Printer Friendly

NOTES ON THE MICROHABITAT ASSOCIATIONS OF THE HISPID COTTON RAT (SIGMODON HISPIDUS) AND WHITE-FOOTED MOUSE (PEROMYSCUS LEUCOPUS) IN THE PIEDMONT PHYSIOGRAPHIC PROVINCE.

ABSTRACT

We examined the microhabitat associations of the hispid cotton rat (Sigmodon hispidus) and white-footed mouse (Peromyscus leucopus). The distribution of sympatric rodent species is influenced by the distribution of preferred microhabitats with which each species is associated. We used stepwise discriminant function analyses to determine how microhabitat characteristics differed between traps that captured cotton rats, traps that captured white-footed mice and traps that did not capture either species. Three variables discriminated between the microhabitat surrounding traps that captured cotton rats and traps that captured white-footed mice: number of stems of black berries (Rubus spp.), number of clumps of broomsedge (Andropogon virginicus) and shrub thickness. Shrub richness was higher and coarse woody debris was more abundant around traps that captured cotton rats than around traps in which neither species was captured. The density of grass was higher and the evenness of the shrub layer was lower around tra ps that captured white-footed mice than around traps in which neither species was captured. Our results support previous studies that indicate the distribution of cotton rats is influenced by the distribution of broomsedge and black berries, and the distribution of white-footed mice is influenced by the distribution of areas with a thick and diverse shrub layer.

Key words: hispid cotton rat, microhabitat, Peromyscus leucopus. Sigmodon hispidus, white-footed mouse.

INTRODUCTION

Populations of small mammals are not randomly distributed across a landscape. The distributions of cotton rats (Sigmodon hispidus) and whitefooted mice (Peromyscus leucopus) are influenced by the distribution of microhabitats they prefer [1, 2]. For example, white-footed mice commonly are associated with a well-developed shrub layer [3, 4, 5]. Cotton rats are most often found in areas dominated by dense herbaceous growth [6]. Odum [7] found the distribution of cotton rats to be associated with broomsedge (Andropogon virginicus) and the broomsedge-shrub stage of succession.

Previous studies indicate the microhabitat preferences of these two species differ; however, a quantitative comparison of the microhabitat associations of these two species in the Piedmont physiographic province has not been conducted. We investigated habitat associations of cotton rats and white-footed mice in the Piedmont physiographic province and tested for differences in microhabitat associations between the two species.

MATERIALS AND METHODS

The study was conducted in Athens-Clarke County, Georgia (33[degrees]53' N, 83[degrees]23' W, Elev. 250 m), during April and May 1998. Two grids were placed in four different habitat types: 1-year-old clearcut, 4-year-old clearcut, 4-year-old seed tree cut and a mature pine stand. Large Sherman live traps (7.6 x 8.9 x 22.9 cm) baited with dried oats were placed at each grid coordinate in the 5 x 5 trapping grid. Traps were set during three, 3-4 day periods (April 5-8, April 12-14 and April 20-22). Traps were checked each morning. Species and sex was recorded for each trapped animal and a uniquely number ear-tag (#1005 Size 1 Monel, National Band and Tag Co., Newport, Kentucky) was attached to all individuals. Animals were released at the point of capture.

In order to test for differences in microhabitat use between cotton rats and white-footed mice, vegetation sampling was conducted. Vegetation sampling consisted of measurements surrounding 20 traps that only captured white-footed mice, 17 traps that only captured cotton rats and 20 traps in which neither species was captured. Recorded in a circular plot (1-m radius) centered on the trap station were a.) total volume of coarse woody debris (CWD); b.) volume of CWD in each of 3 decay classes; density, richness, diversity and evenness of herbaceous vegetation; c.) density, richness, diversity, and evenness of shrub vegetation; d.) herbaceous thickness; e.) shrub thickness (% of ground obscured from 1.5 m above the ground by vegetation [greater than or equal to]50 cm high); f.) grass density; and g.) canopy density. Also, the total number of black berry (Rubus spp.) and muscadine (Vitus rotundifolia) stems and the total number of clumps of broomsedge were recorded in each circular plot.

We measured the diameter and length of all downed logs in the sampling-circle to determine the volume of CWD ([greater than or equal to] 7.5 cm diameter). The stage of decay of CWD was classified on a scale of 1-3. Decay class 1 was assigned to hard CWD with branches remaining and bark intact, decay class 2 was assigned to hard CWD with no branches and exfoliating bark, and decay class 3 was assigned to soft CWD with no branches or bark.

Herbaceous thickness and grass densities were determined using ocular

estimation in two 0.5-[m.sup.2] squares within the circumference of the circle. The placement of the 0.5-[m.sup.2] square was determined by dividing the circle into four quadrants (NE, NW, SE, SW) and placing the square in two of the quadrants selected at random. A corner of the 0.5-[m.sup.2] square was in contact with the flag at the center of the circle. Average grass density and herbaceous thickness were calculated using both estimates from the 0.5-[m.sup.2] squares. The abundance of herbaceous species was examined and recorded within the first of the two quadrants used to estimate grass density and herbaceous thickness.

Shrub thickness of the entire circle was determined by ocular estimation. We used a spherical densiometer to calculate the density of the canopy. The total number of black berry and muscadine stems and clumps of broomsedge in the circle was recorded.

We performed a series of stepwise discriminant function analyses (SDFA; 8) including the habitat variables previously mentioned, to determine how the microhabitat characteristics differed between sampling circles surrounding traps that captured cotton rats and white-footed mice and traps that did not capture either species. We then applied a canonical discriminant function analysis [9] to the same data with the significant variables from the SDFA to determine the placement of these variables along the discriminant function [10]. This analysis provided the total sample standardized canonical coefficients (TSSCC) for each habitat variable. The relative magnitude of the TSSCC measures the contribution of each habitat variable to the discrimination. The sign of the TSSCC indicates whether the relationship with the particular variable is positive or negative. We tested to see which habitat variables not selected for inclusion in the model by the SDFA were correlated to the significant discriminating variables. Co rrelations with r [greater than or equal to] 0.30 were considered significant. We used an alpha level of 0.05 as a rejection criterion for all statistical tests.

RESULTS

A total of 187 captures were made during 2000 trap nights. A total of 17 (8.5%) traps only caught cotton rats (mean: 1.6 captures/trap; range: 1-2 captures/trap), 72 (36%) only caught white-footed mice (mean: 3.1 captures/trap; range: 2-7 captures/trap), and 94 (47%) captured neither species. A total of 20 traps that captured white-footed mice and 20 traps that captured neither species were randomly selected and used for statistical analyses.

Of the 17 variables entered into the SDFA, three discriminated between the microhabitat surrounding traps that captured cotton rats and the microhabitat surrounding traps in which neither species was captured (Table 1). Shrub thickness and the volume of GWD in decay class 3 each accounted for 27% of the variation, while shrub richness accounted for 31% of the variation. The centroid for traps associated with cotton rats was at the positive end of the discriminant analysis, indicating that shrub thickness and volume of CWD in decay class 3 was greater and shrub richness was less around traps where cotton rats were captured. Positive correlations between the significant discriminating variables and other measurements indicate that shrub thickness, number of black berry stems, herbaceous richness, shrub diversity, and shrub evenness were greater where cotton rats were captured (Table II).

Two variables discriminated between the microhabitat surrounding traps that captured white-footed mice and the microhabitat surrounding traps in which neither species were captured (Table I). The average grass density accounted for approximately 23% of the variation and the evenness of the shrub layer accounted for 18% of the variation. The centroid for traps associated with white-footed mice was at the positive end of the discriminant analysis indicating that the grass was more dense and the shrub layer less even around traps that captured white-footed mice than traps where neither species was captured. Positive correlations between the significant discriminating values and other measurements indicate that a greater average grass density and a greater number of clumps of broomsedge surrounded traps that captured white-footed mice (Table II). They also indicate that a lower shrub evenness, a greater number of black berry stems, greater shrub thickness, greater shrub richness and greater shrub diversity chara cterized the microhabitat surrounding traps that captured white-footed mice.

