Printer Friendly

Rodent communities of native woodland, replanted, and secondary succession sites in the Lower Rio Grande Valley, Texas.

Abstract. -- Wildlife habitat is rapidly disappearing in the Lower Rio Grande Valley of Texas and is in critical need of protection. United States Fish and Wildlife Service plans call for protection of 53,420 ha in the area, with the Rio Grande serving as the major corridor linking tracts of native and restored vegetation. Species richness, diversity, density, biomass and similarity of rodent communities were compared among a native woodland, a replanted field, and an unaided secondary succession site to obtain information on the efficacy of vegetation efforts in promoting rodent community diversity. Species diversity varied from H' = 0.0 to 0.65 depending on habitat, grid and season. Species richness ranged from 5 to 9 among the three habitats and a total of 10 species was captured. Density for all species combined ranged from 269 to 388 rodents/ha. Sigmodon hispidus, Peromyscus leucopus, and Liomys irroratus were the most abundant rodents in the native woodland and replanted habitats where they together comprised 88% and 90%, respectively, of the individuals captured. Mus musculus replaced P. leucopus in abundance in the unaided succession habitat. Biomass was greatest where S. hispidus was most abundant. Community similarity was greater between the two successional habitats than either was to the native woodland. The native woodland had fewer species but greater evenness than either successional habitat. The replanting technique used by the United States Fish and Wildlife Service in a formerly cultivated field produced greater diversity of rodents in less time than unaided secondary succession of a fallow field. Variation in application of planting techniques can produce significant differences in vegetation and rodent communities on small replanted areas.

**********

The Lower Rio Grande Valley (LRGV) of Texas is both a political unit and a biogeographic unit (Judd et al. 2002). As a political entity, it includes the four southernmost counties in the state, i.e., Cameron, Hidalgo, Starr and Willacy. Biogeographically, it corresponds closely with the Matamoran District of the Tamaulipan Biotic Province (Blair 1950). This 1.2 million ha region exhibits great biodiversity. More than 500 vertebrate and 170 woody species occur in the LRGV (Judd et al. 2002) and 67 species are listed as threatened or endangered (Jahrsdoerfer & Leslie 1988). Because of its high biodiversity, large number of threatened and endangered species, large number of neotropical species that reach the northern limit of their distribution in the LRGV (Blair 1950; Oberholser 1974; Lonard & Judd 1993), and small amount of native habitat remaining (Purdy 1983; United States Fish and Wildlife Service 1985; Jahrsdoerfer & Leslie 1988), state and federal governments and non-profit organizations are purchasing lands for preservation and revegetation.

It is estimated that less than 5% of the habitat that originally covered the LRGV still remains (Purdy 1983; Jahrsdoerfer & Leslie 1988). Indeed, the United States Fish and Wildlife Service (USFWS), Texas Parks and Wildlife Department (TPWD) and Texas Nature Conservancy have identified the LRGV as an area where wildlife habitat is rapidly disappearing and in dire need of protection. In 1979, the Lower Rio Grande Valley National Wildlife Refuge (LRGVNWR) was established to implement a USFWS Land Protection Plan that calls for protection of 53,420 ha in the LRGV with the Rio Grande serving as the major corridor linking tracts of native and restored vegetation (USFWS 1985). When completed, the Lower Rio Grande Wildlife Corridor will extend 240 km from the mouth of the river in Cameron County to Falcon Dam in Starr County. The refuge currently consists of over 138 tracts comprising about 31,697 ha. To date, about 4,047 ha have been replanted.

Lands acquired for the corridor usually are fields that have been in cultivation. Left alone, these abandoned fields undergo secondary succession. The first plants to become established typically are herbaceous annuals (Vora & Messerly 1990). In time, these colonizing species are gradually replaced by woody species. The rate at which succession proceeds depends, in part, on the ability of mid- and late-successional species to disperse to a site and successfully compete with colonizing species that are already established. Revegetation projects based on the Facilitation Model of succession (Connell & Slatyer 1977) attempt to accelerate succession by introducing climax species into an area (Judd et al. 2002).

Due to conversion of many riparian areas of the southwestern USA to urban and agricultural uses, restoration projects are becoming common. Still, there is a paucity of information regarding the affects of restoration efforts on rodent communities (Anderson 1994; Ellis et al. 1997; Ellison & Van Riper 1998; Patten 1997). Judd et al. (2002) and Sternberg (2003) have assessed success of TPWD and USFWS revegetation efforts in accelerating plant succession and in achieving similar composition and structure as native woodlands, but there is no published information on animal communities of replanted tracts in the LRGV.

This study compares species richness, diversity, density, biomass and similarity of rodent communities among native woodland, replanted field and unaided secondary succession sites in a riparian area. It provides information on the efficacy of revegetation efforts at promoting rodent community diversity in the LRGV by investigating the null hypothesis: there are no significant differences in rodent species richness, diversity, density or biomass at native woodland, replanted field, and unaided secondary succession sites.

