Occasional Papers Museum of Texas Tech University Number 215 21 June 2002 Small Mammal Communities and Habitat Associations in the C fflHUAHUAN Desert Near Fort Bliss, New Mexico Melinda L Clary, Richard E. Strauss, Brian A. Locke, Donna J, Howell, Robert J. Baker, and Robert D. Bradley Desert ecosystems are characterized as some of the most biologically diverse areas in the United States. Many species of flora and fauna are unique to desert habitats. This diversity is in itself phenomenal consid¬ ering that many species maintain exceptionally stable populations despite unpredictable fluctuations in the environment (Zeng and Brown, 1987). Free-standing water rarely is available in these arid ecosystems. Al¬ though the lack of available water would be detrimen¬ tal for organisms living in most other environments, animals in desert habitats have adaptive morphological and physiological characteristics that enable them to thrive in an environment that is unsuitable for other species (Ghobrial and Nour, 1975). Rodent species with such adaptations especially are diverse in desert ecosystems. Partitioning of resources facilitates co¬ existence of rodent species in desert rodent communi¬ ties of southwestern North America; thus desert ro¬ dents are ideal subjects for studies of coexistence, competition, and community structure (Heske et al., 1994). The Chihuahuan Desert is located in the south¬ ernmost portion of the Great American Desert and is bounded by the 100“'and 108“* meridians and the 2P’ and 33"“ parallels (Milstead, 1960). It includes parts of southern New Mexico, all of Texas west of the Pecos River (except for the Guadalupe Mountains), the eastern half of Chihuahua, the western portion of Coahuila, and parts of Durango, Zacatecas, Nuevo Leon, San Luis Potosi, Aguascalientes, and Tamaulipas (Schmidly, 1974). The region has a diverse flora and fauna including eight orders, 24 families, 60 genera, and roughly 119 species of mammals (Schmidly, 1974). Fort Bliss is an U.S. Army installation in the north¬ ern region of the Chihuahuan Desert. It occupies ap¬ proximately 4,523 km^ (452,279 ha), ranging from El Paso County, Texas, to Otero County, New Mexico. This study was conducted on the McGregor Range, which includes the northeastern portion of Fort Bliss. McGregor Range is dominated by desert shrubland, with some desert grasslands, and juniper woodlands 2 Occasional Papers, Museum of Texas Tech University at higher elevations. It is located within the Tularosa Basin, which is typified by lowland valleys, rocky hill¬ sides, and scattered arroyos (Gardner, 1951; Schmidt, 1986; Jorgensen, 1996; Monasmith, 1997). Overall, the region is characterized by a semiarid to arid cli¬ mate. Despite studies previously conducted on Fort Bliss concerning the status of small mammals (Jorgensen, 1996; Jorgensen and Demarais, 1996; Root, 1997; Weeks, 1997), few researchers have com¬ pared mammalian species across different habitat types. Such a study would provide information about habitat utilization of small mammal species on McGregor Range, thus allowing Fort Bliss personnel to consider these associations when planning military training ac¬ tivities. Such activities have been shown to affect ecosystem stability (Baumgardner, 1990; Brattstrom and Bondello, 1983; Carroll et al., 1999; Edwards et al., 1998; Gese et al., 1989; Shaw and Diersing, 1990; and Stephenson et al., 1996) and therefore should be considered when planning military operations. Due to the steady increase in human population and urbanization of surrounding public lands, it is be¬ coming necessary for military personnel to ecologi¬ cally manage the acres of non-developed land on their properties in order to preserve areas for existing wild¬ life and to assure sound stewardship of public lands. Information provided by this project will be used to better understand and manage these biological re¬ sources, and minimize the potential impacts of mili¬ tary training in these environments. For example, such information could be used to avoid areas of high ro¬ dent diversity and density, as well as to identify habi¬ tats with unique rodent compositions. The objectives of this study were to: (1) com¬ pare small mammal composition (rodents) among 12 unique habitats on McGregor Range to determine if certain habitats support rodent communities of higher diversity, (2) compare small mammal densities per habitat and determine any influence of vegetative cover, (3) determine if the amount of rainfall influences den¬ sity or diversity, (4) determine the survivability of each species, and (5) compare the average activity among species and habitats to determine the influence of veg¬ etative density on rodent activities. Methods The research design