Three variables discriminated between microhabitats surrounding traps that captured cotton rats and traps that captured white-footed mice (Table I). Number of black berry stems and the number of clumps of broomsedge accounted for 42% of the variation between the microhabitats surrounding traps that captured cotton rats and traps that captured white-footed mice. Shrub thickness accounted for 31% of the variation. Based on the canonical discriminant function analysis, the centroid for traps associated with cotton rats was at the positive end of the discriminant analysis. The TSSCC's for all three variables were positive, indicating that the number of black berry stems, clumps of broomsedge and thickness of shrubs was greater around traps that captured cotton rats than traps that captured white-footed mice. Positive correlations between the significant discriminating variables and other measurements indicate shrub thickness, average grass density and shrub diversity was greater surrounding traps where cotton ra ts were captured than where white-footed mice were capture (Table II).

DISCUSSION

The results of this study suggest that the microhabitat associations of cotton rats and white-footed mice are influenced by several habitat characteristics. Based on previous studies, we expected cotton rats to be associated with a high average grass density. Instead, like the report made in this physiographic province by Odum [5], the distribution of cotton rats was similar to the distribution of broomsedge and black berry. Our results are similar to previous studies that report that the distribution of white-footed mice is influenced by the distribution of areas with a thick and diverse shrub layer [3, 4, 5]. Given its association with woodland habitat, we expected a greater volume of CWD to be associated with traps that captured white-footed mice. Instead, we found the volume of CWD to be higher where cotton rats were captured. Thus, we found similarities and inconsistencies in the microhabitat associations of cotton rats and white-footed mice previously reported in this and other physiographic regions. W e conclude, therefore, that the microhabitat associations and spatial distribution of the two species are variable across the landscape.

A better understanding of the microhabitat associations of these two species and how the associations differ allow better prediction of how land management decisions will affect populations of cotton rats and white-footed mice. For example, the results of our study suggest site preparation methods that promote the growth of broomsedge and/or black berry in the Piedmont physiographic province would lead to increases in the abundance of cotton rats through promoting the habitat preferred by this species. Also, timber harvest or thinning could enhance the understory and midstory layers, increasing the availability of habitat preferred by both species.

ACKNOWLEDGMENTS

We greatly appreciate the field assistance of T. Craven and N. Blackman. We thank Dr. S. H. Schweitzer for helpful comments on previous versions of this manuscript.

LITERATURE CITED

(1.) McMillan BR and Kaufman DW: Differences in use of interspersed woodland and grassland by small mammals in northeastern Kansas. Prairie Nat 26: 107-116, 1994.

(2.) Peterson SK, Kaufman GA and Kaufman DW: Habitat selection by small mammals of the tall-grass prairie: experimental patch choice. Prairie Nat 17: 65-70, 1985.

(3.) Kaufman DW, Peterson SK, Fristick R and Kaufman GA: Effect of microhabitat features on habitat use by Peromyscus leucopus. Amer Midl Nat 110: 177-185, 1983.

(4.) Drickamer LC: Microhabitat preferences for two species of deermice (Peromyscus) in a northeastern United States deciduous hardwood forest. Acta Theriologica 35: 241-252, 1990.

(5.) Dueser RD and Shugart HH: Microhabitats in forest-floor small mammal fauna. Ecology 59: 89-98, 1978.

(6.) Goertz JW and Long RC: Habitat of five species of rat in Louisiana. Amer Midl Nat 90: 460-465, 1973.

(7.) Odum EP: An eleven year history of a Sigmodon population. J Mammal 36: 368-378, 1955.

(8.) Powers JJ and Ware GO: "Statistical Procedures in Food Research." J.R. Piggott Ed. Elsevier Applied Science, New York, 1950.

(9.) Vonhof MJ: Roost-site preferences of big brown bats (Eptesicus fuscus) and silver-haired bats (Lasionycteris noctivigans) in the Pend d'Oreille Valley in southern British Columbia, pp. 62-80, in "Bats and forest symposium." RMR Barclay and RM Brigham, eds. British Columbia Ministry of Forests, Victoria, British Columbia, Canada, 1996.

(10.) Afifi AA and Clark V: "Computer aided multivariate analysis." Chapman and Hall, Boundary Row, London, 1996.

Summary of stepwise discriminant function analyses comparing environmental variables associated with traps that captured cotton rats (Sigmodon hispidus) with empty traps, traps that captured white-footed mice (Peromyscus leucopus) with empty traps, and traps that captured cotton rats with traps that captured white-footed mice. The centroid for traps that captured white-footed mice in the second discriminant function, and cotton rats in the first and third discriminant functions lie at the positive end of the discriminant axis.
 Order Partial
Variable Included F P R2
Cotton rats vs. empty traps
 Shrub thickness 1 6.736 0.0183 0.2723
 Shrub richness 2 7.597 0.0135 0.3089
 Volume of CWD [**] (decay class 3) 3 6.089 0.0253 0.2756
White-footed mice vs. empty traps
 Average grass density 1 6.161 0.0216 0.2268
 Shrub evenness 3 4.191 0.0547 0.1807
Cotton rats vs. white-footed mice
 Number of black berry 1 5.676 0.0258 0.1979
 (Rubus spp.) stems
 Number of broomsedge 3 5.989 0.0233 0.2219
 (Andropogon virginicus) clumps
 Shrub thickness 4 9.010 0.0070 0.3106
Variable TSSCC [*]
Cotton rats vs. empty traps
 Shrub thickness 1.2677
 Shrub richness -0.7827
 Volume of CWD [**] (decay class 3) 0.4029
White-footed mice vs. empty traps
 Average grass density 1.1210
 Shrub evenness -0.8093
Cotton rats vs. white-footed mice
 Number of black berry 0.5206
 (Rubus spp.) stems
 Number of broomsedge 0.6653
 (Andropogon virginicus) clumps
 Shrub thickness 0.4222
(*.)TSSCC: Total Sample Standardized Canonical Coefficient
(**.)CWD: Coarse Woody Debris


Summary of correlations between significant discriminating variables and other variables included in the discriminant function analyses comparing traps that captured cotton rats (Sigmodon hispidus) with empty traps, traps that captured white-footed mice (Peromyscus leucopus) with empty traps, and traps that captured cotton rats with traps that captured white-footed mice.
Variable Variable
Cotton rats vs. empty traps
 Volume of CWD [**] Shrub richness
 (decay class 3)
Shrub thickness Number of black berry
 (Rubus spp.) stems
Shrub thickness Herbaceous richness
Shrub thickness Shrub diversity
Shrub thickness Shrub evenness
White-footed mice vs. empty traps
Shrub evenness Number of black berry
 (Rubus spp.) stems
Shrub evenness Shrub thickness
Shrub evenness Shrub richness
Shrub evenness Shrub diversity
Average grass density Number of broomsedge
 (Andropogon virginicus) clumps
Cotton rats vs. white-footed mice
Number of black berry Number of broomsedge
 (Rubus spp.) stems (Andropogon virginicus) clumps
Number of black berry Average herbaceous thickness
 (Rubus spp.) stems
Number of black berry Shrub thickness
 (Rubus spp.) stems
Number of broomsedge Average grass density
 (Andropogon virginicus)
 clumps
Number of broomsedge Average herbaceous thickness
 (Andropogon virginicus)
 clumps
Shrub thickness Shrub diversity
Variable r P
Cotton rats vs. empty traps
 Volume of CWD [**] 0.4064 0.0126
 (decay class 3)
Shrub thickness 0.3664 0.025
Shrub thickness 0.3328 0.0441
Shrub thickness 0.6902 0.0001
Shrub thickness 0.5693 0.0008
White-footed mice vs. empty traps
Shrub evenness 0.3827 0.0336
Shrub evenness 0.5335 0.0020
Shrub evenness 0.3860 0.0351
Shrub evenness 0.6899 0.0002
Average grass density 0.5324 0.0004
Cotton rats vs. white-footed mice
Number of black berry 0.3704 0.0240
 (Rubus spp.) stems
Number of black berry 0.4954 0.0018
 (Rubus spp.) stems
Number of black berry 0.4070 0.0124
 (Rubus spp.) stems
Number of broomsedge 0.4306 0.0078
 (Andropogon virginicus)
 clumps
Number of broomsedge 0.4389 0.0066
 (Andropogon virginicus)
 clumps
Shrub thickness 0.5372 0.0022
(**.)CWD: Coarse Woody Debris
COPYRIGHT 1999 Georgia Academy of Science
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Menzel, Michael A.; Carter, Timothy C.; Houston, Ashlie T.; Longe, Robert L.
Publication:Georgia Journal of Science
Geographic Code:1USA
Date:Sep 22, 1999
Words:2737
Previous Article:ZOO ATLANTA'S SCIENTIFIC VISION.
Next Article:OBSERVATIONS OF SATURN IN 1997.
Topics:

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters