Vertebrate Inventory of Richland Creek Wildlife Management Area in Eastern Texas

Terrestrial habitats of Richland Creek Wildlife Management Area, Texas, were surveyed for vertebrate diversity in 1998 and 1999. During the 2-year sampling period, 10 species of amphibians, 20 species of reptiles, and 23 species of mammals were collected or observed. These actions represent an effort of Texas Parks and Wildlife to inventory the fauna of state-owned property. This information will begin to form a baseline to assess future management decisions. RESUMEN Llevamos a cabo un inventario de los vertebrados en hábitats terrestres del Richland Creek Wildlife Management Area, Texas, durante 1998 y 1999. Durante los dos años del periodo de muestreo, observamos o colectamos 10 especies de anfibios, 20 especies de reptiles, y 23 especies de mamı́feros. Estas acciones representan un esfuerzo de Texas Parks and Wildlife para hacer un inventario de la fauna en las propiedades del estado. Esta información será la base para evaluar futuras decisions de manejo de recursos naturales. December 2004 529 Notes The ecological, economic, and esthetic values of biodiversity are recognized widely (Wilson, 1992; Meffe and Carroll, 1997), and managing for biological diversity is a priority for state and federal natural resource and wildlife agencies. One key threat to regional biodiversity relates to trends in land-use practices that can result in fragmented habitat, thus increasing the probability that some components of biodiversity will experience local extirpation in these habitat patches. Although national parks, national wildlife refuges, and state and local management units represent important means of protecting biodiversity, such areas are effectively ‘‘land-bridge islands’’ subject to increased rates of local extinction relative to the area of the reserve and degree of isolation (Newmark, 1995). For large mammals in many national parks in the western United States, there is an apparent increase in extinction that fits the expectations of ‘‘land-bridge islands’’ (Newmark, 1995). Like many western states, Texas is experiencing increased fragmentation of natural areas as land-use trends change, and the rate at which wildlife habitat fragmentation is occurring relates to changes in land ownership resulting in more people owning smaller properties (Wilkins et al., 2003). Unlike many western states, land ownership in Texas is primarily under private control, with considerably less acreage managed by local, state, and federal agencies. As a result of private land ownership, one of the most effective means for offsetting the negative effects of fragmentation and the resulting loss of biodiversity in Texas is the retention and protection of the network of Wildlife Management Areas (WMAs) created by Texas Parks and Wildlife (TPW). Conceptually, these WMAs are maintained for research and for outdoor activities including hunting, fishing, and non-consumptive uses. In recent years, TPW has emphasized the management of all biodiversity, both game and nongame species. As a result of both this new mission and the ever-increasing rate at which wildlife habitat and biodiversity is being lost, it is imperative that all WMAs be inventoried for their components of biodiversity. Such baseline information is absolutely essential for long-term monitoring of changes in biodiversity in response to trends in land use and development. The purpose of this study was to gather baseline information on fishes (Gelwick et al., 2000), amphibians, reptiles, and mammals at Richland Creek WMA and to establish a foundation for future management decisions allowing for long-term monitoring of biodiversity. Established in 1987, Richland Creek WMA (ca. 5,583 ha) is located 58 km southeast of Corsicana, Texas, between Richland-Chambers Reservoir and the Trinity River in Freestone and Navarro counties. This property, comprised of 2 management units separated by U.S. Highway 287, is situated in an ecotone separating the Post Oak Savannah and Blackland Prairie ecological regions, and a large portion of the area lies within the Trinity River and Richland Creek floodplains (Gelwick et al., 2000). The combined average rainfall for Freestone and Navarro counties is about 97 cm/y. The average low temperature is 28C in January, while the average high temperature is 358C in July (Department of the Army, Corps of Engineers, 1982). Soils on Richland Creek WMA consist of periodically flooded Trinity and Kaufman clays, with occasional lenses of Lamar Clay Loam or Silawa Fine Sandy Loam (Department of the Army, Corps of Engineers, 1982). The property varies from fairly well drained uplands and transition zones to large bottomland areas subject to periodic and prolonged flooding. Virtually all woodlands on Richland Creek WMA are bottomland forests characterized by regrowth cedar elm (Ulmus crassifolia), sugarberry (Celtis laevigata), green ash (Fraxinus pennsylvanica), boxelder (Acer negundo), and black willow (Salix nigra). Occasional pockets of bur oak (Quercus macrocarpa), Shumard oak (Q. shumardii), overcup oak (Q. lyrata), water oak (Q. nigra), willow oak (Q. phellos), post oak (Q. stellata), and native pecan (Carya illinoensis) occur on the area. Understory vegetation is diverse, including hawthorn (Crataegus), honey locust (Gleditsia tricanthos), soapberry (Sapindus saponaria), cat briar (Smilax bona-nox), swamp privet (Forrestiera acuminata), and dense vines of poison ivy (Rhus toxocodendron), rattan (Berchemia scanders), Virginia creeper (Parthenocissus quinquefolia), and trumpet creeper (Bignonia radicans). We conducted surveys for terrestrial vertebrates on both management units from March 1998 to August 1999, following procedures outlined by TPW (Simpson et al., 1996) and 530 vol. 49, no. 4 The Southwestern Naturalist adapted for specific habitats in the Richland Creek WMA. Amphibians and reptiles were sampled primarily from May through August using time-constrained searches (0900 to 1200 h and 1600 to 1900 h), timed nocturnal road searches (2000 to 2400 h), baited hoop nets, minnow traps, and Y-shaped drift-fence arrays. Time-constrained searches consisting of groups of 4 to 12 observers actively searching for amphibians and reptiles (herps) were conducted daily during sampling periods of 3 to 4 d throughout the months listed in Table 1. Data were grouped by month (some months containing multiple sampling periods) and expressed as number of herp captures per person-minute to standardize information on numbers of individuals of different species found within a site. We conducted timed nocturnal road searches by automobile 1 to 2 times per sampling period at approximately the same time each night along the same route. Road driving data were expressed as the number of herps observed per minute. We used hoop nets baited with carp (Cyprinella carpio) to trap aquatic turtles in ponds and reservoirs during 1 sampling period in June 1998. Watersnakes (Nerodia), cottonmouths (Agkistrodon piscivorus), and Graham’s crayfish snakes (Regina grahamii) were trapped using Gee Minnow Traps (Forestry Suppliers, Inc., Jackson, Mississippi) placed along shorelines of flooded pastures during 1 sampling period in June 1998 and another in August 1999. Nine Yshaped drift-fence arrays were constructed with a single pitfall trap in the center and funnel traps at the ends of the fences. This design, modified from that of Heyer et al. (1994), allowed us to avoid the problems associated with flooding of buckets in lowland habitats. Sites along habitat edges were selected for drift fence construction to maximize species richness in our sampling effort. Drift-fence arrays were run 2 to 3 nights per sampling period. We expressed turtle trap, minnow trap, and pitfall array capture data as captures per trapnight. Small mammals were captured using Sherman live traps. Seven transects consisted of lines with 50 traps spaced at 6.1-m intervals baited with birdseed and peanut butter. We did not use peanut butter during times of intense activity of imported red fire ants (Solenopsis invicta). Medium-sized mammals were trapped on 1 permanent transect of 10 Tomahawk live traps spaced 30 m apart and baited with canned cat food. Additional carnivore traps were set opportunistically where carnivore activity was judged to be high. Incidental observations of mammals were noted, especially during nocturnal road searches. Voucher specimens, morphological data, supporting genetic materials, and associated ecological and locality data were deposited in the Texas Cooperative Wildlife Collection, Department of Wildlife and Fisheries Sciences, at Texas A&M University. All combined methods of sampling amphibians and reptiles resulted in 533 captures of 10 species of amphibians and 20 species of reptiles (Table 1; Crother et al., 2003). Based on published distributional information and county records, at least 33 additional amphibian and reptile species are expected to occur on the property (Dixon, 2000). Time-constrained searches and night driving together accounted for 465 (87.3%) of the total captures. Time-constrained searches and incidental captures yielded the most species-rich samples, with 20 and 18 species sampled, respectively (Table 2). The timber rattlesnake (Crotalus horridus), the only Federal Category 2 and state-threatened species verified on Richland Creek WMA, was collected using both of these methods. For amphibians and reptiles, nocturnal road searches provided the highest capture rate (1 capture per 3.5 min) among the active sampling methods, and hoop nets provided the highest capture rate (1 capture per 1.1 trap nights) among passive sampling methods (Table 2). Thirty-four animals representing 18 species were observed incidentally (Table 2); 3 of these species were not found with the standardized methods outlined by Simpson et al. (1996). Although 30 species of amphibians and reptiles were captured, 77% of the specimens represented only 5 species (Table 1). The northern cricket frog (Acris crepitans), green treefrog (Hyla cinerea), and southern leopard frog (Rana sphenocephala) exhibited the highest relative abundance, followed by the diamondbacked watersnake (N. rhombifer) and the cottonmouth (A. piscivorus). The false map turtle (Graptemys pseudogeographica), snapping turtle (Chelydra serpentina), eastern box turtle (Terra-