for this study involved sam¬ pling small mammals in 12 distinct habitats. Range botanists from Fort Bliss characterized each habitat based on vegetative composition and density, and la¬ beled them as sandy arroyo scrub, nonstabilized dune, coppice dune, creosote-tarbush scrub, mixed desert scrub, grama grassland, creosote grassland, yucca grassland, swale, acacia hillside, Chilopsis (desert wil¬ low) arroyo, and succulent hillside habitats. Each habi¬ tat type was ranked by range botanists on the basis of percent vegetative cover from 1 (0.0-20.0%) to 5 (80.0- 100.0%). The experimental design for this project was described in Clary et al. (1999). Two grids, each with census and assessment lines, were constructed for each habitat. Census lines, consisting of two par¬ allel trap lines 30 m apart (240 m in length) with trap stations placed at 10 m intervals along each line for a total of 50 traps were used for initial captures. The assessment lines, containing 72 traps, formed a dia¬ mond-like configuration about the census lines and were utilized for recording recaptures on each grid. Each of these lines (census and assessment) was sampled using Sherman live-traps (H. B. Sherman Trap Co., Tallahassee, FL) baited with birdseed and rolled oats during two seasonal periods (Spring and Fall) for two consecutive years (1997 and 1998). Typically, sam¬ pling of the grids occurred on three consecutive nights on the census lines followed by sampling for an addi¬ tional three consecutive nights on the assessment lines, for a total of 35,136 trap nights. We recorded Texas Tech Museum identification number, species, weight, sex, toe-clip number, trap station, and date of capture for each individual captured on the census line. All individuals were released at the site of capture. Cap¬ tures on assessment lines were identified from toe- clip patterns and the date and trap station of capture was recorded. For this study, it was assumed that animals encountered traps randomly in a particular habitat and neither sex, age, or dominance of an indi¬ vidual influenced capture probabilities. Clary et al.— Mammal Communities in the Chihuahuan Desert 3 A reference collection of voucher specimens and tissue samples of at least one adult male and one adult female of each species was prepared and deposited in the Museum, Texas Tech University. In addition, toes obtained during the toe-clipping procedures were pre¬ served in lysis buffer (Longmire et al, 1997) and serve as voucher material. Nomenclature followed Jones et al. (1997) and specimens were identified using keys and characteristics from Davis and Schmidly (1994), Findley (1987), and Findley and Caire (1974). To determine monthly precipitation, a rain gauge was attached to a T-post and placed at the first trap station of each grid. Mineral oil was added to the gauges to prevent evaporation. Rainfall was recorded (in millimeters) at the first of each month from April 1997 to October 1998. The pattern of rainfall for each month was recorded for each habitat and season. In some cases, minor disturbances on the grids (e.g., cattle) affected the reliability of some of the gauges and monthly accumulation could not be recorded. These incidences were not included in the accumula¬ tion averages of the affected grid. All statistical analyses used in this study were generated in the software program Matlab (Matlab 5.2, 1998; www.biol.ttu.edu/faculty/facpages/strauss/ matlab.htm). Species diversity was calculated for each grid and habitat using Simpson’s diversity index (Simpson, 1949). Species diversity values were com¬ pared between and among habitats to determine which habitats possessed the highest diversity. Diversity val¬ ues were compared to the vegetative density (percent cover) of each habitat using correlation analyses. Spe¬ cies composition was compared among the habitats by grouping habitats via a UPGMA cluster analysis of species frequencies for each habitat. The cluster analy¬ sis was bootstrapped for 1,000 iterations and those habitats grouping together at least 50% of the time (bootstrap support value = 0.50) were recognized. Clustering methods similar to those used by Brown and Heske (1990) were used to determine species com¬ position similarities among the habitats. The average rainfall from each trapping period was tabulated and compared with each habitat’s diversity value through correlation analyses to examine the relationship between rainfall and rodent community diversity. Rodent density per habitat, estimated as the num¬ ber of individuals for all species captured per hectare, was compared among habitats and seasons. Percent cover (vegetation) per habitat also was compared with rodent density. In addition,regression analyses were used