MATERIALS AND METHODS

Study area. -- The study sites are located approximately 6 km south of Weslaco, Hidalgo County, Texas, and about 1 km north of the Rio Grande. The McManus Unit of the Las Palomas Wildlife Management Area is 22 ha of relatively undisturbed, thorn woodland managed by TPWD. Coma (Sideroxylon celastrina) and cedar elm (Ulmus crassifolia) are the dominant trees (Sternberg 2003). Coma and granjeno (Celtis pallida) are the dominant shrubs and crucita (Chromolaena odorata) is the dominant species in the ground layer. There is a dense cover of trees and shrubs at this site (Sternberg 2003).

The La Coma tract, managed by the USFWS, includes a 14.99 ha farm field that was replanted in October 1995 with 7,948 seedlings of 31 native shrub and tree species. Seedling density was 592 individuals/ha and plant survivorship was estimated at 85% in July 1996. Sternberg (2003) gives the number of seedlings planted for each of the 31 species. Tenaza (Havardia pallens) was the most abundant with 925 individuals. Tepeguaje (Leucaena pulverulenta) and jara dulce (Baccharis neglecta) were dominant trees at the replanted site and jara dulce was the dominant shrub when the area was sampled from November 1999 to January 2000. Guineagrass (Urochloa maxima) and jara dulce were the dominant species in the ground layer. Cover provided by trees and shrubs was markedly less than at the adjacent native woodland.

The La Coma tract also includes an unaided secondary succession site of about 55 ha. It is a field that was in agricultural production until 1985 when USFWS purchased the land and allowed it to go fallow. Mesquite (Prosopis glandulosa), retama (Parkinsonia aculeata) and lotebush (Ziziphus obtusifolia) were the only trees present at this site and cover provided by trees and shrubs was 6.5 to 19.8 times less than at the native woodland (Sternberg 2003). Mesquite was the dominant shrub and guineagrass was the dominant species in the ground layer (Sternberg 2003).

Rodent trapping. -- Two 25-trap grids were established in each habitat. An edge grid was located within 50 m of the border of a given habitat and an interior grid was placed randomly within a site, but more than 30 m distant from the edge grid. Sherman live-traps (8 cm W by 9 cm H by 23 cm L) were set at 10-m intervals and baited with rolled oats. On 13 December 1998, a temperature decrease of 17.8[degrees]C from 21.7[degrees]C to 3.9[degrees]C killed nine trapped Mexican spiny pocket mice (Liomys irroratus). Thereafter, cotton bedding was added to the traps (Schweiger et al. 2000) and peanut butter was mixed with rolled oats to prevent additional mortality, when temperature was forecast to be lower than 15[degrees]C. No additional deaths of trapped Mexican spiny pocket mice occurred. Rodents were trapped every other week in each of the habitats. Trapping commenced on 1 November 1998 and ended on 1 December 1999. Thus, each grid was trapped 25 times. Edge and interior grids were trapped on separate, but consecutive nights with the starting order alternating among trapping periods. Traps were set at dusk, checked at dawn and left closed and in place between trapping periods.

Rodents were weighed to the nearest 0.5 g using a Pesola[R] spring balance, marked uniquely by toe-clipping and released at the point of capture following guidelines of the Ad Hoc Committee for Acceptable Field Methods in Mammalogy (1987). Individuals were identified to species, gender, and age using criteria in Genoways (1973), Batzli (1977), Holbrook (1979), Davis & Schmidly (1994), McMurry et al. (1994) and Schmidt & Engstrom (1994).

Analyses. -- Species richness, species diversity, and evenness were compared between edge and interior grids of a given habitat using the seasonal values for these parameters as samples in Student t-tests. Species richness, species diversity, and evenness were compared among habitats and seasons using two-way analysis of variance (ANOVA). Student t-tests were used to identify significant differences in pair-wise comparisons of the parameters between habitats.

Species richness was the count of species present. Species diversity was assessed using the Shannon diversity index based on log[.sub.10] (Brower et al. 1998; Krebs 1999). For seasonal comparisons, Winter was 14 November 1998 to 1 March 1999; Spring 2 March to 2 June 1999; Summer 3 June to 2 September 1999; and Fall 3 September to 13 November 1999. The last trapping period was included in the annual data, but was omitted from seasonal comparisons so that each season would have six trapping periods.

Similarity of rodent communities was compared between habitats using the Percent Similarity test (Brower et al. 1998). Density is reported as numbers per hectare. The Jolly-Seber method (Krebs 1999) was used to estimate total rodent density and density of three species (i.e., Mexican spiny pocket mouse, white-footed mouse [Peromyscus leucopus] and hispid cotton rat [Sigmodon hispidus]) that were recaptured frequently enough to permit calculation of the Mean Maximum Distance Moved (MMDM). MMDM was added to the periphery of a grid to calculate species-specific grid size for each of the three species. Density of all rodents combined was compared among habitats using the edge and interior grids within a habitat as samples in a one-way ANOVA. Student t-tests were used in pair-wise comparisons to assess differences in the density of Mexican spiny pocket mice, white-footed mice, and hispid cotton rats among habitats.

