Occasional Papers Museum of Texas Tech University Number 211 5 December 2001 THE MAMMALS OF DEVILS RIVER STATE NATURAL AREA, TEXAS Joel G, Brant and Robert C. Dowler Devils River State Natural Area (DRSNA) is lo¬ cated about 40 miles (68 km) north of Del Rio in Val Verde County, Texas. In 1988, approximately 20,000 acres (8,097 hectares) of land was purchased from the Finegan Ranch, which comprised about one third of the original Fawcett Ranch, to create DRSNA. This state natural area is situated around Dolan Creek and its associated side canyons, including East Canyon, Rough Spring Canyon, and Live Oak Canyon (Figs. 1 and 2). The Devils River composes the southwestern boundary of the area and is fed by several springs, most notably Dolan Springs, which add 22,000 gal¬ lons (83,600 L) of water per minute to the river (Smith, 1975). Annual precipitation in the area averages 18.81 inches (65.84 mm) however, during 1998 and 1999 DRSNA was experiencing a drought with the average annual rainfall of 13.51 inches (47.29 mm) recorded by the National Oceanic and Atmospheric Administra¬ tion (NO A A) from the Del Rio International Airport (Bryan et ah, 2001), Vertical relief in the park aver¬ ages 200-300 feet (67-100 m), with slopes of 25-30 percent (Texas Parks and Wildlife Department, 1990). Devils River State Natural Area is an ecotonal zone at the juncture of three ecogeographic regions: the Edwards Plateau to the north and east, the South Texas Plains to the south, and the Trans-Pecos to the west (Blair, 1950). This confluence of ecoregions creates an area where eastern species might reach the western extent of their range and southern species might attain their northern limits. The Devils River also creates a dispersal corridor for southern species to move northward into west-central Texas, The soils of DRSNA (Fig. 3) are composed mainly of Ector-Rock associations, which are subdi¬ vided into two categories, hilly and very steep. The very steep category is the most abundant and occurs in association with canyon slopes throughout the park. The hilly subdivision is restricted to the plateau be¬ tween the canyons. Other soil types including Olmos very gravelly loams, Dev soils, and Riverwash, occur along Dolan Creek and its tributaries in East Canyon and Speed Canyon (Golden et al., 1982). The first reports of mammals in Val Verde County were included in Vernon Bailey’s biological survey of Texas (1905). Bailey reported 82 species of mammals in the Lower Sonoran Zone that includes the Devils River Valley. The first comprehensive mammalian sur¬ vey for eastern Val Verde County was performed by Boston (1966). He reported the occurrence of 39 spe¬ cies of mammals including county records for the desert shrew (Notiosorex crawfordi) and the hoary bat (Lasiurus cinereus), and sightings of the hooded skunk (Mephitis macroura) in Val Verde County. Scudday and Hanselka (1975) first surveyed the mammals of the Dolan Creek - Dolan Falls area, a good portion of which would later become DRSNA. They reported 39 species of mammals, 29 collected by them and 10 reported in the literature. The most recent study that included this area was the Mammals of the Edwards 2 Occasional Papers, Museum of Texas Tech University Figure I. Location of Devils River State Natural Area in Val Verde County, Texas. Figure 2. Devils River State Natural Area with primary sampling sites la¬ beled. 1) Jumper Slope, 2) East Canyon Slope, 3) Cenizo Slope, 4) Grass¬ land Plateau, 5) Dolan Springs, 6) the Headquarters Building, 7) Fawcett Cave, 8) Radio Towers, 9) First Landing Strip, 10) Second Landing Strip, 11) Jose Maria Springs. Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 3 ] Ector-Rock outcrop assoc iations, hi lly Sc ale in Mile s [jam Ector-Rock outcrop associations, very steep I556&3 Riverwash Olmos very gravelly loam ■■ Dev soils, frequently flooded Figure 3. A soil map of Devils River State Natural Area, adapted from Golden et al. (1982), Plateau, Texas (Goetze, 1998). Many of the species listed by Goetze (1998) should be expected to occur at Devils River State Natural Area. The objectives of this study to describe the mam¬ malian fauna of the Devils River State Natural Area are: 1) to determine the occurrence of mammal spe¬ cies, 2) to estimate the relative abundance of small rodent species, 3) to determine habitat associations for species, 4) to determine seasonal patterns of abun¬ dance for species, and 5) to report on seasonal repro¬ ductive patterns of species. Materials and Methods The primary study sites at Devils River State Natural Area were selected on the basis of a report submitted to the Texas Parks and Wildlife Department (1990) on the vegetation of this area. That study des¬ ignated 13 sites as representative of the vegetation com¬ munities, For this study of the mammalian fauna, five of those 13 sites were selected as representative of the habitat types in the park, as follows: Riparian, Juniper- 4 Occasional Papers, Museum of Texas Tech University Oak Series shrubland, Cenizo Series shrubland, Lechuguilla-Sotol Series desert shrubland, and Curlymesquite-Sideoats Grama Series grassland. One primary sampling site was assigned to each habitat type and sampled seasonally. Primary sampling sites were used to gather data on reproductive status and relative abundance of rodents. The primary sampling sites (Fig, 2) are as follows: Juniper Slope, East Can¬ yon, Cenizo Slope, Grassland Plateau, and Dolan Springs. Supplementary sampling was carried out in addition to the primary sites to gain a comprehensive list of mammals in the area without depleting the pri¬ mary sample sites. Each sampling site was recorded using handheld global positioning system (GPS) re¬ ceivers (Garmin 12xls) with readings in latitude and longitude. The Juniper Slope site (29° 57.3’ N, 100° 57.9* W) is a Juniper - Oak Series shrubland, the most com¬ mon vegetation type encountered at Devils River State Natural Area (Fig. 4). The soil composition is gravelly colluvial loams with exposed bedrock near the top of the slope. East Canyon (29° 56.6* N, 100° 56. T W) is a south-facing slope that supports a xeric Lechuguilla - Sotol Series desert shrubland. The soil is composed of shallow clay loams between boulders and bedrock fractures. Exposed bedrock is common on this slope and creates striations across the slope. This slope was the steepest sampled at 33%. Cenizo Slope (29° 55.9* N, 100° 58.5* W) is very similar to Juniper Slope but faces southeast and the vegetation is Cenizo Series shrubland with fewer juni¬ pers (Fig. 6). The soil is similar to Juniper Slope ex¬ cept that there is little or no exposed bedrock. Grassland Plateau (29° 55.2* N, 100° 57.1* W) is a Curlymesquite - Sideoats Grama Series grassland, the least common vegetation type within DRSN A (Fig. 7). Grassland habitat is only found on ridge tops be¬ tween canyons and only where the area is great enough for soils to accumulate. The soil composition is mod¬ erately shallow, stony loams with no bedrock exposed. The primary sampling site at Dolan Springs (29° 53.8*N,100°59,rW) is a riparian habitat of Netleaf Hackberry - Little Walnut Series deciduous woodland {Fig. 8). The soil composition is silty alluvial loams distal to the springs and limestone bedrock adjacent to the springs. Standard techniques for sampling small to me¬ dium-sized species were utilized to survey the mam¬ malian fauna (Jones et al., 1996: Animal Care and Use Committee, 1998). Sherman live traps were used for rodents, Tomahawk live traps for medium-sized spe¬ cies, mist nets for bats, and pitfall traps for shrews. Large mammals and other species that are difficult to trap were sampled using spotlighting, firearms, and sightings. Voucher specimens were made for the first cap¬ ture of a species on a primary sampling site each sea¬ son. Other individuals of the same species captured during the same season on the same primary sampling site were identified in the field and released after mea¬ surements of tail length, hind foot length, and ear length were taken. These individuals were marked by cut¬ ting a line in the fur on the back and then released at the point of capture. Voucher specimens were pre¬ pared as museum study skins and skeletons and de¬ posited in the Angelo State Natural History Collections (ASNHC). Frozen tissue samples for most specimens were also deposited in the Collection of Frozen Tis¬ sues, ASNHC. Specimen records also were obtained from the following systematic collections: Angelo State Natural History Collections (ASNHC), University of Califor¬ nia, Berkeley (MVZ), Vertebrate Collection, Sul Ross State University (SRSU), Texas Cooperative Wildlife Collection at Texas A&M University (TCWC), Mu¬ seum of Texas Tech University (TTU), and National Museum of Natural History (USNM). Interviews were conducted with residents at DRSNA, Bill and Paula Armstrong and Jim Finegan, to determine the recent history of the area and establish species of mammals formerly occurring at DRSNA. Relative abundance was estimated using capture indices (Conroy, 1996) based on the number of indi¬ vidual animals captured per unit of time and effort (e.g., trap-night, net-hour). Primary sites were sampled using 200 trap-nights per season, for a total of 800 trap-nights per primary sampling site. Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 5 Figure 4* Juniper Slope primary sampling site at Devils River State Natural Area. Figure 5. East Canyon primary sampling site at Devils River State Natural Area. 6 Occasional Papers, Museum of Texas Tech University Figure 6, Cenizo Slope primary sampling site at Devils River State Natural Area. Figure 7. Grassland Plateau primary sampling site at Devils River State Natural Area. Brant and Dowler—The Mammals of Devixs River State Natuxeal Area, Texas 7 Figure 8. Dolan Springs primary sampling site at Devils River State Natural Area. Capture indices were analyzed using a canonical correspondence analysis to determine the effect of habitat on the relative abundance of species within the park (ter Braak, 1986). Environmental variables mea¬ sured were as follows: three vegetation variables (per¬ cent woody vegetation, percent succulent vegetation, and percent herbaceous vegetation), five soil variables (percent soil, percent gravel, percent cobblestone, per¬ cent boulder, percent bedrock), and slope. Measure¬ ment of these variables was performed using 50 m transects. For each transect the percentage of each variable was determined by the total distance (in meters) that the variable occurred divided by the total distance measured (50 m). Transects were surveyed season¬ ally starting in March 1999 and ending in November 1999. Each primary sampling site had transects for a minimum of two seasons. Species Accounts This study records the presence of eight orders and 18 families of mammals at Devils River State Natu¬ ral Area. The following accounts treat 39 species na¬ tive to DRSNA including three county records. All collection localities are recorded in relation to the Head¬ quarters Building (Fig. 2). Brant (2000) has maps showing locations of captures and sightings for each species. The sequence of species accounts conforms to the phylogenetic order presented in Nowak (1999). Scientific names follow Wilson and Reeder (1993), All subspecific designations are according to Hall (1981). Common names follow Davis and Schmidly (1994). ORDER DIDELPHIMORPHIA Family Didelphidae Didelphis Virginian a calif ornica Bennett, 1833 (Virginia Opossum): The opossum occurs in riparian areas of DRSNA but no specimens were collected during this study. Their absence could be a result of the drought this area has been experiencing for the 8 Occasional Papers, Museum of Texas Tech University past several years or a result of the flood that occurred in August 1998. Bill Armstrong, DRSNA manager, reported that few opossums are encountered in the area and that he has only seen four individuals since 1988 (pers. comm.). Scudday and Hanselka (1975) reported that the opossum is fairly common east of the Pecos River and may be encountered in any part of the area. Although no opossums were collected during this study, one individual was sighted crossing the creek bed of Dolan Creek two miles north and one mile east of the Headquarters Building on 30 October 1999. Two specimens in the Angelo State Natural History Collec¬ tions were previously collected from Jose Maria Springs and Dolan Springs on 24 October 1993. Specimens examined (2): Jose Maria Springs, 1 (ASNHC 9610); Dolan Springs, 1 (ASNHC 10973). ORDER XEN ARTHRA Family Dasypodidae Dasypus novemcinctus mexicanus Peters, 1864 (Nine-banded Armadillo): Scudday and Hanselka (1975) reported that the armadillo could be found any¬ where in the area but seemed to prefer riparian habi¬ tats. Two specimens were collected during this study and individuals were sighted throughout the area in all seasons. Juveniles were seen during August and Sep¬ tember at two locations. One male was collected in the Dolan Creek wash on 6 July 1998 and one female was collected near the Devils River on 10 September 1999. Specimens examined (2): % mi. S Headquarters Building (in Dolan Creek bottom), 1 (ASNHC 10651); 3 mi. S, 1 Vi mi. W Headquarters Building, along the Devils River (29° 53.7’ N, 100° 59.7’W), 1 (ASNHC 10974). ORDER INSECTIVORA Family Soricidae Notiosorex crawfordi crawfordi (Cones, 1877) (Desert Shrew): Boston (1966) reported the first speci¬ men of the desert shrew for Val Verde County. Since then only one other specimen (SRSU 3009) has been collected from the county. Scudday and Hanselka (1975) listed this species as possible for the area but unrecorded. Four specimens were collected during this study. Two were collected on a slope above the Plateau Live Oak grove next to Dolan Springs. One of these was hand caught on 15 June 1998, after it was discovered under a rock. The other was a skeleton found in a natural pitfall on 17 June 1998. Two more specimens were collected (on 8 May 1999 and 30 October 1999) in pitfall traps on the north side of DRSNA. None of the specimens collected during this study were repro¬ ductive ly active. Specimens examined (4): 2'A mi. N Headquarters Building (29° 58.4’ N, 100° 58.2 s W), 2 (ASNHC 10652-10653); 3 mi. S, 1 mi. W Headquarters Build¬ ing (29° 53.8 s N, 100° 59.1 s W), 2 (ASNHC 10656- 10657). ORDER CHIROPTERA Family Vespertilionidae Myotis velifer incautus (J. A. Allen, 1896) (Cave Myotis): The cave myotis is the only species of myotis known to occur at DRSNA. Fawcett Cave in the south¬ eastern comer of DRSNA (Fig. 2) supports a large colony during summer months and is possibly a hiber¬ nation site for this and other species of bats. Bat flights were witnessed from May to November with several hundred bats emerging from the cave. Emergence began just after dark and had durations of two hours. Sampling at this site was accomplished using a stan¬ dard hoop-net used for collecting insects. The cave mouth is only about three feet in diameter and descends about 30 to 40 feet. The last bat flight witnessed was on 1 November 1999 with an emergence of about 100 bats. On 14 November 1999 several bats were ob¬ served in flight inside the cave but none emerged. On 20 November 1999 no bat activity was observed. Deal and Fieseler (1975) mapped Fawcett Cave and showed the location of the bat colony on a detailed map. Myotis velifer is the only species listed as inhabiting this cave. Eleven specimens were collected at three sites in DRSNA. Four specimens were collected from Fawcett Cave, two males and two females, in July 1999. Four specimens were collected at Dolan Springs in mist nets set across the springs. A male and female Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 9 were collected in June 1998 and a female and scrotal male were collected in September 1999. Three males were collected from Jose Maria Springs on 23 Octo¬ ber 1993. Specimens examined (11): 1 mi. S, 1 mi. WHead¬ quarters Building, Jose Maria Springs, 3 (ASNHC 10663-10665); 2 mi. S, 3 V 2 mi. E Headquarters Build¬ ing, Fawcett Cave (29° 54.6’ N, 100° 54,7’ W), 4 (ASNHC 10666-10669); 3 mi. S, 1 mi. W Headquar¬ ters Building, Dolan Springs (29° 53.8’ N, 100° 59. T W), 4 (ASNHC 10670-10672, 10981). Additional records-. Dolan Springs, Fawcett Ranch, 5 (SRSU 1508, 1510-1512, 1514); Dolan Springs, 36 mi. N, 6 mi. W Del Rio, 2 (TTU 6166- 6167); Fawcett Cave 36 mi. N Del Rio, 10 (TTU 6579- 6581, 7240-7246). Pipistrellus hesperus maxim us Hatfield, 1936 (Western Pipistrelle): At DRSNAseven specimens of western pipistrelle from two localities were collected. Two females were collected on 17 June 1998 from a pool near Jose Maria Springs. Two males and one female were collected from Dolan Springs on 7 July 1998 and one male was collected from the same site on 17 April 1999. Another specimen was collected from Dolan Springs on 23 June 1998. One female collected from a pool near Jose Maria Springs on 16 April 1999 had two embryos with crown-rump lengths of 7 mm. Specimens examined (7): 1 mi. S, 1 mi. W Head¬ quarters Building, pool near Jose Maria Springs (29° 55.6’ N, 100° 59.3’ W), 2 (ASNHC 10673-10674); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8’N, 100° 59.D W), 5 (ASNHC 10675-10678, 10975), Additional records'. Dolan Springs, Fawcett Ranch, 2 (SRSU 1505, 1513); Dolan Falls, Fawcett Ranch, 2 (SRSU 1998, 2002). Nycticeius humeralis mexicanus Davis, 1944 (Evening Bat): The evening bat ranges throughout eastern and southern Texas reaching the western ex¬ tent of its range in the Hill Country region of Texas, Recent records of the evening bat suggest that it has been expanding the western limits of its range into the Trans Pecos and west-central Texas (Manning et ah, 1987; Dowler et ah, 1992; Dowler et ah, 1999). This species is mainly associated with woodland vegetation around or near streams or rivers (Schmidly, 1991). One specimen of the evening bat was collected at DRSNA. Dowler et al. (1999) reported the collec¬ tion of a male at Jose Maria Springs on 7 July 1998. This specimen is a county record and a range exten¬ sion of about 50 miles {80 km) from the nearest known records in Kinney County, Texas. Specimens examined (1): 1 mi. S, 1 mi. W Head¬ quarters Building, Jose Maria Springs (ASNHC 10105). Lasiurus cinereus cinereus (Palisot de Beauvois, 1796) (Hoary Bat): The hoary bat is a migratory spe¬ cies distributed across the state but few records exist for Val Verde County or the Edwards Plateau (Goetze, 1998). In Val Verde County one individual was col¬ lected in Del Rio near the Rio Grande River (Boston, 1966). Three specimens of the hoary bat were collected at DRSNA. All were collected in mist nets over Dolan Springs. One female was collected on 8 May 1999. One male with testes measuring 7 by 5 mm and one female were collected within minutes of each other on 12 September 1999. These specimens represent the 2 nd - 4 th specimens recorded in Val Verde County Al¬ though they did not collect any specimens, Scudday and Hanselka (1975) reported seeing this species at DRSNA. Specimens examined (3): 3 mi. S, 1 mi. W Head¬ quarters Building, Dolan Springs (29° 53.8’ N, 100° 59 .V W), 3 (ASNHC 10660-10662). Lasiurus seminolus (Rhoads, 1895) (Seminole Bat): The Seminole bat is an eastern species that reaches the limits of its range in east Texas. Recently it has expanded its range to include Burleson County (Lee, 1987), Fayette County (Yancey and Jones, 1996), and Travis County (Brant and Dowler, 2000). This species is often found in association with Spanish moss (Tillandsia usneoides) and prefers forested areas. They can also be found roosting singly behind loose tree bark or under foliage (Schmidly, 1991). 10 Occasional Papers, Museum of Texas Tech University One specimen was collected at DRSNA. A fe¬ male that was not reproductive!y active was taken at Dolan Springs on 10 September 1999, This specimen represents the westernmost record for this species occurring 200 miles (320 km) from the nearest known record for the Seminole bat. This is an important record for a species that was not expected to occur in Val Verde County, Texas. Brant and Dowler (2000) dis¬ cuss the possible explanations for the occurrence of this eastern forest species in a western arid environ¬ ment. Specimens examined( 1): 3 mi. S, 1 mi, W Head¬ quarters Building, Dolan Springs (29° 53.8' N, 100° 59.U W), 1 (ASNHC 10650). Antrozous pallidus pallidus (Le Conte, 1856) (Pallid Bat): This species reaches its eastern distribu¬ tional limits in Texas in the Edwards Plateau where records are sparse (Goetze, 1998). Three specimens were collected from DRSNA. A male and a female were netted at a pool near Jose Maria Springs on 10 September 1999. The male was scrotal with testes measuring 12 by 4 mm. A male was also collected from Dolan Springs on 17 April 1999, Specimens examined (3): 1 mi, S, 1 mi. W Head¬ quarters Building, pool near Jose Maria Springs (29° 55.6’ N, 100° 59,3’ W), 2 (ASNHC 10658, 10976); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8’ N, 100° 59. V W), 1 (ASNHC 10659). Additional records'. Dolan Falls, Fawcett Ranch, 1 (SRSU 2001); Dolan Falls, 0.5 mi. NW camp house, 1 (SRSU 1997); Dolan Springs, Fawcett Ranch, 2 (SRSU 1506-1507). Family Molossidae Tadarida brasiliensis mexicana (Saussure, 1860) (Brazilian Free-tailed Bat): This common bat in Texas has roosts in buildings associated with the DRSNA manager’s residence. None were collected with emerg¬ ing bats at Fawcett Cave. Tadarida brasiliensis are usually found in caves with domed ceilings, a mini¬ mum height of 10 m, and an entrance size of 5 by 10 m (Wilkins, 1989), which could explain their absence from Fawcett Cave. Seventeen specimens of T. brasiliensis were col¬ lected from DRSNA. Six specimens were collected from a pool near Jose Maria Springs on two occa¬ sions, five males on 10 September 1999 and one fe¬ male on 16 April 1999. Ten specimens were collected from Dolan Springs on six occasions in the months of April, June, August, and September. No reproduc- tively active individuals were recorded. Specimens examined (17): 1 mi. S, 1 mi. W Head¬ quarters Building, pool near Jose Maria Springs {29° 55.6’ N, 100° 59.3’ W), 6 (ASNHC 10679-10683, 10977); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8' N, 100<> 59. V W), 10 (ASNHC 10684-10690,10978-10980); Dolan Creek and Devils River, 1 (ASNHC 7606). Additional records : Dolan Falls, Fawcett Ranch, 2 (SRSU 1999-2000); Dolan Springs, Fawcett Ranch, 2 (SRSU 1509, 1519). ORDER CARNIVORA Family Canidae Urocyon cinereoargenteus scottii Meams, 1891 (Common Gray Fox): At DRSNA gray foxes were encountered on rocky slopes containing juniper-oak series vegetation, in rocky washes, and occasionally along the ridge tops. Three specimens of U. cinereoargenteus were collected at DRSNA, Two skulls were collected on different occasions from Jose Maria Springs and along the Devils River. A female that was not reproductively active was collected in a Toma¬ hawk live trap at the Juniper Slope primary sampling site on 7 July 1998. Several other gray foxes were sighted in the washes of Dolan Creek and along the canyon rims. Specimens examined (3): 1 mi. N, l A mi. E Head¬ quarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9'W), 1 (ASNHC 11139); 1 mi. S, 1 mi. W Head¬ quarters Building, Jose Maria Springs, 1 (ASNHC 10315); 3 mi. S, 1 Vi mi. W Headquarters Building, Devils River (29° 53.7’ N, 100° 59.7’ W), 1 (ASNHC 10982). Brant and Dowler—The Mammals of Devils River State Natural Area, Texas 11 Family Procyonidae Bassariscus astutus flavus Rhoads, 1894 (Ring¬ tail): Two specimens of the ringtail were collected at DR SNA, Anonscrotal male was collected on 7 Janu¬ ary 1999 at the East Canyon primary sampling site. A female without embryos was collected on 13 January 1999 at the Juniper Slope primary sampling site. An¬ other individual was sighted on the cliffs along the Devils River in September 1999. Paula Armstrong re¬ ported seeing a ringtail on the equipment bam near the DRSNA headquarters in June 2000 (pers. comm.). Specimens examined (2): 1 mi. N, l A mi. E Head¬ quarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9’ W), 1 (ASNHC 10692); l / 4 mi. N, 2 !4 mi. E Headquarters Building, East Canyon (29° 56.6’ N, 100° 56. l’W), 1 (ASNHC 10693). Procyon lotor fuscipes Mearns, 1914 (Common Raccoon): Raccoons were among the most common carnivores at DRSNA and were collected in all habitat types. Six specimens of P. lotor were collected at DRSNA. Two specimens were collected from Dolan Springs on separate occasions, 7 August 1998 and 13 January 1999. One female was collected from Dolan Creek in the northern portion of DRSNA on 28 March 1999. Two specimens were collected during May 1999, one male from the Juniper Slope primary sam¬ pling site on the 9 th and an individual from the southern ridge above East Canyon on the 20 th . A male was col¬ lected along the Dolan Creek Road just past campsite four on 12 September 1999. Many other individuals were sighted or captured then released throughout DRSNA. Two scrotal males were recorded in August 1998 and September 1999 with testes measurements of 23 by 15 mm and 30 by 26 mm respectively. No reproductively active females were recorded during this study. Specimens examined (6): 2 l A mi. N Headquar¬ ters Building (29° 58.4’ N, 100° 58.2’ W), 1 (ASNHC 11142); 1 mi. N, l A mi. E Headquarters Building, Juni¬ per Slope (29° 57.3’ N, 100° 57.9’ W), 1 (ASNHC 10694); !4 mi. S, 3 Vi mi. E Headquarters Building, 1 (ASNHC 10301); 1 !4 mi. S, 3 / 4 mi. W Headquarters Building, 1 (ASNHC 11143); 3 mi. S, 1 mi. W Head¬ quarters Building, Dolan Springs (29° 53.8’ N, 100° 59.l’W), 2 (ASNHC 10695, 11144). Additional records: Dolan Springs, 36 mi. N, 6 mi. W Del Rio, 1 (TTU 6665). Family Mustelidae Taxidea taxus berlandieri Baird, 1858 (Ameri¬ can Badger): Taxidea taxus ranges throughout Texas except in eastern portions of the state. It is most com¬ monly associated with open habitats such as grass¬ lands and deserts. Because badgers may avoid areas with shallow, stony soils (Goetze, 1998), the lack of habitat or deep soil at DRSNA make the probability of encountering this species low. No specimens of the badger were collected at DRSNA but one individual was sighted in the area on 29 July 1999 at about 2300 hrs. This individual was seen along the East Canyon Road as it ascends to the canyon rim. Occasionally, Bill Armstrong will find a badger hole on the property, but has never seen an individual (pers, comm,). Jim Finegan, who was raised in the area, has only seen two badgers in the area (pers.comm.). Lindzey (1982) suggested that young dispersing badgers often cross areas of habitat that seem unfavorable, which might account for the oc¬ currence of T. taxus in DRSNA. Family Mephitidae Spilogale gracilis leucoparia Merriam, 1890 (Western Spotted Skunk): The rocky slopes, washes, and canyons of DRSNA provide ample habitat for this species, which usually occurs in rocky areas in the Edwards Plateau (Goetze, 1998). One female S. gra¬ cilis was collected at DRSNA on 30 October 1999 at the Cenizo Slope primary sampling site. This speci¬ men represents the 4 th record for this species in Val Verde County. The nearest records are one specimen from Comstock collected in 1890 and two specimens from Langtry collected in 1903 (Bailey, 1905). Bos¬ ton (1966) reported seeing one near Sycamore Creek in southeastern Val Verde County. Specimens examined (1): 14 mi. S Headquarters Building, Cenizo Slope (29° 55.9’ N, 100° 58.5’ W), 1 (ASNHC 11141). Mephitis mephitis varians Gray, 1837 (Striped Skunk): Mephitis mephitis was found in all habitats at 12 Occasional Papers, Museum of Texas Tech University DRSNA. Four specimens of the striped skunk were collected from DRSNA. A male was collected on 22 October 1993 in the area. Two males were collected from near the Headquarters Building, one on 14 Janu¬ ary 1999 and the other on 21 November 1999. A fe¬ male was collected in the Grassland Plateau primary sampling site on 14 November 1999. Three other in¬ dividuals were sighted along Dolan Creek, Specimens examined (4): Headquarters Building (29° 56.4 s N, 10G o 58.3’W),2 (ASNHC11140,11165); 1 1/3 mi. S, 1 mi, E Headquarters Building, Grassland Plateau (29" 55.2’ N, 100" 57.1’ W), 1 (ASNHC 10691); Devils River State Natural Area, 1 (ASNHC 11166), Conepatus mesoleucus mearnsi Merriam, 1902 (Common Hog-nosed Skunk): Two specimens of C. mesoleucus were collected in DRSNA. A male was collected two miles south and one mile west of the Headquarters Building on 10 September 1999. A male was collected along the East Canyon Road in the East Canyon primary sampling site on 19 November 1999. Other individuals were sighted at the Grassland Pla¬ teau primary sampling site and near the airstrip in the southeastern comer of the area, Paula Armstrong has seen only one individual near her house in the last ten years (pers. comm.). Specimens examined (2): l A mi. N, 2!4mi. E Head¬ quarters Building, East Canyon (29° 56.6’ N, 100" 56.1 ’ W), 1 (ASNHC 11163); 2 mi. S, 1 mi. W Headquarters Building (29" 54.6’ N, 100" 59.4’ W), 1 (ASNHC 11164). Family Felidae Lynx rufus texensis J. A. Allen, 1895 (Bobcat): No specimens of L. rufus were collected at DRSNA. Two individuals were sighted at different locations in the park and bobcat sign, such as feces and tracks, were found near Dolan Springs. Bill Armstrong re¬ ported repeatedly sighting a female with two young along a ridge two miles north and one mile east of the Headquarters Building. In December 1999, he saw a bobcat along East Canyon Road (pers. comm.). Puma concolor Stanley ana (Goldman, 1936) (Mountain Lion): No specimens of P. concolor were collected at DRSNA, Scudday and Hanselka (1975) listed this species as occasionally passing through the area. On 15 June 1992, a female R concolor (ASNHC 9658) was collected from the Russell Ranch near Loma Alta, to the east of DRSNA. Hunters sighted an indi¬ vidual on the plateau between East Canyon and Rough Spring Canyon in November 1998. In July 1999 a felid scream was heard by J. G. Brant at about 2200 hrs, in the direction of the south rim of East Canyon. This scream seemed to be too deep for L . rufus but the identity of the source was not confirmed. Some visitors to DRSNA reported seeing a mountain lion next to the Bunk House in September 1999 (Bill Armstrong, pers, comm.), ORDER ARTIOD ACT YL A Family Tayassuidae Pecan tajacu angulatus (Cope, 1889) (Collared Peccary): Three specimens of P. tajacu were col¬ lected from DRSNA. Two were skulls found at Dolan Springs, one on 27 March 1999 and one on 19 No¬ vember 1999. Several individuals were sighted at Dolan Springs and a herd of about 15 individuals with three young was sighted crossing Dolan Creek Road two miles south and one mile west of the Headquarters Building in November 1999. Bill Armstrong reported that collared peccaries were abundant in 1988 with sightings concentrated around Dolan Creek. In the last four years he has not seen them and believes they are decreasing in abundance (pers. comm.). Accord¬ ing to Jim Finegan the population seemed to crash around 1989 or 1990. He only knows of two current groups, one in Oliver Canyon and one near Jose Maria Springs (pers, comm.). Specimens examined (3): 3 mi. S, 1 mi. W Head¬ quarters Building, Dolan Springs (29° 53.8’ N, 100" 59. U W), 2 (ASNHC 10696,10983); Devils River State Natural Area, 1 (ASNHC 9664). Family Cervidae Odocoileus virginianus texanus (Meams, 1898) (White-tailed Deer): No specimens of white-tailed deer were collected from DRSNA, but they were sighted in every habitat type. Annual deer hunts are a source of income for the property with two gun hunts and one bow hunt sold each year. About seven or eight white¬ tailed deer are harvested annually from the combined hunts (Bill Armstrong, pers.comm.). Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 13 ORDER RODENTIA Family Sciuridae Ammospermophilus interpres (Memam, 1890) (Texas Antelope Squirrel): No specimens of A . interpres were collected from DRSNA, Several individuals were sighted in East Canyon and in Dolan Creek north of the Headquarters Building in November 1998 and in the spring of 1999. Spermophilus mexicanus parvidens Meams, 1896 (Mexican Ground Squirrel): One female S. mexicanus was collected from DRSNA on 7 July 1998 while crossing the Dolan Creek Road. No other indi¬ viduals were sighted in DRSNA but one was sighted in May 1999 just outside the area on private land. Specimens examined (1): 1 mi, S, Headquarters Building (ASNHC 10697). Spermophilus variegatus grammurus (Say, 1823) (Rock Squirrel): In DRSNA rock squirrels are usually found in association with rocky slopes. Two specimens of S. variegatus were collected from DRSNA. One female was collected on 14 November 1998 and another female was collected from a differ¬ ent locality on 15 November 1998. One individual was sighted at the base of the East Canyon primary sampling site on 20 November 1999 and another indi¬ vidual was seen near the radio towers on the southern rim of East Canyon. Paula Armstrong reported seeing a rock squirrel near campsite six (pers. comm.). Specimens examined (2): 2 14 mi, N Headquar¬ ters Building (29° 58.4* N, 100° 58.2’ W), 1 (ASNHC 11146); % mi. SW Headquarters Building, 1 (ASNHC 10698). Additional records : Dolan Falls, 1 mi. W Dolan Falls, 1 (SRSU 1537). Sciurus niger limitis Baird, 1855 (Eastern Fox Squirrel): No specimens of fox squirrels were col¬ lected from DRSNA. One individual was sighted at Jose Maria Springs during the summer of 1998 and another individual was seen along the Devils River in June 1999. Bill Armstrong reported eastern fox squir¬ rels inhabiting the oak grove near Dolan Springs (pers. comm.). Additional records : 36 mi. N, 6 mi. W Del Rio, 1 (TTU 6671). Family Castoridae Castor canadensis mexicanus V. Bailey, 1913 (American Beaver): No specimens of the beaver were collected from DRSNA. Scudday and Hanselka (1975) do not list this species as occurring in the area but state that beavers occasionally occur in Dolan Creek and disappear after a major flood, Boston (1966) listed this species as occurring throughout Val Verde County in every major river and stream. The nearest record for C. canadensis is about 5 miles north of Del Rio on the Devils River. A stump that possibly could have been cut down by a beaver was found near Dolan Springs in June 1998. According to Bill Armstrong (pers. comm.), beavers were fairly prolific in DRSNA as recently as 1996. Beaver Lake is further upstream on the Devils River near the town of Juno. This im¬ plies that beaver were common in this stretch of the Devils River. Family Heteromyidae Perognathus merriami gilvus Osgood, 1900 (Merriam’s Pocket Mouse): At DRSNA this species was most commonly found in relatively loose, grav¬ elly soil. Eleven specimens of P, merriami were col¬ lected from DRSNA. Two specimens were collected from the Grassland Plateau primary sampling site on 7 January 1999 and 19 August 1999. One specimen was collected from the Juniper Slope primary sam¬ pling site on 9 May 1999. One specimen was col¬ lected at the Headquarters Building on 12 September 1999, Seven specimens were collected in funnel traps set in association with drift fences for a reptile survey from an area 214 miles north of the Headquarters Build¬ ing on several occasions. Specimens examined (11): 2 A mi. N Headquar¬ ters Building (29° 58.4’ N, 100° 58.2’ W), 7 (ASNHC 10720-10721, 11004-11008); 1 mi. N, ‘Arm. E Head¬ quarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9 s W), 1 (ASNHC 11009); Headquarters Building (29° 56.4 s N, 100° 58.3 S W), 1 (ASNHC 11010); 1 1/3 mi. S, 1 mi. E Headquarters Building, Grassland Pla¬ teau (29° 55.2 s N, 100° 57.1 s W), 2 (ASNHC 10721, 11011 ), 14 Occasional Papers, Museum of Texas Tech University Chaetodipus nelsoni canescens (Merriam, 1894) (Nelson’s Pocket Mouse): At DRSNAthis species is encountered in slope habitats. This species was the second most frequently encountered rodent at DRSNA with 41 specimens collected from the area, Chaetodipus nelsoni was the second most abundant rodent at each primary sampling site with the excep¬ tion of Grassland Plateau, where it was not encoun¬ tered. Specimens were collected in every season with this species being the most abundant summer rodent at DRSNA. Scudday and Hanselka (1975) listed this species as possibly occurring in the area but unveri¬ fied. Reproductive individuals were recorded from January to July with three scrotal males collected from January to June and six reproductive females collected from June to July Specimens examined (41): 1 mi. N, 14 mi. E Head¬ quarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9’ W), 4(ASNHC 10699-10700,10984-10985); !4 mi. N, 214 mi. E Headquarters Building, East Canyon (29° 56.6’ N, 100° 56.1 ’ W), 3 (ASNHC10701,10986- 10987); 14 mi. S Headquarters Building, Cenizo Slope (29°55.9’N, 100° 58.5’ W), 3 (ASNHC 10704,10988- 10989); 1 mi. S Headquarters Building, 6 (ASNHC 10702-10703, 10707-10708, 10717, 10990); 1 mi. S, 1 mi. W Headquarters Building, Jose Maria Springs (29° 55.6’N, 100° 59.3’ W), 3 (ASNHC 10705-10706, 11003); 214 mi. S, 1 Vi mi. W Headquarters Building, 3 (ASNHC 10709-10710,10991); 3 mi. S, 1 mi. W Head¬ quarters Building, Oak Grove Campground, 6 (ASNHC 10992-10997); 3 mi. S, 1 mi. W Headquarters Build¬ ing, Dolan Springs (29° 53.8’ N, 100° 59.1 ? W), 11 (ASNHC 10711-10716, 10998-11002); Devils River State Natural Area, 2 (ASNHC 10718-10719). Additional records-. 