to examine the relationship between rainfall and density. Survivorship was estimated for each species and comparisons were made using the proportion of indi¬ viduals (per species) recaptured from the total initial captures at six-month intervals (6,12, and 18 months). In addition, survivorship rates were estimated by log- log regression of number of recaptures since first cap¬ ture. It should be noted that this estimate does not distinguish between death and emigration. The average activity of individuals per species was estimated by converting the trap stations and traps of capture per grid to points on an x,y point-coordi¬ nate system. The centroid (median) of activity for each recaptured individual was generated based on the assemblage of captures and deviations from the cen¬ troid to each trap of capture were calculated. The squared deviations were then averaged per species (mean squared deviations, MSD) and compared among the species. In addition, the squared deviations were compared across habitats to determine whether dif¬ ferences in vegetative density (percent cover) among the habitats affected the amount of activity among the residing species. An analysis of variance (ANOVA) was used to determine significant differences among the species and habitat activity averages. Post-hoc multiple comparisons based on sequential Bonferroni probabilities (Rice, 1989) were used to determine sig¬ nificant pairwise differences within the species and habitat groups. 4 Occasional Papers, Museum of Texas Tech University Results Diversity.- Nineteen species were captured from the 12 habitats. The sandy arroyo scrub habitat pos¬ sessed the highest rodent diversity (D^ = 0.872), al¬ though several other habitats possessed similar diver¬ sity values, including Chilopsis arroyo (0.828), Acacia hillside (0.816), creosote-tarbush scrub (0.816), suc¬ culent hillside (0.803), swale (0.802), Yucca grassland (0.796), creosote grassland (0.742), and mixed desert scrub (0.707) habitats (Table 1). Species diversity was lowest in the coppice dune habitat (0.384), whereas nonstabilized dunes (0.645) and grama grasslands (0.472) had low to medium diversity values. Species diversity in the habitats was positively correlated (r = 0.74, p < 0.001) with percent vegetative cover. Species diversity for most grids remained rela¬ tively unchanged between trapping periods (Spring 1997, Fall 1997, Spring 1998, and Fall 1998) and be¬ tween years (1997, 1998). However, small mammal diversity increased in the nonstabilized dune habitat during 1998 (Spring and Fall), decreased in the cop¬ pice dune habitat during 1998 (Spring and Fall), de¬ creased in the mixed desert scrub habitat during Spring 1998, and was lowest in the yucca grassland habitat during Spring 1997. All interpretations on community diversity as a function of rainfall were based on direct observation and correlation analyses (Figure 1). Average rainfall in Spring 1998 was the lowest (5 mm) recorded during Table 1. The Simpson ^ diversity index (Simpson, 1949), used to estimate species diversity per season for each of the 12 habitats on Fort Bliss Military Base. Total diversity values for each habitat for all four seasons are provided in the last column. Habitat Spring 1997 FaU1997 Spring 1998 FaU 1998 Total Nonstablized Dune 0.552 0.541 0.735 0.696 0.645 Sandy Arroyo Scrub 0.876 0.881 0.865 0.807 0.872 Mixed Desert Scrub 0.738 0.779 0.480 0.607 0.707 Swale 0.772 0.768 0.891 0.799 0.802 Succulent Hillside 0.788 0.804 0.690 0.766 0.803 Chilopsis Arroyo 0.865 0.823 0.776 0.749 0.828 AcaciaWiWside 0.831 0.858 0.802 0.821 0.816 Grama Grassland 0.393 0.452 0.521 0.540 0.472 Creosote Grassland 0.747 0.663 0.813 0.745 0.742 Creosote-Tarbush Scrub 0.835 0.793 0.755 0.754 0.816 Coppice Dune 0.423 0.532 0.228 0.244 0.384 Yucca Grassland 0.599 0.870 0.818 0.708 0.796 Clary et al.— Mammal Communities in the Chihuahuan Desert 5 X I Q m o <0 Species Diversity per Habitat and Season Spring 1997 Fall 1998 Season —Nonstabitized Dune ' ■ . Sandy Arroyo Scrub —Mixed Desert Scrub (0.61) Swale -w-Succulent Hillside (0.72) .»'■ ■ Chilopsis Arroyo Figure 1. Species diversity values per habitat and season compared with season rainfall averages. Positive coefficients are shown in parentheses. 6 Occasional Papers, Museum of Texas Tech University the study; whereas Spring 1997, Fall 1997, and Fall 1998 averaged approximately 50 mm of rainfall simi¬ lar. Seasonal diversity values were positively corre¬ lated with seasonal rainfall in the mixed desert scrub (r = 0.76, p = 0.23), succulent hillside (r = 0.90, p = 0.10), Acacia hillside (r = 0.64, p = 0.35), creo- sote-tarbush scrub (r = 0.46, p = 0.53), and coppice dune (r = 0.47, p = 0.53) habitats. However, none of these positive correlations were significant (p > 0.05). The UPGMA cluster analysis revealed certain habitats grouping together by species composition simi¬ larities, with bootstrap support values of at least 0.50 (Figure 2). The coppice dune and mixed desert scmb habitats clustered with a bootstrap support value of 0.55 and the three grasslands (creosote grassland, grama grassland, and Yucca grassland) clustered with a bootstrap value of 0.58. In addition, the acacia hill¬ side, sandy arroyo scrub, and succulent hillside habi¬ tats clustered with a bootstrap support value of 0.71 with two of the habitats, acacia hillside and sandy ar¬ royo scrub, clustering with a bootstrap support value of 0.85. Density — Rodent densities were estimated within habitat by calculating the number of captured individu¬ als per hectare (Figure 3). The swale habitat exhibited the highest rodent density of 39.2 individuals/ha (Fig¬ ure 3). Other habitats having relatively high rodent densities were the Acacia hillside (38.8 individuals/ha), Chilopsis arroyo (33.4 individuals/ha), and creosote grassland (30.0 individuals/ha) habitats. The lowest rodent density (10.0 individuals/ha) was found in the coppice dune habitat. The grama and Yucca grassland habitats also contained relatively low rodent densities (17.9 and 21.3 individuals/ha, respectively). Relationships between vegetation and rodent den¬ sities for each habitat were examined by correlation analyses. Percent cover of vegetation and rodent den¬ sity were positively, but not significantly correlated across all habitats (r = 0.65, p > 0.05). In addition, regression analyses were used to estimate the influ¬ ence of rainfall accumulation on rodent densities per season. Although there appeared to be a decrease in density within some of the habitats during low accu¬ mulation periods, none of the regressions were sig¬ nificant. Survivorship- The proportion of individuals re¬ captured from the total number of initial captures was calculated at six-month intervals for each species (Fig¬ ure 4). Individuals of 16 species captured in the initial census were recaptured after six months (Figure 4). Due to deficient sample sizes, Spermophilus spilosoma, Perognathus flavescens, Chaetodipus hispidus, and Reithrodontomys montanus were excluded from the regression analysis. The percentage of recaptures per species ranged from 3.3% for Neotoma micropus to 26.1% for Dipodomys merriami. Individuals of 13 species were recaptured at 12 months, with the re¬ capture rates per species ranging from 0.7% for Sigmodon hispidus to 11.1 % for Dipodomys spectabilis. Six of the 13 species were from the family Heteromyidae and possessed the highest recapture rates (3.4-11.1%). Individuals of 10 species were recap¬ tured after 18 months with six of the species from the family Heteromyidae. Recapture rates ranged from 0.7% for Peromyscus leucopus and Peromyscus maniculatus to 11.1% for £). spectabilis. Results from the regression analysis indicated N. micropus had the highest exponential rate of decline (d = - 0.56) from the beginning to the completion of the study. Other species with high exponential rates of decline were Onychomys arenicola (d = - 0.39), S. hispidus (d = - 0.37), Reithrodontomys megalotis (d == - 0.35). D. merriami had the lowest exponential rate of decline (d = - 0.18) followed by Chaetodipus intermedins (d = - 0.185), Dipodomys ordii (d == - 0.19), Perognathus flavus (d = - 0.21), and Chaetodipus eremicus {d = - 02\). Activity- When calculating the average activi¬ ties by individuals of each species, six were eliminated on the basis of inadequate sample size: P flavescens, C. hispidus, D. spectabilis, R. montanus, N. micropus, and S. spilosoma. The MSD were calculated from the centroid (spatial mean of activity to each trap station of capture) and averaged for each species per habitat type (Figure 5). The averages among the remaining 13 species differed significantly (F = 3.65, df = 18, p = < 0.001). Onychomys leucogaster had the widest activity range (3.94 MSD) followed by FI maniculatus (3.87 MSD), D. merriami (3.81 MSD), O. arenicola (3.72 MSD), and D. ordii (3.71 MSD). The lowest average activity range of 2.50 MSD was exhibited by Clary et al.— Mammal Communities in the Chihuahuan Desert ■D c m m m 2 O B o s "O c CQ 15 m 2 O CQ E CQ o o •o c jg 15 0) 2 O S o 3 >• -Q (_ O CO -C w 3 n )L. CQ % (/> O 2 o 0 CO CO 0 T3 i«Ml ig X CQ O CQ O < 3 O CO o >* •o c CQ CO 0 •o c 0 3 O O 3 CO 0 c 3 a 0 O a a o O n 3 Q CO t: 0 0 0 o “O 0 >< o >% 2 < t ns; -c o 0 c Q XJ 0 15 CO 0 c o z 6 ^3 u § o B o o o (D cx I 8 o a o T3 (D OD 8 §■ I I 4-» I I cx "O -2 < r4 8 .§} 12 a bootstrap support value of > 50% are shown in bold. Rainfall (mm) Density (# captires/ha) Density (# captire^ha) 8 Occasional Papers, Museum of Texas Tech University 30.0 25.0 20.0 15.0 10.0 5.0 0.0 .....^Coppioe Dune ....fl-Grama Grassland ..^Nonstabilized Dune -^Mixoi Desert Scrub -^Creosote Grassland ' Yucca Grassland Hi Spring 1997 Fall 1997 Spring 1998 Fall 1998 Figure 3. Rodent density, defined as the number of captures per hectare, for each of the four trapping periods at Fort Bliss. 400 Clary et al.— Mammal Communities in the Chihuahuan Desert 9 indicates species having a significant value (p < 0.05). Mean Squared Deviaticans Mean squared deviations Occasional Papers, Museum of Texas Tech University • C. emmicus ® C. intermedlus ^ P. ffavus S D. merriami ■ D. ordii ■ P. emmicus ■ P leucopus ■ P, mamculatus M O. arenicofa ■ O. leucogaMer 0 R imgaidtis ■ S. hispidm MN. afbigutd N. Dune S A Scrub M.D. Scrub Swale s. mm® C. Arroyo A. Hillside O. Grassland C. Grassland C. T. Scrub C. Dune Y. Grassland Habitat Figure 5. Average species movements shown as the mean squared deviation from the centroid activity for each species (top) and habitats (bottom). An analysis of variance detected significant differences for species (p < 0.00) and for habitats (p = 0.002). Clary et al.— Mammal Communities in the Chihuahuan Desert 11 S. hispidus. Other low values were found for R. megalotis and P. flavus with averages of 2.65 and 2.74 MSD, respectively. The averaged MSD across species also was com¬ pared among the 12 habitats (Figure 5). Values dif¬ fered significantly (2.68,11, p = 0.002) with the high¬ est recorded for small mammals from the acacia hill¬ side habitat (3.96 MSD). Similar results were found for rodents from the grama grassland and mixed desert scrub habitats with 3.80 and 3.79 MSD, respectively. Results indicated that the lowest average activity value was for small mammals from the succulent hillside habitat with 2.66 MSD. Results were similar in the swale and creosote grassland habitats (3.06 and 3.26 MSD, respectively). Despite the significant differences found from the ANOVA generated across species and habitat groups, no significant differences resulted in the pairwise comparisons within each group. Discussion With few exceptions, there appeared to be little difference among diversity values in reference to trap¬ ping period or year within the same habitat type. This may be expected due to the short duration of the study and overall lack of long-term data for this environ¬ ment. The low diversity seen in coppice dune habitats (which contained mainly D. merriami and D. ordii) was possibly a result of kangaroo rats being more ef¬ ficient or successful foragers in open habitats when compared to other species. Overall, these findings are similar to those of Harris (1984), Kotler (1984), and Kotler and Brown (1988). Unfortunately, with the exception of Jorgensen (1996) few comparative stud¬ ies are available for Chihuahuan Desert communities. In terms of overall species diversity, habitats characterized by high percent vegetative cover (> 60%) including the swale, Chilopsis arroyo, sandy arroyo scrub, and Acacia hillside habitats possessed the high¬ est species diversity. Brown (1975) suggested that habitats with higher productivity show less competi¬ tion among species. Conversely, the coppice dune habitat (< 20% vegetative cover) consistently demon¬ strated a paucity of small mammals. There is little cover and fewer seed-producing plants associated with this habitat, and seed production is the key determi¬ nant of rodent species diversity in North American desert ecosystems (Brown, 1975). Diverse commu¬ nities of small mammals have been found to exist more frequently in densely vegetated habitats than in sparsely vegetated habitats (Jorgensen, 1996; Brown and Zeng, 1989). In a study of desert rodent communities in the Mojave Desert, species diversity increased with in¬ creasing vegetative cover (Hafner, 1977). However, in our study we did not examine seed production, there¬ fore we could not discern between the effects of veg¬ etative cover in terms of protection from predators or as a means of food production. Also, seed size and utilization by small mammals may effect species dis¬ tribution and density. Although the results in the correlation analyses were not significant, the decrease in rodent diversity observed in some of the habitats during Spring 1998 may be due to the low average precipitation recorded within that season. Fluctuations in rainfall directly af¬ fect water and forage availability. Water accessibility may play a significant role in coexistence of desert rodents. The effects of available moisture, vegetative density, and habitat complexity are intricately involved in determining the diversity of rodent communities (Christian, 1980; Hafner, 1977). In addition to exhibiting high rodent diversity, habitats characterized by dense vegetation (> 60% vegetative cover) also contained higher densities of rodents. The swale habitat had the highest percent vegetative cover (80-100%) and also exhibited the high¬ est rodent density. Although not significant, results from the correlation analyses revealed a positive cor¬ relation between rodent density and percent cover of vegetation. High densities of rodents in densely veg¬ etated habitats may be due to an increase in forage (seed-producing plants, vegetation, or insects). As noted by Thompson (1982), the mean densities of seeds and the variation in those densities are greater beneath vegetation. In addition, the ample amount of vegeta¬ tion may be beneficial as adequate cover to serve as protection from predators (Kotler and Brown, 1988). 12 Occasional Papers, Museum of Texas Tech University A positive correlation was revealed between sea¬ sonal rodent density and rainfall. The drought period in Spring 1998 seemed to have an adverse effect on rodent density in most of the habitats. For example, the mixed desert shrub and creosote-tarbush scrub habitats showed a decline in total number of species present during Spring 1998. Eleven species present during 1997 were absent from these habitats in Spring 1998, most notably absent were O. leucogaster, O. arenicola, P. eremicus, P. leucopus, P. maniculatus, and 5. hispidus. As discussed above, water availabil¬ ity is crucial to most rodent species. Although heteromyids do not drink water, times of low resource availability resulting from low rainfall can cause a slower rate of reproduction for some species. Only 2.0 % of the total number of individuals captured across species were recaptured after 18 months. The probability of survival may vary with individual characteristics and also as a function of vari¬ ous environmental variables (Lebreton et al., 1992). For example, six out of the 10 species of the family Heteromyidae had individuals recaptured after 18 months (ranging from 1.6% to 11.1% of initial cap¬ tures). Their success may be due to certain morpho¬ logical and metabolic adaptations characteristic of this family. For example, individuals of the genus Dipodomys have many attributes, such as inflated au¬ ditory bullae and bipedality, which enhance senses and speed for avoidance of predators in addition to forag¬ ing efficiency. Other members of the family Heteromyidae, including Chaetodipus sp. and Perognathus sp., are known to aestivate during times of low seed availability. Interestingly, species of Neotoma were seldom recaptured. It is unclear if woodrats became trap-shy (due to their large size) or whether their survival rates were low. Higher averages of activity were present in the habitats characterized by relatively sparse vegetative cover (< 40% cover). These sites were primarily domi¬ nated by bipedal species {Dipodomys sp.), while more densely vegetated sites (> 40.0% cover) contained primarily quadrepedal species. Some microhabitat theo¬ ries suggest that morphological adaptations associated with locomotion allow individuals to utilize specific microhabitats (Price, 1978; Price and Brown, 1983; Kotler, 1984). Large, bipedal kangaroo rats {Dipodomys sp.) are associated with open microhabi¬ tats, while small, quadrepedal pocket mice {Perognathus sp.) are associated with shrubby micro- habitat (Harris, 1984; Root 1997). For D. merriami, the presence of open areas is the most important fac¬ tor affecting its distribution (Congdon, 1974). Alter¬ natively, more dense habitats such as swales and sandy arroyo scrub may be critical in maintaining popula¬ tions of N. micropus, P. leucopus, and S. hispidus. The high average activities in Onychomys sp. may be the result of their feeding guild (carnivorous). Hav¬ ing a diet consisting of primarily insects may require members of this genus to travel greater lengths in pur¬ suit of food in comparison to coexisting herbivores. The low average activity found in S. hispidus may be a result of its herbivore characteristics as well. This herbivorous species predominately was found within the swale habitat, which contained the lowest average activity across species and is characterized by abun¬ dant, dense vegetation (> 80% vegetative cover). Stable population dynamics cannot be attributed to any single combination of traits (Brown and Zeng, 1989). Likewise, variation in species diversity, rodent density, and survivorship cannot be attributed to a single environmental factor. Vegetative composition and den¬ sity as well as rainfall have been shown to be factors affecting rodent community ecology. Not surprisingly, ecological components of the desert ecosystem are intertwined, and the organization of a community and its patterns of temporal change reflect the fluctuations of species populations that comprise the community (Brown and Heske, 1990). Clary et al.—Mammal Communities in the Chihuahuan Desert 13 Acknowledgments Funding for this project was provided by a De¬ partment of Defense grant (MIPR W52EU251606913) and was administered by William Whitworth (US AC ERL). We also thank the following personnel at Fort Bliss (Kelly Fischer, Shane Offutt, Will Roach, and Keith Landreth) as well as the Davis Dome staff for assistance during this project. Special thanks to Darin Bell, Darin Carroll, Cody Edwards, Kristina Halcomb, Ted Jolley, Oleksiy Knyazhnitskiy, Nicole Lewis-Oritt, Stacy Mantooth, Roslyn Martinez, Cole Matson, Anton Nekrutenko, Mark O’Neill, Lottie Pep¬ pers, Calvin Porter, Heather Roberts, Brenda Rodgers, Irene Tiemann-Boege, Jeff Wickliffe, and Frank Yancey, II, for assistance in collection and preparation of specimens. Jody Martin and Nick Parker of the Texas Cooperative Fish and Wildlife Research Unit, Texas Tech University provided important administra¬ tive support during this study. Specimens were col¬ lected under scientific collecting permit #2865 issued by the New Mexico Department of Game and Fish. Literature Cited Baumgardner, G. D. 1990. Mammal surveys on land condi¬ tion trend plots at Fort Hood Texas. Unpub¬ lished report for U. S. Army Construction Engi¬ neering Research Laboratory. Department of Wildlife & Fisheries Sciences, Texas A&M Uni¬ versity, College Station, Texas. Brattstrom, B. H., and M. C. Bondello. 1983. Effects of off¬ road vehicle noise on desert vertebrates. Pp. 167-206, in Environmental effects of off-road ve¬ hicles; impacts and management in arid areas (R. H. Wenn and H. G. Wilshire, eds.). Springer- Verlag, New York, New York. Brown, J. H. 1975. Geographical ecology of desert ro¬ dents. Pp. 315-341, in Ecology and Evolution of Communities (M. L. Cody & J. M Diamond, eds.). The Belknap Press of Harvard University Press, Cambridge, Massachusetts. Brown, J. H., and E. J. Heske. 1990. Temporal changes in a Chihuahuan desert rodent community. Oikos, 59:290-302. Brown, J. H., and Z. Zeng. 1989. Comparative population ecology of eleven species of rodents in the Chihuahuan Desert. Ecology, 70:1507-1525. Carroll, D. S., R. C. Dowler, and C. W. Edwards. 1999. Estimates of relative abundance of the medium¬ sized mammals at Fort Hood, Texas, using scent station visitation. Occasional Papers, Museum of Texas Tech University, 188:1-10. Christian, D. P. 1980. Vegetative cover, water resources, and microdistributional patterns in a desert ro¬ dent community. Journal of Animal Ecology, 49:807-816. Clary, M. L., D. M. Bell, C. W. Edwards, T. W. Jolley, O. Knyazhnitskiy, N. Lewis-Oritt, S. J. Mantooth, L. L. Peppers, I. Tiemann-Boege, F. D. Yancey, II, D. J. Howell, B. A. Locke, R. J. Baker, and R. D. Bra¬ dley. 1999. Checklist of mammals from twelve habitat types at Fort Bliss Military Base; 1997- 1998. Occasional Papers, Museum of Texas Tech University, 192; i +1-16. Congdon, J. 1974. Effects of habitat quality on distribu¬ tions of three sympatric species of desert rodents. Journal of Mammalogy, 55:659-662. Davis, W. B., and D. J. Schmidly. 1994. The Mammals of Texas. Texas Parks and Wildlife Department, Austin, X + 388 pp. Edwards, C. W., R. C. Dowler, and D. S. Carroll. 1998. Assessing medium-sized mammal abundance at Fort Hood military installation using live-trap- ping and spotlight counts. Occasional Papers, Museum of Texas Tech University, 185:1-23. Findley, J. S. 1987. The Natural History of New Mexican Mammals. University of New Mexico Press, Al¬ buquerque, x+150 pp. Findley, J. S., and W. Caire. 1974. The status of mammals in the northern region of the Chihuahuan Desert. Pp. 127-140, in Transactions of the Symposium on the Biological Resources of the Chihuahuan Desert region United States and Mexico (D. H. Riskind and R. H. Wauer, eds.). Sul Ross State University, Alpine, Texas. Gardner, J. L. 1951. Vegetation of the Creosotebush area of the Rio Grande valley in New Mexico. Eco¬ logical Monographs, 21:379-403. 14 Occasional Papers, Museum of Texas Tech University Gese, E. M., O. J. Rongstad, and W. R. Mytton. 1989. Change in coyote movements due to military ac¬ tivity. Journal of Wildlife Management, 53:334- 339. Ghobrial, L. L, and T. A. Nour. 1975. The physiological adaptations of desert rodents. Pp. 413-444, in Rodents in Desert Environments (I. Prakash and P. K. Ghosh, eds.). The Hague. W. Junk, Publish¬ ers. Hafner, M. S. 1977. Density and diversity in Mojave Desert rodent and shrub communities. Journal of Animal Ecology, 46:925-938. Harris, J. H. 1984. An experimental analysis of desert foraging ecology. Ecology, 65:1579-1584. Heske, E. J., J. H. Brown, and S. Mistry. 1994. Longterm experimental study of a Chihuahuan Desert ro¬ dent community: 13 years of competition. Ecol¬ ogy, 75:438-445. Jones, C., R. S. Hoffmann, D. W. Rice, M. D. Engstrom, R. D. Bradley, D. J. Schmidly, C. A. Jones, R. J. Baker. 1997. Revised checklist of North American mam¬ mals north of Mexico, 1997. Occasional Papers, Museum of Texas Tech University, 173:1-119. Jorgensen, E. E. 1996. Small mammal and herptofauna communities and habitat associations in foot¬ hills of the Chihuahuan Desert. Unpublished Ph.D. dissertation, Texas Tech University, Lub¬ bock, Texas. Jorgensen, E. E., and S. Demarais, 1996. Final report: small mammal and herpetofauna habitat associations and communities on the McGregor Range, Fort Bliss; Sacramento Mountain foothills. Director¬ ate of the Environment, Fort Bliss, El Paso, Texas. Kotler, B. P. 1984. Risk of predation and the structure of desert rodent communities. Ecology, 65:689-701. Kotler, B. P, and J. S. Brown. 1988. Environmental hetero¬ geneity and the coexistence of desert rodents. Annual Review of Ecology and Systematics, 19:281-307. Lebreton, J., K. P. Burnham, J. Clobert, and D. R. Ander¬ son. 1992. Modeling survival and testing bio¬ logical hypotheses using marked animals: a uni¬ fied approach with case studies. Ecological Monographs, 62:67-118. Longmire, J. L., M. Maltbie, and R. J. Baker. 1997. Use of “lysis buffer” in DNA isolation and its implica¬ tion for museum collections. Occasional Papers, Museum of Texas Tech University, 163:1-13. Matlab v. 5.2. 1998. The Math Works, Inc. Milstead, W. W. 1960. Relict species of the Chihuahuan Desert. Southwestern Naturalist, 5:75-88. Monasmith, T. J. 1997. Fire effects on small mammals and vegetation of the northern Chihuahuan Desert. Unpublished Master’s thesis, Texas Tech Uni¬ versity, Lubbock. Price, M. V. 1978. The role of microhabitat in structuring desert rodent communities. Ecology, 60:4-49. Price, M. V., and J. H. Brown. 1983. Patterns of morphol¬ ogy and resource use in North American desert rodent communities. Great Basin Naturalist Mem¬ oirs, 7:117-134. Rice, W. R. 1989. Analyzing tables of statistical test. Evolution, 43:223-225. Root, J. J. 1997. Microsite and habitat boundary influ¬ ence on small mammal capture, diversity, and movements. Unpublished Master’s Thesis, Texas Tech University, Lubbock, Texas. Schmidly, D. J. 1974. Factors governing the distribution of mammals in the Chihuahuan Desert region. Pp. 163-192, in Transactions of the Symposium on the Biological Resources of the Chihuahuan Desert region United States and Mexico (D. H. Riskind and R. H. Wauer, eds.). Sul Ross State University. Schmidt, R. H. 1986. Chihuahuan climate. Pp. 40-63, in Second symposium on Resources of the Chihuahuan Desert region. (J. C. Barlow, A. M. Powell, B. N. Timmermann, eds.). Chihuahuan Desert Institution. Alpine, Texas. Shaw, R. B., and V. E. Diersing. 1990. Tracked vehicle impacts on vegetation at the Pinon Canyon ma¬ neuver site, Colorado. Journal of Environmental Quality, 19:234-243. Simpson, E. H. 1949. Measurement of diversity. Nature, 163:688. Stephenson, T. R., M. R. Vaughn, and D. E. Andersen. 1996. Mule deer movements in response to mili¬ tary activity in southeast Colorado. Journal of Wildlife Management, 60:777-787. Thompson, S. D. 1982. Microhabitat utilization and for¬ aging behavior of bipedal and quadrupedal heteromyid rodents. Ecology, 63:1303-1312. Weeks, B. E. 1997. Niche partitioning mechanisms of desert heteromyid rodents. Unpublished Master’s Thesis, Texas Tech University, Lub¬ bock, Texas. Zeng, Z., and J. H. Brown. 1987. Population ecology of a desert rodent: Dipodomys merriami in the Chihuahuan Desert. Ecology, 68:1238-1340. Clary et al.— Mammal Communities in the Chihuahuan Desert 15 Addresses of authors: Melinda L. Clary Department of Biological Sciences Box 43131 Texas Tech University Lubbock, TX 79409-3131 Current Address: Environmental Project Manager LOPEZGARCIA GROUP 1825 Market Center Blvd., Ste. 500 Dallas. TX 75207 e-mail: mclary@wendylopez.com Richard E. Strauss Department of Biological Sciences Texas Tech University Lubbock. TX 79409-3131 e-mail: rich.strauss@ttu.edu Robert J. Baker Department of Biological Sciences and Museum Texas Tech University Lubbock. TX 79409-3131 e-mail: rjbaker@ttu. edu Brian A. Locke Senior Wildlife Ecologist Directorate of Environment Fort Bliss, TX 79916-6812 e-mail: lockeb@emhlO.bliss.army.mil Robert D. Bradley Department of Biological Sciences and Museum Texas Tech University Lubbock. TX 79409-3131 e-mail: rbradley@ttacs.ttu.edu Donna J. Howell HCR 64 Box 4 Ledoux, NM 87732 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 79409-3191. 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