Biomass was the sum of the average weights of individuals of a given species captured at least three times on a given grid. After applying the MMDM to the species-specific grid size, biomass per hectare for each species was calculated. Defined in this way, biomass estimates were not possible for six species represented by very few individuals and captures. Total biomass of resident rodents was compared among habitats using grids within a habitat as samples in a one-way ANOVA.

RESULTS

Species richness, evenness, and diversity. -- A total of 3,750 trap-nights resulted in 2,122 captures of 923 individuals. Species richness per grid ranged from 5 to 9, and 10 species were captured among all grids (Table 1). Grids in the same habitat generally had similar species richness, but edge grids in the replanted and unaided succession habitats had two more species than the interior grids. Edge grids in the replanted and unaided succession habitats had four (44.5%) more species than either grid in the native woodland. At the native woodland, three species (hispid cotton rat, white-footed mouse, and Mexican spiny pocket mouse) accounted for 88.0% of the individuals captured. At the replanted site these three species accounted for 90.3% of the individuals captured. At the unaided succession site, hispid cotton rat, Mexican spiny pocket mouse, and house mouse (Mus musculus) comprised 93.1% of the individuals captured. Thus, house mice replaced white-footed mice in abundance in the unaided succession habitat.

Species richness, species diversity, and evenness were compared between edge and interior grids to determine if there was significant variation between the grids of a habitat on an annual basis (Table 2). There was no significant difference in species richness between grids of the native woodland (t = 0.658, 6 df, P > 0.5), replanted (t = 0.918, 6 df, P > 0.2), or unaided succession (t = 0.343, 6 df, P > 0.5) habitats. Likewise, there was no significant difference in species diversity between grids of the native woodland (t = 0.934, 6 df, P > 0.2), replanted (t = 0.934, 6 df, P > 0.2), or unaided succession (t = 0.389, 6 df, P > 0.5) habitats. Similarly, evenness did not exhibit significant variation between grids of the native woodland (t = 0.723, 6 df, P > 0.4), replanted (t = 0.625, 6 df, P > 0.5) or unaided succession (t = 0.634, 6 df, P > 0.5) habitats.

Species richness was relatively stable among seasons at the native woodland and most variable among seasons in the unaided succession habitat (Table 2). There was significant variation in species richness among seasons and habitats, and due to interaction (Table 3). Between habitat differences were due to values in spring and summer (Table 2). The mean for these two seasons combined was significantly greater for the replanted habitat (n = 4, [bar.x] = 5.75, SD = 0.5) than either the native woodland (n = 4, [bar.x] = 4.75, SD = 0.5) (t = 2.829, 6 df, P < 0.05) or the unaided succession habitat (n = 4, [bar.x] = 2.75, SD = 1.5) (t = 3.795, 6 df, P < 0.01). Species richness of the native woodland also was greater in spring and summer than in the unaided succession habitat (t = 2.530, 6 df, P < 0.05).

Differences in the number of individuals captured per species were reflected in evenness values (Tables 1 and 2). Evenness was greatest in the native woodland and lowest in the unaided succession habitat where hispid cotton rats alone accounted for 79.9% of the individuals captured. The only source of significant variation for evenness was among habitats (Table 3). Comparison of mean annual values showed that the native woodland (n = 8, [bar.x] = 0.8222, SD = 0.084) had a significantly greater mean than the replanted (n = 8, [bar.x] = 0.6548, SD = 0.698) (t = 4.326, 14 df, P < 0.001) and unaided succession habitats (n = 8, [bar.x] = 0.3555, SD = 0.180) (t = 6.639, 14 df, P < 0.001). Furthermore, the replanted habitat had significantly greater mean annual evenness than the unaided succession habitat (t = 4.382, 14 df, P < 0.001).

Species diversity was consistently lower at the unaided succession site than at the replanted or native woodland habitats (Table 2) and markedly so in spring, summer and fall. Variation was significant among habitats and seasons (Table 3). There was no consistent pattern of seasonal variation in species diversity. Species diversity was greatest in spring in the native woodland and in winter in the replanted and unaided succession habitats. Species diversity was least in fall in the native woodland and replanted habitats, and in spring in the unaided succession habitat. Mean annual species diversity was significantly greater in the native woodland (n = 8, [bar.x] = 0.5534, SD = 0.054) than in the replanted (n = 8, [bar.x] = 0.4669, SD = 0.097) (t = 2.212, 14 df, P < 0.05) or unaided succession habitats (n = 8, [bar.x] = 0.2362, SD = 0.155) (t = 5.478, 14 df, P < 0.01). The replanted habitat also had significantly greater mean annual species diversity than the unaided succession habitat (t = 3.571, 14 df, P < 0.01).

