond pythons, Morelia spilota spilota (Serpentes: Boidae): A radio telemetric study. Aust. Wild. Res., 15:515-531. Stull, O. G. 1929. The description of a new subspecies of Pituophis melanoleucus from Louisiana. Occas. Papers Mus. Zool. Univ. Michigan, 205:1-3. Sweet, S. S. & W. S. Parker. 1990. Pituophis melanoleucus. Catalogue of American Amphibians and Reptiles, 474:1-8. Thomas, R., B. J. Davis & M. R. Culbertson. 1976. Notes on variation and range of the Louisiana pine snakes, Pituophis melanoleucus ruthveni, Stull (Reptilia, Serpentes, Colubridae). J. Herpetol.. 10(3): 252-254. U. S. Fish and Wildlife Service. 1991. Animals proposed for review. Federal Register, 56(225) :58804-588 1 3 . Weatherhead, P.J. & F. W. Anderka. 1984. An improved radio transmitter and implantation technique for snakes. J. Herpetol., 18(3):264-269. Young, R. A. & T. L. Vanderventer. 1988. Recent observations on the Louisiana pine snakes, Pituophis melanoleucus ruthveni (Stull). Bull. Chicago Herp. Soc., 23:203-207. MJE at: ealy@wcnet.net TEXAS J. SCI. 56(4): 395-404 NOVEMBER, 2004 ARBOREAL BEHAVIOR IN THE TIMBER RATTLESNAKE, CROTALUS HORRIDUS , IN EASTERN TEXAS D. Craig Rudolph, R. R. Schaefer, D. Saenz and R. N. Conner Southern Research Station, JJSD A Forest Service 506 Hay ter Street, Nacogdoches, Texas 75965 Abstract.— There have been several recent reports, and anecdotal observations extending back at least to J. J. Audubon, suggesting that the timber rattlesnake ( Crotalus horridus) is one of the most arboreal members of the genus. Most previous records are of snakes located at heights of less than 5 m. Telemetry studies in eastern Texas have documented more frequent arboreal activity (16.1% of locations of sub-adult snakes) and at greater heights (up to 14.5 m) than previously reported. Unlike previous reports, observations of arboreal activity were restricted to sub-adult snakes (<90 cm SVL), possibly because adult snakes in the current study area are considerably larger than those in other areas where arboreal activity has been documented. Increasing body size and mass may preclude arboreal behavior in larger individuals of this species. Despite considerable speculation on the motivation(s) for arboreal activity in this species, the factors involved remain unclear. Arboreal behavior in snakes is increasingly recognized as an important aspect of snake ecology (Lilly white & Henderson 1993). Anecdotal accounts of arboreal activity by timber rattlesnakes ( Crotalus horridus) date back at least to Audubon (Klauber 1972). In a well known painting by Audubon a timber rattlesnake is depicted attacking Northern Mockingbirds ( Mimas polyglottos) in a shrub. This painting has elicited considerable discussion concerning the arboreal proclivities of timber rattlesnakes (Klauber 1972). In recent years, increasing use of radio-telemetry to investigate the biology of timber rattlesnakes has resulted in a proliferation of reports and citations of arboreal activity (Saenz et al. 1996; Coupe 2001; Fogel et al. 2002, Sealy 2002, Bartz & Sajdak 2004). During an ongoing study of C. horridus in eastern Texas, Saenz et al. (1996) reported several observations of arboreal behavior. Observations subsequent to the Saenz et al. (1996) report suggest that arboreal behavior, at least by sub-adult individuals, is more frequent in eastern Texas and involves greater heights than previously reported. A detailed understanding of arboreal behavior in C. horridus is limited by the paucity of published records. Saenz et al. (1996) suggested that increasing snake size may limit arboreal behavior in C. horridus . Other authors have suggested that arboreal behavior may be 396 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 related to basking, avoiding flood waters, ecdysis and foraging (Klauber 1972; Coupe 2001; Fogel et al. 2002, Sajdak & Bartz 2004), and that females may exhibit more frequent arboreal activity than males (Coupe 2001). Additional observations reported here will help to clarify aspects of the arboreal behavior of C. horridus. Study Area and Methods The study area was on and adjacent to the floodplain of the Angelina River in Nacogdoches Co. , Texas. Specific study sites were the Stephen F. Austin Experimental Forest located 12 km SW of Nacogdoches (31° 30’N, 94° 47’ W), and the Loco Bayou Hunt Club located 15 km WSW of Nacogdoches (31° 31’N, 94° 50’W). Habitat at both sites consisted of bottomland hardwood forest dominated by oaks ( Quercus sp.), sweetgum ( Liquidamber styraciflua) and hickories {Cary a sp.); and adjacent upland forest dominated by loblolly and shortleaf pines {Pinus taeda and P. echinata ), oaks {Quercus sp.) and a diverse array of other species. Portions of the bottomland habitats were subject to winter and spring flooding in most years. Crotalus horridus were captured as encountered during the course of the study, transported to the laboratory, and implanted with S1-2T transmitters (Holohil Systems Ltd.). Transmitters were implanted subcutaneously following the general procedures of Reinert & Cundall (1982) and Weatherhead & Anderka (1984). Snakes were retained in the laboratory, with access to a heating pad, for approximately 7 d following surgery to facilitate healing. Transmitters were replaced at approximately 18 mo intervals. Following release, snakes were relocated at irregular intervals, GPS locations recorded, and a series of habitat measurements and other data recorded as required for ongoing studies. In instances where individuals were located in arboreal situations, snake height, plant species, diameter at breast height (dbh) of supporting tree, presence of vines and other pertinent observations were noted. A series of climbing trials using C. horridus were conducted on selected trees. Lengths of muscadine grape {Vitis rotundifolia) vines 3-6 cm in diameter were occasionally attached to tree trunks to simulate situations noted during climbing events. Observation of subsequent climbing behavior provided some indication of the arboreal abilities of C. horridus. A series of feeding trials were also conducted using Brown-headed RUDOLPH ET AL. 397 Cowbirds ( Molothrus ater) . Cowbirds were captured in mist nets or box traps, placed in cages with individual C. horridus of various sizes, and the snakes’ subsequent behavior recorded. Results Thirty four C. horridus (60-140 cm SVL) were radio- tracked between 1993 and 2000 yielding more than 500 relocations. During this period 12 sub-adult snakes <90 cm SVL and with a mass <510 g were relocated a total of 218 times. Eight of these 12 snakes were located in arboreal situations a total of 35 times (Table 1). Each of the four snakes <90 cm SVL never found in an arboreal location were individuals represented by less than 10 relocation points. Snakes larger than 90 cm SVL, range 90-140 cm, were never observed in arboreal situations, with one exception. An adult male (136 cm SVL) was located in a shrub at heights ranging from 0.5 to 1.2 m on three occasions during a 15 day period. This individual had uncharacteristically occupied a hibernacu- lum in a bottomland hardwood site prone to flooding. In each arboreal observation the snake had been forced out of the hibernaculum and into the shrub by rising water. This observation is not included in the analyses that follow. The 35 instances of arboreal behavior represent 16.1% (35 of 218) of total observations of snakes <90 cm SVL and 17.9% (35 of 196) of observations of those individuals located in arboreal situations at least once. Arboreal behavior was observed in all months from March to October, the general activity period of C. horridus in eastern Texas. Of the minimum of 21 separate climbing events, females were involved in 11, males in 10. Contingency table comparison of arboreal relocations vs. total relocations for females (18 of 147, 12.2%) and males (17 of 71, 23.9%) showed a slight, but significant bias favoring males (x2 = 3.88. P < 0.05). The heights at which C. horridus were located ranged from 0.8 - 14.5 m with a mean of 5.9 m based on the 23 distinct arboreal locations represented. Individual snakes were relocated in the same tree (n = 9), occasionally with minor movements (n = 2) , during subsequent reloca¬ tions ranging from three to 24 days. There is no way of knowing whether these individuals returned to the ground between observations. Instances where snakes were relocated in the same arboreal location on subsequent days were typically those located at greater heights, however the irregularity of the relocation schedule makes detailed comparisons difficult. In all cases where visual evaluation was possible, snakes were Table 1. Snake measurements and arboreal behavior data for timber rattlesnakes ( Crotalus horridus) in eastern Texas. 398 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 W) O O On ON •nOOinco^oininOO •o in m oo rr o o »n no it) in oo cn (N ri ri A ^ = 5 cm cm r- co 3 Tf w CO s v GO in "'t co NO ON CO rf co oo m m cn 't t (S v ^ no co c- in CO CM & J< a a H CO JD 4 a a § <3 § d S ^ c £ a a Si cu Q n in d o 00 ON o o o o o ON ON ON ON ON ON ON ON in 'Cf NO r- OO 00 r-> 00 00 00 00 00 00 r- r- r- in in m in >n 00 00 jO 3 Q ON ON ON 3 3 ON ON 'Ct ON ON ON >» ON S' r- n O O u 3 s *3 *3 O °0 -ct ON r- iH 1 00 -ct CO °0 o hO O -ct m c- CM to in (N CM CM »— 1 CM r- ON • 00 ON 3 < d D 00 CO r- ON OO r-» ON ^ L s s <-i co ts in n in Ste 5t CM Mfc CO =«; -ct =*te -ct £ £ £ £ >n n CM CM CM CM CM CM CO JO jo JO JO U JO JO JO JO JO 5t 5fc Mfc Mfc Mt: Mfc 3 "3 "3 "3 ”3 *3 "3 "3 'rt "3 (O 2.5 m, often much greater. Vines, smaller diameter trees with low branches, loose bark and leaning trunks potentially facilitated the climbing in six of these instances. However, in the remaining nine instances the snakes were located in canopy or sub¬ canopy trees (14 - 48 cm DBH) at heights of 4.5 - 14.5 m without obvious characteristics that would facilitate climbing. In the most extreme case, a C. horridus was located at a height of 14.5 m at the first major fork of a laurel oak ( Quercus p hellos). The trunk was vertical, with a clear bole, and no vines to facilitate climbing. Access to this site was limited to climbing the vertical trunk or via the canopies of adjacent trees. Climbing trials with C. horridus <90 cm SVL demonstrated limited climbing ability compared to other species ( Elaphe sp., Masticophis flagellum ) that typically exhibit arboreal behavior. In cases where smaller branches were available C. horridus were able to maneuver slowly along horizontal or inclined branches, bridge between branches, and coil around branches to maintain a stable hold. However, it was not possible to elicit climbing of vertical, or nearly vertical, branches of any diameter, or boles of trees. Throughout these trials snakes gave the impression of awkwardness and hesitancy. Eighteen laboratory trials were conducted in which birds ( Molothrus ater) were presented to C. horridus of various sizes (range 75 - 104 cm SVL), and subsequent prey capture occurred. In all instances following the initial strike, the snakes maintained a hold on the bird until death of the bird. Time until apparent death of the cowbirds ranged from 54-364 sec with a mean of 188 sec. Feathers appeared to present a substantial impediment to fang penetration, and in several instances the snakes were observed to manipulate the cowbird between their jaws without releasing the bird, often for several min, until they were able to penetrate the feathers with a fang. Smaller snakes that did not immediately achieve an effective bite often had the anterior portion of their body moved 400 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 around the cage by the struggles of the cowbirds. The overall behavior of the snakes striking birds was distinctly different from observations of these same snakes preying on a variety of mammalian species where prey was struck and immediately released. Discussion Previously published accounts (Saenz et al. 1996; Coupe 2001; Fogel et al. 2002) and included references and communications, Sealy 2002; Sajdak & Bartz 2004; Bartz & Sajdak 2004) suggest that C. horridus consistently exhibits arboreal behavior and vindicates portions of Audubon’s early observations. However, much remains to be learned about arboreal behavior in C. horridus , including prevalence, onto¬ genetic variation, geographic variation and motivation. Size appears to limit arboreal behavior in C. horridus . Published accounts (Saenz et al. 1996; Coupe 2001; Fogel et al. 2002; Sajdak & Bartz 2004; Bartz & Sajdak 2994; this study) report only five individuals >90 cm SVL demonstrating arboreal behavior: two individuals (99.5 and 112.5 cm SVL) reported by Coupe (2001) without specific details, two individuals (100.5 and 98.0 cm SVL) reported by Bartz & Sajdak (2004) engaged in courtship approximately 1 m above the ground, and the adult male individual reported in this study at modest heights after being forced from its hibernaculum by rising water. The relationship between size and arboreal behavior has not been reported previously, with the exception of Saenz et al. (1996) preliminary report of this study, presumably due to the relatively small adult size of the more northern populations involved in most previous reports. This study documents more extensive arboreal behavior by C. horridus , at least sub-adults, than previously reported (Coupe 2001; Fogel et al. 2002; Sajdak & Bartz 2004; Bartz & Sajdak 2004). Al¬ though Klauber (1972) characterized C. horridus as “among the more persistent climbers,” arboreal behavior has been described as uncommon (Fogel et al. 2002), and characterized as frequent, rare, numerous instances, rarely observed (communications in Coupe 2001) without specific details. Only Coupe (2001) provides more specific data, stating that C. horridus were observed in arboreal situations during 13.2% of relocations; however, this figure is based on the subset of individuals observed in such situations at least once. In this study sub- adults were located in arboreal situations during 16.1% of relocations, and restricting the data to only those individuals observed in arboreal situations at least once (comparable to Coupe’s 2001 data) raises this figure to 17.9%. RUDOLPH ET AL. 401 Obviously, these data are not directly comparable, primarily because C. horridus in the more northern populations rarely reach body lengths at which arboreal behavior becomes extremely rare in eastern Texas. This study, including the preliminary observations reported by Saenz et al. (1996), is the first to report arboreal activity at substantial heights. Most previous reports are of individuals located at modest heights of 3 m or less, with a maximum of 5 m (Coupe 2001; Fogel et al. 2002; Sajdak & Bartz 2004). In eastern Texas the mean height of arboreal locations was 5.9 m with a maximum of 14.5 m, considerably higher than previously reported for this species. Sub- adult C. horridus were regularly located in the lower portions of tree canopies. Arboreal behavior in C. horridus in eastern Texas appears to be more frequent and involve greater heights than is the case in more northern populations. It is important to realize, however, that this comparison is based on sub-adult individuals in eastern Texas, individuals comparable in size to most adults in more northern populations. These comparisons suggest that arboreal behavior is more prevalent in the more southern portions of the range of C. horridus. Additional data from a wider geographic range would be desirable. Coupe (2001) suggested that arboreal behavior might be more prevalent among females. In eastern Texas, males were more frequently observed in arboreal situations based on percent of observations. Over¬ all, currently available data do not demonstrate a consistent difference in arboreal behavior between females and males. The motivation leading to arboreal behavior in C. horridus has elicited considerable speculation but little insight. Of the 23 individuals involved in a minimum of 41 separate climbs and observed on a total of 107 separate days reported in Coupe (2001), Fogel et al. (2002), Sealy (2002), Sajdak & Bartz (2004), Bartz & Sajdak (2004), and this study, two were associated with flood waters, three with ecdysis, one with basking by a gravid female, and four (2 pairs) with courtship. All of these observations were of individuals at heights <5 m, generally <3 m. Attaining a preferred thermal regime (basking) could conceivably be associated with several of the above observations and unrecognized in others. However, in Texas obvious basking behavior is rare. Individu¬ als are generally exposed on the forest floor but do not seek open areas, track sun flecks, or show other behaviors that could be associated with basking. Even gravid females, which typically seek heavy cover (hollow logs, debris piles), do not need to bask given the relatively high 402 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 average temperatures in the region. Consequently, basking and other activities noted above can only account for a minority of the observations, and do not appear to be involved in the observations at more extreme heights. These considerations may not even represent the primary motivation that led to the initial climbing activity in all cases. Avoidance of terrestrial predators is potentially a factor leading to arboreal behavior. If restricted to periods when active foraging is not occurring (ecdysis, post- feeding periods after mobility is regained) benefits might result. However, data or observations that support this hypothesis are not available. Arboreal foraging is a possibility mentioned by Klauber (1972), Saenz et al. (1996), Fogel et al. (2002), and Coupe (2001). Arboreal foraging was verified in one instance (Sajdak & Bartz 2004) when a was observed capturing a Yellow-bellied Sapsucker ( Sphyrapicus varius) at a height of 4.5-6 m. Verification of arboreal foraging behavior is difficult because definitive foraging postures in arboreal situations, analogous to those described in terrestrial situations (Reinert et al. 1984), have not been recognized. Crotalus horridus preys primarily on endotherms (Clark 2001). Consequently, potential prey available in arboreal situations in Texas are restricted to numerous species of birds, southern flying squirrels ( Glaucomys volans) , squirrels ( Sciurus sp.) and a limited variety of other small mammals. In a recent compilation of the prey of C. horridus , Clark (2001) reported that approximately 1% of recorded prey items were birds, although a substantial number of those identified to species were primarily terrestrial. Squirrels of the genus Sciurus, the primary prey of adult C. horridus in eastern Texas are often abundant in arboreal situations. However Sciurus sp. , and in many cases even G. volans, are too large for C. horridus, of the sizes that typically climb, to handle. Birds would seem to be the most likely prey of C. horridus in arboreal situations. Climbing and predation on birds has been observed in other pitvipers. The shedao pitviper ( Gloydius shedaoensis) in China, a relatively thick-bodied pitviper where adults average 60-70 cm SVL, actively climbs trees and shrubs and ambushes birds primarily during periods of avian migration (Shine et al. 2002). Striking and holding avian prey, presumably a secondarily acquired trait in Crotalids that prey regularly on mammals (Martins et al. 2002; Stiles et al. 2002), may increase the efficiency of predation on birds. Striking and holding onto avian prey was the strategy used in the report of Sajdak & Bartz (2004), even during a minimum vertical fall of 3 m to a lower branch. Mam- RUDOLPH ET AL. 403 malian prey that is potentially more dangerous to C. horridus is typically released immediately after striking (Chiszar et al. 1982; Stiles et al. 2002). Strike and release would present significant difficulties in trailing prey that could fly, even for short distances, and would presumably be extremely difficult from arboreal situations (Martins et al. 2002). Observations of prey taxa, that present little potential danger or are potentially difficult to trail or handle, that various Crotalids strike and hold include scorpions, fishes, frogs, lizards and birds (Parker & Stotz 1977; Rubio 1998; Hayes & Duvall 1991; Reiserer 2002; Stiles et al. 2002). The limited climbing abilities of C. horridus may limit the possibilities of arboreal foraging to smaller snakes. The apparent lack of behaviors such as coiling around limbs for support, or specialized support postures used by other heavy bodied arboreal species would appear to limit the ability of C. horridus to capture and handle prey items in arboreal situa¬ tions. The report by Sajdak & Bartz (2004) of the C. horridus falling to a lower branch during prey capture supports this view. Despite these limitations, foraging remains the most likely general explanation for arboreal behavior in C. horridus. Acknowledgments We thank J. A. Matos, L. McBrayer, R. E. Thill, B. Parresol and R. R. Fleet for constructive comments and suggestions on an early draft of this manuscript. S. J. Burgdorf was a primary contributor to all aspects of this research, and J. G. Dickson was intimately involved during the early years of this study. M. Duran, T. Trees and J. C. Tull provided invaluable field assistance. The U. S. Fish and Wildlife Service and Texas Parks and Wildlife Department provided the necessary permits. We also thank J. Mast and the members of the Loco Bayou Hunt Club for access to study sites. The mention of trade names does not constitute endorsement by the U. S. Department of Agriculture. Literature Cited Bartz, A. D. & R. A. Sajdak. 2004. Crotalus horridus (timber rattlesnake). Arboreality, courtship. Herpetol. Rev., 35:61. Chiszar, D., C. Andren, G. Nilson, B. O’Connell, J. S. Mestas, Jr., H. M. Smith & C. W.Radcliffe. 1982. Strike-induced chemosensory searching in old world vipers and new world pit vipers. Anim. Learn. Behavior, 10(2): 121-125. Clark, R. W. 2001. Diet of the timber rattlesnake, Crotalus horridus. J. Herpetol., 36(3):494-499. Coupe, B. 2001. Arboreal behavior in timber rattlesnakes ( Crotalus horridus ). Herpetol. 404 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Rev., 32(2):83-85. Fogel, D. D., T. J. Leonard & J. D. Fawcett. 2002. Crotalus horridus horridus (Timber Rattlesnake). Climbing. Herpetol. Rev., 33(3):212. Hayes, W. K. & D. Duvall. 1991. A field study of prairie rattlesnake predatory strikes. Herpetologica, 47(1):78-81. Klauber, L. M. 1972. Rattlesnakes: Their habits, life histories, and influence on mankind. University of California Press. Berkeley, CA. 1536 pp. Lillywhite, H. B. & R. W. Henderson. 1993. Behavioral and functional ecology of arboreal snakes. Pp. 1-48, in Snakes: ecology and behavior (R. A. Seigel & J. T. Collins, eds.) McGraw Hill, Inc., New York, NY. 414 pp. Martins, M., O., A. V. Marques & I. Sazima. 2002. Ecological and phylogenetic correlates of feeding habits in neotropical pitvipers of the genus Bothrops. Pp. 307-328, in Biology of the vipers (G. W. Schuett, M. Hoggren, M. E. Douglas, & H. W. Greene, eds.) Eagle Mountain Publ. Eagle Mountain, UT. 580 pp. Parker, S. A. & D. Stotz. 1977. An observation on foraging behavior of the Arizona ridge-nosed rattlesnake, Crotalus willardi willardi, (Serpentes:Crotalidae). Bull. Maryland Herpetol. Soc., 13:123. Reinert, H. K. & D. Cundall. 1982. An improved surgical implantation method for radio-tracking snakes. Copeia, 1 982(3) :702-705. Reinert, H. K., D. Cundall & L. M. Bushar. 1984. Foraging behavior of the timber rattlesnake, Crotalus horridus. Copeia, 1 984(4) :976-981. Reiserer, R. S. 2002. Stimulus control of caudal luring and other feeding responses: a program for research on visual perception in vipers. Pp. 361-384, in Biology of the vipers (G. W. Schuett, M. Hoggren, M. E. Douglas, & H. W. Greene, eds.) Eagle Mountain Publ. Eagle Mountain, UT. 580 pp. Rubio, M. 1998. Rattlesnake, portrait of a predator. Smithsonian Institution Press. Washington and London. 239 pp. Sajdak, R. A. & A. W. Bartz. 2004. Crotalus horridus (timber rattlesnake). Arbo reality, diet. Herpetol. Rev. 31(1):60-61. Saenz, D., S. J. Burgdorf, D. C. Rudolph & C. M. Duran. 1996. Crotalus horridus (Timber Rattlesnake). Climbing. Herpetol. Rev., 27(3): 145. Sealy, J. B. 2002. Ecology and behavior of the timber rattlesnake ( Crotalus horridus ) in the upper Piedmont of North Carolina: identified threats and conservation recommendations. Pp. 561-578, in Biology of the vipers (G. W. Schuett, M. Hoggren, M. E. Douglas, & H. W. Greene, eds.) Eagle Mountain Publ. Eagle Mountain, UT. 580 pp. Shine, R. L., E. Z. Sun & X. Bonnet. 2002. A review of 30 years of ecological research on the shedao pitviper, Gloydius shedaoensis. Herpetol. Nat. Hist., 9(1): 1-14. Stiles, K., P. Stark, D. Chiszar & H. M. Smith. 2002. Strike-induced chemosensory searching (SICS) and trail-following behavior in copperheads (. Agkistrodon contortrix ). Pp. 413-418, in Biology of the vipers (G. W. Schuett, M. Hoggren, M. E. Douglas, & H. W. Greene, eds.) Eagle Mountain Publ. Eagle Mountain, UT. 580 pp. Weatherhead, P. J. & F. W. Anderka. 1984. An improved radio transmitter and implantation technique for snakes. J. Herpetol., 18(3):264-269. DCR at: crudolph01@fs.fed.us TEXAS J. SCI. 56(4):405-414 NOVEMBER, 2004 NESTING HABITAT OF EASTERN WILD TURKEYS {MELEAGRIS GALLOPAVO SYLVESTRIS) IN EAST TEXAS Bobby G. Eichler* and R. Montague Whiting, Jr. Arthur Temple College of Forestry Stephen F. Austin State University Nacogdoches , Texas 75962 * Current address: Texas Parks and Wildlife Department Mount Pleasant, Texas 75455 Abstract. — Eastern wild turkeys ( Meleagris gallopavo sylvestris) captured in Iowa and Georgia were relocated to the Pineywoods of east Texas where they were radio-marked and released. During the 1995 and 1996 nesting seasons, nest sites of radio-marked hens were located and characteristics of the habitat surrounding the sites and of randomly selected sites in the same vegetation type were evaluated using paired r-tests. Of 24 nest located, 6 were successful. Most nests were in mature pine-hardwood stands or pine regeneration areas. Nest sites had higher densities of living and dead grasses and higher screening cover values than did random sites {P < 0.05). Other habitat characteristics did not differ between nest and random sites ( P > 0.05). These results suggest that herbaceous ground cover is the most important habitat variable which hens use when selecting nest sites. Habitat character¬ istics surrounding nests located in this study were similar to those documented in other studies in the southeast. Although nesting habitat probably is adequate in east Texas, land managers could increase such habitat by mowing utility rights-of-way on a two to three-year schedule, implementing a three to five-year prescribed burning regime, thinning pine stands at or before canopy closure, retaining slash after logging operation, and delaying site preparation in regeneration areas until after the nesting season. In Texas, the eastern wild turkey {Meleagris gallopavo sylvestris) originally ranged over approximately 12, 145,000 hectares in 40 counties in the Pineywoods Ecological Region (Newman 1945). The birds occupied river bottom and upland forest communities. During the 1800’s, Texas settlers thought eastern wild turkey populations to be inexhaustible (Carpenter 1959). However, commercial hunting and extensive land clearing led to declining turkey numbers throughout the early 1900’s (Carpenter 1959). In 1941, the Texas Legislature closed the turkey season throughout the Pineywoods, but the action came much too late; by 1942, less than 100 native eastern wild turkeys remained in Texas (Newman 1945). Records indicate that wild turkey restoration efforts by the Texas Game, Fish and Oyster Commission began in east Texas as early as 1924 (Newman 1945; Carpenter 1959). Many unsuccessful restocking attempts were made during the next 40 to 50 years; most failed attempts 406 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 used pen-reared eastern wild turkeys or the Rio Grande subspecies (M. gallop avo intermedia) . However, between 1979 and 1981, wild-trapped eastern wild turkeys were released on two east Texas sites, and populations flourished (Swank et al. 1985). In 1987, the Texas Parks and Wildlife Department, in cooperation with the National Wild Turkey Federation, initiated a large-scale eastern wild turkey restoration program in east Texas. The program used wild- trapped eastern turkeys acquired from southeastern and mid western states. Some restockings were successful, but others failed and populations remained low in many areas (I. D. Burk, per. comm.). Wild turkey populations are sustained by annual brood productivity (Seiss et al. 1990). Thus, nesting habitat is critical to the well-being of the species (Badyaev 1995). In order to increase wild turkey productivi¬ ty in east Texas, suitable nesting habitat needs to be identified. The objectives of this study were to describe vegetative characteristics surrounding nest sites of wild turkey hens and to compare these charac¬ teristics to vegetative characteristics surrounding random sites. Methods Four study areas in Tyler County, Texas were stocked with wild- trapped eastern wild turkeys relocated from Iowa and Georgia during January and February of 1994. Twelve hens and three gobblers were released at each site; equal numbers of hens and gobblers were from Iowa and Georgia. Prior to release the turkeys were aged, banded and fitted with back-pack style radio transmitters. An attempt was made to radio-locate the birds daily for the first two weeks after release. If mortality occurred during that period, the bird was replaced. There¬ after, the birds were radio-located at least once a week, and up to three times a week. During February 1995, eight wild turkey hens were captured on a study area in Trinity County; the birds were aged, banded, fitted with transmitters and released at the point of capture. In January of 1996, an additional 15 wild- trapped hens from Iowa were fitted with transmitters and released on that study area. Beginning on 1 April of 1995 and 1996 hens were radio-located three to five times per week. When a hen exhibited very localized daily movements, it was assumed she had initiated a nest. Once a hen was radio-located three times in the same place, it was assumed incubation had begun, and she was radio-located daily. After approximately 10 days of incubation, the nest location was estimated using triangulation, azimuths, and estimated observer-to-nest distances. After the hen had EICHLER & WHITING 407 left the nest area for at least one day, an attempt was made to locate and determine the fate of the nest; Nests were classified as successful ( > one egg hatched) or unsuccessful (depredated or abandoned). Macro and micro-habitat characteristics were evaluated at each nest location. The macro-habitat variables were forest type and tree size class. Forest type of the stand surrounding each nest was classified as either pine, pine-hardwood, riparian or opening; openings included food plots, rights-of-way, pastures and seedling ( < 1.4 m tall) pine planta¬ tions. Tree size classes (trees > 1.4 m tall) were based on diameter at breast height (DBH) of dominant trees in the area surrounding the nest site. Size classes used were sapling (< 12.7 cm DBH), pole (12.7 to 27.9 cm DBH) and sawtimber (> 27.9 cm DBH) (Stoddard & Stoddard 1987). Chi-square tests were used to determine if nesting hens selected macro-habitats according to availability. Habitat composition data from a study by George (1997) were used with Chi-square tests for Tyler County nests. In that study, macro-habitats were classified as pure pine forests, pine-hardwood forests, riparian forests or openings. Habitat composition data for the Trinity County study area were gathered from the Temple-Inland Forest Products Corporation five-year plan for the area; macro-habitats were categorized the same as the George (1997) study. Micro-habitat data were collected in the area immediately surrounding the nest site. Micro-habitat characteristics measured included basal area of pine, basal area of hardwood, total basal area, distance to nearest man-made edge, distance to nearest natural edge, percent canopy closures, relative screening cover of the understory and relative densities of the ground cover. Basal areas were measured from the center of the nest using a 10- factor prism. Distances to nearest man-made and natural edges were measured using a 23 -m logger’s tape. Canopy closures of the under¬ story, midstory and overstory were evaluated using a modified point- quadrat technique (Smeins & Slack 1982). Understory was vegetation < 2 m tall, midstory 2 to 15 m tall, and overstory vegetation > 15 m tall. With the nest as the plot center, a 10-m transect was established in each cardinal direction. Along each transect, five subpoints were spaced at 2-m intervals; the first subpoint on each transect was 2 m from the nest site. Canopy closure data were gathered at each subpoint using a sighting tube (Whiting & Fleet 1987; Parsons 1994). At each subpoint, 408 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 an observer looked straight up or down through the sighting tube, and for each height class, if vegetation obstructed the crosshairs, "yes" was recorded. From this procedure, a percent closure score could be calcu¬ lated for each canopy layer. Under story vegetation cover was evaluated in five strata between ground level and 1.7 m above ground level. Each of the lower four strata were 30 cm wide; the top stratum was 50 cm wide. Screening cover data were gathered using a vegetation profile board (VPB). The VPB was 1.7 m tall, 8.9 cm wide and divided into five alternating red and white colored sections which corresponded to the five strata evalu¬ ated. The board was placed at the nest center and the percent of each section obscured was estimated from a distance of 15 m and a height of approximately 46 cm (Nudds 1977). Scores were based on a scale of one to five and reflected the percentage of the board which was ob¬ scured by vegetation. Scores of one, two, three, four and five, indicated 0 to 20%, 21 to 40%, 41 to 60%, 61 to 80% and 81 to 100% obscurity, respectively. Screening cover scores were estimated from each cardinal direction by stratum. These values were then averaged to provide percent screening cover for each stratum. Relative density of ground cover (i.e., living or dead vegetation) was evaluated using a point quadrat technique. A 10-pin frame was used to sample ground cover within 60 cm of the ground (Parsons 1994). The pin frame measured 80 cm high and 1 10 cm long; pins were centered at 10-cm intervals along the frame. Data were gathered at five subpoints around each nest site. For the first subpoint, the pin frame was centered on the actual nest bowl. The remaining four subpoints were 15 m from the nest in the cardinal directions. At each subpoint, the pins were lowered from approximately 60 cm and each pin-to-plant hit was record¬ ed by plant category (i.e., living woody, herbaceous, grass or dead grass). These data were used to derive an index of relative density for each plant category. This index was simply the number of hits by category per 10 pins (i.e., an index of 31 for living grasses would indicate that the 10 pins made 31 contacts with living grasses). The last hit recorded for each pin was either litter or soil. As each pin could have only one contact with either litter or bare soil, numbers of hits were converted to percentages. Average height of ground cover, as bracketed by the pin frame, also was measured at each subpoint. Immediately after micro-habitat measurements of a nest location were completed, micro-habitat data were collected from a random location in the same macro-habitat type (forest type and stand class). Standing at EICHLER & WHITING 409 the nest, the observer glanced at the second hand of his wristwatch and used the direction it was pointing as a random direction. Using a compass, the observer then paced a distance which had been previously taken from a random numbers table; minimum and maximum acceptable distances were 100 m and 250 m, respectively. Data gathered at random locations were the same as those gathered at nest locations. Differences in micro-habitat variables between nest sites and random sites were evaluated using paired Mests; all tests were performed at a 0.05 alpha level. Results At the beginning of the 1995 and 1996 nesting seasons, there were 37 and 44 hens, respectively, with active transmitters. Although eight hens died or were lost between the 1995 and 1996 nesting seasons, 15 additional Iowa hens were released on the Trinity County site, thus increasing the sample size by seven. During the two nesting seasons, 24 nests were located, 11 in 1995 and 13 in 1996; six nests were successful and 18 were unsuccessful. Twelve nests were in Trinity County, and 12 nests were in Tyler County; the six successful nests were in Tyler County. The majority of nests were in pine-hardwood habitat types (11) and openings (8) (Eichler 1999:20). In both counties, there were differences between habitat availability and habitats selected for nesting (Table 1). In Tyler County, openings made up only 21.0% of the study area, yet six hens (50.0%) selected this habitat type in which to nest. Converse¬ ly, only one nest was in a pine-hardwood stand and this habitat type made up 27.2% of the study area (x2 = 23.00, 3 df, P = 0.001) (Table 1). In Trinity County, ten of 12 (83.3%) nests were in pine-hardwood stands which comprised 55.0% of the study area. Although riparian forests comprised 36.0% of the area, no hens nested in that habitat (x2 = 52.23, 3 df,P = 0.001). Thirteen of the 24 nests were in sawtimber stands; 11 were in pine-hardwood forests and two were in riparian forests. Eight nests were in openings; of these, four were in pine seedling stands < two years old, two in an abandoned field, one in a grazed field, and one in the thick vegetation (i.e., rough) bordering a food plot. The remaining three nests were in sapling and pine pole stands. Four successful nests were in openings (all were pine seedling stands < two years old), one in a pine sapling stand, and one in a pine pole stand. 410 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Table 1 . Habitat availability and use of habitat types and stand classes by nesting eastern wild turkey hens in east Texas, spring 1995 and 1996. Habitat composition differed from utilization rates at Tyler County sites ( X 2 = 23.00, 3 df, PC0.001) and Boggy Slough sites (X2 = 52.23, 3 df, P< 0.001). Tyler County _ Boggy Slough Habitat Habitat composition Nests composition Nests All nests Habitat <%) (No.) (%> (%) (No.) (%) (No.) (%) Habitat type Pure pine 30.7 3 25.0 4.0 0 00.0 3 12.6 Pine-hardwood 27.2 1 8.3 55.0 10 83.3 11 45.8 Riparian 21.1 2 16.7 36.0 0 00.0 2 8.3 Opening 21.0 6 50.0 5.0 2 16.7 8 33.3 Total 100.0 12 100.0 100.0 12 100.0 24 100.0 Stand class Opening 6 50.0 2 16.7 8 33.3 Sapling 2 16.7 0 00.0 2 8.3 Pole 1 8.3 0 00.0 1 4.2 Sawtimber 3 25.0 10 83.3 13 54.2 Total 12 100.0 12 100.0 24 100.0 Some micro-habitat variables differed between nest sites and random sites (Table 2). In all strata, screening cover values at nest sites were higher than those at random sites; the differences were significant in the 0.31 to 0.60 m and the 1.21 to 1.70 m strata and approached signifi¬ cance in the 0.00 to 0.30 m stratum (Table 2) . The largest difference was in the 0.31 to 0.60 m stratum where screening cover averaged about 14% higher at nest than at random sites. Ground cover densities were greater at nest than at random sites for all except the herbaceous category. Ground cover densities for both living grass and dead grass were significantly higher at nest sites than at random locations (Table 2). Although not statistically significant, canopy closures in the midstory were more open at nest sites than random locations (Table 2). Discussion In this study, hens selected pine habitat types and openings in which to nests. Previous studies in the Southeast have shown similar results (Campo et al. 1989; Seiss et al. 1990; Sisson et al. 1990; Still & Baumann 1990). In a previous east Texas study, 89% of the nests were in upland pine forest types; however, as opposed to this study, those nests were equally distributed among size classes of timber (Campo et al. 1989). In South Carolina, Still & Baumann (1990) found 21 of 37 nests in pine habitats and in Georgia, Sisson et al. (1990) found 83% of all nests in pine stands. However, in Mississippi, mature pine stands EICHLER & WHITING 411 Table 2. Results of paired r-tests (24 df) comparing micro-habitat characteristics of eastern wild turkey nest sites to random sites ( n = 24) in east Texas, 1995-1996. Habitat component Nest sites Random sites t P Distance from edge (m) Natural 39.3 42.7 0.298 0.769 Manmade 40.0 35.7 0.449 0.658 Basal area (m2/ha)a Pine 11.2 12.4 -0.834 0.413 Hardwood 5.7 5.20 -0.290 0.775 Total 17.0 17.6 -1.000 0.327 Canopy coverage (%)a Understory ( < 2 m) 57.8 56.7 -0.249 0.805 Midstory (2-15 m) 53.9 60.8 -2.029 0.054 Overstory (> 15 m) 42.2 44.7 -1.334 0.195 Screening cover (%) 0.00 - 0.30 m 93.3 86.7 2.205 0.055 0.31 - 0.60 m 79.4 65.4 2.893 0.008 0.61 - 0.90 m 64.4 53.1 1.450 0.161 0.91 - 1.20 m 58.1 49.6 1.228 0.232 1.21 - 1.70 m 46.9 33.8 2.328 0.029 Ground cover density (hits / 10 pins) Living grass 31.7 25.3 2.559 0.018 Dead grass 10.4 3.1 2.119 0.045 Herbaceous 6.5 6.7 -0.073 0.942 Woody species 13.2 10.9 1.239 0.228 Litter (%) 88.4 89.0 -0.187 0.854 Height of ground cover (cm) 25.8 23.5 0.965 0.354 a Only tested in sapling, pole and sawtimber stands. contained the most nests (18 of 38) but were used according to availa¬ bility (Seiss et al. 1990). In that study, other habitats in which hens nested included bottomland hardwoods and pine and hardwood regenera¬ tion areas. Use of early successional habitats for nesting is also similar to findings of other studies (Everett et al. 1981; Campo et al. 1989; Seiss et al. 1990; Still & Baumann 1990). Seiss et al. (1990) found 36.8% of all nests in regeneration areas whereas that type made up only 12.5% of available habitat. In South Carolina, Still & Baumann (1990) found 10 of 37 (27%) nests in seed-tree cuts or clearcuts < 10 years old, and in a previous east Texas study, 26% of the hens nested in pine regeneration stands one to seven years old (Campo et al. 1989); in this study 25.0% of hens nested in regeneration areas. Everett et al. (1981) found that rights-of-way with roughs one to three years old were preferred nesting habitat in Alabama. 412 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Prescribed burning may be an important factor for nest site selection. In Trinity County, 10 of the 12 nests were in pine uplands which were burned on a three to five-year regime. These results are similar to those of Sisson et al. (1990) who found that most nests (74%) were in upland pine stands on a one to three- year burn rotation; in that study, such stands comprised only 7.2% of the area. Conversely, in Alabama, Exum et al. (1987) found 89% of nesting hens used areas left unburned for three or more years. In 1996, three of seven nests in Trinity County were in pine sawtimber stands which had been thinned < two months prior to nesting season. All three nests were concealed by logging slash. Previous studies have shown that hens prefer nesting in thinned stands (Hillestad 1973; Lutz & Crawford 1987; Campo et al. 1989) and use logging slash as concealment (Martin 1984; Lutz & Crawford 1987; Swanson 1993). Hillestad (1973) found that four of seven hens selected recently cut-over loblolly pine, shortleaf pine, or sweetgum stands in which to nest. In Oregon, Lutz & Crawford (1987) found that nesting hens used thinned conifer stands more frequently than expected (P < 0.05) and that nests were commonly adjacent to slash. In this study, nests sites had abundant screening cover in the 0.00 - 0.30 and 0.31 - 0.60 m strata with values ranging from 80-95%. Nests sites in pole and sawtimber stands had lower basal area and canopy closure values than did random sites (Eichler 1999), and higher densities of living grasses and woody seedlings. These results parallel those of other studies which have shown that nest sites normally have lower densities of overstory trees, basal areas and canopy closures, and higher screening concealment than do random locations (Lazarus & Porter 1985; Holbrook et al. 1987; Lutz & Crawford 1987; Campo et al. 1989, Still & Baumann 1990, Swanson 1993; Lopez 1996). In the Post Oak Savannah Region of east Texas, nest sites occurred in areas with relatively high coverage of forbs in the understory and ground layers (Lopez 1996). Still & Baumann (1990) found that nesting hens pre¬ ferred low to moderately stocked stands, suggesting that ground cover was important. Holbrook et al. (1987) found that cover below the 2-m level was more dense around nests than at random locations. Characteristics of habitats used by nesting hens in this study were very similar to those in other studies in the Southeast. These results suggest that nesting habitat is adequate in east Texas (Eichler 1999). EICHLER & WHITING 413 However, there are several practices which land managers could use to increase nest success. This study indicated that herbaceous ground cover is the most important habitat variable hens use when selecting nest locations. In forested stands, a three to five-year burning regime would seem to be appropriate to stimulate and maintain herbaceous densities for nesting throughout the Piney woods of east Texas. Additionally, thinning pole and sawtimber stands would allow for this same type of ground cover. After logging operations, slash and tree tops should be left as is to provide cover at least until after the nesting season. Likewise, in newly created regeneration areas, site preparation practices should be delayed until after the nesting season when possible. Lastly, utility rights-of-way should be mowed on a two to three-year schedule; a mosaic of two to three-year roughs would allow for nesting habitat and discourage the growth of brush thickets. Acknowledgments We thank Jason Hoffman, Phillip LeWallen and Stacy Roland for field assistance. Jim George and Jacky Chen aided with data analysis. Thanks to John Burk (Texas Parks and Wildlife Department, Eastern Wild Turkey Program Specialist) for his knowledge of east Texas restocking efforts. International Paper Company and Temple-Inland Forest Products Corporation allowed us access to their lands. This project was funded by the National Wild Turkey Federation, Texas Parks and Wildlife Department, and the Arthur Temple College of Forestry at Stephen F. Austin State University. Literature Cited Badyaev, A. V. 1995. Nesting habitat and nesting success of eastern wild turkeys in the Arkansas Ozark Highlands. Condor, 97( 1 ) :22 1 -232. Campo, J. J., C. R. Hopkins & W. G. Swank. 1989. Nest habitat use by eastern wild turkeys in eastern Texas. Proc. Annu. Conf. Southeast. Assoc. Fish and Wildl. Agencies, 43:350-354. Carpenter, C. 1959. Return of an exile. Texas Game and Fish, 1 7( 1) :24-25 . Eichler, B. G. 1999. Nesting and brook rearing habitat of eastern wild turkeys in east Texas. Unpublished M.S. Thesis, Stephen F. Austin State University, Nacogdoches, Texas, 48 pp. Everett, D. D., D. W. Speake & W. K. Maddox. 1981. Use of rights-of-way by nesting wild turkeys in north Alabama. Environmental Concerns in Rights-of-Way Management, 2:64-1-6. Exum, J. H., J. A. McGlincy, D. W. Speake, J. L. Buckner & F. M. Stanley. 1987. Ecology of the eastern wild turkey in an intensively managed pine forest in southern Alabama. Tall Timbers Res. Sta. Bull. 23, Tallahassee, Florida, 70 pp. George, J. R. 1997. Comparison of eastern wild turkey broodstock relocated into the Pineywoods Ecological Region of Texas. Unpublished M.S. Thesis, Stephen F. Austin 414 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 State University, Nacogdoches, Texas, 90 pp. Hillestad, H. O. 1973. Movements, behavior, and nesting ecology of the wild turkey in eastern Alabama. Pages 109-123, in G. C. Sanderson and H. C. Schultz, editors. Wild turkey management: current problems and programs. University of Missouri Press, Columbia, 355 pp. Holbrook, H. T., M. R. Vaughan & P. T. Bromley. 1987. Wild turkey habitat preference and recruitment in intensively managed Piedmont forests. J. Wildl. Manage., 5 1 ( 1 ) : 1 82- 1 87. Lazarus, J. E. & W. F. Porter. 1985. Nest habitat selection by wild turkeys in Minnesota. Proc. Natl. Wild Turkey Symp., 5:67-81. Lopez, R. R. 1996. Population dynamics of eastern wild turkeys relocated into the Post Oak Savannah of Texas. Unpublished M.S. Thesis, Texas A&M University, College Station, Texas, 59 pp. Lutz, R. S. & J. A. Crawford. 1987. Reproductive success and nesting habitat of Merriam’s wild turkeys in Oregon. J. Wildl. Manage., 51(4):783-787. Martin, D. J. 1984. The influence of selected timber management practices on habitat use by wild turkeys in east Texas. Unpublished M.S. Thesis, Texas A&M University, College Station, Texas, 129 pp. Newman, C. C. 1945. Turkey restocking efforts in east Texas. J. Wildl. Manage. 9(4): 279-289. Nudds, T. D. 1977. Quantifying the vegetative structure of wildlife cover. Wildl. Soc. Bull. 5(3): 113-1 17. Parsons, D. S. 1994. Nesting, brood habitat, and utilization of planted food plots by northern bobwhite in east Texas. Unpublished M.S. Thesis, Stephen F. Austin State University, Nacogdoches, Texas, 115 pp. Seiss, R. S., P. S. Phalen & G. A. Hurst. 1990. Wild turkey nesting habitat and success rates. Proc. Natl. Wild Turkey Symp., 6:18-24. Sisson, D. C., D. W. Speake, J. L. Landers & J. L. Buchner. 1990. Effects of prescribed burning on wild turkey habitat preference and nest site selection in south Georgia. Proc. Natl. Wild Turkey Symp., 6:44-50. Smeins, F. E. & R. D. Slack. 1982. Fundamentals of ecology laboratory manual. Second edition. Kendall/Hunt Publ., Dubuque, Iowa, 140 pp. Still, H. R., Jr. & D. P. Baumann, Jr. 1990. Wild turkey nesting ecology on the Frances Marion National Forest. Proc. Natl. Wild Turkey Symp., 6:13-17. Stoddard, C. H. & G. M. Stoddard. 1987. Essentials of forestry practice. Fourth edition. John Wiley & Sons, New York, 432 pp. Swank, W. G., D. J. Martin, J. J. Campo & C. R. Hopkins. 1985. Mortality and survival of wild-trapped eastern wild turkeys in Texas. Proc. Natl. Wild Turkey Symp., 5:113-120. Swanson, D. A. 1993. Population dynamics of the wild turkey in West Virginia. Unpublished Ph. D. Dissertation, West Virginia University, Morgantown, 168 pp. Whiting, R. M. , Jr. & R. R. Fleet. 1987. Bird and small mammal communities of loblolly-shortleaf pine stands in east Texas. Pages 49-66, in H. A. Pearson, F. E. Smeins, and R. E. Thill, compilers. Ecological, physical, and socioeconomic relationships within southern national forest. United States Forest Service General Technical Report SO-68, 293 pp. BGE at: bobby. eichler@tpwd. state. tx. us TEXAS J. SCI. 56(4):4 15-426 NOVEMBER, 2004 THE RED-COCKADED WOODPECKER: INTERACTIONS WITH FIRE, SNAGS, FUNGI, RAT SNAKES AND PILEATED WOODPECKERS Richard N. Conner, Daniel Saenz and D. Craig Rudolph Wildlife Habitat and Silviculture Laboratory Southern Research Station, USD A Forest Service, 506 Hay ter St. Nacogdoches, Texas 75965-3556 Abstract. — Red-cockaded woodpecker (Pico ides borealis ) adaptation to fire-maintained southern pine ecosystems has involved several important interactions: (1) the reduction of hardwood frequency in the pine ecosystem because of frequent fires, (2) the softening of pine heartwood by red heart fungus ( Phellinus pirn) that hastens cavity excavation by the species, (3) the woodpecker’s use of the pine’s resin system to create q barrier against rat snakes (Elaphe sp.), and (4) the woodpecker as a keystone cavity excavator for secondary-cavity users. Historically, frequent, low-intensity ground fires in southern pine uplands reduced the availability of dead trees (snags) that are typically used by other woodpecker species for cavity excavation. Behavioral adaptation has permitted red-cockaded woodpeckers to use living pines for their cavity trees and thus exploit the frequently burned pine uplands. Further, it is proposed that recent observations of pileated woodpecker (Dryocopus pileatus ) destruction of red-cockaded woodpecker cavities may be related to the exclusion of fire, which has increased the number of snags and pileated woodpeckers. Red-cockaded wood¬ peckers mostly depend on red heart fungus to soften the heartwood of their cavity trees, allowing cavity excavation to proceed more quickly. Red-cockaded woodpeckers use the cavity tree’s resin system to create a barrier that serves as a deterrent against rat snake predation by excavating small wounds, termed resin wells, above and below cavity entrances. It is suggested that red-cockaded woodpeckers are a keystone species in fire-maintained southern pine ecosystems because, historically, they were the only species that regularly could excavate cavities in living pines within these ecosystems. Many of the more than 30 vertebrate and invertebrate species known to use red-cockaded woodpecker cavities are highly dependent on this woodpecker in fire-maintained upland pine forests. The red-cockaded woodpecker ( Picoides borealis ) evolved in a landscape where frequent, low-intensity fires burned within upland southern pine ecosystems. The fires reduced the numbers of hardwoods, and it is suggested that they also reduced the numbers of dead trees (snags) relative to their abundances in hardwood stands along riparian areas and bottomlands (Conner et al. 2001a). Hardwood snags, which serve as typical cavity trees for many woodpecker species in this scenario, were probably scarce. It was in this landscape that the red-cockaded woodpecker adapted to excavating cavities in live pine trees. The extended length of time required to excavate cavities in live pines 416 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 and the subsequent rarity of completed cavities in this ecosystem appear to be closely linked to the evolution of cooperative breeding in the red- cockaded woodpecker (Walters et al. 1988; 1992; Conner & Rudolph 1995). Cavities for nesting and roosting in living pines require a long time to excavate (Conner & Rudolph 1995; Harding & Walters 2002) and are so rare across the pine forest landscape that it is to the advan¬ tage of young woodpeckers, particularly young males, to forego dis¬ persal and defer breeding until a breeding slot opens up in their natal cluster of cavity trees or a nearby cavity-tree cluster (Walters et al. 1992). These young woodpeckers from previous nesting efforts remain with the breeding pair and assist in subsequent nesting efforts by incu¬ bating eggs, feeding and brooding young, excavating cavities, and helping to defend the group’s territory (Ligon 1970; Walters et al. 1988; Conner et al. 2001a). In this paper a scenario is suggested by which historically frequent, low-intensity ground fires in southern pine uplands reduced the availa¬ bility of dead trees (snags) that are typically used by woodpeckers for cavity excavation. Standing dead trees were more abundant in the more mesic hardwood sites where other species of woodpeckers are abundant. Behavioral adaptations permitted red-cockaded woodpeckers to excavate cavities into living pines for nesting and roosting. Thus, red-cockaded woodpeckers exploited the frequently burned pine uplands (Conner et al . 2001a), where the rarity of more typical cavity-excavation sites in dead branches and dead trees historically excluded or decreased the abundance of other woodpecker species in the southeastern United States because they typically do not make cavities in live pines (Conner et al. 1975; Kilham 1983). Discussion is also presented on how the woodpecker’s adaptation to pine ecosystems has benefited other species by creating cavities in a relatively cavity-barren landscape. The Interaction of Fire with Upland Pine Landscapes Fossil pollen records indicate that fire-maintained pine ecosystems began to spread from peninsular Florida approximately 12,000 years ago and arrived at the western extreme of their distribution in Texas about 4,000 years ago (Webb 1987). This expansion was permitted by the retreat of the Laurentide ice sheet of the Wisconsin glaciation to the north (Conner et al. 2001a). Bartram (1791) described the original longleaf pine ( Pinus palustris ) forests as nearly unbroken expanses of widely spaced pines within a sea of grass. Fire, which burned in both CONNER ET AL. 417 the winter and growing season, was an integral part of the spread of pine ecosystems (Bonnicksen 2000; Conner et al. 2001a). Historically, frequent fires were ignited primarily during dry periods by lightning, Native Americans, and early settlers (Catesby 1731; Michaux 1802). The frequent fires burned day and night and meandered across the land¬ scape until they encountered sites too isolated or too wet to burn (Frost 1993; Glitzenstein et al. 1995). The fires killed invading hardwoods in the upland pine ecosystem and maintained the herbaceous ground cover that consisted primarily of grasses and forbs (Jackson et al. 1986; Glitzenstein et al. 1995). Throughout the South, fallen pine needles and dried grasses served as fuel for the ground fires, which burned every one to three-plus years (Landers 1991; Glitzenstein et al. 1995; Bonnicksen 2000). Michaux’ s (1802) observations indicate that longleaf pine forests which occupied seven- tenths of the landscape in the Carolinas were burned annually. Because hardwoods were rare in well-burned pine uplands (Chapman 1909; Platt et al. 1988; Frost 1993), live pines and pine snags were the primary sources of potential nest sites for woodpeckers. Although low- intensity ground fires may burn existing snags created by lightning and bark beetle ( Dendroctonus sp., Ips sp.) infestation, they typically do not generate sufficient heat to kill pines, which would create new snags (Conner 1981; Conner et al. 2001a). Therefore, it is suggested that even pine snags may have been scarce in southern pine ecosystems. Interaction of Red-cockaded Woodpeckers with Fungi The use of living pines as sites to excavate cavities for nesting and roosting resulted in an increase in the length of time required for the woodpeckers to make a cavity. Most woodpecker species in eastern North America can excavate a new cavity in a dead, decayed snag in two to four weeks (Conner et al. 1975; 1976; Kilham 1983). Pileated woodpeckers (. Dryocopus pileatus ) can excavate a cavity in 23 days in the eastern United States, but excavation time can take three to six weeks in the Pacific Northwest (Bull & Jackson 1995). Downy wood¬ peckers ( Picoides pubescens) can excavate a complete cavity in two weeks, whereas hairy woodpeckers ( Picoides villosus ) can take up to four weeks (Kilham 1983). Red-bellied woodpeckers ( Melanerpes carolinus) typically can excavate a completed cavity within two weeks (Shackelford et al. 2000) and red-headed woodpeckers ( Melanerpes erythrocephalus) within three weeks (Jackson 1976). Cavity excavation 418 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 by northern flickers (Colap tes auratus) can take up to four weeks (Burns 1900). Lawrence (1967) observed that average cavity excavation time for northern flickers was 12.1 days, hairy woodpeckers 19.7 days, downy woodpeckers 16.0 days, and yellow-bellied sapsuckers (Sphyrapicus varius) 19.7 days. Because red-cockaded woodpeckers use living pines for cavity trees, where the heartwood is often not decayed (Conner & Locke 1982), cavity excavation may require numerous years (Conner & Rudolph 1995). Unlike snags, which often have decayed sapwood and heart- wood, the sapwood of live pines is not decayed (Conner & Locke 1982), and red-cockaded woodpeckers have to excavate through 8 to 16 cm of solid wood (Conner et al. 1994). Increasing sapwood thickness and the presence of flowing pine resin that seeps from the wound caused by cavity excavation further complicates the process and slows the rate of excavation (Conner et al. 1994; Conner & Rudolph 1995; Conner et al. 2001a). If resin flow is abundant, the woodpeckers typically must wait for the resin to crystallize before recommencing excavation, again, increasing the time required for cavity excavation (Conner & Rudolph 1995). Cavity excavation rates in red-cockaded woodpeckers may be influenced by the availability of suitable cavities (Harding & Walters 2002). As the need for cavities increases within a group of wood¬ peckers, the birds may accelerate their excavation activities (Conner et al. 2002). Although red-cockaded woodpeckers can excavate a completed cavity into a pine with undecayed heartwood and sapwood (Conner & Locke 1982), the presence of red heart fungal ( Phellinus pini ) decay in the heartwood has an influence on the time required to excavate a complete cavity (Conner & Rudolph 1995). Red-cockaded woodpeckers are able to detect the presence of the fungus within the boles of the pines and actively select pines with red heart fungal decay for cavity trees (Conner & Locke 1982). Red heart fungus enters the heartwood of pines via broken branch stubs (Conner & Locke 1982; Conner et al. 2004). After gaining access to the heartwood of a pine, at least 15 to 20 years of growth and decay within the heartwood are required before the fungus produces a sporophore (conk) on the bole of the pine (Conner et al. 2004). This same 15- to 20-year time period is required for the fungus to decay a minimally sufficient diameter of heartwood (12 cm; Conner et al. 2004) for a woodpecker cavity. Although the age of the pine appears to be the primary factor associated with increasing frequency of heartwood decay (Conner et al. 1994), tree spacing and growth rate also CONNER ET AL. 419 have an influence (Conner et al . 2004) . Older pines tend to have higher frequencies of heartwood decay and pines growing slowly in diameter prune lower branches more slowly and appear to have higher frequency of heartwood decay (Conner et al. 2004). Increased time during the natural limb pruning process allows more time for spores to infect wood tissue. As red heart fungus decays the heartwood it softens the wood, and decayed heartwood is more easily excavated than sound heartwood. The presence of decayed heartwood can decrease the time required for cavity excavation by 1.3 years (Conner et al. 1994). Even with heartwood decay present in many cavity trees, an average of 1.8 years in loblolly ( Pinus taeda) (n = 9 excavations), 2.4 years in shortleaf pines ( P . echinata) (n = 12 excavations), and 6.3 years in longleaf pines (n = 12 excavations) is required to fully excavate a cavity (Conner & Rudolph 1995). Many red-cockaded woodpecker cavity trees are lost annually to bark beetles, lightning, wind action, and enlargement by pileated woodpeckers (Conner et al. 1991). Thus, the availability of pines infected with red heart fungus may determine whether red-cockaded woodpeckers have a sufficient number of useable cavity trees available for nesting and roosting in a given year. Interaction of Red-cockaded Woodpeckers with Resin and Rat Snakes Adaptation to contending with resin that flows from living pines when cavities are excavated has affected the interaction between red-cockaded woodpeckers and rat snakes ( Elaphe sp.) and enhanced the survival of the woodpecker. Southern pines produce and maintain pine resin (gum) within an elaborate system of canals and ducts that extends from the pine’s needles down into its roots. Resin is a mixture of primarily light resin oils (monoterpenes), which serve as solvents, and the heavier resin acids (diterpenes) , which give the resin its viscous and sticky nature (Hodges et al. 1977). The resin system in pines has evolved as their primary defense against bark beetles (Hodges et al. 1979). When bark beetles attack, the pine flushes the wound with resin and if sufficient resin is present, the attacking beetles are “pitched out.” A similar response occurs when red-cockaded woodpeckers initiate cavity excavation. If resin flow is very high, it will temporarily interfere with cavity excavation as noted previously. 420 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Red-cockaded woodpeckers nesting and roosting in living pines are extremely vulnerable to predation by rat snakes (Neal et al. 1993). Predictable, long-term use of individual cavities allows the local snake population to learn the location of cavities (Neal et al. 1993), and living pines with intact bark are easily climbed by rat snakes (Rudolph et al. 1990b). However, red-cockaded woodpeckers derive substantial protec¬ tion from rat snakes by taking advantage of resin produced by pines to establish a resin barrier that prevents access to cavities by rat snakes. As cavities approach completion, red-cockaded woodpeckers excavate a series of small (1-2 cm) wounds into the cambium on the pine’s bole around and above and below their cavity entrance. These wounds, termed resin wells, are pecked daily by the woodpeckers and the repeat¬ ed pecking causes continuous wounding of the xylem-cambial boundary, keeping a stream of clear, fresh pine resin flowing from the wells and down the pine’s bole. Multiple resin wells on a healthy cavity tree create a substantial barrier of sticky fresh resin that serves as a deterrent to climbing rat snakes (Ligon 1970; Jackson 1974; Rudolph et al. 1990b). However, repeated wounding of cavity trees over several years can decrease the ability of the pines to produce resin (Conner et al. 2001b) and pines with inadequate resin flow are abandoned by the woodpeckers (Conner & Rudolph 1995). Red-cockaded woodpeckers must continue to excavate new cavities to replace cavities with inade¬ quate resin barriers and cavity trees lost to mortality factors or cavity enlargement by other woodpeckers. Red-cockaded woodpeckers can detect how much resin a pine can produce (Conner et al. 1998). The socially dominant breeding male red-cockaded woodpecker selects the cavity tree that produces the most resin for his roost cavity. It is the breeding male’s roost tree that usually becomes the breeding pair’s nest tree. By selecting the cavity tree with the highest resin yield, the nesting effort of the breeding pair seems to receive the highest protection possible from rat snake predation (Conner et al. 1998). Red-cockaded Woodpeckers as a Keystone Cavity Excavator In the historic fire-maintained upland pine ecosystems of the South where pines existed nearly as a tree monoculture (Chapman 1909; Platt et al. 1988; Frost 1993), red-cockaded woodpeckers were the only woodpeckers able to excavate complete cavities in living pines regularly (Ligon 1970; Conner et al. 2001a). Reports of other North American species of woodpecker excavating cavities in live portions of living pines CONNER ET AL. 421 in the eastern United States are extremely rare or nonexistent (Bent 1939; Reller 1972; Conner et al. 1975; Jackson 1976; Kilham 1983). Red-cockaded woodpeckers historically were and continue to be a keystone species because they are the primary woodpecker species to provide cavities for more than 30 other wildlife species within fire- maintained pine ecosystems of the South (Table 1). If dead trees were rare because they were consumed by the frequent ground fires, other woodpecker species and cavities created by them were likely also rare. Data on woodpecker species use of well-burned open pine habitats versus mixed pine-hardwood habitats support the argument that other woodpecker species were less abundant in the historic fire-maintained pine forests of the South than in habitats where hardwoods were present (Shackelford & Conner 1997). Detections of pileated woodpeckers (mean number detected per 3.5 ha plot sector) were 33 % higher (0.85 per plot visit versus 0.64) in infrequently burned pine-hardwood forest habitats than in more regularly burned longleaf pine habitats. Detections of red-bellied woodpeckers and northern flickers were 24% higher (1.56 per plot visit versus 1.26) and 75% higher (0.35 per plot visit versus 0.20), respectively, in pine-hardwood versus open pine habitats. The differences in the abundance of other Picoides were even more extreme. Detections of hairy and downy woodpeckers were 350% higher (0.27 per plot visit versus 0.06) and 2300% higher (0.24 per plot visit versus 0.01), respectively, in pine-hardwood versus open pine habitats. In contrast, a mean of 0.46 red-cockaded woodpeckers were detected per plot visit in the open pine habitats whereas none was detected in the pine-hardwood habitats (Shackelford & Conner 1997). Support for this suggestion that red-cockaded woodpeckers likely were and continue to be a keystone cavity provider for other cavity nesters in well-burned, fire-maintained southern pine ecosystems comes from the abundance of observations of other species using red-cockaded wood¬ pecker cavities. Numerous vertebrate and invertebrate species are known to use red-cockaded woodpecker cavities (Table 1). Because so many other cavity-nesting species are dependent on red-cockaded wood¬ peckers for cavities, forest biodiversity would suffer substantially in the absence of this endangered woodpecker in fire-maintained pine eco¬ systems of the South. Several species, such as red-bellied and red¬ headed woodpeckers and southern flying squirrels appear to compete actively with red-cockaded woodpeckers for intact cavities (Jackson 1978; Neal et al. 1992; Kappes & Harris 1995). The fact that red- 422 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Table 1. Vertebrate and invertebrate species observed using unenlarged and enlarged red-cockaded woodpecker cavities in the southeastern United States. Cavity occupant References for observation Birds American kestrel ( Falco sparverius) Brown-headed nuthatch (Sitta pusilla ) Carolina chickadee ( Poecile carolinensis) Eastern bluebird ( Sialia sialis ) Eastern screech-owl ( Otus asio ) European starling (Sturnus vulgaris) Great crested flycatcher (Myiarchus crinitus) Northern flicker ( Colaptes auratus ) Pileated woodpecker ( Dryocopus pileatus ) Red-bellied woodpecker (Melanerpes carolinus ) Red-headed woodpecker (M. erythrocephalus ) Tufted titmouse (Baeolophus bicolor) White-breasted nuthatch (Sitta carolinensis) Wood duck (Aix sponsa) (Rudolph et al. 1990a) (Jackson 1978) (Beckett 1971) (Baker 1971; Jackson 1978) (Baker 1971; Conner et al. 1997) (Dennis 1971; Jackson 1978) (Baker 1971; Conner et al. 1997) (Baker 1971; Dennis 1971) (Baker 1971; Jackson 1978) (Dennis 1971; Jackson 1978) (Baker 1971; Beckett 1971) (Baker 1971; Beckett 1971) (Baker 1971) (Baker 1971) Mammals Eastern gray squirrel (Sciurus carolinensis) Evening bat (Nycticeius humeralis) Fox squirrel ( Sciurus niger) Raccoon (Procyon lotor) Southern flying squirrel (Glaucomys volans) (Dennis 1971; Jackson 1978) (Rudolph et al. 1990a) (Baker 1971; Jackson 1978) (Loeb 1993) (Baker 1971; Beckett 1971) Reptiles and amphibians Broad-headed skink ( Eumeces laticeps) Five-lined skink (Eumeces fasciatus) Gray treefrogs (Hyla versicolor & H. chrysoscelis) Rat snake (Elaphe obsoleta) (Conner et al. 1997) (Jackson 1978) (Jackson 1978; Conner et al. 1997) (Baker 1971; Dennis 1971) Arthropods Ants Honey bee (Apis mellifera) Moths (Lepidoptera) Mud daubers (Sphecidae) Paper wasps (3 Polistes sp.) Spiders (Conner et al. 1997) (Dennis 1971; Jackson 1978) (Conner et al. 1997) (Conner et al. 1997) (Dennis 1971; Rudolph et al. 1990a) (Conner et al. 1997) headed and red-bellied woodpeckers, two woodpeckers that normally are primary excavators, regularly use red-cockaded woodpecker cavities for nesting over a wide geographic area (Neal et al. 1992) provides compel¬ ling evidence of the keystone role red-cockaded woodpeckers play in upland pine ecosystems. Red-bellied woodpeckers have been reported using red-cockaded woodpecker cavities more than any other species of bird throughout the South. Pileated woodpeckers enlarge the entrance to red-cockaded wood¬ pecker cavities such that they are no longer useable by the endangered woodpecker (Carter et al. 1989). Red-cockaded woodpeckers likely do not use these enlarged cavities because of their increased vulnerability CONNER ET AL. 423 to predators and competitors. Once a cavity entrance is enlarged, however, larger secondary cavity users, such as the American kestrel, eastern screech-owl, northern flicker, fox squirrel, raccoon, and wood duck, are able to use the cavity (Table 1). Anthropogenic forces have greatly altered the southern forest land¬ scape over the past 150 years (Frost 1993; Conner et al. 2001a). Exclu¬ sion and suppression of fire from fire-maintained ecosystems and con¬ version of pine forests to other land uses have occurred south wide. Such changes have permitted hardwood species to invade the previously open pine uplands and likely increased the availability of dead trees across the previously pine-dominated landscape. Snags do not always ignite under modern day prescribed fire conditions, especially when nearly all burns are conducted during winter under cool, humid condi¬ tions when the risk of wildfire is low. These changes have permitted other species of woodpeckers to be in closer proximity to red-cockaded woodpeckers than they were historically (Saenz et al . 2002) . A serious consequence of this change is the high rate of damage done to red- cockaded woodpecker cavities by pileated woodpeckers (Conner et al. 1991; Conner & Rudolph 1995; Saenz et al. 1998; 2002). The rate of damage is so severe that many red-cockaded woodpecker populations suffer an annual net loss of useable cavities. In Texas, red-cockaded woodpecker populations on the Angelina National Forest averaged an annual net loss of 4.6 useable cavities over a 10 year period (Conner et al. 1991; Conner & Rudolph 1995). The loss of cavities to tree death (57 cavity trees) was roughly equal to the loss due to pileated wood¬ pecker enlargement (55 cavity trees). Red-cockaded woodpeckers could not have evolved in the fire-main¬ tained pine ecosystems of the South if they suffered such a loss rate historically. They would have lost cavities faster than they could have excavated them. Pileated woodpecker abundance and their current rate of cavity destruction likely are elevated above what occurred in the South in the historic fire-maintained pine ecosystems of pre-Columbian times. Testing this hypotheses would be somewhat problematic in present day landscapes. Because of the large home range of a pileated woodpecker pair and red-cockaded woodpecker group, large tracts (5,000+ ha) of unbroken well-burned longleaf pine forest that are not fragmented from a timber-type and land-use perspective and still con¬ tained populations of red-cockaded woodpeckers would be needed to test the hypotheses. Such landscape conditions are now only a historic memory (Frost 1993; Conner et al. 2001a). 424 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Acknowledgments We thank J. A. Jackson, N. E. Koerth, C. E. Shackelford, and J. R. Walters for constructive comments on an early draft of the manuscript. Research on the red-cockaded woodpecker was done under U.S. Fish and Wildlife Service federal permit TE832201-0 to Richard N. Conner. Literature Cited Baker, W. W. 1971. Progress report on life history studies of the red-cockaded woodpecker at Tall Timbers Research Station. Pp. 44-60, in The ecology and management of the red-cockaded woodpecker (R. L. Thompson, ed.), Bureau of Sport Fisheries and Wildlife and Tall Timbers Research Station, Tallahassee, FL, 188 pp. Bartram, W. 1791 [1955]. Travels through North and South Carolina, Georgia, east and west Florida. Dover Publications, New York, xi + 414 pp. Beckett, T. 1971 . A summary of red-cockaded woodpecker observations in South Carolina. Pp. 87-95, in The ecology and management of the red-cockaded woodpecker (R. L. Thompson, ed.), Bureau of Sport Fisheries and Wildlife and Tail Timbers Research Station, Tallahassee, FL, 188 pp. Bent, A. C. 1939. Life histories of North American woodpeckers. United States National Museum Bulletin 174, xii + 334 pp. Bonnicksen, T. M. 2000. America’s ancient forests. John Wiley & Sons, Inc., New York, xii 4- 594 pp. Bull, E. L. & J. A. Jackson. 1995. Pileated woodpecker ( Dryocopus pileatus). No. 148 in The Birds of North America (A. F. Poole & F. B. Gill, eds.), The American Ornithologists’ Union and the Academy of Natural Sciences of Philadelphia, Washington, D.C., and Philadelphia, PA, 24 pp. Burns, F. L. 1900. A monograph of the flicker ( Colaptes auratus). Wilson Bulletin, No. 31, New Series, 7(2): 1-83. Carter, J. H., Ill, J. R. Walters, S. H. Everhart & P. D. Doerr. 1989. Restrictors for red-cockaded woodpecker cavities. Wildl. Soc. Bull., 17(l):68-72. Catesby, M. 1731. The natural history of Carolina, Florida and the Bahama Islands. Vol. I, M. Catesby, London, UK, xliv + 100 pp. Chapman, H. H. 1909. A method of studying growth and yield of longleaf pine applied in Tyler Co., Texas. Proc. Soc. Am. For., 4:207-220. Conner, R. N. 1981. Fire and cavity nesters. Pp. 61-65, in Prescribed fire and wildlife in southern forests (G. W. Wood, ed.), Belle W. Baruch Forest Science Institute of Clemson Univ., Georgetown, SC, 170 pp. Conner, R. N., R. G. Hooper, H. S. Crawford & H. S. Mosby. 1975. Woodpecker nesting habitat in cut and uncut woodlands in Virginia. J. Wildl. Manage., 39(1): 144-150. Conner, R. N. & B. A. Locke. 1982. Fungi and red-cockaded woodpecker cavity trees. Wilson Bull., 94(l):64-70. Conner, R. N., O. K. Miller, Jr. & C. S. Adkisson. 1976. Woodpecker dependence on trees infected by fungal heart rots. Wilson Bull., 88(4):575-581 . Conner, R. N. & D. C. Rudolph. 1995. Excavation dynamics and use patterns of red-cockaded woodpecker cavities: relationships with cooperative breeding. Pp. 343-352, in Red-cockaded woodpecker: recovery, ecology and management (D. L. Kulhavy, R. G. Hooper & R. Costa, eds.), Center for Applied Studies in Forestry, College of Forestry, Stephen F. Austin State Univ., Nacogdoches, TX, xvii + 551 pp. Conner, R. N., D. C. Rudolph, D. L. Kulhavy & A. E. Snow. 1991. Causes of mortality of red-cockaded woodpecker cavity trees. J. Wildl. Manage., 55(3) :53 1-537. CONNER ET AL. 425 Conner, R. N., D. C. Rudolph, D. Saenz & R. R. Schaefer. 1994. Heartwood, sapwood, and fungal decay associated with red-cockaded woodpecker cavity trees. J. Wildl. Manage., 58(4):728-734. Conner, R. N., D. C. Rudolph, D. Saenz & R. R. Schaefer. 1997. Species using red-cockaded woodpecker cavities in eastern Texas. Bull. Texas Ornithol. Soc., 30(1 ) : 1 1 - 1 6. Conner, R. N., D. C. Rudolph & J. R. Walters. 2001a. The red-cockaded woodpecker, surviving in a fire-maintained ecosystem. Univ. Texas Press, Austin, xv + 363 pp. Conner, R. N., D. Saenz, D. C. Rudolph, W. G. Ross & D. L. Kulhavy. 1998. Red-cockaded woodpecker nest-cavity selection: relationships with cavity age and resin production. Auk, 1 1 5(2) :447-454. Conner, R. N., D. Saenz, D. C. Rudolph, W. G. Ross, D. L. Kulhavy & R. N. Coulson. 2001b. Does red-cockaded woodpecker excavation of resin wells increase risk of bark beetle infestation of cavity trees? Auk, 1 18(1):21 9-224 . Conner, R. N., D. Saenz, D. C. Rudolph & R. R. Schaefer. 2002. Does the availability of artificial cavities affect cavity excavation rates in red-cockaded woodpeckers? J. Field Ornithol., 73(2): 125-129. Conner, R. N., D. Saenz, D. C. Rudolph & R. R. Schaefer. 2004. Extent of Phellinus pini decay in loblolly pines and red-cockaded woodpecker cavity trees in Texas. Mem. New York Bot. Garden, 89:315-321. Dennis, J. V. 1971. Species using red-cockaded woodpecker holes in northeastern South Carolina. Bird-Banding, 42(2): 79-87. Frost, C. C. 1993. Four centuries of changing landscape patterns in the longleaf pine ecosystem. Proc. Tall Timbers Fire Ecol. Conf., 18:17-43. Glitzenstein, J. S., W. J. Platt & D. R. Streng. 1995. Effects of fire regime and habitat on tree dynamics in north Florida longleaf pine savannas. Ecol. Monogr., 65(4):441-476. Harding, S. R. & J. R. Walters. 2002. Processes regulating the population dynamics of red-cockaded woodpecker cavities. J. Wildl. Manage., 66(4): 1083-1095. Hodges, J. D., W. W. Elam & W. F. Watson. 1977. Physical properties of the oleoresin system of the four major southern pines. Can. J. For. Res., 7(3): 520-525. Hodges, J. D., W. W. Elam, W. F. Watson & T. E. Nebeker. 1979. Oleoresin characteristics and susceptibility of four southern pines to southern pine beetle (Coleoptera: Scolytidae) attacks. Can. Entomol., Ill (8) :889-896. Jackson, J. A. 1974. Gray rat snakes versus red-cockaded woodpeckers: predator-prey adaptations. Auk, 91 (2): 342-347. Jackson, J. A. 1976. A comparison of some aspects of the breeding ecology of red-headed and red-bellied woodpeckers in Kansas. Condor, 78(l):67-76. Jackson, J. A. 1978. Competition for cavities and red-cockaded woodpecker management. Pp. 103-1 12, in Endangered birds: management techniques for endangered species (S. A. Temple, ed.), University of Wisconsin Press, Madison, WI, xxiii + 466 pp. Jackson, J. A., R. N. Conner & B. J. S. Jackson. 1986. The effects of wilderness on the endangered red-cockaded woodpecker. Pp. 71-78, in Wilderness and natural areas in the eastern United States: a management challenge (D. L. Kulhavy & R. N. Conner, eds.), Center for Applied Studies, College of Forestry, Stephen F. Austin State Univ., Nacogdoches, TX, 416 pp. Kappes, J., Jr. & L. D. Harris. 1995. Interspecific competition for red-cockaded woodpecker cavities in the Apalachicola National Forest. Pp. 389-393, in Red-cockaded woodpecker: recovery, ecology and management (D. L. Kulhavy, R. G. Hooper & R. Costa, eds.), Center for Applied Studies in Forestry, College of Forestry, Stephen F. Austin State Univ., Nacogdoches, TX, xvii + 551 pp. Kilham, L. 1983. Life history studies of woodpeckers of eastern North America. Nuttall Ornithological Club, No. 20. Cambridge, MA, viii + 240 pp. 426 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Landers, J. L. 1991. Disturbance influences on pine traits in the southeastern United States. Proc. Tall Timbers Fire Ecol. Conf . , 17:61-98. Lawrence, L. De K. 1967. A comparative life-history study of four species of woodpeckers. Omithol. Monogr., No. 5, 156 pp. Ligon, J. D. 1970. Behavior and breeding biology of the red-cockaded woodpecker. Auk 87(2) :255-278. Loeb, S. C. 1993. Use and selection of red-cockaded woodpecker cavities by southern flying squirrels. J. Wildl. Manage. , 57(2):329-335. Michaux, F. A. 1802. Early western travels, Vol. 3: Andre Michaux’s travels into Kentucky, 1793-1796; Francois Andre Michaux’s travels west of the Alleghany Mountains, 1802. (R. G. Th waites, ed.), A. H. Clark Co., Cleveland, OH, 1904, 382 pp. Neal, J. C., W. G. Montague & D. A. James. 1992. Sequential occupation of cavities by red-cockaded woodpeckers and red-bellied woodpeckers in the Ouachita National Forest. Proc. Ark. Acad. Sci., 46(1): 106-108. Neal, J. C., W. G. Montague & D. A. James. 1993. Climbing by black rat snakes on cavity trees of red-cockaded woodpeckers. Wildl. Soc. Bull., 21(2): 160-165. Platt, W. J., G. W. Evans & S. L. Rathburn. 1988. The population dynamics of a long-lived conifer. Am. Nat., 131 (4) : 49 1 -525 . Reller, A. W. 1972. Aspects of behavioral ecology of red-headed and red-bellied woodpeckers. Am. Midi. Nat., 88(2):270-290. Rudolph, D. C., R. N. Conner & J. Turner. 1990a. Competition for red-cockaded woodpecker {Picoides borealis) roost and nest cavities: the effects of resin age and cavity entrance diameter. Wilson Bull., 102(l):23-36. Rudolph, D. C., H. Kyle & R. N. Conner. 1990b. Red-cockaded woodpeckers versus rat snakes: the effectiveness of the resin barrier. Wilson Bull., 102(1): 14-22. Saenz, D., R. N. Conner & J. R. McCormick. 2002. Are pileated woodpeckers attracted to red-cockaded woodpecker cavity trees? Wilson Bull . , 1 1 4(3) :29 1 -296. Saenz, D. R. N. Conner, C. E. Shackelford & D. C. Rudolph. 1998. Pileated woodpecker damage to red-cockaded woodpecker cavity trees in eastern Texas. Wilson Bull., 1 10(3):362-367. Shackelford, C. E., R. E. Brown & R. N. Conner. 2000. Red-bellied woodpecker (Melanerpes carolinus) . No. 500, in The Birds of North America (A. F. Poole & F. B. Gill, eds.), The American Ornithologists’ Union and the Academy of Natural Sciences of Philadelphia, Washington, D.C., and Philadelphia, PA, 24 pp. Shackelford, C. E. & R. N. Conner. 1997. Woodpecker abundance and habitat use in three forest types in eastern Texas. Wilson Bull., 109(4): 61 4-629. Walters, J. R., C. K. Copeyon & J. H. Carter, III. 1992. Test of the ecological basis of cooperative breeding in red-cockaded woodpeckers. Auk, 109(1 ):90-97. Walters, J. R., P. D. Doerr & J. H. Carter, III. 1988. The cooperative breeding system of the red-cockaded woodpecker. Ethology, 78(2):275-305. Webb, T., III. 1987. The appearance and disappearance of major vegetational assemblages: long-term vegetational dynamics in eastern North America. Vegetatio, 69(1): 177-187. RNC at: c_connerrn@titan . sfasu . edu TEXAS J. SCI. 56(4): 427-440 NOVEMBER, 2004 FEEDING HABITS OF SONGBIRDS IN EAST TEXAS CLEARCUTS DURING WINTER Donald W. Worthington, R. Montague Whiting, Jr. and Janies G. Dickson Arthur Temple College of Forestry, Stephen F. Austin State University Nacogdoches , Texas 75962 and U. S. Forest Service, Southern Research Station Nacogdoches, Texas 75965 Abstract.— This east Texas study was undertaken to determine the importance of seeds of forbs, grasses, and woody shrubs to songbirds wintering in young pine plantations which had been established utilizing the clearcut regeneration system. The feeding habits and preferences of four species of songbirds, northern cardinals ( Cardinalis cardinalis), song sparrows {Melospiza melodia), dark-eyed juncos {Junco hyemalis), and white-throated sparrows ( Zonotrichia albicollis ) were examined from November to February of 1980-81, 1981-82, and 1982-83. Differences in consumption percentages were compared among bird species using AN OVA and Duncan’s multiple range tests. Paired /-tests were used to compare seeds consumed to seeds available by bird species. Differences ( P < 0.05) existed among bird species in consumption percentages of seeds of various genera. Northern cardinals selected seeds of Callicarpa, Croton, Datura , and Galactia. Song sparrows used seeds of Ambrosia, Panicum, and Seteria in excess of abundance. Dark-eyed juncos also selected Ambrosia as well as Eragrostis and Parietaria over seeds of other genera. Ambrosia, Parietaria, Aristida, and Viola were preferred by white-throated sparrows. Of the 4.7 million ha of commercial forest land in East Texas, 1.8 million are owned by forest industry (McWilliams & Lord 1988). Most such lands are intensively managed for pine on a short rotation ( < 50 years), evenage basis. A common practice on industrial forest lands is to clearcut the marketable timber at rotation age, prepare the site, and plant pine seedlings. After site preparation, growth and seed production of grasses and forbs are stimulated by decreased competition for nutri¬ ents, water, and sunlight. In the winter months, seeds of such plants are a valuable food source for birds. Few data exist on food habits and preferences or food availability to free-ranging songbirds wintering in young southern pine plantations. Therefore, the objectives of this study were to analyze winter foods of northern cardinals ( Cardinalis cardinalis ), song sparrows ( Melospiza melodia ), dark-eyed juncos (, Junco hyemalis ), and white-throated sparrows {Zonotrichia albicollis) collected on areas which had been recently clearcut, site prepared, and planted to pine seedlings, and to 428 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 determine if these species were selecting seeds of certain genera or if their feeding habits were dependent on seed availability. Methods Two study areas in the Piney woods Ecological Region of east Texas were selected, one in Nacogdoches County and another in Angelina County. Although the areas were in different counties, they were less than 20 km apart. Both areas had been clearcut, then residual vegetation sheared and along with debris, raked into long piles called windrows. The windrows were burned on the Angelina County study area. Both areas were planted with one-year-old pine seedlings during the study period. With one exception, soils on both study areas were well-drained fine sandy loams or loamy sands. A small part of the Angelina County study area was nearly level, thus poorly drained (Worthington 1984). Northern cardinals, dark-eyed juncos, and song sparrows were collected on the study areas during November, December, January, and February of 1980-81, 1981-82, and 1982-83; white-throated sparrows were collected in 1982-83 only. Efforts were made to collect five individuals of each species on each study area per month. All birds were collected in the morning. Each collected bird was immediately weighed to the nearest 0.5 grams. The digestive tract (esophagus, proventriculus, gizzard) was then removed and injected with 1 CC of 10% formalin (Dillery 1965) to stop the digestive process. The tract was placed in a self-sealing plastic bag along with an identification number. The location where the bird was first observed was marked with plastic flagging bearing the bird’s identification number. Upon returning from the field, each digestive tract was frozen and stored. In the laboratory, the contents of each digestive tract were dried at 38 °C for 48 hours, then weighed to the nearest 0.0001 g. Digestive tract contents were then separated into four groups, namely plant seeds, insect parts, grit, or unidentified material. Seeds were then separated to genus using keys (Musil 1963; Landers & Johnson 1976) and a U.S. Forest Service reference seed collection. Seeds not identified were kept separate, labeled unknown, and assigned a number. Many of these unknown seeds were later identified. All food materials were then redried at 38 °C for 48 hours and weighed to the nearest 0.0001 g. Seeds on the ground, presumably available to the collected birds, WORTHINGTON, WHITING & DICKSON 429 were sampled during the 1981-82 and 1982-83 study periods, usually the same day the birds were collected. Seeds were sampled on five 10 cm radius subplots in the area where each bird was first observed. The first subplot was where the bird was originally observed and the others were in each cardinal direction, 2 m from the first subplot. Food materials were collected using a hand-held power vacuum. Seeds on standing vegetation directly above the subplots also were collected. In the laboratory, availability samples were frozen for 48 hours to kill insects, then coarse debris was removed. The remaining material was passed through a series of sieves to sort seeds by size class and remove fine debris. A binocular dissecting scope was used when separating seeds from fine debris. The seeds were sorted, dried, and weighed in the same manner as were seeds in the digestive tracts of the birds. The five subplot samples were combined to form a single availability sample for analyses. For each bird species, the number of individuals that consumed each seed genus was determined by study area. Each value was then divided by the total number of birds of that species to obtain frequency of occurrence. Differences in frequencies of occurrence were tested among bird species by study area using two-by-four Chi-square tests. Due to differences in body weights and total digestive tract content weights among the four bird species (Worthington 1984:61), actual weights of seed genera consumed were not compared among the bird species. Instead, weights of all identified and unidentified seeds in each bird’s digestive tract were summed and the weight of each genus was converted to a percent of that sum. These values reflected consumption percentages and were compared among the four bird species. Seeds available to the birds were evaluated similarly. The conversion of actual weights to percentages also allowed for comparisons between consumed and available seeds. Insect parts, grit, and unidentified material were not compared. Differences among bird species in seed consumption percentages were tested using ANOVA with Duncan’s multiple range tests. Differences in seed availability percentages were tested in the same manner. For each genus, paired t- tests were used to compare percentages of seeds consumed to percentages of seeds available by bird species and study 430 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 area. As seed availability data were not collected in 1980-81, seed consumption data from that year were not used when comparing seeds consumed to those available. In order for a seed genus to be included in the statistical comparisons of digestive tracts, it had to average at least 2% of consumed seeds, by weight, for at least one bird species. To be included in comparisons of availability data, a genus had to comprise at least 4% of the available seeds, by weight, for at least one bird species. Throughout the study, the null hypothesis used was that of no difference among groups being tested. The rejection level was set at 0.05 for all tests. Results Ninety-five northern cardinals, 59 song sparrows, and 86 dark-eyed juncos were collected during the three winters; 45 white- throated sparrows were collected in the winter of 1982-83. Unidentified material comprised 64.2 , 67.0, 65.7, and 72.9% of weights of digestive tract contents of northern cardinals, song sparrows, dark-eyed juncos, and white- throated sparrows, respectively; identifiable seeds made up 23.5, 16.5, 18.2, and 14.3% of digestive tract content weights of the four species, respectively. With one exception, small amounts of greenery, insects, and grit made up the remainder; one northern cardinal had con¬ sumed a ground skink (Scincella lateralis). Most unidentified material was in the gizzard. It was assumed that proportions of unidentifiable material in that organ were the same as those identifiable (West 1973). Seeds consumed Seeds of 38 genera were identified and recorded in digestive tracts of the birds (Worthington 1984:64-74). Eight groups of seeds could not be identified, but only one was consumed in greater than trace (i.e., < 1.0%) quantities. With one exception, seeds of all identifiable genera recorded in digestive tracts were also recorded in availability samples; no Datura seeds were recorded in availability samples. In Nacogdoches County, differences existed among bird species in frequencies of occurrence of seeds of 10 genera (Table 1). A higher proportion of the northern cardinal digestive tracts contained seeds of Callicarpa, Croton, and Datura than did those of the other three bird species. Conversely, Ambrosia occurred in a lower proportion of northern cardinals than in the other three species. Eragrostis and WORTHINGTON, WHITING & DICKSON 431 Table 1 . Numbers of birds and frequency of occurrence of seeds in digestive tracts of northern cardinals (NOCA), song sparrows (SOSP), dark-eyed juncos (DEJU), and white-throated sparrows (WTSP, 1982-1983 only) collected in eastern Texas during winter 1980-81, 1981-82, and 1982-83. Within a row, a different letter indicates different frequencies of occurrence among bird species at the 0.05 level. Seed genera NOCA n % SOSP n % DEJU n % WTSP n % X2 P Nacogdoches County Amaranthus 7 12.3a 2 18.2a 24 51.1b 5 23.8a <0.001 Ambrosia 4 7.0a 5 45.5b 26 55.3b 15 71.4b <0.001 Callicarpa 23 40.4a 1 9.1b 0 0.0b 2 9.5b <0.001 Carex 4 7.0 1 9.1 0 0.0 0 0.0 0.153 Croton 28 49.1a 0 0.0b 0 0.0b 1 4.8b <0.001 Cyperus 2 3.5a 3 27.3b 8 17.0b 0 0.0a 0.008 Datura 29 50.1a 2 18.2b 3 6.4b 0 0.0b <0.001 Digitaria 12 21.1 4 36.4 13 27.7 1 4.8 0.120 Eragrostis 0 0.0a 5 45.5c 12 25.5b 0 0.0a <0.001 Panicum 2 3.5a 6 54.6c 11 23.4b 0 0.0a <0.001 Parietaria 0 0.0a 1 9.1a 20 42.6b 12 57.1b <0.001 Paspalum 4 7.0 2 18.2 2 4.3 0 0.0 <0.201 Phytolacca 20 35.1a 2 18.2ab 3 6.4b 0 0.0b <0.001 Rudbeckia 0 0.0 0 0.0 2 4.3 0 0.0 <0.284 Sample size 57 11 47 21 Angelina County Amaranthus 0 0.0a 2 4.2a 12 30.8b 1 4.5a <0.001 Ambrosia 4 10.5a 15 31.3b 17 43.6b 12 50.0b 0.003 Callicarpa 14 36.8a 0 0.0b 0 0.0b 0 0.0b <00001 Carex 12 31.6a 16 33.3a 1 2.6b 4 16.7b 0.002 Croton 13 34.2a 1 2.1b 1 2.6b 0 0.0b <0.001 Cyperus 1 2.6a 8 16.7ab 10 25.6b 6 25.0b 0.031 Datura 15 39.5a 3 6.3b 0 0.0b 0 0.0b <0.001 Digitaria 0 0.0a 7 14.6b 7 18.0b 1 4.5ab 0.029 Eragrostis 0 0.0 3 6.3 4 10.3 1 4.5 0.252 Panicum 1 2.6a 38 79.2b 31 79.5b 8 33.3c <0.001 Parietaria 0 0.0 1 2.1 1 2.6 0 0.0 0.693 Paspalum 2 5.3a 2 4.2a 8 20.5b 0 0.0a 0.008 Phytolacca 0 0.0 4 8.3 2 5.1 0 0.0 0.172 Rudbeckia 0 0.0a 1 2.1a 10 25.6b 0 0.0a <0.001 Sample size 38 48 39 24 432 THE TEXAS JOURNAL OF SCIENCE— VOL. 56(4), 2004 Panicum seeds were found in greater proportions of song sparrow digestive tracts than in those of the other three species and in more dark-eyed junco tracts than in northern cardinal or white- throated sparrow tracts. Amaranthus occurred in a higher proportion of dark¬ eyed juncos than in the other three bird species. A majority of white- throated sparrows consumed Ambrosia and Parietaria. In Angelina County, Callicarpa, Croton , and Datura were recorded in higher proportions of northern cardinal digestive tracts than in those of the other three species (Table 1). Conversely, Ambrosia and Panicum were found in lower proportions of northern cardinal digestive tracts than in digestive tracts of the other species. Ambrosia was found in half of white- throated sparrows, and Panicum occurred in almost 80% of the song sparrows and dark- eyed juncos. Finally, higher proportions of dark-eyed juncos than the other species consumed Amaranthus , Paspalum, and Rudbeckia. Seeds of 23 genera comprised at least 2% of the total weight of seeds in the digestive tracts of one or more bird species (Worthington 1 984: 64-74) . Percent consumption of 1 2 of these genera differed among bird species (Table 2). Combined, Croton and Datura comprised approximately 64 and 43 % of the weight of seeds in the digestive tracts of northern cardinals collected on the Nacogdoches County and Angelina County study areas, respectively. On both study areas, these combined percentages were higher than those of the other three bird species (Table 2). Song sparrows and dark-eyed juncos consumed relatively large quantities of Ambrosia , Digitaria , and Panicum on both study areas. Percent consumption of these genera by song sparrows and dark-eyed juncos were generally higher than for northern cardinals and white- throated sparrows, except for Ambrosia which made up a higher per¬ centage of the digestive tract contents of white-throated sparrows than of the other species (Table 2). On the Nacogdoches County study area, white-throated sparrows also consumed relatively more Parietaria than did the other species. Seeds available. — Eighty-two genera of seeds were collected on the two study areas, 72 on the Nacogdoches County study area and 61 on the Angelina County study area (Worthington 1984:62-63). Fifty-one genera were common to both areas; 21 and 10 were exclusive to Nacogdoches County and Angelina County, respectively. However, WORTHINGTON, WHITING & DICKSON 433 Table 2. Weights (in percent) of seeds recorded in digestive tracts of northern cardinals (NOCA), song sparrows (SOSP), dark-eyed juncos (DEJU), and white-throated sparrows (WTSP, 1982-1983 only) collected in eastern Texas during winter 1980-81, 1981-82, and 1982-83. Only genera for which there were significant differences among bird species are shown. Within a row by study area, a different letter denotes different proportions among bird species at the 0.05 level. Nacogdoches County Angelina County Seed genera NOCA SOSP DEJU WTSP NOCA SOSP DEJU WTSP Amaranthus 0.39a 2.07ab 26.46c 13.30b 0.00a 1.43a 10.50b 3.72ab Ambrosia 0.35a 24.94b 26.42b 49.05c 4.18a 11.64a 9.98a 39.41b Callicarpa 8.08a 4.55ab 0.00b 1.99b 17.21a 0.00b 0.00b 0.00b Carex 1.28 0.79 0.00 0.00 5.58ab 10.18b 1.87a 4.48ab Croton 34.40a 0.00b 0.00b 2.94b 21.00a 1.50a 0.93b 0.00b Datura 30.04a 2.29b 0.65b 0.00b 22.50a 0.44b 0.00b 0.00b Digitaria 0.30a 15.24b 8.12b 1.14a 0.00a 0.92ab 3.31b 0.80ab Eragrostis 0.00a 7.01b 5.50b 0.00a 0.00 1.25 0.74 0.11 Panicum 0.07a 6.50b 3.84b 0.00a 0.27a 42.74b 34.81b 13.91a Parietaria 0.00a 8.54ab 15.34b 28.05c 0.00 1.35 0.55 0.00 Paspalum 1.99 6.79 1.91 0.00 0.25a 1.07a 7.10b 0.00a Phytolacca 11.53a 10.67a 1.53b 0.00b 0.00 3.28 1.40 0.00 Total (%) 88.44 89.39 89.77 96.47 71.26 75.80 71.19 62.43 Sample size 57 11 47 21 38 48 39 0.55 24 0.00 only 16 genera each contributed a minimum of 4% of the seeds available to at least one bird species. The genera of frequently occurring seeds included Andropogon, Digitaria, Panicum, Phytolacca , Rhus, Solidago, and Uniola. On the Nacogdoches County study area, there were differences among bird species in seed availability frequencies of five commonly occurring genera (Worthington 1984:34). However, only Ambrosia, Digitaria, and Eragrostis comprised at least 2% of the weight of seeds consumed. Ambrosia occurred more frequently in white-throated sparrow and song sparrow food availability samples than in those of northern cardinals, and Digitaria was recorded in higher percentages of song sparrow and dark-eyed junco than white- throated sparrow food availability samples. Eragrostis was found in a higher percentage of song sparrow food availability samples than in those of the other species (Worthington 1984:34). For the Angelina County study area, frequencies of only two seed genera differed among food availability samples (Worthington 1984:35). Neither of these, Eupatorium and Heterotheca , could be considered important food items to the collected birds. 434 THE TEXAS JOURNAL OF SCIENCE- VOL. 56(4), 2004 Table 3. Weights (in percent) of seeds available to northern cardinals (NOCA), song sparrows (SOSP), dark-eyed juncos (DEJU), and white-throated sparrows (WTSP, 1982-1983 only) collected in eastern Texas during winter 1981-82 and 1982-83. Genera shown are those for which there were differences in percent availability and/or percent consumption. Within a row by study area, a different letter denotes different proportions at the 0.05 level. Nacogdoches County Angelina County Seed genera NOCA SOSP DEJU WTSP NOCA SOSP DEJU WTSP Amaranthus 8.78a 4.54a 25.92c 17.28b 0.56 3.11 3.40 1.64 Ambrosia 0.89 3.13 4.36 3.51 0.14a 5.79ab 8.25b 1.31ab Callicarpa 7.68 6.21 3.74 1.81 2.30 0.10 0.00 0.79 Car ex 0.02 0.00 0.03 0.00 0.09 4.27 0.47 3.32 Croton 4.95 0.49 4.74 1.65 0.30 0.02 0.24 0.51 Digitaria 1.13a 14.23b 3.41a 0.31a 0.00a 0.34a 1.96b 0.49a Eragrostis 0.15 0.75 0.26 0.00 1.09 0.71 0.13 3.08 Eupatorium 1.69 3.25 0.52 2.06 2.33 7.90 1.88 3.89 Galactia 0.28a 5.94b 1.25a 0.65a 0.00 0.17 0.00 0.00 Heterotheca 0.95 5.01 1.28 3.18 7.72 2.29 6.96 12.83 Panicum 5.18 7.74 2.63 0.44 10.54a 25.99b 26.00b 1 1 .02a Parietaria 0.00 0.21 0.10 0.35 0.00 0.00 0.00 0.00 Paspalum 0.74 0.93 0.06 0.17 1.57 0.92 0.00 1.28 Phytolacca 14.16a 5.33a 10.30a 30.11b 1.13 2.17 2.43 0.18 Rhus 29.26 33.77 21.88 28.93 21.75a 5.14b 15.94ab 24.42a Uniola 0.43 0.00 0.63 0.64 7.28 9.42 9.56 8.26 Sample size 42 7 37 21 29 39 24 24 There were some differences in weights (in percent) of seeds available to the bird species in each county (Table 3). In Nacogdoches County, there were differences among species for Amaranthus, Digitaria, Galactia, and Phytolacca. There was a higher proportion of Amaran¬ thus seeds in dark-eyed junco availability samples than in those of the other species, and a higher proportion in white-throated sparrow samples than in northern cardinal or song sparrow samples. Song sparrow availability samples contained higher proportions of Digitaria and Galactia seeds than did samples for the other species, and Phytolacca seeds ranked higher in white-throated sparrow samples than in samples for the other species (Table 3). In Angelina County, there were differences in seed availability percentages of Ambrosia , Digitaria , Panicum , and Rhus among bird species. Both Ambrosia and Panicum seeds were less available to northern cardinals than to the other species. Digitaria seeds ranked higher for dark- eyed juncos than for the other species, but made up less WORTHINGTON, WHITING & DICKSON 435 than 2% of the food available to that species. Rhus, which comprised large proportions of the seeds available on both study areas (Table 3), was not an important food source to any species. Seeds selected.— Callicarpa, Croton, Datura, Galactia, and Phytolacca comprised 90% of the seeds consumed by northern cardinals in Nacogdoches County during the winters of 1981-82 and 1982-83 (Table 4); Croton, Datura, and Galactia were consumed in excess of availability. The same was true of Callicarpa and Croton in Angelina County. Phytolacca availability exceeded consumption in Nacogdoches County but was not recorded in any Angelina County digestive tracts (Table 4). Only seven song sparrows were collected in Nacogdoches County, thus statistical comparisons are weak at best. However, almost 35% of the seeds identified in the digestive tracts of those birds were Ambrosia. Seeds of that genus, Care) c, Panicum, and Seteria were dominant in Angelina County song sparrows. Consumption percentages of the two latter genera were greater than availability percentages (Table 4). For darked-eyed juncos from Nacogdoches County, consumption of Ambrosia, Eragrostis, and Parietaria exceeded availability. Amaran- thus, which was readily available on that study area, comprised slightly over 25% of the seeds consumed. In Angelina County, seeds of Amaranthus, Ambrosia, Digitaria, and Panicum comprised almost 70% of identifiable seeds in dark-eyed junco digestive tracts; consumption and availability percentages of these genera were similar (Table 4). White-throated sparrows were collected only in 1982-83. In both counties, Ambrosia compromised the largest proportion of identifiable seeds. Consumption of that genus and Parietaria exceeded availability in Nacogdoches County. In Angelina County, Ambrosia, Aristida, and Viola demonstrated similar trends. Amaranthus in Nacogdoches County and Cy perns and Panicum in Angelina County were important food items for which consumption and availability percentages did not differ (Table 4). Rhus seeds were recorded in two white- throated sparrows in Angelina County. Discussion Although identifiable seeds comprised relatively small proportions of digestive tracts, this study provided strong evidence that northern cardinals, song sparrows, dark-eyed juncos, and white-throated sparrows 436 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Table 4. Comparisons of percent seed availability and percent seed consumption for northern cardinals, song sparrows, dark-eyed juncos, and white-throated sparrows in eastern Texas during winter 1981-82 and 1982-83. Paired /-tests values also are shown. Nacogdoches County Angelina County Seed genera Pet. Avail. Pet. Cons P - value Pet. Avail. Pet. Cons. P- value Northern cardinals n = 42 n = 29 Amaranthus 8.78 0.19 0.006 0.57 0.00 0.194 Callicarpa 7.68 9.22 0.481 2.63 19.09 0.030 Carex 0.02 1.91 0.230 1.03 6.36 0.162 Croton 4.95 46.01 <0.001 0.31 26.81 0.001 Datura 0.00 24.35 <0.001 0.00 6.66 0.113 Galactia 0.28 5.16 0.041 0.00 0.00 1.000 Heterotheca 0.95 0.01 0.049 7.69 0.00 0.047 Myrica 2.17 0.00 0.274 5.16 0.00 0.153 Panicum 5.18 0.03 0.018 10.77 0.33 0.024 Phytolacca 14.16 5.31 0.016 1.15 0.00 0.179 Rhus 29.26 0.00 <0.001 22.92 0.00 0.002 Uniola 0.43 0.00 0.310 7.91 3.66 0.381 Song sparrows n = 1 n = 39 Amaranthus 4.54 0.87 0.252 3.11 1.77 0.615 Ambrosia 3.28 34.91 0.053 5.79 11.41 0.227 Carex 0.00 1.23 0.356 4.27 8.47 0.265 Digitaria 14.23 19.29 0.734 0.34 0.49 0.880 Eupatorium 3.25 0.00 0.352 7.90 0.00 0.001 Panicum 7.74 2.15 0.343 25.99 44.48 0.010 Phytolacca 5.33 0.00 0.120 2.17 2.60 0.894 Rhus 33.74 0.00 0.071 5.14 0.00 0.042 Seteria 0.00 0.00 1.000 0.00 6.35 0.044 Uniola 0.00 0.00 1.000 9.42 0.03 0.004 Dark-eyed juncos n — 37 n = 24 Amaranthus 26.10 25.63 0.912 3.40 11.27 0.110 Ambrosia 4.33 29.60 <0.001 8.25 13.88 0.454 Digitaria 3.37 5.93 0.337 1.96 5.25 0.312 Eragrostis 0.27 5.62 0.044 0.13 1.12 0.169 Heterotheca 1.29 0.00 0:049 6.96 0.00 0.035 Panicum 2.63 3.37 0.641 25.95 39.23 0.166 Parietaria 0.10 19.49 <0.001 0.00 0.89 0.094 Phytolacca 10.20 0.30 <0.001 2.43 2.28 0.658 Rhus 21.75 0.00 <0.001 15.91 0.00 0.025 Uniola 0.63 2.28 0.135 9.54 2.22 0.117 White-throated sparrows* n — 21 n — 24 Amaranthus 17.28 13.30 0.645 1.64 3.72 0.637 Ambrosia 3.51 49.05 <0.001 1.31 39.41 0.001 Aristida 0.00 0.00 1.000 0.00 11.99 0.041 Cyperus 0.13 0.00 0.892 3.20 8.88 0.409 Eupatorium 2.06 0.00 0.134 3.89 0.00 0.106 Heterotheca 3.18 0.14 0.181 12.83 0.00 0.016 Panicum 0.44 0.00 0.014 11.02 13.91 0.948 Parietaria 0.35 28.05 0.002 0.00 0.00 1.000 Phytolacca 30.11 0.00 <0.001 0.18 0.00 0.319 Rhus 28.93 0.00 0.001 24.42 1.28 0.008 Uniola 0.64 0.00 0.329 8.26 0.40 0.057 Viola 0.00 2.35 0.126 0.00 6.28 0.020 * Collected in winter 1982-83 only. WORTHINGTON, WHITING & DICKSON 437 selected seeds of some genera over those of others. Korschgen (1980) noted that if a food item occurred in high numbers of individuals and in high volume within the individuals, the food was of high quality or preference. In this study, three or four genera met these criteria for each bird species. For most of these genera, consumption exceeded availability. Seeds utilized by northern cardinals were very different from those used by the other species. With study areas combined, Callicarpa, Croton , and Datura comprised approximately 69 % of the seeds identi¬ fied in northern cardinal digestive tracts. These genera made up only trace proportions in digestive tracts of the other bird species. The importance of Croton and Callicarpa to northern cardinals is well- documented (Martin et al. 1951 ; Halkin & Linville 1999). Carex, Rhus , Setaria , and Panicum have also been classified as important to northern cardinals (Halkin & Linville 1999). Although seeds of these genera were collected on the study areas, they made up minor portions of northern cardinal diets, and no Rhus was recorded in any northern cardinal. No mention of northern cardinals consuming Datura was found in the literature. Reasons for the absence of Datura seeds in availability samples are unknown; Datura plants were present on both study areas. Although there were similarities in diets of song sparrows, dark-eyed juncos, and white-throated sparrows, the relative rank of the important genera varied among species. For song sparrows, Panicum made up 38% of identifiable seeds; Ambrosia (15%) ranked second and Carex (7%) third. Neither Martin et al. (1951) nor Arcese et al. (2002) listed Panicum as an important food source for song sparrows. Results of this study contradict those findings, and it is possible that the low number of song sparrows collected in Nacogdoches County was due to the lack of Panicum. Ambrosia seeds are an important winter food item for song sparrows (Martin et al. 1951), as are those of Amaranthus, Digitaria , and Setaria (Arcese et al. 2002). In this study, seeds of these three genera comprised relatively minor proportions of song sparrows diets. In dark-eyed junco digestive tracts, Ambrosia (23%), Amaranthus (20%), Panicum (18%), and Parietaria (12%) made up almost three- fourths of the identifiable seeds. Judd (1901) and Nolan et al. (2002) noted the importance of Ambrosia and Amaranthus to dark-eyed juncos. 438 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 Martin et al. (1951) also found seeds of Ambrosia and various grasses to be important food items for the species. White-throated sparrows were abundant on both study areas during winter 1982-83. With data from study areas pooled, Ambrosia , (43%) comprised a higher proportion of that species diet than did any genera in diets of the other species. Falls & Kopachena (1994) noted the importance of Ambrosia to white- throated sparrows. However, they also stated that fruits of Rhus were important to the species. During this study, numerous white- throated sparrows were observed foraging in Rhus , and it was assumed that they were eating Rhus fruit. Several of those were birds were collected, yet Rhus seeds comprised a very minor proportion of the diet. Halls (1977) noted that birds cannot sustain weight on a heavy diet of Rhus and that it is normally eaten with other foods. The very small amount of Rhus consumed by birds collected in this study support Halls’ comments and indicate that birds observed foraging in Rhus were either seeking other food items or were consum¬ ing minute quantities of that genus. Conclusions In this study, each bird species consumed seeds of several genera in excess of availability. Also, availability percentages exceeded consump¬ tion percentages for some genera and did not differ for others. Al¬ though seeds of all genera were available to each species, the differences among species may have been due to differences in habitat selection within the clearcuts. Virtually all northern cardinals were first observed in or adjacent to the relatively dense vegetation of the windrows or small riparian zones which were present on both study areas. Song sparrows were usually in dense grassy areas between rows of planted pine seed¬ lings. Dark-eyed juncos were in similar areas, but at higher elevations where ground cover was less dense. White-throated sparrows were collected in areas similar to those of northern cardinals. These results demonstrate that when properly administered, the cl earcutting method of regeneration creates excellent habitat for ground- foraging, seed-eating birds which winter in the southern United States. This method creates openings in the forest and, combined with site preparation techniques that scarify both the soil and dormant seeds, promotes the establishment of seed-bearing forbs and grasses. WORTHINGTON, WHITING & DICKSON 439 Acknowledgments We appreciate the field assistance of Steve Best and Nolan Smith and the laboratory assistance of Tracy Flavins, Kathleen Kroll, and Karen Hoza- Wilson. We are indebted to John Roese for much of the statistical analyses. Rhonda Barnwell, Crystal Linebarger, and Ashley Sample provided manuscript preparation. This project was funded by the U.S. Forest Service, Southern Experiment Station, and the Arthur Temple College of Forestry at Stephen F. Austin State University. Literature Cited Arcese, P., M. K. Sogge, A. B. Marr & M. A. Patten. 2002. Song sparrow. In The birds of North America, No. 704 (A. Pool and F. Gill, editors). The Academy of Natural Sciences, Philadelphia, Pennsylvania and the American Ornithologists’ Union, Washington, D.C., pp. 1-39. Dillery, D. G. 1965. Post-mortem digestion of stomach contents in the savannah sparrow. Auk, 82(2):281. Falls, J. B. & J. G. Kopachena. 1994. White-throated sparrow. In The birds of North America, No. 128 (A. Pool and F. Gill, editors). The Academy of Natural Sciences, Philadelphia, Pennsylvania and the American Ornithologists’ Union, Washington, D.C., pp. 1-30. Halkin, S. L. & S. U. Linville. 1999. Northern cardinal. In The birds of North America, No. 128 (A. Pool and F. Gill, editors). The Academy of Natural Sciences, Philadelphia, Pennsylvania and the American Ornithologists’ Union, Washington, D.C., pp. 1-29. Halls, L. K., ed. 1977. Southern fruit-producing woody plants used by wildlife. U. S. Dept. Agric., For. Serv. Gen. Tech. Rep. S0-16, 235 pp. Judd, S. D. 1901. The relation of sparrows to agriculture. U.S. Dept. Agric., Biol. Surv. Bull. No. 15, 98 pp. Korschgen, L. J. 1980. Procedures for food habit analyses. Pp. 113-127, in Wildlife management techniques manuel (D. D. Schemnitz, editor). The Wildlife Society, Washington, D.C., 686 pp. Landers, J. L. & A. S. Johnson. 1976. Bobwhite food habits in the southeastern United States with a seed key to important foods. Misc. Publ. No. 4, Tall Timbers Res. Stn., Tallahassee, Florida, 90 pp. Martin, A. C., H. S. Zim & A. L. Nelson. 1951. American wildlife and plants: a guide to wildlife food habits. McGraw Hill, New York, New York, 499 pp. McWilliams, W. H. & R. G. Lord. 1988. Forest resources of East Texas. U.S. Dept. Agric., For. Serv. Resour. Bull. SO-136, 61 pp. Musil, A. F. 1963. Identification of crop and weed seeds. U. S. Dept. Agric., Agriculture Handbook No. 219, 171 pp. Nolan, V., Jr., E. D. Ketterson, D. A. Cristol, C. M. Rogers, E. D. Clotfelter, R. C. Titus, S. J. Schoech & E. Snajdr. 2002. Dark-eyed junco. In The birds of North America, No. 716 (A. Pool and F. Gill, editors). The Academy of Natural Sciences, Philadelphia, Pennsylvania and the American Ornithologists’ Union, Washington, D.C., pp. 1-42. 440 THE TEXAS JOURNAL OF SCIENCE-VOL. 56(4), 2004 West, G. C. 1973. Foods eaten by tree sparrows in relation to availability during summer in northern Manitoba. J. Arctic Institute of North Am., 26(1):7-21. Worthington, D. W. 1984. Winter songbird feeding habits on east Texas clearcuts. Unpublished M.S. thesis, Stephen F. Austin State University, Nacogdoches, Texas, 83 pp. RMW at: mwhiting@sfasu.edu TEXAS J. SCI. 56(4):44 1-451 NOVEMBER, 2004 INDEX TO VOLUME 56 (2004) THE TEXAS JOURNAL OF SCIENCE Sandra L. Woods Department of Biology, Angelo State University San Angelo, Texas 76909 This index has separate subject and author sections. Words, phrases, locations, proper names and the scientific names of organisms are followed by the initial page number of the article in which they appeared. The author index includes the names of all authors followed by the initial page number of their respective article(s). SUBJECT INDEX A Abiotic Factors 35 Abiotic stress 335 Acacia berlandieri 253 Acacia farnesiana 253 Acacia rigidula 253 Acacia schaffheri 253 Acari 369 Accuracy 149 Acid Sulfate conditions 91 Activity patterns 383 Adult foraging behavior 141 Acheta domesticus 141 Allelopathic component 3 Ambrosia 427 American beech, decline in 285 American Fisheries Society 63 American Ornithologists’ Union 1957, 1998 77 Amistad Reservoir 223, 237 Aquilla Lake 187 Araneida 369 Arboreal behavior 395 Aristida 427 Arizona 267 Arkansas 73, 273 Arsenic 91 Artesia Wells 237 Arthropod assemblage 369 Asexual reproduction 175 Aspidoscelis gulaxis 237 Aspidoscelis laredoensis 237 Asteraceae 15 Asymmetrical Avifauna 197 Axis movement 149 B Baird’s pocket gophers 383 Big Thicket National Preserve 299 442 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 Big Thicket Science Conference 281 Biomass 175 Bivalvia 63, 223 Brackish marshes 103 Branch elongation 253 Brittle Star 175 Broadcasting 231 Burning regime 405 Brush-Grassland 15 Burrows 383 C Caenidae 123 Caenis latipennis 123 California 55 Callicarpa All Canopy 35 Cardinalis cardinalis All Carolina Chickadee 187 Cartago Province 81 Catarina 237 Caudata 273 Channel Islands 175 Chapparal Wildlife Management Area Chinese tallow tree 335, 357, 369 Chloraleucon ebano 253 Clearcuts 427 Clonal complex 237 Clutch size 81 Cnemidophus gularis 237 Cnemidophorus laredoenss 237 Coahuila 223 Coastal-marshes 15 Coelomocytes 175 Coffee Snake 81 Colonized 157 Colorado River 223 Colubridae 267, 383 Commercial dog and cat chow 141 Community similarity 103 Continental shelf 157 Cooke County 73 Costa Rica 81 Crickets 141 Critical Thermal Maximum 123 Crotalus horridus 395 Crotalus cerastes 55 Croton All Cupressacae 3 D Dallas county 73 Dark-eyed j uncos 427 Datura All Demographics of occupancy 131 Density 187 Departure from neutrality 157 Devils River 223 Diel activity patterns 383 Diet 77 Diptera 369 Diversity 187 131 Dominant species 103 Dry oak-pine forest 299 Dryocopus pileatus 415 Durophagus shark 215 E Early olfactory experience 141 Eastern wild turkeys 405 Echinodermata 175 Ecological resistance 237 Ecological notes 263 Edwards Plateau 35 Emberizidae 77 Encinal 237 INDEX 443 Endoparasites 273 Ephemeroptera 123 Equal sex ratio 131 Eragrostris 427 Europe 263 Eurymerodesmidae 73 Eurymerodesmus mundus 73 Eurymerodesmus angularis 73 Evolutionary lineage 273 F Fabaceae 15 Fagus grandifolia 285 Falcon Reservoir 223 Fall 197, 253 Feeding habits of songbirds 427 Feeding regimes 141 Fertilization 357 Fire 299, 319, 415 Fissiparous species 175 Flood plain 267 Flood plain forests 335 Florida 73, 263 Follicles 268 Food-borne olfactory cues 141 Forbes 427 Fossil record 215 Freshwater mussels 63 Freshwater marshes 103 Fuel characteristics 319 Full-factorial experiment 149 Fungi 415 Fusconaia askewi 63 G Galactia 427 Galapagos Islands 175 Genetic drift 157 Genetic polymorphism 157 Geographic shift 179 Germination rates 347 Glacial retreat 157 Gober Chalk 215 Grasses 427 Grasshopper Mouse 141 Grayson county 73 Growth 335 Gulf of Mexico 237 H Haplotype variation 157 Hardin County, Texas 285 Hemoglobin 175 Hemiptera 369 Herbaceous ground cover 405 Herbicide 347 Heterodon nasicus 267 Histological examination 268 Hogna carolinensis 141 Holotype 215 Homogeneity 157 Honduras 81 Hornshell 223 Hurricane damage 285 Hybridization 237 I Illinois 73, 267 Impoundment 187 Impoverished diet 141 Incubation 207 Independent variables 149 Insect fauna 369 Intercanopy 35 Invasion 299, 335 444 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 J Johnson county 73 Junco hyemalis 427 Junior synonym 215 Juniperus ashei 3, 35 K Kansas 215, 268 Keystone species 415 Known prey odor 141 L Lakes 91 Lampsilis satura 63 Late cretanaceous 215 Limnodrilus 263 Linear model 149 Liquidambar styraciflua 357 Longleaf pine 319 Louisiana 73, 265 Louisiana pine snake 383 Lower Rio Grande Valley 253 Lutjanus campechanus 157 M Macleod & Slaughter 215 Mate guarding 207 Mayfly 123 Mealworms 141 Mean error 1,49 Mearns 141 Median intrapair distance 207 Meleagris gallopavo sylvestris 405 Melospiza melodia 427 Mesic species 299, 319 Mexico 197, 223, 237, 267 Middens 131 Milliped 73 Mimus polyglottos 207 Mismatch distribution 157 Mississippi 73 Missouri 73 Mitochondrial DNA 157 Mockingbird 207 Monogamous passerine 207 Monotypic genus 73 Mudstone 91 Mulch depth 347 Muridae 131, 141 Mutation 157 Mycorrhizal fungi, inoculum 357 N Native trees 335 Nebraska 73 Neches River Estuary 265 Neotoma micropus 131 Nest building 207 Nesting attempts 179 Nesting habitat 405 Neuroptera 369 Nevada 197 Nicaragua 81 Ninia maculate 8 1 North America 263 North American Rattlesnake 55 North Carolina 73, 263 Northern cardinal 187, 427 Northern Gulf of Mexico 157 Northern Mockingbird 207 Novel prey odors 141 Novel pure chemical odor 141 Nucleotide-site 157 INDEX 445 Nuevo Leon 197 Nyssa sylvatica 357 O Oak 197 Oak forrest 197 Obovaria jacksoniana 63 Odor choice test 141 Odor preferences 141 Oklahoma 73, 273 Olfactory imprinting 141 Oligochaeta 263 Onychomys arenicola 141 Open Woodland 267 Ophiactis simplex 175 Ophiuroidea 175 Ortheroptera 369 Oxidation 91 Oxygen concentrations 123 P PABNHS 15 pH 123 Panama 77, 81, 175 Panicum 427 Parabloids 231 Parietaria 427 Parthenogenetic 237 Passeriformes 77 Paternity assurance behavior 207 Payload 149 Peloscolex 263 Peripheral populations 237 Periplaneta americana 141 Physiological Tolerance ranges 123 Picoides borealis 415 Pileated woodpeckers 415 Pine 197 Pine oak 197 Pine snake 383 Pineywoods Ecological Region 405 Pirns palustris 319 Pinus taeda 357 Pituophis ruthveni 383 Plethodon glutinous 273 Plethodontidae 273 Plethodon Sequoyah 273 Pleurobema riddellii 63 Poecile caroinensis 187 Poaceae 15 Polydesmida 73 Popenaias popeii 223 Postal notice 455 Prairie 267 Prairie restoration 347 Precipitated Fe(OH)3 91 Prescribed fire 319 Ptychodontidae 215 Ptychodus martini Williston 215 Ptychodus connellyi 215 Pyrite 91 Q Quadrula mortoni 63 Quercus alba 357 Quercus nigra 357 R Radars 231 Radio-telemetry 383 Rainfall 253 Radial error variability 149 Radio- marked 405 Random component 149 Rasacas 15 446 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 Rat snakes 415 Red-cockaded woodpecker 415 Red Snapper 157 Red tube feet 175 Regressed testes 55 Regression model 149 Repeatability 149 Reproduction 81, 171 Reproductive cycle 268 Reproductive data 268 Reservoirs 91 Reservoirs impoundment 187 Revolution 231 Rio Grande 73, 237 Rio Grande Delta 103 Rio Grande Valley 77, 179 Rio Sabinas 223 River centered zone 237 Roaches 141 Robotics 149 Roden tia 131, 141 Roxton Limestone member 215 S Salamander 273 Salt flats 15 Salt marshes 103 San Luis Potosf 267 Sandy substrate 237 Sapium sebiferum 335, 357, 369 Sapling growth and mortality 299 Savannah restoration 347 Searching behavior 141 Secondary vilellogensis sensu 268 Sediments 91 Seed germination 347 Seedlings 3 Seedling demography 35 Seeds 427 Semidester habitat 268 Serpentes 267 Sequoyah slimy salamander 273 Serpentes 81 Seteria 427 Sevier County, Arkansas 273 Shade- tolerant species 299 Shannon’s Diversity Index 197 Shredding mowers 347 Single breeding season 179 Snags 415 Snake 268 Social structure 131 Soil fertility 357 Solar energy collection 23 1 Songbirds, feeding habits 427 Song sparrows 427 South Concho River 223 Southern Canada 267 Southern mixed forests 285 Southern Plains Woodrat 131 Species composition 103 Species diversity 103 Species evenness 103 Species richness 103, 369 Speed 149 Sporophila torqueola 77 Spring 197, 253 Squamata 237 Standard deviation 149 Stem density 319 Stepwise fluctuations 123 Structural changes 319 Summer 197 Sulfur 91 Survivorship 123 Syntopy 237 Systematic 263 INDEX T Tamaulipas State, Mexico 237 Tanyard Branch Creek 123 Tanks 15 Taught position 149 Taxonomic 263 Teeth 215 Teiidae 237 Telemetry studies 395 Temporal signature 157 Temperature 347 Tenebrio molitor 141 Terrestrial avian communities 187 Testicular cycle 55 Texas 15, 73, 77, 91, 131, 175, 179, 215, 223, 237, 263, 253 Texas Counties: Cameron 15 Cooke 73 Dallas 73 Dimmit 237 Fannin 215 Grayson 73 Hardin 285 Hildago 237, 253 Johnson 73 Lasalle 237 Starr 237, 253 Walker 123 Webb 237 Zapata 237 Texas Ebony 253 Thorn scrub vegetation 237 Thysanoptera Timber rattlesnake 395 Trans-Pecos 179 Tree litter 3 Trophic structure 369 Tubificidae 263 Tubificoides heterochaetus 263 U United States 73 Unionidae 63, 223 Uplland communities 319 Upper Lower Campanian 215 V Vasa deferentia 55 Vascular plants 15 Vegetational communities 197 Vermiculite Control 3 Village Creek basin 63 Viola 427 187, Virginia 263 W Walker County 123 Water quality tolerance 263 Water regines, varying 335 Waves 231 Weches Formation 91 Western hognose snake 267 Western interior sea 215 White-collared seedeater 77 White-throated sparrows 427 White- winged dove 179 Wier Woods, Texas 285 Winter 197, 253, 427 Wolf Spider 141 Woodland overstory 35 Woody plant communities 319 Woody shrubs 427 Wright-Fisher model 157 Y Y-maze olfactometer 141 Yolk deposition 55, 268 448 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 Z Zaragoza 197 Zenaida asiatica 179 Zonotrichia albicollis All INDEX 449 AUTHOR INDEX Amir-Moez, A. R. 231 Baccus, J. T. 179 Barker, C.A. 91 Bodily, R. Y. 207 Bordelon, V. L. 63 Bradley, R. D. 131 Brink, J. 149 Burridge, C. P. 157 Bursey, C. R. 273 Butterfield, B. J. 335 Christensen, A. B. 175 Cook, J. L. 123 Conner, R. N. 395, 415 Contreras-Balderas, A. J. 197 Cordes, J. E. 237 Correa-Sandoval, A. 223 DeWalt, S. 369 Dickson, J. G. 427 Donahue, C. 347 Ealy, M. J. 383 Eddy, M. R. 253 Eichler, B. G. 405 Eitniear, J. C. 77 Elsik, I. S. 285 Fleet, R. R. 383 Fulhorst, C. F. 131 Fulton, M. R. 285, 299 Gold, J. R. 157 Goldberg, S. R. 55, 81, 171, 267 Gonzalez-Rojas, J. I. 197 Hall, R. W. 299 Hamm, S. A. 215 Haney, A. 149 Harcombe, P. A. 285, 299, 319 Harrel, R. C. 63, 263 Hartley, M. K. 369 Hinds, B. 149 Howells, R. G. 223 Jha, S. 285 Judd, F. W. 103, 253 Judy, K. 91 Kerstupp, A. O. 197 Knox, R. G. 319 Ledger, E. B. 91 Lin, J. 299 Liu, C. 319 Lonard, R. I. 15, 103 McAllister, C. T. 73, 273 McKinley, D. 3 Moore, D. I. 73 Neudorf, D. L. H. 207 Nijjer, S. 357 Olalia- Kerstupp, A. 197 Paulissen, M. A. 237 Puckett, R. T. 123 Punzo, F. 141 Ransom, Jr., D. 187 Richard, N. L. 15 Richardson, A. T. 15 Rogers, W. E. 335, 347, 357, 369 Rudolph, D. C. 383, 395, 415 Ruthven, III, D. C. 131 Ruvalcuba-Ortega, I. 197 450 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 4, 2004 Saenz, D. 395, 415 Schaefer, R. R. 395 Schaefer, C. L. 179 Shelley, R. M. 73 Shimada, K. 215 Siemann, E. 335, 347, 357, 369 Slack, R. D. 187 Small, M. F. 179 Strenth, N. E. 223 Suchecki, J. R. 131 Van Auken, O. W. 3, 35 Walker, J. M. 237 Wayne, R. 35 Welch, R. D. 179 Whiting, Jr., R. M. 405, 427 Worthington, D. W. 427 INDEX 451 REVIEWERS The Editorial staff wishes to acknowledge the following indi- iduals for serving as reviewers for those manuscripts considered for publication in Volume 56. Without your assistance it would not be possible to maintain the quality of research results published in this volume of the Texas Journal of Science. Abbott, J. Allison, T. Anderson, J. Arnold, K. Baskin, J. Bestgen, K. Bidwell, T. Bray, S Breshears, D. Brush, T. Bryan, Jr, A. Burk, J. Cameron, G. Cecil, D. Ciccimuri, D. Clark, W. Connor, W. Cook, T. Curran, S. Dinsmore, S. Divine, D. Ernst, C. Everitt, J. Farrish, K. Foster, C. Gelwick, F. Graves, J. Harcombe, P. Harper, C. Harper, D. Harrel, R. Harveson, L. Hathcock, C. Henke, S. Henry, B. Hicks, D. Highton, R. Holley, A. Howells, R. Hurst, G. Jones, R. Judd, F. Jurena, P. Krauss, K. Lonard, R. MacFadden, B. Mathewson, C. Maxwell, T. McAllister, C. McDonald, H. McGregor, K. Monfredo, W. Montagna, P. Murray, H. Nieland, D. Norwine, J. Ortego, B. Painter, C. Parker, W. Persans, M. Rayor, L. Ribble, D. Richardson, A. Riskind, D. Robertson, P. Rupert, J. Schmidlin, T. Schwertner, T. Schwimmer, D. Siemann, E. Singer, F. Smeins, F. Smith, E. Stangl, Jr., F. Taylor, E. Upton, S. Wake, D. Walker, E. Wallace, M. Walley, H. Welbourn, W. Winne, C. Wittrock, D. Woodin, M Zaidan, F. THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 453 IN RECOGNITION OF THEIR ADDITIONAL SUPPORT OF THE TEXAS ACADEMY OF SCIENCE DURING 2004 Patron Members Ali R. Amir-Moez Deborah D. Hettinger Don W. Killebrew David S. Marsh John Sieben Ned E. Strenth Sustaining Members James Collins Dovalee Dorsett Stephen R. Goldberg Donald E. Harper, Jr. Norman V. Horner Michael Looney Sammy M. Ray Fred Stevens William F. Thomann Milton W. Weller Supporting Members David A. Brock Frances Bryant Edens Michael Halbouty Patrick Donegan Hollis George D. McClung Jimmy T. Mills Jim Neal Gary Powell Judith Schiebout Lynn Simpson 454 THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 THE TEXAS ACADEMY OF SCIENCE www . texasacademy ofscience . org Membership Information and Application MEMBERSHIP.— Any person or member of any group engaged in scientific work or interested in the promotion of science is eligible for membership in The Texas Academy of Science. (Please print or type) Name Last First Middle Mailing Address City State Zip Ph FAX E-mail: Regular Member $30.00 Supporting Member $60.00 Sustaining Member $100.00 Patron Member $150.00 Student-Undergraduate $15.00 Student- Graduate $15.00 Joint $35.00 Emeritus $10.00 Corporate Member $150.00 Affiliate $5.00 (list name of organization) Contribution AMOUNT REMITTED SECTIONAL INTEREST AREAS: A. Biological Science I. Geology B. Botany K. Mathematics C. Chemistry L. Physics D. Computer Science M. Science Education E. Conservation & Management O. Systematics & Evolutionary Biology F. Environmental Science P. Terrestrial Ecology G. Freshwater & Marine Sciences Q. Anthropology H. Geography R. Threatened or Endangered Species Please indicate your Sectional interest(s) below: 1. 2. 3. Send Application Form and Check or Money Order to: Dr. Fred Stevens, Executive Secretary The Texas Academy of Science CMB 5980 Schreiner University Kerrville, Texas 78028-5697 Please photocopy this Application Form THE TEXAS JOURNAL OF SCIENCE-VOL. 56, NO. 4, 2004 455 United States Postal Service Statement of Ownership, Management, and Circulation 1. Publication Title The Texas Journal of Science 2. Publication Number 3. Filing Date 1 October 2004 0 0 4 0 4 4 03 4. Issue Frequency Quarterly 5. Number of Issues Published Annually 4 6. Annual Subscription Price $30 Membership $50 Subscription 7. Complete Mailing Address of Known Office of Publication (Not printer) (Street, city, county, state, andZiP+4) Biology Department, Angelo State University 2601 West Avenue N, Tom Green County, San Angelo, TX 76909 Contact Person N.E. Strenth Telephone 325.942.2189 ex. 247 8. Complete Mailing Address of Headquarters or General Business Office of Publisher (Not printer) Dr. Ned E. Strenth, Biology Department Angelo State University San Angelo, TX 76909-5069 _ 9. Full Names and Complete Mailing Addresses of Publisher. Editor, and Managing Editor (Do not leave blank) _ _ Publisher (Name and complete mailing address) Dr. Ned E. Strenth, Biology Department Angelo State University San Angelo, TX 76909-5069 _ Editor (Name and complete mailing address) Dr. Ned E. Strenth, Biology Department Angelo State University San Angelo, TX 76909-5069 _ Managing Editor (Name and complete mailing address) Dr. Ned E. Strenth, Biology Department Angelo State University San Angelo, IX 76909-5069 _ 10. Owner (Do not leave blank. If the publication is owned by a corporation, give the name and address of the corporation immediately followed by the ' names and addresses of all stockholders owning or holding 1 percent or more of the total amount of stock It not owned by a corporation, give the names and addresses of the individual owners, if owned by a partnership or other unincorporated firm, give its name and address as well as those of each individual owner. If the publication is published by a nonprofit organization, give its name and address.) Full Name Complete Mailing Address Texas Academy of Science Angelo State University Department of Biology 2601 West Avenue N San Angelo, TX 76909-5069 11. Known Bondholders. Mortgagees, and Other Security Holders Owning or Holding 1 Percent or More of Total Amount of Bonds, Mortgages, or Other Securities. If none, check box - — - - B None Full Name Complete Mailing Address 1 2. Tax Status (For completion by nonprofit organizations authorized to mail at nonprofit rates) (Check one) The purpose, function, and nonprofit status of this organization and the exempt status for federal income tax purposes: CX Has Not Changed During Preceding 12 Months □ Has Changed During Preceding 1 2 Months (Publisher must submit explanation of change with this statement) PS Form 3526, October 1999 (See Instructions on Reverse) 456 THE TEXAS JOURNAL OF SCIENCE— VOL. 56, NO. 4, 2004 13. Publication Title The Texas Journal of Science 14. Issue Date for Circulation Data Below November 2004 Extent and Nature of Circulation Average No. Copies Each Issue During Preceding 12 Months No. Copies of Single Issue Published Nearest to Filing Date a. Total Number of Copies (Net press run) 1100 1100 b. Paid and/or Requested Circulation (1) Paid/Requesled Outside-County Mall Subscriptions Stated on Form 3541. (Include advertisers proof and exchange copies) 808 965 (2) Paid In-County Subscriptions Stated on Form 3541 (Include advertisers proof and exchange copies) 15 15 (3) Sales Through Dealers and Carriers, Street Vendors, Counter Sales, and Other Non-USPS Paid Distribution 157 157 (4) Other Classes Mailed Through the USPS 0 0 Total Paid and/or Requested Circulation K / Sum of 15b. (1). (2), (3), and (4)] T 980 980 dFree Distribution by Man (Samples, compliment ary, and other free) (1) Outside-County as Stated on Form 3541 0 0 (2) In-County as Stated on Form 3541 0 0 (3) Other Classes Mailed Through the USPS 0 0 e- Free Distribution Outside the Mail (Carriers or other means) 0 0 Total Free Distribution (Sum of 15d. and 15e.) ^ 0 0 a. Total Distribution (Sum of 15c. and 151) ► 980 980 h. Copies not Distributed 120 120 I- k. Total (Sum of ISg. andh.) ^ 1100 1100 I- Percent Paid and/or Requested Circulation (15a divided by 15g. times 100) 89% 89% 16. Publication of Statement of Ownership "Vol 56 #4 Publication required. Will be printed in the ’ issue of this publication. □ Publication not required. 17. Signature and Title of Editor, PubSel lerfBusiness Manager, or Owner Date 29 Sept. 2004 I certify that all information furnished on this form Is true and complete. I understand that anyone who furnishes false or misleading Information on this form or who omits material or information requested on the form may be subject to criminal sanctions (including Tines and imprisonme nt) and/or civil sanctions (including civil penalties). ' Instructions to Publishers 1 . Complete and file one copy of this form with your postmaster annually on or before October 1 . Keep a copy of the completed form for your records. 2. In cases where the stockholder or security holder is a trustee, include in items 10 and 11 the name of the person or corporation for whom the trustee is acting. Also include the names and addresses of individuals who are stockholders who own or hold 1 percent or more of the total amount of bonds, mortgages, or other securities of the publishing corporation. In item 11. if none, check the box. Use blank sheets if more space is required. 3. Be sure to furnish all circulation information called for in item 15. Free circulation must be shown in items 15d, e, and f. 4. Item 15h., Copies not Distributed, must include (1) newsstand copies originally stated on Form 3541 , and returned to the publisher, (2) estimated returns from news agents, and (3), copies for office use, leftovers, spoiled, and all other copies not distributed. 5. If the publication had Periodicals authorization as a general or requester publication, this Statement of Ownership, Management, and Circulation must be published; it must be printed in any issue in October or, if the publication is not published during October, the first issue printed after October. 6. In item 16, indicate the date of the issue in which this Statement of Ownership will be published. 7. Item 17 must be signed. Failure to file or publish a statement of ownership may lead to suspension of Periodicals authorization. PS Form 3526, October 1999 (Reverse) THE TEXAS ACADEMY OF SCIENCE, 2004-2005 OFFICERS President : President Elect : Vice-President : Immediate Past President : Executive Secretary : Corresponding Secretary : Managing Editor. Manuscript Editor : Treasurer : AAAS Council Representative : John A. Ward, Brook Army Medical Center Damon E. Waitt, Lady Bird Johnson Wildflower Center David S. Marsh, Angelo State University John T. Sieben, Texas Lutheran University Fred Stevens, Schreiner University Deborah D. Hettinger, Texas Lutheran University Ned E. Strenth, Angelo State University Robert J. Edwards, University of Texas-Pan American James W. Westgate, Lamar University Sandra S. West, Southwest Texas State University DIRECTORS 2002 Sushma Krishnamurthy, Texas A&M International University Raymond D. Mathews, Jr., Texas Water Development Board 2003 Hudson R. DeYoe, University of Texas-Pan American Cynthia Contreras, Texas Parks and Wildlife Department 2004 Benjamin A. Pierce, Baylor University Donald L. Koehler, Balcones Canyonlands Preserve Program SECTIONAL CHAIRPERSONS Anthropology : Roy B. Brown, Institute Nacional de Antropologia y Historia Biological Science : Francis R. Horne, Texas State University Botany : Herbert D. Grover, Harden-Simmons University Chemistry: Mary Kipecki-Fjetland, St. Edward’s University Computer Science: Laura J. Baker, St. Edwards University Conservation and Management: Felipe Chavez-Ramirez, Platte River Whooping Crane Trust Environmental Science: William Thomann, University of the Incarnate Word Freshwater and Marine Science: Sharon Conry, Baylor University Geology and Geography: Carol Thompson, Tarleton State University Mathematics: Hueytzen J. Wu, Texas A&M University-Kingsville Physics: David Bixler, Angelo State University Science Education: Jimmy Hand, Austin, Texas Systematics and Evolutionary Biology: Kathryn Perez, University of Alabama Terrestrial Ecology: Jerry Cook, Sam Houston State University Threatened or Endangered Species: Alice L. Hempel, Texas A&M University-Kingsville COUNSELORS Collegiate Academy: William J. Quinn, St. Edward’s University Junior Academy: Vince Schielack, Texas A&M University Nancy Magnussen, Texas A&M University THE TEXAS JOURNAL OF SCIENCE PrinTech, Box 43151 Lubbock, Texas 79409-3151 PERIODICAL POSTAGE PAID AT LUBBOCK TEXAS 79402 RETURN SERVICE REQUESTED 6543 SMITHSONIAN INSTITUTION LIBRARIES NHB25MRC154 PO BOX 37012 WASHINGTON, DC 20013-7012