36 mi. N, 6 mi. W Del Rio, 1 (Jones and Manning, 1991). Family Muridae Neotoma albigula albiguta Hartley, 1894 (White-throated Woodrat): Only one specimen of the white-throated woodrat was recorded from DRSNA during this study. A subadult female was collected on Dolan Creek Road on the night of 11 September 1999, Scudday and Hanselka (1975) list this species as prob¬ ably occurring in the area but unverified. Specimens examined (1): 214 mi. S Headquar¬ ters Building (ASNHC 11147). Additional records : 36 mi, N, 6 mi. W Del Rio, 1 (TTU 6909). Peromyscus maniculatus blandus Osgood, 1904 (Deer Mouse): Six specimens of the deer mouse were collected from DRSNA. One male was collected from Jose Maria Springs on 18 April 1999. Three speci¬ mens were collected at Dolan Springs, one female in June 1998 and two scrotal males in November and January 1999. The county record for this species was collected by Boston (1966) at several localities in Val Verde County including near Loma Alta. Specimens examined (6): 2 14 mi. N Headquar¬ ters Building (29° 58.4’ N, 100° 58.2’ W), 1 (ASNHC 11015); 2 mi. N Headquarters Building (29° 58.3’ N, 100° 58.2’W), 1 (ASNHC 11013); 1 mi. S, 1 mi. W Headquarters Building, Jose Maria Springs (29° 55.6’ N, 100° 59.3’ W), 1 (ASNHC 11014); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8’ N, 100° 59.T W), 3 (ASNHC 10730-10731, 10780). Peromyscus leucopus texanus (Woodhouse, 1853) (White-footed Mouse): Twelve specimens of the white-footed mouse were collected from DRSNA, One individual was collected from the Grassland Pla¬ teau on 7 January 1999. One male was collected from the Cenizo Slope primary sampling site on 23 June 1998. Ten specimens were collected from Dolan Springs on the limestone bedrock near brush. Six scro¬ tal males with testes measuring larger than 11 by 8 mm were collected from August to January. Two pregnant females both with four embryos were re¬ corded in August and November. Scudday and Hanselka (1975) listed this species as possibly occur¬ ring in the area but unverified. Specimens examined (12): 14 mi. S Headquar¬ ters Building, Cenizo Slope (29° 55.9’ N, 100° 58.5’ W), 1 (ASNHC 10769); 1 1/3 mi. S, 1 mi. E Head¬ quarters Building (29° 55.2’ N, 100° 57.1’ W), 1 (ASNHC 10778); 3 mi. S, 1 mi. W Headquarters Build¬ ing, Dolan Springs (29° 53.8’ N, 100° 59.1’ W), 10 (ASNHC 10723-10729, 11012, 11017, 11149), Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 15 Peromyscus pectoralis lacianus V. Bailey, 1906 (White-ankled Mouse): The white-ankled mouse was the most commonly encountered rodent at DRSNA with SO specimens collected from the area. It was the most abundant rodent at every primary sampling site except for Dolan Springs where it was third in abun¬ dance. Peromyscus pectoralis were collected in every season with this species being the most abundant ro¬ dent in autumn, winter, and spring. These findings are in agreement with Scudday and Hanselka (1975), Twenty-one scrotal males with testes measuring larger than 11 by 7 mm were recorded from October to March. Three females with an average of four em¬ bryos and two females with placental scars were re¬ corded from January to March. Non-reproductive in¬ dividuals were recorded from November to June. Specimens examined (80): 2Ya mi. N, Ya mi. E Headquarters Building (29° 58.4’ N, 100° 56.9’ W), 7 (ASNHC 10732-10736,11020,11150); 1 mi. N, 'Ami. E Headquarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9’ W), 10 (ASNHC 10737-10742, 11025- 11027,11151); Ami.N, 2%mi. E Headquarters Build¬ ing, East Canyon (29° 56.6’ N, 100° 56.1’ W), 17 (ASNHC 10743-10754,10771,11019, 11030, 11152- 11153); Ya mi. N, IVa mi. E Headquarters Building, Vegetation Site 9, 2 (ASNHC 11040-11041); 'Ami. N, Ya mi. E Headquarters Building, Airstrip (29° 56.T N, 100°57.4’ W), 1 (ASNHC 11018); Ami. SHeadquar¬ ters Building, Cenizo Slope (29° 55.9’ N, 100° 58.5’ W), 20 (ASNHC 10755-10769, 10770, 11031-11032, 11154); 1 mi, S, A mi. W Headquarters Building, Veg¬ etation Site 1, 2 (ASNHC 11038-11039); lmi. S, A mi. W Headquarters Building, Vegetation Site 12, 5 (ASNHC 11033-11037); 1 mi. S, 1 mi. W Headquar¬ ters Building, Jose Maria Springs (29° 55.6’ N, 100° 59.3’ W), 3 (ASNHC 11021, 11024); 1% mi. S, VA mi. E Headquarters Building, Campsite A, 2 (ASNHC 10777-10778); 1 1/3 mi. S, 1 mi. E Headquarters Build¬ ing, Grassland Plateau (29° 55.2 s N, 100° 57.1 s W), 7 (ASNHC 10772-10776, 11155-11156); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8’ N, 100° 59.1 s W), 4 (ASNHC 10779, 11028-11029, 11157); Devils River State Natural Area, 2 (ASNHC 10781-10782). Additional records : 36 mi. N, 6 mi. W Del Rio, 1 (TTU 6916); Dolan Creek, Along Dolan Creek, 2 (SRSU 1861-1862); Dolan Falls, Fawcett Ranch, 1 (SRSU 1889); Dolan Springs, 4 (SRSU 1515-1518). Reithrodontomys mo tt tan us griseus V. Bailey, 1905 (Plains Harvest Mouse): One specimen of the plains harvest mouse was collected at DRSNA, Anon- scrotal male was collected from the Grassland Plateau primary sampling site on 7 January 1999. The grass¬ land habitat type is the only suitable habitat at DRSNA in which this species would be expected. This speci¬ men represents the county record for R. montanus in Val Verde County, Texas, and is a range extension of more than 20 miles (>32 km) from the nearest record in Edwards County, Texas. Specimens examined (1): 1 1/3 mi. S, 1 mi. E Headquarters Building, Grassland Plateau (29° 55.2 J N, 100° 57.1 s W), 1 (ASNHC 10784). Reithrodontomys fulvescens lacyei J. A. Allen, 1896 (Fulvous Harvest Mouse): One nonscrotal male R. fulvescens was collected at Dolan Springs on 20 November 1999. This specimen was caught in a Sherman live trap that was set in a bush about 50 cm off the ground. This species was not listed in Scudday and Hanselka (1975) as expected in the area. The only other known records of this species in Val Verde County are two specimens (USNM 18355, 18358) collected in February 1890 from Del Rio (Bailey, 1905), Specimens examined (1): 3 mi. S, 1 mi, W Head¬ quarters Building, Dolan Springs (29° 53.8 s N, 100° 59.1 s W), 1 (ASNHC 10783). Sigmodon hispidus berlandieri Baird, 1855 (His¬ pid Cotton Rat): Five specimens of S. hispidus were collected from DRSNA, all from a single locality. Three males and two females were collected from Dolan Springs over four separate occasions, 8 August 1998, 13 January 1999, 25 May 1999, and 20 November 1999. One scrotal male and one female with five em¬ bryos were recorded during August. Specimens examined (5): 3 mi. S, 1 mi. W Head¬ quarters Building, Dolan Springs (29° 53.8 s N, 100° 59.1 s W), 5 (ASNHC 10785-10789). 16 Occasional Papers, Museum of Texas Tech University Family Erethizontidae Erethizon dorsatum couesi Mearns, 1897 (Por¬ cupine): Two specimens of E . dorsatum were col¬ lected from DRSNA. One road kill was collected from 1 A mile north of the Headquarters Building on the Dolan Creek Road in June 1998. Another individual was trapped at the Juniper Slope primary sampling site on 12 January 1999. Other porcupines were sighted in the area, mostly around Dolan Creek. This species has been increasing in abundance since it was first sighted on the property by Jim Finegan and his grand¬ father around 1964 (pers.comm.). Specimens examined (2): 114 mi. N Headquar¬ ters Building, byroad, 1 (ASNHC 10790); 1 mi. N, % mi. E Headquarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9’ W), 1 (ASNHC 11044). ORDER LAGOMORFHA Family Leporidae Sylvilagus floridanus chapmani (J. A. Allen, 1899) (Eastern Cottontail): Twelve specimens of S. floridanus were collected at DRSNA. A male and a pregnant female with two embryos (crown-rump length of 60mm) were collected from Jose Maria Springs on 6-7 July 1998. A scrotal male with testes measuring 54 by 14 mm was collected from l A mile south and three miles east of the Headquarters Build¬ ing on 19 November 1999. A female was collected from the Cenizo Slope primary sampling site on 8 Au¬ gust 1999 and a scrotal male was collected from that site on 14 November 1998. A scrotal male was col¬ lected from the Grassland Plateau primary sampling site on 13 January 1999. Specimens examined (12): 14 mi. S Headquarters Building, Cenizo Slope (29° 55.9’ N, 100“ 58.5’ W), 5 (10314, 10793, 11050-11051, 11064); A mi. S, 3 3 A mi. E Headquarters Building (29° 53.8’ N, 100° 54.1’ W), 3 (ASNHC 10795, 11052, 11063); 1 mi. S, 1 mi. W Headquarters Building, Jose Maria Springs (29° 55.6’ N, 100° 59.3’ W), 1 (ASNHC 10792); 1 1/3 mi. S, 1 mi. E Headquarters Building, Grassland Plateau (29° 55.2’ N, 100° 57.1’ W), 1 (ASNHC 10794); 1 Vi mi. S, 3 A mi. W Headquarters Building, Campsite 2 (29“ 55.5’ N t 100° 59.1 s W), 1 (ASNHC 11053); 3 mi. S, 1 mi. W Headquarters Building, Dolan Springs (29° 53.8’ N, 100° 59.1’ W), 1 (ASNHC 11062). Sylvilagus auduboniiparvulus (J. A. Allen, 1904) (Desert Cottontail): One desert cottontail was col¬ lected from DRSNA. A subadult male was collected from Jose Maria Springs on 7 July 1998. Scudday and Hanselka (1975) reported this species as occur¬ ring in the Dolan Creek area but no voucher speci¬ mens could be located. Field identification of cotton¬ tails was based on a ratio of ear length to hind foot length (Davis and Schmidly, 1994) and recognized five specimens as S. audubonii. Subsequent examination of skulls revealed the presence of only one representa¬ tive of this species. This specimen’s identity is fur¬ ther complicated by the fact that it is a subadult. The use of ear to hind foot length ratios did not consis¬ tently provide the correct identification of cottontails. Measurements from dried skins were more reliable than those from specimen tags; however, neither were in¬ fallible for correct identification of specimens. Specimens examined (1): 1 mi. S, 1 mi, W Head¬ quarters Building, Jose Maria Springs (29° 55,6’ N, 100“ 59.3’ W), 1 (ASNHC 10791). Lepus californicus texianus Waterhouse, 1848 (Black-tailed Jackrabbit): Jackrabbits were observed in all sampled areas of DRSNA and four specimens were collected. One specimen was collected from the Headquarters Building on 22 June 1998. A specimen from the primary sampling site Grassland Plateau was collected on 7 January 1999. One specimen was col¬ lected near Juniper Slope on 11 September 1999. On 19 November 1999 a specimen was collected from the ridge top near the radio towers. Other individuals were sighted in many areas of DRSNA. Specimens examined (4): 1 mi. N, 'Ami, E Head¬ quarters Building, Juniper Slope (29° 57.3’ N, 100° 57.9’ W), 1 (ASNHC 11046); Headquarters Building (29°56,4’N, 100°58.3’W), 1 (ASNHC 11047); Ami. S, 214 mi. E Headquarters Building, 1 (ASNHC 11048); 1 1/3 mi. S, 1 mi. E Headquarters Building, Grassland Plateau (29° 55.2’ N, 100“ 57.1’ W), 1 (ASNHC 11049). Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 17 Introduced Species Eight species of non-native mammals represent¬ ing two orders occur at DRSNA. The previous own¬ ers of the property intentionally introduced the seven representatives of the order Artiodactyla in the 1970s (J. Finegan, pers. comm.). The one representative of the order Rodentia is an accidental introduction result¬ ing from human activity. There was no evidence of any other introduced species in the area, including fe¬ ral cats and feral hogs. ORDER ARTIODACTYLA Family Cervidae Dama dama (Linnaeus, 1758) (Fallow Deer): A native of southern Europe and the Mid-east, the fallow deer has been introduced into 93 Texas counties (Davis and Schmidly, 1994). Fallow deer were introduced into DRSNA in the mid-1970s in an effort to establish exotic game hunting (Scudday and Hanselka, 1975). According to Bill Armstrong (pers. comm.) these deer did not last long. The environment was too harsh for this species and there was not enough cover for them. Axis axis (Erxleben, 1777) (Axis Deer): A native of India, the axis deer was introduced into Texas in the 1930s. Davis and Schmidly (1994) reported herds of this species in 94 counties in Texas, making it the most abundant exotic ungulate in Texas. There is a population of A, axis in East Canyon that was intro¬ duced in the 1970s (Scudday and Hanselka, 1975). This population consists of about 20 individuals and was sighted on a number of occasions. They prob¬ ably spend most of their time at the mouths of Speed Canyon, East Canyon, and Escondido Canyon. Cervus nippon Temminck, 1838 (Sika Deer): Originally from oriental countries such as China, Ja¬ pan, Korea, and Manchuria, the sika deer has been introduced into 77 Texas counties (Davis and Schmidly, 1994). Sika deer have persisted at DRSNA since they were introduced in the mid-1970s (Scudday and Hanselka, 1975). No individuals were sighted during this study and Bill Armstrong has not seen any since 1996 (pers. comm.). This species most likely no longer occurs in the area with the possible exception of one or two stragglers. Cervus elaphus Linnaeus, 1758 (Wapiti or Elk): The wapiti is native to Europe, Asia, and the Rockies of North America. They were introduced into the Trans-Pe^os in 1928 and currently there are herds in five Trans-Pecos counties (Davis and Schmidly, 1994). Elk were introduced into DRSNA in the 1970s and persisted for almost 20 years. Bill Armstrong reports that the last of the elk population was extirpated in 1991 (pers. comm.). Family Bovidae Antilope cervicapra (Linnaeus, 1758) (Blackbuck): Blackbucks are native to Pakistan and India and were introduced into Texas in 1932 on the Edwards Plateau (Davis and Schmidly, 1994). They were introduced into DRSNA in the mid-1970s in an effort to establish exotic game hunting (Scudday and Hanselka, 1975). This species persisted until 1991 when the last of the DRSNA population was extir¬ pated (Bill Armstrong, pers, comm,). Ammotragus lervia (Pallas, 1777) (Barbary Sheep): The barbary sheep is endemic to North Africa but has been introduced into the Llano Estacado, Trans- Pecos, Edwards Plateau, and southern Texas (Davis and Schmidly, 1994). They were introduced to DRSNA at about the same time as the axis deer in the 1970s (Scudday and Hanselka, 1975) and still occur in the area near cliffs and in areas with high relief. A skull of A. lervia was collected from DRSNA and barbary sheep were sighted on the plateaus between canyons and near the top of steep slopes or cliffs. Like the white-tailed deer, hunting permits are issued for this species and make up a large portion of the income for the area. Bill Armstrong estimated that between 20 and 30 barb ary sheep are harvested each year (pers. comm.). Specimens examined (1): 'A mi. S, 3 mi. E Head¬ quarters Building (ASNHC 11167). Ovis orientalis Linnaeus, 1758 (Mouflon): Mou- flon are native to Turkey, the Middle Eastern coun¬ tries, and the European islands of Corsica and Sardinia (Nowak, 1999). Scudday and Hanselka (1975) re¬ ported that mouflon were introduced in the 1970s to 18 Occasional Papers, Museum of Texas Tech University stimulate hunting in the area. Bill Armstrong (pers, comm.) says that there was a relatively large popula¬ tion at DRSNA but he has not seen them for several years. Hunters killed two rams during the November 1999 hunt. ORDER RODENT1A Family Myocastoridae Myocastor coypus (Molina, 1782) (Nutria): The nutria ranges throughout eastern, southern, and central Texas and reaches its western distributional limits in the Trans-Pecos, west-central, and north-cen¬ tral Texas. They were introduced to clean up over¬ vegetated lakes and as a possible furbearer (Davis and Schmidly, 1994). No specimens ofM. coypus were collected from DRSNA. One individual was sighted on the Devils River. Several people visiting DRSNA to fish reported hearing nutria along the banks in the evening. Scudday and Hanselka (1975) also reported this species occurring along the Devils River. Species of Unverified Occurrence Nineteen species of mammals representing four orders have been recorded from Val Verde County, Texas but have not been recorded from DRSNA. Fu¬ ture research efforts might reveal the presence of these species at DRSNA. Cryptotis parva (Say, 1823) (Least Shrew): The least shrew ranges throughout eastern and southern Texas and most of the Panhandle (Davis and Schmidly, 1994). It reaches its western distributional limits in west-central Texas (Simpson and Maxwell, 1989; Dowlerand Boyd, 1996) and southwest Texas (Bailey, 1905) but does not occur on the Edwards Plateau (Goetze, 1998). This shrew prefers grassland habi¬ tats and might occur in the grassland areas between canyons; however, the lack of continuous suitable habi¬ tat from the nearest known record in Del Rio (Bailey, 1905) makes it unlikely that this species occurs in DRSNA. Myotis yumanensis (H. Allen, 1864) (Yuma Myotis): The Yuma myotis occurs throughout the Rio Grande valley and Trans-Pecos as a migrant species. A single record from Starr County, Texas, justifies the extent of the range of this species into south Texas (Schmidly, 1991). This species is most commonly associated with watercourses and prefers lower el¬ evations. In Texas it mostly has been collected near the Rio Grande (Schmidly, 1977) but the range of the species does include DRSNA. Scudday and Hanselka (1975) include this species as possibly occurring at DRSNA in their survey of the area. The cliffs along the Devils River provide ample roost sites for this spe¬ cies. The closest records of the Yuma myotis to DRSNA are from Del Rio and the mouth of the Pecos (Schmidly, 1991). Pipistrellus subjlavus (F. Cuvier, 1832) (East¬ ern Pipistrelle): Eastern pipistrelles range throughout eastern and southern Texas and reach the extent of their range in the Panhandle (Schmidly, 1991) and west- central regions of Texas (Dowler et ah, 1992). This species has recently expanded its range into the Trans- Pecos (Yancey et al., 1995) and currently includes DRSNA. Pipistrellus subflavus most likely would occur near Dolan Springs where it has ample roost sites in the surrounding trees with thick growths of ball moss (Tillandsia recurvirostra), which is related to Spanish moss (Tillandsia usneoides ), a preferred roost site for the eastern pipistrelle (Menzel et al., 1999). The closest records are from Del Rio and the Rio Grande River (Schmidly, 1991). Lasiurus borealis (Muller, 1776) (Eastern Red Bat): The eastern red bat is a migratory species dis¬ tributed across Texas (Schmidly, 1991), On the Edwards Plateau they are most commonly found in riparian and riverine areas (Goetze, 1998). Their range includes DRSNA and there are records for just north of Del Rio on the Devils River (Schmidly, 1991), and therefore it is likely that they are in the area. They would probably occur at Dolan Springs or Jose Maria Springs where there are large trees for roosting. Plecotus townsendii Cooper, 1837 (Townsend’s Big-eared Bat): Townsend’s big-eared bat occurs Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 19 throughout the Trans-Pecos, Panhandle, and Edwards Plateau regions of Texas (Schmidly, 1991). In the Trans-Pecos it is most commonly found in rocky, can¬ yon areas where there are plenty of crevices and caves for roosting (Schmidly, 1977), All of DRSNAis well within the range of this species and provides plenty of appropriate habitat, so it is reasonable to expect this species to occur there. The closest record of this species (TTU 6921) occurs just 3 mi. S, 12 mi. E of DRSNA between this area and Loma Alta (Schmidly, 1991). Eumops perotis (Schinz, 1821) (Western Mas¬ tiff Bat): Eumops perotis ranges along the Rio Grande River in the Trans-Pecos and into Val Verde County. They prefer roost sites in rocky cliffs and canyons and need a large (>2 m) unobstructed drop from their roost site in order to take flight (Schmidly, 1991). The topography along Dolan Creek and the Devils River provides ample roost sites for this species. DRSNA is at the northern distributional limit of this species but does have suitable habitat Scudday and Hanselka (1975) listed this rare species as possibly occurring in the area due to a record from Del Rio. This species has also been taken in Val Verde County from Langtry (Schmidly, 1991). Cams latrans Say, 1823 (Coyote): This species has never been reported from DRSNA despite other records from Val Verde County. Bill Armstrong esti¬ mates that they are absent for about 20 miles (32 km) in any direction (pets, comm.), Boston (1966) re¬ ported that the coyote is not common in eastern Val Verde County due to extensive trapping by ranchers. The nearest record for this species is a specimen col¬ lected in 1905 near Del Rio (Bailey, 1905), Mephitis macroura Lichtenstein, 1832 (Hooded Skunk): The hooded skunk occurs in the southern portion of the Trans-Pecos reaching its northern dis¬ tributional limits in Ward County, Texas (Davis and Schmidly, 1994). This species most likely does not occur at DRSNA, Boston (1966) reported sighting a hooded skunk in the Edwards Plateau portion of north¬ ern Val Verde County, but no specimens have been recorded from the county. The nearest records for this species are over 100 miles (160 km) away in Pecos and Brewster counties (Schmidly, 1977). Spermophitus spilosoma Bennett, 1833 (Spot¬ ted Ground Squirrel): Tire spotted ground squirrel occurs throughout the Panhandle and Trans-Pecos reaching its eastern distributional limits on the Edwards Plateau, Rolling Plains, and southern Texas (Davis and Schmidly, 1994). This species probably does not oc¬ cur in DRSNA, Neither Boston (1966) nor Scudday and Hanselka (1975) listed this species as occurring in this area. Hall (1981) recorded a specimen from the mouth of the Pecos River. Thomomys bottae (Eydoux and Gervais, 1836) (Botta’s Pocket Gopher): Botta's pocket gopher ranges throughout the Trans-Pecos and reaches its eastern distributional limits on the Edwards Plateau (Davis and Schmidly, 1994). Scudday and Hanselka (1975) listed this species as possibly occurring at DRSNA but fail¬ ure to find any sign of this species after a concen¬ trated effort suggests that it may not occur here. Bos¬ ton (1966) recorded this species from south of Loma Alta on Highway 277 and near Juno. Chaetodipus hispidus (Baird, 1858) (Hispid Pocket Mouse): The hispid pocket mouse ranges throughout Texas with the exception of approximately eight counties in extreme eastern Texas (Davis and Schmidly, 1994). This species would most likely be encountered on the ridge tops between the many can¬ yons at DRSNA, The closest records are from Del Rio and Comstock. Scudday and Hanselka (1975) and Boston (1966) did not list the hispid pocket mouse as occurring in the area. Chaetodipus eremicus (Mearns, 1898) (Chihuahuan Desert Pocket Mouse): The Chihuahuan Desert pocket mouse ranges throughout the Trans- Pecos reaching its eastern distributional limits on the Edwards Plateau and Permian Basin (Davis and Schmidly, 1994). Lee et al. (1996) elevated the popu¬ lations of C. penicillatus eremicus , the scientific des¬ ignation of this species since 1900, in Texas to the specific level. The presence of this species at DRSNA is unlikely. Davis and Schmidly (1994) stated that the Chihuahuan Desert pocket mouse has not been en¬ countered in rocky or gravelly soils. The streambeds of Dolan Creek and the Devils River are both com¬ posed of primarily boulders, rocks, and gravel, Bos¬ ton (1966) and Scudday and Hanselka (1975) did not 20 Occasional Papers, Museum of Texas Tech University include this species as occurring in the area. The two records of the Chihuahuan Desert pocket mouse for Val Verde County were collected from Del Rio and the Pecos River at the beginning of the century (Bailey, 1905). Dipodomys merriami Meams, 1890 (MerrianTs Kangaroo Rat): Merriam’s kangaroo rat ranges throughout the Trans-Pecos reaching its eastern dis¬ tributional limits on the Edwards Plateau and Llano Estacado (Davis and Schmidly, 1994). This species most likely does not occur at DRSNA. Scudday and Hanselka (1975) did not list Merriam’s kangaroo rat as a possibly occurring species, nor did Boston (1966) list this species as occurring east of the Pecos River in Val Verde County. The two records of this species for Val Verde County were collected in 1903 from Langtry (Bailey, 1905). Neotoma micropus Baird, 1855 (Southern Plains Woodrat): The southern plains woo drat ranges throughout western, southern, and north-central Texas (Davis and Schmidly, 1994). Where Neotoma micropus and N. albigula are sympatric, this species is usually found in valleys between rocky slopes (Goetze, 1998). Habitat exists to support this species along Dolan Creek and in the canyon bottoms of East Canyon. The area at the mouths of East Canyon, Speed Canyon, and Escondido Canyon also would be suitable for this spe¬ cies. The closest records of this species are from near the Devils River in Amistad National Recreation Area (Ditton and Schmidly, 1977) and near Del Rio (Bailey, 1905; Boston, 1966). Baiomys taylori (Thomas, 1887) (Northern Pygmy Mouse): The northern pygmy mouse ranges throughout Texas with the exception of the Trans- Pecos (Davis and Schmidly, 1994). This species would most likely be encountered around Dolan Springs. Both Sigmodon hispidus and Reithrodontomys fulvescens were collected at Dolan Springs and are often found in connection with B. taylori (Davis and Schmidly, 1994). The closest record for this species is from near Juno, about 15 miles (24 km) northwest of DRSNA (Goetze, 1998). Onychomys leucogaster (Wied-Neuwied, 1841) (Northern Grasshopper Mouse): The northern grass¬ hopper mouse ranges throughout western Texas reach¬ ing its eastern distributional limits in north-central Texas, southern Texas, and the Edwards Plateau (Davis and Schmidly, 1994). The probability of encountering this species at DRSNA is low, as its the preferred habi¬ tat of sandy soils or open brushland is either lacking or patchy in this part of Val Verde County. Scudday and Hanselka (1975) and Boston (1966) did not list this species as possibly occurring in the area. The nearest records are from Comstock in 1890 (Bailey, 1905). Peromyscus eremicus (Baird, 1858) (Cactus Mouse): The cactus mouse ranges throughout the Trans-Pecos reaching its eastern distributional limits along the Pecos River and the lower Rio Grande valley (Davis and Schmidly, 1994). This species would most likely be encountered in DRSNA at the base of canyon walls or on cliff faces. Scudday and Hanselka (1975) did not list this species as occurring in the area. The records for Val Verde County are from the turn of the century near Comstock and Langtry (Bailey, 1905). Mus musculus Linnaeus, 1758 (House Mouse): No specimens of this introduced species were col¬ lected at DRSNA. Scudday and Hanselka (1975) listed this species as occurring in the area but no voucher specimens could be located. The closest records of this species are from Del Rio (Boston, 1966). Rattus rattus (Linnaeus, 1758) (Roof Rat): It is unlikely that black rats occur at DRSNA due to the paucity of human structures in the area. Scudday and Hanselka (1975) did not list this species as possibly occurring in the area. Boston (1966) recorded that this species was only found within the city limits of Del Rio. Brant and Dowler—The Mammals of Devils River State Natural Area, Texas 21 Community Interactions Capture indices for bats were estimated using the number of bats captured divided by the number of net-hours that mist nets were operating (Table 1). Dolan Springs and a pool near Jose Maria Springs were the only two sampling sites used in this analysis. Fawcett Cave was excluded because mist nets or other constant effort methods were not used to sample this site. The pool near Jose Maria Springs is located one mile south and one mile west of the Headquarters Build¬ ing (29° 55.6’ N, 100° 59.3* W). A total of 94 net- hours were sampled at this site. Three species were captured at this site: Tadarida brasHiensis, PipistreUus hesperus , and Antrozous pallidus (Table 1), Both 71 brasiliensis and P hesperus were the most abundant bats at this site with a capture index of 0.032. Antrozous pallidus had a capture index of 0.021. Dolan Springs was sampled for a total of 72.5 net-hours. Four species were captured at this site: Tadarida brasiliensis, Myotis velifer , Lasiurus cinereus ^ and Las turns seminolus (Table 1). Tadarida brasiliensis was the most abundant bat at this site with a capture index of 0.097, followed by M. velifer , L. cinereus , and L. seminolus. The total capture indices for the year were esti¬ mated using 166.5 net-hours (Fig, 9). Tadarida brasiliensis is the most abundant bat species occur¬ ring at DRSNA with a capture index of 0.060. Myotis velifer , P. hesperus, and L. cinereus each have a cap¬ ture index of 0.018. Antrozous pallidus and L. seminolus are fifth and sixth in abundance with cap¬ ture indices of 0.012 and 0.006 respectively. With the inclusion of the bats at Fawcett Cave, the most abun¬ dant bat would be M. velifer due to the roosting colony located inside the cave. A total of 197 captures were recorded for ro¬ dents from the five primary sampling sites at DRSNA. Total trap success for these sites was 4.9% out of 4000 trap nights. Capture indices were calculated for each species by primary sampling site for the year using 800 trap nights (Table 2). Dolan Springs was the most diverse site within DRSNA with a total of six species of small rodents (Table 2). Peromyscus leucopus was the most abun¬ dant small rodent at this site with a capture index of 0,024 followed by Chaetodipus nelsoni, Peromyscus pectoralis i Sigmodon hispidus, Peromyscus maniculatus , and Reithrodontomys fulvescens . Table 1, The capture indices for each species of bat from Dolan Springs, near Jose Maria Springs, and the total for both sites together. Bold-faced numbers represent the capture index for the most abundant species per sampling site and numbers in parentheses represent the total number of captures for that species per site. Dolan Springs Jose Maria Springs Total Tadarida brasiliensis 0.097 (7) 0.032 (3) 0.060 (10) Myotis velifer 0.041 (3) 0.018(3) PipistreUus hesperus 0.032 (3) 0.018(3) Lasiurus cinereus 0.041 (3) 0,018(3) Antrozous pallidus 0.021 (2) 0.012(2) Lasiurus seminolus 0.014(1) 0.006(1) 22 Occasional Papers, Museum of Texas Tech University iii Tadarida Myotis Pipistrel!us Lasiurus Antrozous Lasiurus bras Hie ns is velifer hesperus cinereus pallidus seminoius Figure 9. Graphic representation of the total capture indices for the bats at two sites in Devils River State Natural Area. Cenizo Slope had a species composition of two species (Table 2): P. pectoralis (0.035) and C. nelsoni (0.016), Juniper Slope, which was similar in habitat to Cenizo Slope, also had these two species in similar abundances (Table 2) and a third species, Perognathus merriami. East Canyon, which had different vegeta¬ tion and soil composition, had similar species abun¬ dances to these two slope sites (Table 2). Peromyseus pectoralis was again the most abundant with C. nelsoni ranking second. Grassland Plateau differed from the other sites in that C. nelsoni did not occur there (Table 2). Peromyseus pectoralis was once again the most abun¬ dant small rodent with a capture index of 0.019 fol¬ lowed by Perognathus merriami. This site was the only site where Reithrodontomys montanus was captured. The total capture indices for the year were cal¬ culated for each species using 4000 trap nights (Table 2). Peromyseus pectoralis was by far the most abun¬ dant small rodent at Devils River State Natural Area (Fig. 10), being encountered in every sampling site and having a capture index of 0.032. Chaetodipus nelsoni was the second most abundant small rodent, being collected at every site that had a slope while absent from sites without a slope. Sigmodon hispidus was abundant at Dolan Springs, but was not captured anywhere else in the park. This implies that the hispid cotton rat is locally abundant but likely not an impor¬ tant part of the rodent communities in other parts of DRSNA, Perognathus merriami probably has a greater abundance than evidenced at the five primary sam¬ pling sites. Several individuals were caught at other sites with loose soils, particularly in an area 2 l A miles north of the Headquarters Building (29° 58.4’ N, 100° 58.2 ? W). Peromyseus maniculatus, Reithrodontomys fulvescens, and R. montanus each were captured only a few times at a single site. Abundance of small rodents was also examined seasonally (Fig. 11). Capture indices were estimated using 1000 trap nights per season. Rodent capture diversity was greatest during the winter sampling pe¬ riod at DRSNA. Seven species were captured with P. pectoralis being the most abundant with a capture in¬ dex of 0.0900, followed by P . leucopus , S. hispidus, C. nelsoni, P. maniculatus, Perognathus merriami , and R. montanus. The capture diversity for spring was five species. Peromyseus pectoralis was the most fre¬ quently captured rodent during this period with a cap¬ ture index of 0.0230, followed by C, nelsoni, P . leucopus , S. hispidus, and Perognathus merriami . Only C. nelsoni and S. hispidus were captured more fre¬ quently in spring. Peromyseus maniculatus and R. montanus were not captured during this period. Brant and Doyvler— The Mammals of Devils River State Natural Area, Texas 23 Table 2. The capture indices for each species of small rodent from the primary sampling sites at DRSNA. Bold-faced numbers represent the capture index for the most abundant species per sampling site and the number in parentheses represents the total number of captures for that species at that site. Dolan Springs Cenizo Slope Juniper Slope East Canyon Grassland Plateau Total Peromyscus pectoralis 0.013 (10) 0.035 (28) 0.031 (25) 0.060 (48) 0.019 (15) 0.032 (126) Chaetodipus nelsoni 0.014(11) 0.016(13) 0.010(8) 0.025 (2) — 0.009 (34) Peromyscus leucopus 0.024 (19) — — — — 0.005(19) Sigmodon hispidus 0.010(8) — — — — 0.002(8) Perognathus merriami — - 0.001 (1) — 0.005 (4) 0.005 (5) Peromyscus maniculatus 0.004(3) ■— — — — <0.001 (3) Reithrodontomysfulvescens 0.001 (1) — — — <0.001 (1) Reithrodontomys montanus — — — — 0.001 (1) <0.001 (1) Peromyscus Chaetodipus Peromyscus Sigmodon Perognathus pectoralis nelsoni leucopus hispidus merriami Figure 10. Graphic representation of the total capture indices for the five most abundant rodents in Devils River State Natural Area. Capture diversity during summer was the low¬ est of any season with only four species captured (Fig. 11). Chaetodipus nelson (0.0090) was the most abun¬ dant rodent followed by P. leucopus , Perognathus merriami , and Peromyscus pectoralis with lower cap¬ ture frequencies, Sigmodon hispidus was not cap¬ tured in summer. Capture diversity for autumn was six species with P. pectoralis (0.0120) again being the most abundant rodent. Peromyscus leucopus, C. nelsoni , S . hispidus y P. maniculatus , and R.fulvescens were captured less frequently. 24 Occasional Papers, Museum of Texas Tech University 0,0250 3 0.0100 CL m O 0.0050 0.0000 *? c o ■§i 5 g Co I Winter B Spring I Summer I Autumn a- £ SI la s § > 3 ll 3 =3 Q) Oj 3 3 —* c Co 3 CD 3! O' 5 3 3 3. ^ tu c0 S3 C Co Co =0 CD 3? 3 Cl O 3! S' 3 ■5 a CD 5 I § 3 3- Q. CD O =5 c Co 3j O 3 CO Figure 11, Graphic representation of the seasonal capture indices for small rodents at Devils River State Natural Area. Each primary sampling site was analyzed using a 50m transect to determine the percent composition of each of the environmental variables. Juniper Slope was characterized by a high influence of herbaceous vegetation and cobblestone substrates (Table 3). East Canyon had a high percentage of succulent vegetation and boulder sized substrates, while Cenizo Slope was more influenced by woody vegetation and cobblestone (Table 3). Grassland Plateau was characterized by a high percentage of herbaceous vegetation and soil (Table 3). A high percentage of bedrock with approxi¬ mately equal percentages of woody and herbaceous vegetation characterized Dolan Springs (Table 3), Brant (2000) performed a canonical correspon¬ dence analysis (CCA) using all environmental variables to describe the habitat associations among rodents at DRSNA. When Monte Carlo permutation tests were performed to test the significance using all environ¬ mental variables, the results were insignificant. An ordination diagram and discussion of the results are presented in Brant (2000). Using only vegetation vari¬ ables and capture indices, a CCA was performed pro¬ ducing three axes that explained 100 percent of the variation. The first axis is defined by percent succu¬ lent vegetation (Table 4), whereas the second axis is defined by percent woody vegetation. The first two axes explain 94.2 percent of the variation. The third axis is defined by percent herbaceous vegetation and explains the remaining 5.8 percent of the variation (Table 5). An ordination diagram (Fig. 12) shows a graphic representation of habitat associations based on veg¬ etation variables. Sites are represented by dots and species are represented by triangles. Vegetation vari¬ ables are represented as arrows with the length of the arrow representing the importance of that variable. The closer a species is to the origin, the more generalized that species is, according to the vegetation variables sampled. The same is true for sample sites. The smaller the degree of the angle between a vegetation vector and a species or sampling site, the more that vector influences the species or site. Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 25 Table 3. The environmental variable percentages for each sampling site. Juniper Slope Cenizo Slope East Canyon Grassland Plateau Dolan Springs % Woody 0.5671 0.6191 0,4088 0.0384 0,4032 % Herbaceous 0.7466 0.4404 0.4345 0.9616 0.4092 % Succulent 0.2208 0.2904 0.5228 0.1108 0.0360 % Soil 0.0160 0.0327 — 0.9604 0.1828 % Gravel 0.3373 0.1864 — 0.6304 0.0224 % Cobblestone 0.6035 0.6951 0.0055 0.0424 — % Boulder 0.0511 0.0611 0.8745 0,0280 0.0644 % Bedrock 0,0217 0.0049 0.1200 — 0,5532 % Slope 0.1830 0.2267 0.3300 — _ Table 4. The vegetation variable canonical coefficient component loadings for each canonical axis . The values in boldface represent high component loadings for that canonical axis. Axis 1 2 3 % Woody 0.0084 -0.9370 0.2816 % Herbaceous -0.4566 0.6875 0.4591 % Succulent -0.7915 -0.4211 -0.4346 Table 5. A summary of the canonical correspondence analysis using only vegetation variables with eigenvalues, percentage of variation explained, and correlation coefficients listed for each canonical axis. Axis 1 2 3 4 Total Inertia Eigenvalues 0.540 0.185 0.045 0.058 0.828 Species-environment correlations 0,969 0.995 0.838 0,000 Cumulative percentage values of species data 65.2 87.6 93.0 100.0 of species-environment relation 70.1 94.2 100.0 0.000 Sum of all unconstrained eigenvalues 0.828 Sum of all canonical eigenvalues 0.770 26 Occasional Papers, Museum of Texas Tech University A Monte Carlo permutation test was performed to analyze the significance of the habitat associations predicted by the vegetation variables. The first axis alone was not significant (F= 1.871, P= 0.2000). However, all the axes together were significant (F= 4.420, P= 0.0350). Habitat associations between spe¬ cies and sampling site thus can be differentiated using the vegetation variables sampled. Discussion Fieldwork, museum surveys, and literature searches have recorded the presence of 47 species of mammals occurring in the immediate vicinity of DRSNA for the last 30 years. Eight of the 47 species were introduced into the area within the last 50 years. These species represent eight orders, comprised of one didelphimorph, one xenarthran, one insectivore, seven chiropterans, nine carnivores, nine artiodactyls, 16 rodents, and three lagomorphs. Another 19 spe¬ cies possibly occur in the area, including one insecti¬ vore, five chiropterans, two carnivores, and 11 ro¬ dents. The mammal fauna at DRSNA is similar to that of other regions in this part of Texas. Schmidly (1977) reported 96 species of mammals occurring in the Trans-Pecos region of Texas. Thirty-four of those species reported for the Trans-Pecos occur at DRSNA. Goetze (1998) listed 76 species of mammals occur¬ ring on the Edwards Plateau of which 38 occur at DRSNA. The only species occurring at DRSNA that was not listed by Goetze (1998) was Lasiurus seminolus . The mammalian fauna of DRSNA seems to have a closer affinity with that of the Edwards Plateau biotic province, with which it shares 37 species, com¬ pared with the Trans-Pecos biotic province, with which it shares 33 species. The mammals occurring at DRSNA are compa¬ rable to other sites within Val Verde County, Texas. Boston (1966) listed 39 species of mammals as oc¬ curring east of the Pecos River in Val Verde County. Thirty-one of the species occurring at DRSNA were listed by Boston (1966). Scudday and Hanselka (1975) reported 29 species as occurring in the Dolan Falls area. Twenty-eight of the species they listed were verified by this study. The exception is Mus muscu¬ ius, which was listed by Scudday and Hanselka (1975) but was never encountered from June 1998 through November 1999. The rodent diversity and density at most sites within DRSNA were low during this study, with a to¬ tal capture success of only 4.9 percent. The drought affecting the area during this study might have resulted in reduced rodent populations. Three species of rodents, two bats, and one car¬ nivore were each represented by only one specimen. The three county records for Val Verde County result¬ ing from this project are included in these six speck mens along with Reithrodontomys fulvescens, which has not been collected in this county for over one hun¬ dred years. The environmental variable that had the most in¬ fluence on species composition of rodents in grass¬ land habitat types at DRSNA was the percent soil present (Brant, 2000). Of the three rodent species that occur in this habitat type in DRSNA, one is a habitat generalist ( Peromyscus pectoralis) and one is only found at this site { Reithrodontomys montanus). Only Perognaihus merriami demonstrates the influ¬ ence of this environmental variable on habitat associa¬ tion. Perognathus merriami was found in two habitat types but is more associated with the grassland habitat due to the presence of soil, and to a lesser degree her¬ baceous vegetation and gravel (Fig. 12). Another site located 2% miles north of the Headquarters Building was also sampled with some regularity and was simi¬ lar to the Grassland Plateau primary sampling site. A high number of R merriami were encountered at this site, thus supporting the habitat association described by the CCA. The three slope sampling sites represent three habitat types that are similar in many respects and were plotted in the same quadrant of the ordination diagram (Fig. 12). The most important environmental variable for these habitats appears to be succulent vegetation. Woody vegetation could also be considered important, Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 27 * - p r ■ i t - 1 i i r " — " r - ■ ■ 1 r - u — i-1-■- r -■ — i i t —— - Figure 12. Ordination diagram of a canonical correspondence analysis describing habitat associa¬ tions of small rodents at Devils River State Natural Area using capture indices for 8 species of small rodents and vegetation variables for 5 sites. Species are represented by triangles (A) and primary sampling sites are represented by dots (•). Arrows represent the vectors of the environmental variables. The longer the arrow is, the more important the variable. The closer to the origin a species is, the more generalized that species is according to the environmental variables sampled. The smaller the degree of the angle between an environmental vector and a species or sampling site, the more that vector influences the species or site. especially for the Cenizo Slope primary sampling site. The species occurring in these habitats, Peromyscus pectoralis and Chaetodipus nelsoni, are habitat gener¬ alists and represent the two most abundant rodents in DRSNA. These two species are encountered in every habitat type with one exception; C. nelsoni is not found in grassland habitats and is positioned in the opposite quadrant from this habitat type on the ordination dia¬ gram (Fig. 12). The most influential environmental variable for rodents in the riparian habitat at Dolan Springs was the presence of bedrock (Brant, 2000). For this study four species are described as riparian inhabitants 28 Occasional Papers, Museum of Texas Tech University (Peromyscus leucopus, Sigmodon hispidus, Peromyscus maniculatus, and Reithrodontomys fulvescens) „ The other two species occurring in this habitat type are the two habitat generalists found in other sites {Peromyscus pectoralis and Chaetodipus nelsoni). Peromyscus pec- tor alis has its lowest abundance at this site and is po¬ sitioned in the opposite quadrant from the riparian habi¬ tat on the ordination diagram. These species were most often encountered on or at the base of the slope distal to the spring, rather than in the vegetation imme¬ diately adjacent to the spring. Of the eight species of small rodents sampled at five primary sampling sites, the six murid species showed seasonal abundance peaks in the cooler por¬ tions of the year. The two heteromyid rodents, Chaetodipus nelsoni and Perognathus merriami, showed seasonal abundance peaks in the warmer por¬ tions of the year. Peromyscus pectoralis is more likely to be encountered in autumn or spring than in the sum¬ mer. Chaetodipus nelsoni is more likely to be encoun¬ tered in the spring or summer than the fall or winter and Perognathus merriami is more likely to be cap¬ tured in the summer than at any other time of year. Reproductive activity seems to be related to sea¬ sonal abundance for two of the four above rodent spe¬ cies. Chaetodipus nelsoni , which is more abundant in the spring and summer, is reproductively active in the spring and summer with scrotal males occurring from January to June and pregnant females encountered in June and July. Peromyscus pectoralis is less abundant in summer than at any other time of year and is repro¬ ductively active from late autumn to early spring with pregnant females encountered from January to March and scrotal males encountered from October to March. Peromyscus leucopus seems to be a year-round breeder with reproductive males and females encoun¬ tered in January, August, and November during this study. Sigmodon hispidus has a seasonal abundance peak in the winter but is reproductively active in the summer. Two reproductive individuals (a male and a female) were found in August We collected another female with placental scars in May. Cameron and Spencer (1981) reported a year-round breeding sea¬ son for Sigmodon hispidus, as did Lackey et al. (1985) for Peromyscus leucopus . The zoogeographic affinities of DRSNA are typi¬ cal of the Edwards Plateau. Goetze (1998) recog¬ nized seven faunal elements for the mammals of the Edwards Plateau. Based on his results the mammalian fauna at DRSNA can be divided into five faunal ele¬ ments: Chihuahuan, Widespread, Neotropical, East¬ ern, and Campestrian. The Chihuahuan faunal element contains species with distributions centered in the Mexican states of Chihuahua, Coahuila, and Durango. Seventeen of the species occurring at DRSNA are included in this fau¬ nal element: Notiosorex crawfordi, Myotis velifer, Pipistrellus hesperus, Antrozous pallidus, Lepus californicus, Ammospermophilus interpret, Spermophilus mexicanus, Spermophilus variegatus, Perognathus merriami, Chaetodipus nelsoni, Reithrodontomys fulvescens, Peromyscus pectoralis, Neotoma albigula, Bassariscus astutus, Spilogale gra¬ cilis , Conepatus mesoleucus, and Sylvilagus audubonii. Two of these species, Ammospermophilus inter pres and Chaetodipus nelsoni, reach their eastern distribu¬ tional limits in or near Val Verde County. Almost half (44 percent) of the faunal composition at DRSNA origi¬ nates from the Chihuahuan faunal element. Goetze (1998) found that 36 percent of the Edwards Plateau species originated from the Chihuahuan faunal element. The Widespread faunal element is comprised of those species with large enough distributions that it is difficult to assign them to any of the other faunal ele¬ ments. Twelve species occurring at DRSNA are des¬ ignated as part of the Widespread faunal element. These species include: Lasiurus cinereus. Castor canadensis, Peromyscus leucopus, Peromyscus maniculatus, Erethizon dorsatum, Vrocyon cinereoargenteus, Pro- cyon lotor, Taxidea taxus, Mephitis mephitis, Puma concolor, Lynx rufus, and Odocoileus virginianus. Thirty-one percent of the species occurring at DRSNA are accounted for by the Widespread faunal element as compared to the 26 percent of species accounted for on the Edwards Plateau (Goetze, 1998). The Neotropical faunal element is comprised of species with distributions centered in southern Mexico and Central America. Five species at DRSNA are in¬ cluded in the Neotropical faunal element: Didelphis virginiana, Tadarida brasiliensis, Dasypus Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 29 novemcinctus, Sigmodem hispidus, and Pecari tajacu . Goetze (1998) determined that this faunal element ac¬ counts for 11 percent of the species occurring on the Edwards Plateau. The Neotropical faunal element was responsible for 13 percent of the species at DRSNA, Species with distributions centered in the east¬ ern United States comprise the Eastern faunal element. Three species from this faunal element occurring at DRSNA are at their western distributional limits and can only be found in riparian habitats such as Dolan Springs or Jose Maria Springs. This faunal element includes the following four species: Lasiurus seminolus, Nycticeius humeralis, Sciurus niger, and Sylvilagus floridanus. The Eastern faunal element accounts for 10 percent of the species occurring at DRSNA, whereas it accounts for 11 percent of the Edwards Plateau fauna (Goetze, 1998), The Campestrian faunal element contains spe¬ cies with distributions centered on the Great Plains of North America. The only species occurring at DRSNA that is included in this faunal element is Reithrodontomys montanus , which reaches its south¬ ern distributional limits in this area of Texas and can only be found in grassland habitats. Three percent of the mammals occurring at DRSNA originate from the Campestrian faunal element. Nine percent of the mam¬ mals on the Edwards Plateau originate from the Campestrian faunal element (Goetze, 1998), The mammalian fauna at DRSNA is composed primarily of western Edwards Plateau mammals, which include many Chihuahuan species. The area is near the western distributional limits of three eastern spe¬ cies, the eastern distributional limits of two western species, and the southern distributional limits of one northern species. The two most abundant rodents are Peromyscus pectoralis and Chaetodipus nelsoni , both Chihuahuan in origin, and the most abundant bat, Tadarida brasiliensis, is a Neotropical species. The larger mammals are mostly comprised of Widespread species. ACKNO Wl .EDGMENTS This study was funded through a grant from the Natural Resources Program, Texas Parks and Wildlife Department (State Park Scientific Study Permit 43- 98) and is based on research in partial fulfillment of the requirements for a M. S. degree in the Department of Biology, Angelo State University for J. G. Brant. We appreciated the assistance and patience of Bill and Paula Armstrong and Jim Finegan, as well as their knowledge of Devils River State Natural Area. Terry C. Maxwell, Brian J. May, J. Kelly McCoy, and Connie Russell provided advice, statistical analysis assistance, tech support, and field assistance during this study. Darin Carroll, Jill DeVito, Cody Edwards, Richard Humbertson, Eddie Lyons, Shalaija Marion, Brandy Martin, Kathryn Perez, Marcy Revelez, and Marisol Salazar assisted with field collection and specimen preparation. The volunteer program of the Angelo State Natural History Collections also assisted in specimen preparation. Thanks to these museums and museum person¬ nel for providing lists of current holdings in their re¬ spective institutions of mammals from Val Verde County, Texas: The Museum of Vertebrate Zoology, University of California, Berkeley (MVZ); James M. Mueller at the Vertebrate Collection at Sul Ross State University (SRSU); Duane A. Schlitter at the Texas Cooperative Wildlife Collection, Texas A&M Univer¬ sity (TCWC); R. Richard Monk at the Museum of Texas Tech University (TTU); and the United States National Museum of Natural History (USNM). 30 Occasional Papers, Museum of Texas Tech University Literature Cited Animal Care and Use Committee. 1998. Guidelines for the capture, handling, and care of mammals as ap¬ proved by the American Society of Mammalo- gists. J.Mamm,, 79(4): 1416-1431. Bailey, V. 1905. Biological survey of Texas. North Ameri¬ can Fauna 25:1-222. Blair, W. F. 1950. The biotic provinces of Texas. Texas J. Sci., 2(2): 93-117. Boston, C. L. 1966. Mammals of Val Verde County, Texas, east of the Pecos River. Unpublished M.S. the¬ sis, Sul Ross State University, Alpine, TX. 126 PP- Brant, J. G. 2000, A survey of the mammals of Devils River State Natural Area and relative abundance of small mammals in the area. Unpublished M. S. thesis, Angelo State University, San Angelo, TX. 101 pp. Brant, J. G, and R. C. Dowler. 2000. Noteworthy record of the Seminole bat, Lasiurus seminolus (Chiroptera: Vespertilionidae), in Val Verde County, Texas, Texas J. Sci., 52(4):353-355, Bryan, K. B., L. K. Hedges, andT. E. Roberts. 2001. Devils River State Natural Area resource management plan. Unpublished draft report on file at the Natu¬ ral and Cultural Resources Program Office, Texas Parks and Wildlife Department, Fort Davis and Austin. Cameron, G, N,, and S. R. Spencer. 1981. Sigmodon hispidus. Mammalian Species, 158:1-9. Conroy, M. J. 1996. Abundance indices. Pp. 179-192 in: Measuring and monitoring biological diversity: standard methods for mammals. Wilson, D, E,, F. R. Cole, J. D. Nichols, R. Rudran, and M.S. Fos¬ ter, eds. Smithsonian Institution Press, Wash¬ ington, D. C. 409 pp. Davis, W. B., and D. J. Schmidly. 1994. The mammals of Texas. Texas Parks and Wildlife Press, Austin. 338 pp. Deal, D. E., and R. G Fieseler, 1975. Cave resources in the Devils River-Dolan Creek area. Pp. 29-35. in: Devils River: a natural area survey. Part VI of VII. Division of Natural Resources and Environment, The University of Texas, Austin, TX. Ditton, R. B., and D. J. Schmidly. 1977, Amistad: a user- resource analysis of Amistad Recreation Area (Texas), Report prepared for the Office of Natu¬ ral Resources, Southwest Region, National Park Service, Santa Fe, NM. Dowler, R. C., and R. A. Boyd. 1996 . A range extension for the least shrew ( Cryptotis parva ) in west-central Texas. Texas J. Sci., 48(2): 168-170, Dowler, R. C., R. C. Dawkms, and T. C. Maxwell. 1999. Range extension for the evening bat (Nycticeius humeralis) in west Texas, Texas J. Sci., 51(2): 193- 195. Dowler, R. C., T. C. Maxwell, and D. S, Marsh. 1992. Note¬ worthy records of bats from Texas. Texas J. Sci., 44(1): 121-123. Goetze, J. R. 1998. The mammals of the Edwards Plateau, Texas. Spec. Pubs. Mus., Texas Tech Univ., 41:1- 263. Golden, M. L„ W. J. Gabriel, and J. W. Stevens. 1982. Soil survey of Val Verde County, Texas. Soil Conser¬ vation Service, United States Department of Ag¬ riculture. 138 pp. Hall, E. R. 1981. The mammals of North America. John Wiley & Sons, Inc., New York, 1-1181 + 90 pp. Jones, C,, W. J. McShea, M. J. Conroy, and T. H. Kunz, 1996. Capturing mammals. Pp. 115-155 in: Mea¬ suring and monitoring biological diversity: stan¬ dard methods for mammals. Wilson, D. E., F. R. Cole, J. D. Nichols, R. Rudran, and M. S. Foster, eds. Smithsonian Institution Press, Washing¬ ton, D. C. 409 pp. Jones, J. K., Jr., and R, W. Manning. 1991. Comments on distribution of two species of pocket mice (Ge¬ nus Chaetodipus ) along the Pecos River, Texas. Texas J. Sci„43(l):99-101. Lackey, J. A., D. G. Huckaby, and B, G. Ormiston, 1985. Peromyscus leucopus. Mammalian Species, 247:1-10. Lee, T. E., Jr. 1987, Distributional record of Lasiurus seminolus (Chiroptera: Vespertilionidae). Texas J. Sci., 39(2): 193. Lee, T. E., Jr., B. R. Riddle, and P. L. Lee. 1996. Speciation in the desert pocket mouse (Chaetodipus penicillatus Woodhouse). J. Mamm., 77(1):58- 68 , Brant and Dowler— The Mammals of Devils River State Natural Area, Texas 31 Lindzey, F. G. 1982. Badger. Pp. 653-663. in: Wild mam¬ mals of North America. Chapman, J, A., and G A. Feldhamer, eds. Johns Hopkins Univ. Press, Bal¬ timore. 1147pp. Manning, R. W., J. K. Jones, Jr., R. R. Hollander, and C. Jones. 1987, Notes on the distribution and natu¬ ral history of some bats on the Edwards Plateau and in adjacent areas of Texas. Texas J. Sci., 39(3):279-2S5. Menzel, M. A., D. M. Krishoo, T, C. Carter, and J. Laemi. 1999. Notes on tree roost characteristics of the northern yellow bat (, Lasiurus intermedius), the Seminole bat ( L , seminolus), the evening bat (Nycticeius humeralis), and the eastern pipistrelle {Pipistrellus subflavus). Florida Scient., 62(3/ 4): 185-193. Nowak, R. M., ed. 1999. Walker’s mammals of the world. 6 th ed, Baltimore: Johns Hopkins University Press. 1936 pp. Schmidly, D. J. 1977. The mammals of Trans-Pecos Texas. Texas A&M University Press: College Station. 225 pp. -. 1991, The bats of Texas. Texas A&M Univ. Press: College Station. 188 pp. Scudday, J. F., and C. W. Hanselka. 1975. A preliminary survey of the vertebrate fauna of the Dolan Falls- Dolan Creek area. Pp. 58-75. in: Devils River: a natural area survey. Part VI of VII. Division of Natural Resources and Environment. The Uni¬ versity of Texas, Austin, TX. Addresses of authors: Joel G, Brant Department of Biology Angelo State University San Angelo , TX 76909 e-mail: jbrant@ttacs.ttu.edu Simpson, L. A., and T. C. Maxwell. 1989. The mammal fauna of Coke County, Texas, Texas J. Sen, 41 (2): 177-192. Smith, G, Jr. 1975. Impressions of the Devils River. Pp. 1- 3. in: Devils River: a natural area survey. Part VI of VII. Division of Natural Resources and Envi¬ ronment. The University of Texas, Austin, TX, ter Braak, C. J. F. 1986. Canonical correspondence analy¬ sis: a new eigenvector technique for multivariate direct gradient analysis. Ecology, 67:1167-1179. Texas Parks and Wildlife Department. 1990. Devil ’s River State Natural Area summary of representative plant communities. Department Lands Inven¬ tory, unpublished document. Wilkins, K. T. 1989. Tadarida bra si lien sis. Mammalian Species, 331:1-10. Wilson, D. E., and D. M. Reeder (eds.). 1993. Mammal Species of the World: a taxonomic and geographic reference. Smithsonian Institution Press, Wash¬ ington, D.C., xvii + 1206 pp. Yancey, F, D., II, and C, Jones. 1996. New county records for ten species of bats (Vespertilionidae and Molossidae) from Texas. Texas J. Sci,, 48(2): 137- 142. Yancey, F. D., II, C. Jones, and R. W. Manning. 1995. The eastern pipistrelle, Pipistrellus subflavus (Chiroptera: Vespertilionidae), from the Big Bend region of Texas. Texas J. Sci., 47(3):229-23l. Current address: Department of Biological Sciences Texas Tech University Lubbock TX 79409-3131 Robert C. Dowler Department of Biology Angelo State University San Angelo, TX 76909 e-mail: robert.dowler@angelo. edu Publications of the Museum of Texas Tech University It was through the efforts of Horn Professor J Knox Jones, as director of Academic Publications, that Texas Tech University initiated several publications series including the Occasional Papers of the Museum. This and future editions in the series are a memorial to his dedication to excellence in academic publications. Professor Jones enjoyed editing scientific publications and served the scientific community as an editor for the Journal of Mammalogy, Evolution, The Texas Journal of Science, Occasional Papers of the Museum, and Special Publications of the Museum. It is with special fondness that we remember Dr. J Knox Jones. Institutional subscriptions are available through the Museum of Texas Tech University, attn: NSRL Publications Secretary, Box 43191, Lubbock, TX 79409-319L Individuals may also purchase separate num¬ bers of the Occasional Papers directly from the Museum of Texas Tech University. ISSN 0149-175X Museum of Texas Tech University, Lubbock , TX 79409-3191