Community similarity. -- Information needed for calculating Percentage Similarity values is provided in Table 1 and the resulting values are compared between the two grids of the same habitat and among grids of differing habitats in Table 4. Edge and interior grids of the native woodland and unaided succession habitats had a high degree of similarity (86.3% and 91.2% respectively), but the grids of the replanted habitat had only a modest degree of similarity (59.6%). Unaided succession grids showed lower similarity with native woodland grids than did the replanted grids (Table 4). The replanted edge grid had greater similarity with both the unaided succession grids than it did with the replanted interior grid.

Density. -- Estimates of total rodent density ranged from 262 to 387 rodents per ha and the 95% confidence intervals were broad (Table 5). Variation in total rodent density among habitats was not significant (F = 0.178, 2 & 3 df, P > 0.75). Only five white-footed mice and 23 Mexican spiny pocket mice were captured on the unaided succession grids (Table 1), and there were too few recaptures of these individuals to permit calculation of density estimates in this habitat. There was no significant difference in the density of white-footed mice in the native woodland and replanted habitats (t = 0.811, 2 df, P > 0.49), but the density of Mexican spiny pocket mice was significantly greater (t = 8.732, 2 df, P < 0.02) in the native woodland (Table 5). ANOVA of hispid cotton rat density among the three habitats indicated no significant variation (F = 1.863, 2 & 3 df, P > 0.25), but t-tests revealed that the density of hispid cotton rats was significantly greater (t = 5.150, 2 df, P < 0.05) in the unaided succession than in the native woodland.

Biomass. -- Total rodent biomass ranged from 3.2kg/ha on the replanted interior grid to 12.1 kg/ha on the replanted edge grid (Fig. 1). The variation in total rodent biomass between the two replanted site grids was greater than that among all the other grids and as a consequence the variation among habitats was not significant (F = 0.120, 2 & 3 df, P > 0.75). Both grids of the unaided succession habitat had greater rodent biomass than the native woodland grids and the replanted interior grid. This was due to the abundance of hispid cotton rats on the unaided succession grids. Similarly, the high biomass on the replanted edge grid was a result of the high abundance of hispid cotton rats there. The low total biomass of the replanted interior grid was due to the low numbers of hispid cotton rats and the high abundance of the white-footed mouse, which is a relatively small rodent. An adult hispid cotton rat weighs about 110g while an adult white-footed mouse weighs approximately 22g. Thus hispid cotton rats are about five times heavier than white-footed mice.

[FIGURE 1 OMITTED]

DISCUSSION

The null hypothesis that there were no significant differences in rodent species richness, species diversity, density, or biomass among the three habitats was falsified for species richness, species diversity and density for two of the three most abundant species. However, the findings reported here do not necessarily extend to all riparian woodland communities of the Lower Rio Grande Valley because Lonard & Judd (2002) reported that at least three different riparian communities exist between the mouth of the river and Falcon Dam. The habitats investigated in this study were stages in succession of the same Mid-Valley riparian woodland community.

Maintaining and restoring species diversity is a major management objective of the USFWS and it was a point of interest in this study. Two lines of evidence suggest that the replanting technique used by USFWS achieves greater rodent species diversity in less time than unaided secondary succession of fallow fields. First, the replanted habitat had 10 fewer years for development (replanted in 1995) than the unaided secondary succession (fallow since 1985) yet it had significantly greater rodent species diversity. Second, there was no significant difference in rodent species diversity of the replanted habitat and the native woodland that had been undisturbed for more than 40 years. It is important to note that the replanted site had only three years of progress when this study was begun, so it was far from being "restored" (i.e., at or near climax).

The replanted and unaided succession habitats had identical species richness (nine species). Thus, the difference in species diversity of the two habitats was due to greater evenness in the replanted habitat. There are no published studies of rodent species diversity in the LRGV of Texas, and Blair (1952) is the only author to provide information on rodent species richness in LRGV woodlands. The same seven species that he captured were all captured in this study.

Two studies report rodent species diversity at sites in southern Texas north of the LRGV. Nolte & Fulbright (1997) found rodent species diversity values ranging from H' = 0.39 to 0.79 in mesquite grasslands 280 km north of the LRGV in San Patricio County, and Windberg (1998) reported a rodent species diversity of H' = 0.63 in Webb County 230 km northwest of the LRGV. Rodent species diversity in this study varied from H' = 0.0 to H' = 0.65 depending on habitat and season. However, most values ranged from H' = 0.20 to H' = 0.65 and, consequently, were relatively similar to those reported by Nolte & Fulbright (1997) and Windberg (1998).

There are no studies reporting rodent species diversity of replanted sites in southern Texas, but Joule & Cameron (1974) found a rodent species diversity of H' = 0.40 in coastal prairie south of Houston where hispid cotton rats were the dominant species. The unaided secondary succession habitat in this study was principally grassland where hispid cotton rats were the dominant species, but it achieved a species diversity of this magnitude only in winter. Grasses also dominated the replanted edge grid (Sternberg 2003) and hispid cotton rats were the dominant species. It had H' values above 0.40 in all four seasons. Likewise, the native woodland had rodent species diversity values above 0.40 in all seasons.

Greater species richness in the replanted and unaided secondary succession habitats than in the native woodland may have been due to greater density of grasses in the successional habitats. Greater grass cover might have allowed the presence of species such as fulvous harvest mouse (Reithrodontomys fulvescens), Coues' rice rat (Oryzomys couesi) and marsh rice rat (Oryzomys palustris) which require habitats with dense grass cover (Davis & Schmidly 1994).

The comparatively low similarity of rodent communities of the edge and interior grids of the replanted habitat was likely due to difference in the vegetation of the two grids and the numbers of cotton rats and white-footed mice on the grids. The replanted edge grid was dominated by grasses in the ground layer (Mean Percent Cover [MPC] = 73.1) while the replanted interior grid was dominated by jara dulce in all three vegetation layers, i.e., tree (MPC = 61.5), shrub (MPC = 58.6) and ground (MPC = 22.1) layers (C. Best pers. comm.; Sternberg 2003). Hispid cotton rats were numerically dominant on the edge grid where grasses dominated the ground cover and white-footed mice were numerically dominant where the woody jara dulce was the principal vegetation. Likewise, the greater similarity in the rodent communities of the unaided succession and the replanted habitat than either was in comparison to the native woodland was likely due to the presence of grassland habitat in the two successional communities and the low abundance of grass in the native woodland. Absence of jara dulce on the edge of the replanted habitat was probably due to a delay in the plowing of this area prior to planting seedlings. Thus, unlike the interior area of the habitat, a bed was not prepared on the edge of the habitat for dispersing jara dulce seeds (Sternberg 2003). Clearly, small variations in revegetation techniques can lead to significant differences in rodent communities. The differences may be temporary, however, as the rodent community of the replanted habitat is likely to change as the vegetation develops.

Seasonal variation in habitat use by rodent species is well known (Fleharty et al. 1972; Larrison & Johnson 1973; Whitford 1976; Kitchings & Levy 1981; Turner & Grant 1987; Foster & Gaines 1991; Heske et al. 1997; Hanley & Barnard 1999). Thus, seasonal variation in species diversity, such as is reported here, is expected. Some species such as house mice were present on a grid only for several trapping periods and never seen again on the same grid. This may have been due to a dispersal event resulting from crop harvesting of nearby agricultural fields. Clearly, studies conducted for less than a year will likely yield biased estimates of species diversity and richness.

Hispid cotton rat density is correlated positively with grass cover (Cameron 1977; Cameron & Kincaid 1982; Kincaid et al. 1983; Turner & Grant 1987). High hispid cotton rat density also occurred in this study on the unaided succession grids and the replanted edge grid where grass cover was abundant. Cameron (1977; 2003) reported mean Minimum Number Alive values of 6 and 15 hispid cotton rats per hectare (respectively) in eastern Texas, while grids in this study supported hispid cotton rat densities two to 40 times greater.

Similarly, estimates of white-footed mouse density in the native woodland (71 to 81/ha) of this study were markedly greater than mean estimates reported by Wilkins (1995) for a wooded area of east central Texas (16.6 to 32.2/ ha). Wolf & Batzli (2004) suggest that forest edges have lower habitat quality for white-footed mice due to higher predation. Interior grids in this study had higher densities of white-footed mice, Mexican spiny pocket mice, and hispid cotton rats in all habitats except for the hispid cotton rat in the replanted edge grid.

Density of the Mexican spiny pocket mouse has not been reported previously. Estimates of density for this species here are limited to the native woodland and replanted habitats as Mexican spiny pocket mice were few in number and not recaptured frequently enough to yield reliable density estimates for the unaided succession habitat.

Total rodent community biomass has not been reported previously for any habitat in southern Texas. Values obtained in this study ranged from 3.2 to 12.1 kg/ha depending upon habitat and grid, and all but one grid (replanted interior) had total rodent biomass above 6.5kg/ha. These values were markedly greater than the 0.45 to 0.90kg/ha reported by Henke & Bryant (1999) for rodent communities in west Texas and the value of 4.7kg/ha reported by Grant et al. (1985) for east central Texas. Clearly, the habitats of subtropical LRGV support high rodent biomass.

The purpose of the LRGV reforestation effort is to establish a wildlife corridor along the Rio Grande from its mouth to Falcon Dam. Wildlife monitoring on remnant native woodlands should be initiated to provide baseline information on species distribution and density. Also, future revegetation efforts in the LRGV should establish control areas that are contemporaneous with the planted areas but that are not planted. Doing so will facilitate comparisons of the success of species recruitment to the two treatment areas and the development of communities on them so that the efficacy of revegetation efforts can be assessed.

ACKNOWLEDGMENTS

This paper is part of a Master's thesis by M. Sternberg submitted to the Department of Biology at the University of Texas-Pan American. Thanks go to E. Flores and M. V. Sternberg for assistance in the field and S. Benn, C. Best, D. Blankinship, K. Merritt, J. Rupert, and G. Waggerman for access to the McManus Unit and La Coma tract. We are grateful to the staff at the Lower Rio Grande Valley National Wildlife Refuge for use of Sherman live-traps and a vehicle. This work was supported, in part, by two James-Ware-Foltz scholarships to M. Sternberg. We thank K. Wilkins and two anonymous reviewers for insightful comments, which improved the manuscript. The views expressed in this manuscript are those of the authors and not necessarily those of the United States Fish and Wildlife Service.

LITERATURE CITED

Ad hoc Committee on Acceptable Field Methods in Mammalogy. 1987. Acceptable field methods in mammalogy: preliminary guidelines approved by the American Society of Mammalogists. Journal of Mammalogy 68, Supplement: 1-18.

Anderson, D. C. 1994. Demographics of small mammals using anthropogenic desert riparian habitat in Arizona. Journal of Wildlife Management, 58:445-454.

Batzli, G. O. 1977. Population dynamics of the white-footed mouse in floodplain and upland forests. American Midland Naturalist, 97:18-32.

Blair, W. F. 1950. The biotic provinces of Texas. Texas Journal of Science, 2(1):93-117.

Blair, W. F. 1952. Mammals of the Tamaulipan Biotic Province in Texas. Texas Journal of Science, 4(2):230-250.

Brower, J. E., J. H. Zar, & C. N. von Ende. 1998. Field and Laboratory Methods for General Ecology. McGraw-Hill, Boston, Massachusetts, 273 pp.

Cameron, G. N. 1977. Experimental species removal: demographic responses by Sigmodon hispidus and Reithrodontomys fulvescens. Journal of Mammalogy, 58:488-506.

Cameron, G. N. 2003. Prey use by a pair of Barn Owls (Tyto alba) in a Texas coastal habitat. Texas Journal of Science, 55(3):219-224.

Cameron, G. N. & W. B. Kincaid. 1982. Species removal effects on movements of Sigmodon hispidus and Reithrodontomys fulvescens. American Midland Naturalist, 108:60-67.

Connell, J. H. & R. O. Slayter. 1997. Mechanisms of succession in natural communities and their role in community stability and organization. American Naturalist, 111:1119-1144.

Davis, W. B. & D. J. Schmidly. 1994. The Mammals of Texas. Texas Parks and Wildlife Department. Austin, Texas. 338 pp.

Ellis, L. M., C. S. Crawford & M. C. Molles, Jr. 1997. Rodent communities in native and exotic riparian vegetation in the Middle Rio Grande Valley of central New Mexico. Southwestern Naturalist, 42(1):13-19.

Ellison, L. E. & C. Van Riper, III. 1998. A comparison of small mammal communities in a desert riparian floodplain. Journal of Mammalogy, 79:972-985.

Fleharty, E. D., J. R. Choate & M. A. Mares. 1972. Fluctuations in population density of the hispid cotton rat: factors influencing a "crash". Bulletin of the Southern California Academy of Sciences, 71:132-138.

Foster, J. & M. S. Gaines. 1991. The effects of a successional habitat mosaic on a small mammal community. Ecology, 72:1358-1373.

Genoways, H. H. 1973. Systematics and evolutionary relationships of spiny pocket mice, Genus Liomys. Special Publications of The Museum of Texas Tech University, 5:1-368.

Grant, W. E., P. E. Carothers & L. A. Gidley. 1985. Small mammal community structure in the Post Oak Savannah of East-Central Texas. Journal of Mammalogy, 66:589-594.

Hanley, T. A. & J. C. Barnard. 1999. Spatial variation in population dynamics of Sitka mice in floodplain forests. Journal of Mammalogy, 80:866-879.

Henke, S. E. & F. C. Bryant. 1999. Effects of coyote removal on the faunal community in western Texas. Journal of Wildlife Management, 63:1066-1081.

Heske, E. J., D. L. Rosenblatt & D. W. Sugg. 1997. Population dynamics of small mammals in an oak woodland-grassland-chaparral habitat mosaic. Southwestern Naturalist, 42(1): 1-12.

Holbrook, S. J. 1979. Habitat utilization, competitive interactions, and coexistence of three species of cricetine rodents in east central Arizona. Ecology, 60:758-769.

Jahrsdoerfer, S. E. & D. M. Leslie, Jr. 1988. Tamaulipan brushland of the Lower Rio Grande Valley of south Texas: description, human impacts, and management implications. U. S. Fish & Wildlife Service. Biological Report, 88 (36). 63 pp.

Joule, J. & G. N. Cameron. 1974. Field estimation of demographic parameters: influence of Sigmodon hispidus population structure. Journal of Mammalogy, 55:309-318.

Judd, F. W., R. I. Lonard & G. L. Waggerman. 2002. Evaluation of facilitated succession at Las Palomas Wildlife Management Area in South Texas. Texas Journal of Science, 54(2): 163-176.

Kincaid, W. B., G. N. Cameron & B. A. Carnes. 1983. Patterns of habitat utilization in sympatric coastal prairie rodents. Ecology, 64:1471-1480.

Kitchings, J. T. & D. J. Levy. 1981. Habitat patterns in a small mammal community. Journal of Mammalogy, 62:820-824.

Krebs, C. J. 1999. Ecological methodology. 2nd ed. Addison Wesley Longman, Menlo Park, California, 620 pp.

Larrison, E. J. & D. R. Johnson. 1973. Density changes and habitat affinities of rodents of shadscale and sagebrush associations. Great Basin Naturalist, 33:255-264.

Lonard, R. I. & F. W. Judd. 1993. Phytogeography of the woody flora of the Lower Rio Grande Valley, Texas. Texas Journal of Science, 45(2):133-147.

Lonard, R. I. & F. W. Judd. 2002. Riparian vegetation of the lower Rio Grande. Southwestern Naturalist, 47(3):420-432.

McMurry, S. T., R. L. Lochmiller, J. F. Boggs, D. M. Leslie, Jr. & D. M. Engle. 1994. Demographic profiles of populations of cotton rats in a continuum of habitat types. Journal of Mammalogy, 75:50-59.

Nolte, K. R. & T. E. Fulbright. 1997. Plant, small mammal, and avian diversity following control of honey mesquite. Journal of Range Management, 50:205-212.

Norusis, M. J. 1990. SPSS Base System User's Guide. SPSS Inc., Chicago, Illinois. 528 pp.

Oberholser, H. C. 1974. The bird life of Texas. University of Texas Press, Austin, Texas, 781 pp.

Patten, M. A. 1997. Reestablishment of a rodent community in restored desert scrub. Restoration Ecology, 5:156-161.

Purdy, P. C. 1983. Agricultural, industrial, and urban development in relation to the eastern white-winged dove. Unpubl. M.S. thesis, Colorado State University, Fort Collins, Colorado, 117 pp.

Schmidt, C. A. & M. D. Engstrom. 1994. Genic variation and systematics of rice rats (Oryzomys palustris species group) in southern Texas and northeastern Tamaulipas, Mexico. Journal of Mammalogy, 75:914-928.

Schweiger, E. W., J. E. Diffendorfer, R. D. Holt, R. Pierotti & M. S. Gaines. 2000. The interaction of habitat fragmentation, plant, and small mammal succession in an old-field. Ecological Monographs, 70:383-400.

Sokal, R. R. & F. J. Rohlf. 1995. Biometry: the principles and practices of statistics in biological research. W. H. Freeman and Company, New York, 887 pp.

Sternberg, M. A. 2003. Comparison of native brushland, replanted, and unaided secondary succession plant communities in the Lower Rio Grande Valley of Texas. Texas Journal of Science, 55(2):129-148.

Turner, C. L. & W. E. Grant. 1987. Effect of removal of Sigmodon hispidus on microhabitat utilization by Baiomys taylori and Reithrodontomys fulvescens. Journal of Mammalogy, 68:80-85.

United States Fish & Wildlife Service. 1985. Land protection plan for Lower Rio Grande National Wildlife Refuge in Cameron, Hidalgo, Starr, and Willacy Counties, Texas. United States Fish & Wildlife Service, Albuquerque, NM, 12 pp.

Vora, R. S. & J. F. Messerly. 1990. Changes in native vegetation following different disturbances in the Lower Rio Grande Valley, Texas. Texas Journal of Science 42(2):151-158.

Whitford, W. G. 1976. Temporal fluctuations in density and diversity of desert rodent populations. Journal of Mammalogy, 57:351-369.

Wilkins, K. T. 1995. The rodent community and associated vegetation in a tallgrass blackland prairie in Texas. Texas Journal of Science, 47(4):243-262.

Windberg, L. A. 1998. Population trends and habitat associations of rodents in southern Texas. American Midland Naturalist, 140:153-160.

Wolf, M. & G. O. Batzli. 2004. Forest edge--high or low quality habitat for white-footed mice (Peromyscus leucopus)? Ecology, 85:756-769.

MAS at: mitch_sternberg@fws.gov

Mitchell A. Sternberg* and Frank W. Judd

Department of Biology, The University of Texas--Pan American

Edinburg, Texas 78541-2999

*Current address:

Lower Rio Grande Valley National Wildlife Refuge

Route 2, Box 202-A, Alamo, Texas 78516
Table 1. Comparison of species richness and total number of individuals
captured on edge and interior grids at native woodland, replanted, and
unaided succession sites in Hidalgo County, Texas.

 Native Woodland Replanted Unaid. Succession
Species Edge Interior Edge Interior Edge Interior

Sigmodon hispidus 38 24 114 34 125 142
Peromyscus leucopus 39 41 35 88 2 3
Liomys irroratus 45 33 19 16 16 7
Mus musculus 0 0 8 5 7 14
Neotoma micropus 8 16 1 0 1 0
Rattus rattus 3 3 4 3 1 3
Oryzomys couesi 0 0 4 3 5 3
Reithrodontomys 0 0 2 2 1 3
 fulvescens
Chaetodipus hispidus 0 0 0 0 1 0
Oryzomys palustris 0 0 1 0 0 0
Total individuals 133 117 188 151 159 175

Table 2. Comparison of species richness (R), evenness (J'), and
Shannon-Wiener diversity (H') of rodents on edge and interior grids from
Winter, Spring, Summer, and Fall at native woodland, replanted, and
unaided succession habitats in Hidalgo County, Texas.

 Native Woodland Replanted Unaid. Succession
Parameters Edge Interior Edge Interior Edge Interior

Winter
 R 5 5 7 6 7 6
 J' 0.8186 0.7107 0.7153 0.6956 0.4704 0.5657
 H' 0.5722 0.4968 0.6460 0.5413 0.4248 0.4402

Spring
 R 5 5 6 6 2 4
 J' 0.9114 0.8350 0.7203 0.5830 0.2109 0.3282
 H' 0.6370 0.5836 0.5035 0.4537 0.0635 0.1976

Summer
 R 4 5 6 5 4 1
 J' 0.9552 0.8258 0.5409 0.6029 0.4830 0
 H' 0.5751 0.5772 0.4209 0.4214 0.2908 0

Fall
 R 4 4 4 3 4 4
 J' 0.7158 0.8053 0.7086 0.6747 0.4263 0.3596
 H' 0.5004 0.4849 0.4266 0.3219 0.2567 0.2165

Table 3. Two-way ANOVAs for species richness, species diversity and
evenness. Source = source of variation, df = degrees of freedom, SS =
sums of squares, MS = mean square, F = ANOVA value, P = probability
level. Each is a Model I ANOVA.

Parameter Source df SS MS F P

Species Seasons 3 16.34 5.45 7.26 <0.005
Richness Habitats 2 7.59 3.80 5.06 <0.05
 Interaction 6 14.40 2.40 3.20 <0.05
 Error 12 9.00 0.75
Species Seasons 3 0.087 0.029 5.07 <0.025
Diversity Habitats 2 0.430 0.215 37.72 <0.001
 Interaction 6 0.099 0.016 2.88 >0.05
 Error 12 0.069 0.006
Evenness Seasons 3 0.028 0.009 0.35 >0.75
 Habitats 2 0.895 0.447 16.76 <0.001
 Interaction 6 0.119 0.020 0.74 >0.50
 Error 12 0.320 0.027

Table 4. Comparison of community similarity (%) among grids. NWE =
Native Woodland Edge, NWI = Native Woodland Interior, RE = Replanted
Edge, RI = Replanted Interior, USE = Unaided Succession Edge, USI =
Unaided Succession Interior.

Grids NWI RE RI USE USI

NWE 86.3 59.9 64.4 41.2 27.9
NWI 51.8 68.1 33.1 27.9
RE 59.6 80.1 75.1
RI 40.4 36.2
USE 91.2

Table 5. Comparison of Jolly-Seber density estimates (number/ha) among
three habitats in Hidalgo County, Texas. Each habitat had an edge and
interior grid. Density estimates are the means of 23 trapping periods
between 14 November 1998 and 1 December 1999. Ninety-five percent
confidence intervals for the estimates are in parenthesis. An asterisk
(*) indicates there were too few captures to calculate a density
estimate.

 Native Woodland Replanted
Category Edge grid Interior grid Edge grid

Total rodents 322 (122-414) 270 (108-1052) 388 (255-1084)
Liomys irroratus 94 (69-118) 90 (52-128) 19 (3-36)
Peromyscus leucopus 81 (59-103) 71 (45-97) 64 (30-99)
Sigmodon hispidus 81 (43-118) 47 (34-60) 230 (188-272)

 Replanted Unaided Succession
Category Interior grid Edge grid Interior grid

Total rodents 263 (185-734) 311 (175-585) 269 (171-538)
Liomys irroratus 2 (0-5) * *
Peromyscus leucopus 193 (169-217) * *
Sigmodon hispidus 27 (13-41) 258 (171-345) 203 (162-245)
COPYRIGHT 2006 Texas Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Author:Sternberg, Mitchell A.; Judd, Frank W.
Publication:The Texas Journal of Science
Geographic Code:1U7TX
Date:May 1, 2006
Words:6524
Previous Article:Parasites of the Brazilian free-tailed bat, Tadarida brasiliensis (Chiroptera: Molossidae), from southwestern Arkansas.
Next Article:In vitro and in vivo studies of the effects of aluminum on wheat (Triticum aestivum L.) root cell wall peroxidases.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters