HABITAT-SPECIFIC DEMOGRAPHY OF SYMPATRIC VOLE POPULATIONS OVER 25 YEARS

Journal of Mammalogy, 86(3):561 568, 2005 HABITAT-SPECIFIC DEMOGRAPHY OF SYMPATRIC VOLE POPULATIONS OVER 25 YEARS LOWELL L. GETZ,* MADAN K. OLI, JOYCE E. HOFMANN, AND BETTY MCGUIRE Department of Animal Biology, University of Illinois, 505 S Goodwin Avenue, Urbana, IL 61801, USA (LLG) Department of Wildlife Ecology and Conservation, 110 Newins-Ziegler Hall, University of Florida, Gainesville, FL 32611, USA (MKO) Illinois Natural History Survey, 607 E Peabody Drive, Champaign, IL 81820, USA (JEH) Department of Biological Sciences, Smith College, Northampton, MA 01063, USA (BM) We studied the influence of preferred food, forbs, and vegetative cover on survival and reproduction at different population densities of Microtus ochrogaster and M. pennsylvanicus in alfalfa, bluegrass, and tallgrass prairie habitats in east-central Illinois for 25 years. Population densities of M. ochrogaster were greatest in alfalfa, least in tallgrass, and intermediate in bluegrass, whereas those of M. pennsylvanicus were greatest in tallgrass, least in alfalfa, and intermediate in bluegrass. For both species, preferred food availability was greatest in alfalfa, intermediate in bluegrass, and least in tallgrass. Vegetative cover was relatively sparse in alfalfa, especially in winter, and dense throughout the year in bluegrass and tallgrass. Variation in survival emerged as the most important factor explaining population differences between the 2 species. Reproduction had little differential impact on abundance of either species in any of the 3 habitats. Survival of M. ochrogaster was higher in alfalfa than in bluegrass or tallgrass; survival of M. pennsylvanicus was higher in tallgrass than in alfalfa or bluegrass. Differential survival among habitats and between species was influenced primarily by amount of vegetative cover. We suggest that M. ochrogaster is less susceptible than M. pennsylvanicus to predation by raptors and large carnivores (predators that hunt from above vegetative cover), whereas M. pennsylvanicus is less susceptible than M. ochrogaster to snakes and small carnivores (predators that hunt under vegetative cover). Key words: demography, meadow vole, Microtus ochrogaster, Microtus pennsylvanicus, prairie vole, vole habitats Population density of arvicoline rodents in specific habitats results from the balance of positive effects of resources and negative effects of mortality (Batzli 1992). Vegetation characteristics impact directly upon resources and mortality. For herbivorous species, quality and quantity of food availability are related to species composition of the vegetation. Vegetation also provides physical cover, which affords protection from predators and moderates environmental stresses. Structure and density of cover impacts directly on the success of natural predators. Abundance of given species of food plants, as well as the amount and structure of the vegetation, vary among habitats and vary seasonally within habitats. An understanding of factors affecting population success (i.e., population density) within and among habitats is necessary for evaluations of population dynamics of species with potential for cyclicity of fluctuations (Batzli 1996; Korpimäki et al. 2002; Krebs 1996). Survival and reproduction are most responsible * Correspondent: l-getz@life.uiuc.edu Ó 2005 American Society of Mammalogists www.mammalogy.org for temporal and spatial differences in population success of arvicoline rodents (Batzli 1992, 1996; Krebs and Myers 1974; Lin and Batzli 2001), whereas emigration and immigration do not appear to be important factors influencing population densities (Dueser et al. 1981; Gaines and McClenaghan 1980; Lin and Batzli 2001; Verner and Getz 1985). In this paper we use the results of a 25-year study of 2 unmanipulated sympatric vole species (prairie vole [Microtus ochrogaster] and meadow vole [M. pennsylvanicus]) in alfalfa (Medicago sativa), bluegrass (Poa pratensis), and tallgrass prairie (a mixture of big bluestem [Andropogon gerardii], Indian grass [Sorghastrum nutans], and switch grass [Panicum]) habitats to evaluate the influence of food and cover on survival and reproduction of the 2 species. Although grasses are eaten by M. ochrogaster, this species cannot subsist successfully on a diet of graminoids alone; forbs are required for successful reproduction and survival (Cole and Batzli 1978, 1979; Haken and Batzli 1996). Although M. pennsylvanicus preferentially selects forbs over grasses, individuals can subsist and reproduce successfully on graminoids (Haken and Batzli 1996; Lindroth and Batzli 1984; Thompson 1965; Zimmerman 1965). Both species preferentially select sites where vegetative cover is greatest; however, M. pennsylvanicus is more reliant 561

562 JOURNAL OF MAMMALOGY Vol. 86, No. 3 than M. ochrogaster on dense cover (Birney et al. 1976; Getz and Hofmann 1999; Klatt and Getz 1987; Lin and Batzli 2001). MATERIALS AND METHODS Study Sites The study sites were located in the University of Illinois Biological Research Area ( Phillips Tract ) and Trelease Prairie, both 6 km NE of Urbana, Illinois (408159N, 888289W). We trapped voles in study sites (0.5 2.0 ha) in 2 restored tallgrass prairies, 1 located in Trelease Prairie, the other in Phillips Tract, from March 1972 through May 1997 (Getz et al. 2001). Study sites (0.8 1.0 ha) established within a former bluegrass pasture located in Phillips Tract that had been released from grazing in spring 1971 were trapped January 1972 May 1997 (Getz et al. 2001). Two adjacent sites (1.0 1.4 ha) in Phillips Tract planted with alfalfa were trapped May 1972 May 1997. Voles were trapped at a site until invading forbs and grasses began to crowd out the M. sativa. One year before trapping was terminated at 1 site, M. sativa was planted at the other site so that the plants would be fully developed when trapping commenced in that site. There were 4 changes in M. sativa sites during the course of the study; hence, we sampled a total of 5 sites, which are described in detail by Getz et al. (1979, 1987). A grid system with 10-m intervals was established on all study sites. One wooden multiple-capture live trap (Burt 1940) was placed at each station. Each month a 2-day prebaiting period was followed by a 3-day trapping session; traps were set about 1500 h and checked at 0800 h and 1500 h for the next 3 days. All animals were toe-clipped at 1st capture for individual identification (maximum of 2 toes on each foot). All procedures were approved by the University of Illinois Laboratory Animal Resource Committee and meet the guidelines recommended by the American Society of Mammalogists (Animal Care and Use Committee 1998). At each capture we recorded species, grid station, individual identification, sex, reproductive condition (males: testes abdominal or descended; females: vagina open or closed, pregnant as determined by palpation, or lactating), and weight to the nearest 1 g. For analysis, we grouped voles as adult (30 g) or young (29 g). Although there was conspicuous seasonal variation within habitats, the relative proportion of the vegetation comprised of forbs did not differ among the 3 habitats during the course of the study. Previous studies have provided evidence of relative abundance of forbs within the 3 habitats during the growing season (Cole and Batzli 1978, 1979; Getz et al. 1979; Lindroth and Batzli 1984). We obtained additional data to quantify relative abundance of forbs in late winter when food resources were the lowest for the year. The spot-check was conducted in late February 1991, the 2nd year of usage of the alfalfa field being trapped at that time and 2 years after the previous burning of the tallgrass sites. The data thus represent abundance of forbs in the winter during the midperiod of usage of each of the 5 alfalfa sites, after vegetation in tallgrass sites had recovered from burns, and that typical of bluegrass in winter. Ten 10-m line-intercept transects, with markers every 10 cm, were located at random in each habitat. Numbers of forbs (rosettes or overwintering shoots at the soil surface) under the 10-cm markers of each 10-m transect were recorded. Qualitative observational comparisons of vegetative cover in the 3 habitats were made throughout the year; height of the vegetation and density of the vegetative cover were estimated visually. We also measured density of the vegetative cover before the start of vegetation growth in early spring 1974 for a more precise comparison of cover during the winter. The measurements were made the 2nd year the 1st alfalfa field was in use and 3 years after the previous burning of the tallgrass site; these data are representative of vegetative cover present during the winter during the midpoint of usage of an alfalfa site, after vegetative cover in tallgrass had recovered from a burn, and essentially at all times in bluegrass. A sunlight penetration meter (Getz 1968, 1970a, 1970b) was used to obtain a comparative estimate of cover (percentage full sunlight reaching the soil surface) in the 3 habitats. Ten measurements were made adjacent to each of 10 randomly selected grid stations (total of 100 measurements in each habitat) within each habitat. Demographic Variables Monthly survival rates were estimated as the proportion of the voles present from one month to the next month. Survival rates represent total population survival. Persistence of voles 1st captured as young, and presumed to have been born on the study site, also was calculated. We assumed voles recorded as young in a given trapping session were born midway between that trapping session and the previous session. Voles that disappeared from a site were presumed to have done so midway between the last session they were captured and the subsequent session. Thus, young voles captured in only 1 month were given a persistence of 1 month. Persistence as defined herein includes both in situ mortality and emigration; the former is presumed to be the most prevalent cause of disappearance (Verner and Getz 1985). Because of small monthly sample sizes, persistence data for the 2 sexes were combined and recorded only by season. We calculated the proportions of reproductively active adult females for each month. Only females were used in our analysis because reproductive condition of females can be more accurately determined than that of males. We limited our comparisons of body mass to adult males to avoid bias from variation in the proportion of the population composed of young voles and from variation in the reproductive status of females (i.e., changes in body mass due to pregnancy, parturition, and lactation Ostermeyer 1983). Adult body mass was used as an indirect indicator of physiological condition of voles, and in turn, as an indirect indication of effects of food availability on a population (Lin and Batzli 2001; Norrdahl and Korpimäki 2002). In our analysis of body mass, we avoided bias from inclusion of the same individual in subsequent months by using data for each adult male only once, the 1st month it appeared as an adult in the capture data. Because of small sample sizes, body mass data were compiled by season. Data Analysis Population densities of the 2 species in the 3 habitats (see Getz et al. 2001) were used as a covariate to control for the effect of population density on survival and reproduction. For seasonal analyses we allocated all observations to spring (March May), summer (June August), autumn (September November), or winter (December February). We used general-linear models (analysis of covariance [ANCOVA] of SAS procedure GLM [SAS Institute Inc. 1999]) to investigate effects of species, population density, habitat, and season on survival and reproduction (Slade et al. 1997). We 1st fitted a model with all main effects, and all 3-way and 2-way interactions. Then, we sequentially removed nonsignificant (a ¼ 0.05) interaction effects, starting with the highest interaction term with the largest P value. We then refitted the model, removed another highest order interaction term with the largest P value, and repeated the process until all nonsignificant interaction terms were removed. The final general-linear model contained main effects of all variables, and all significant 2- and 3-way interaction effects. By using the final model, we estimated leastsquares means for the highest-order significant interaction term, and tested for the significance of pairwise differences in least-squares means (SAS Institute Inc. 1999). Least-squares means allow one to evaluate within-group means appropriately adjusted for other effects in the model. Body mass and persistence of voles were modeled similarly, except that population density was not used as a covariate. Survival and

June 2005 GETZ ET AL. HABITAT-SPECIFIC DEMOGRAPHY OF VOLES 563 proportion of reproductive females were arcsine-square root transformed, and body mass and persistence data were log-transformed before statistical analyses. We used SPSS 10.0.7 for Macintosh (SPSS, Inc. 2001) or SAS (SAS Institute Inc. 1999) for all statistical analyses. Least-squares means for the highest-order significant interaction effects and associated significance levels for pairwise comparisons are presented for survival (Internet files JM 01 and JM 02), proportion reproductive females (JM 03 and JM 04), persistence (JM 05 and JM 06), and body mass (JM 07 and JM 08); all of these files are available on the Web page http://www.life.uiuc.edu/getz/. Files for all original capture data from this study also are available at this Web site. RESULTS Demography Mean population densities of M. ochrogaster over the 25 years of the study were 50 voles/ha in alfalfa, 18 voles/ha in bluegrass, and 7 voles/ha in tallgrass (P, 0.05 Getz et al. 2001). We recorded 13 population fluctuations in alfalfa (mean peak density, 202 voles/ha; range, 77 638 voles/ha), 11 in bluegrass (mean, 67 voles/ha; range, 25 156 voles/ha), and 5 in tallgrass (mean, 59 voles/ha; range, 34 92 voles/ha). Mean population density of M. pennsylvanicus was highest in tallgrass (30 voles/ha), intermediate in bluegrass (14 voles/ha), and lowest in alfalfa (7 voles/ha; P, 0.05 Getz et al. 2001). We documented 5 population fluctuations in alfalfa (mean, 53 voles/ha; range, 29 79 voles/ha) and 9 fluctuations in bluegrass (mean, 56 voles/ha; range, 35 91 voles/ha). There were extensive periods when M. pennsylvanicus was absent or present in very low population densities in both alfalfa and bluegrass. Population densities of M. pennsylvanicus were erratically high in tallgrass with no distinct fluctuation. Habitat Quality In our late February samples we recorded an average of 22.0 6 2.3 green forbs/10 m, almost all of which were M. sativa, along 10-m transects in alfalfa, 9.7 6 1.9 green forbs/ 10 m in bluegrass, and 2.4 6 0.7 (mean 6 SE) green forbs/10 m in tallgrass. Availability of forbs also was greatest in alfalfa and least in tallgrass during the growing season (Getz et al. 1979; Lindroth and Batzli 1984). Vegetative cover in alfalfa was more dense from midspring through early winter than at other times. During this period, alfalfa plants were,0.5 m tall with only scattered taller plants providing cover up to 1.0 m above the surface. Vegetative cover, 25 cm above the surface was relatively sparse throughout the year and especially so during the first 2 or 3 winters a site was used. By the 3rd or 4th year, invading grasses provided relatively dense, but spotty, cover throughout the winter. Before invading grasses became prominent, penetration of light to the surface in alfalfa in early spring was 62.9 6 5.1% (mean 6 SE) of full sunlight. In bluegrass, vegetative cover near the surface (5 25 cm above), including a mat of grass litter, was dense throughout the year; vole activity appeared to be restricted to small surface runways under the vegetation. Larger forbs provided considerable cover up to 1 2 m above the surface during summer late TABLE 1. Results of analysis of covariance of monthly survival of Microtus ochrogaster and M. pennsylvanicus. See text for definition of sources of variability. Source d.f. F P Species 1 18.50,0.0001 Habitat 2 1.12 0.3265 Season 3 1.06 0.3639 Density 1 42.93,0.0001 Species habitat 2 14.95,0.0001 autumn. Penetration of light to the surface in early spring was only 3.2 6 0.2% of full sunlight. Vegetative cover in tallgrass was dense throughout the year. During spring early summer, grasses formed cover 0.5 1.0 m high; from late summer early winter, there was dense vegetative cover 1.0 1.5 m above the surface. Beginning the winter after a burn, recumbent dead grasses formed a dense layer about 25 cm above the surface throughout the year. Most of the area between the bases of tallgrass clumps was relatively open up to 10 cm above the surface at all times. Penetration of light to the surface in early spring was 4.2 6 0.3% of full sunlight. Demographic Variables Monthly survival. In the final general-linear model (3-way ANCOVA with density as a covariate) model, the main effects of species and population density on monthly survival were significant; although the main effect of habitat and season was not significant, the former variable was involved in a significant 2-way interaction. This interaction (species habitat) underscores that species had different relative survival in the 3 habitats studied (Table 1). Thus, we computed leastsquares mean survival for each level of the variables involved in the significant species habitat interaction, and tested for pairwise differences. Survival of M. ochrogaster (Fig. 1) was greater in alfalfa than in bluegrass (P ¼ 0.0156) or tallgrass (P ¼ 0.0026); survival in bluegrass did not differ from that in tallgrass (P ¼ 0.2945). Survival of M. pennsylvanicus (Fig. 1) was greater in tallgrass than in alfalfa (P, 0.0001) or bluegrass (P ¼ 0.0011). Survival in bluegrass was greater than that in alfalfa, although not significantly so (P ¼ 0.0647). Survival of M. pennsylvanicus was greater than survival of M. ochrogaster in bluegrass (P, 0.0001) and tallgrass (P ¼ 0.0075), but did not differ from that in alfalfa (P ¼ 0.1242). Proportion of females reproductive. In the final generallinear model, the main effects of species, habitat, season, and population density on proportion of reproductive females were significant. Two-way interactions involving species and both season and habitat were significant, as was the 3-way interaction effect involving these 3 variables (Table 2). Thus, we computed least-squares mean proportion of reproductive females for the significant 3-way interaction effect (species, habitat, and season), and tested for pairwise differences. The only consistent within-habitat seasonal difference in proportion of females reproductive (Fig. 2) was lower proportions of both species during winter than other seasons in all 3 habitats (M. ochrogaster: P, 0.0001 in all 3 habitats;

564 JOURNAL OF MAMMALOGY Vol. 86, No. 3 FIG. 1. Seasonal monthly survival (mean proportion þ SE of the voles present from one month to the next month) of Microtus ochrogaster and M. pennsylvanicus in 3 habitats in east-central Illinois over 25 years. Seasons: spring, March May; summer, June August; autumn, September November; winter, December February. See text for description of significant differences. M. pennsylvanicus: alfalfa, P ¼ 0.0003; bluegrass, P ¼ 0.0226; tallgrass, P ¼ 0.0045). The only among-habitat differences in proportions of female M. ochrogaster reproductive were greater proportion reproductive during spring in alfalfa than in bluegrass (P ¼ 0.0154) or tallgrass (P ¼ 0.0418). The proportion of reproductive female M. pennsylvanicus was greater in alfalfa than in bluegrass in winter (P ¼ 0.0019) or TABLE 2. Results of analysis of covariance of proportion of reproductive female Microtus ochrogaster and M. pennsylvanicus. See text for definition of sources of variability. Source d.f. F P Species 1 45.98,0.0001 Habitat 2 12.18,0.0001 Season 3 76.39,0.0001 Density 1 5.96 0.0148 Species season 3 5.52 0.0009 Species habitat 2 6.64 0.0014 Species habitat season 12 2.25 0.0083 FIG. 2. Seasonal proportion of reproductive adult females (mean þ SE) of Microtus ochrogaster and M. pennsylvanicus in 3 habitats in east-central Illinois over 25 years. Seasons: spring, March May; summer, June August; autumn, September November; winter, December February. See text for significant differences. than in tallgrass during summer (P, 0.0001), autumn (P ¼ 0.0306), or winter (P ¼ 0.0001). The proportion of reproductive females in bluegrass tended to be intermediate to that in alfalfa and tallgrass in all seasons. The only significant difference between species in the proportion of reproductive females was a greater proportion of reproductive M. ochrogaster than M. pennsylvanicus in bluegrass and tallgrass during summer (bluegrass, P ¼ 0.0440; tallgrass, P, 0.0001) and winter (bluegrass, P, 0.0001; tallgrass, P, 0.0001). Persistence of young. In the final general-linear model, the main effects of species, habitat, and season on persistence of young were all significant. Two-way interactions involving species and both season and habitat were significant, as was the 3-way interaction effect involving these 3 variables (Table 3). Thus, we computed least-squares mean persistence for the significant 3-way interaction effect (species, habitat, and season). No seasonal difference was found in persistence of young of either species within habitats (Fig. 3). Among habitats, the only difference in persistence of young M. ochrogaster was greater persistence in bluegrass than in tallgrass when born during winter (P ¼ 0.0009). Persistence of young M. pennsylvanicus born during summer (P ¼ 0.0003), autumn (P, 0.0001), and winter (P ¼ 0.0010) was greater in tallgrass than in alfalfa; persistence of young born in bluegrass tended to be intermediate between that in tallgrass and in alfalfa during all seasons

June 2005 GETZ ET AL. HABITAT-SPECIFIC DEMOGRAPHY OF VOLES 565 TABLE 3. Results of analysis of covariance of persistence of Microtus ochrogaster and M. pennsylvanicus. See text for definition of sources of variability. Source d.f. F P Species 1 13.92,0.0002 Habitat 2 4.56 0.0105 Season 5 2.93 0.0324 Species habitat 2 25.25,0.0001 Species season 3 12.39,0.0001 Species habitat season 10 2.63 0.0016 (Fig. 3). Persistence of young M. pennsylvanicus was greater than that of M. ochrogaster when born in bluegrass and tallgrass in summer (bluegrass, P ¼ 0.0312; tallgrass, P ¼ 0.0326) and winter (bluegrass, P, 0.0001; tallgrass, P, 0.0001), whereas persistence of young M. ochrogaster was greater than that of M. pennsylvanicus when born in autumn in alfalfa and bluegrass (P, 0.0001 and P ¼ 0.0004, respectively). Body mass of adult males. In the final general-linear model, the main effects of species, habitat, and season on body mass were significant. Two-way interactions involving species, season, and habitat were significant, as was the 3-way interaction effect involving these 3 variables (Table 4). Thus, we computed least-squares mean proportion reproductive for the significant 3- way interaction effect (species, habitat, and season). Body masses of adult male M. ochrogaster (Fig. 4) were lower during winter than the other 3 seasons in alfalfa (all differences, P, 0.0001) and during winter than autumn and spring in bluegrass or tallgrass (all, P, 0.0001). Among habitats, body masses in alfalfa were greater than those in bluegrass and tallgrass in all seasons (all, P, 0.0001). Body masses of adult male M. pennsylvanicus were lower during winter than autumn and spring in alfalfa and bluegrass, and than all 3 other seasons in tallgrass (all, P, 0.0001). Among habitats, body masses of M. pennsylvanicus were greater in alfalfa than bluegrass during autumn and greater in alfalfa than in tallgrass during autumn and winter (all, P, 0.0001). Body masses of adult male M. pennsylvanicus were greater than those of M. ochrogaster in alfalfa during autumn (P, 0.0001), in bluegrass during spring (P, 0.0001), autumn (P, 0.0001), and summer (P ¼ 0.0002), and all seasons in tallgrass (autumn, spring, and summer, P, 0.0001; winter, P ¼ 0.0249). DISCUSSION Population densities of M. ochrogaster were highest in alfalfa and lowest in tallgrass, whereas those of M. pennsylvanicus showed the opposite pattern (Getz et al. 2001). M. ochrogaster was more abundant than M. pennsylvanicus in alfalfa, whereas the reverse was true in tallgrass; densities of the 2 species did not differ in bluegrass (Getz et al. 2001). Differences in monthly survival rates and persistence of young were more closely related to differences in population densities of the 2 species among the 3 habitats than were proportion of reproductive females. Monthly survival and persistence of young M. ochrogaster in this study generally were greater in alfalfa than in tallgrass and bluegrass. Monthly survival rates and FIG. 3. Seasonal comparisons (mean months þ SE) of persistence (time elapsed from month 1st captured until the individual disappeared from the site) of Microtus ochrogaster and M. pennsylvanicus 1st caught as young in 3 habitats in east-central Illinois over 25 years. Seasons: spring, March May; summer, June August; autumn, September November; winter, December February. See text for significant differences. persistence of young M. pennsylvanicus, on the other hand, were greatest in tallgrass and least in alfalfa. Only greater proportion of reproductive female M. ochrogaster in alfalfa than in either bluegrass or tallgrass during spring was consistent with higher densities in alfalfa than in the latter 2 habitats. Further, no difference was found in proportion of females reproductive in TABLE 4. Results of analysis of covariance of body mass of Microtus ochrogaster and M. pennsylvanicus. See text for definition of sources of variability. Source d.f. F P Species 1 117.28,0.0001 Habitat 2 49.05,0.0001 Season 3 87.71,0.0001 Habitat season 6 3.92 0.0006 Species habitat 2 5.50 0.0041 Species season 3 16.06,0.0001 Species habitat season 6 9.60,0.0001

566 JOURNAL OF MAMMALOGY Vol. 86, No. 3 FIG. 4. Seasonal body masses (g, mean þ SE) of adult male Microtus ochrogaster and M. pennsylvanicus in 3 habitats in eastcentral Illinois over 25 years. Seasons: spring, March May; summer, June August; autumn, September November; winter, December February. See text for significant differences. alfalfa and tallgrass during winter. On the other hand, proportion of reproductive female M. pennsylvanicus during winter was greater in alfalfa than in tallgrass and bluegrass. In addition, the proportion of reproductive female M. ochrogaster was greater than that of M. pennsylvanicus during winter in tallgrass. Although the relationship between vegetative cover and population densities of arvicoline rodents has been well documented (Adler and Wilson 1989; Getz 1985; Klatt and Getz 1987; Lin and Batzli 2001; Peles and Barrett 1996), evidence for causative mechanisms is meager. Both vegetative cover (through its effect on predation risk) and food availability may influence survival and reproduction of voles (Desy and Batzli 1989; Klemola et al. 1998; Korpimäki et al. 1994; Lin and Batzli 2001; Norrdahl and Korpimäki 2000; Peles and Barrett 1996). Lin and Batzli (2001) found similar responses of population densities of M. ochrogaster and M. pennsylvanicus to food and cover, but they observed no relationship between food availability and reproduction in either species when differential dispersal also was considered. However, survival rates of M. pennsylvanicus have been shown to be positively correlated with vegetative cover (Adler and Wilson 1989; Lin and Batzli 2001; Peles and Barrett 1996). In our study, survival of M. pennsylvanicus appeared to be more closely correlated to dense vegetative cover than to abundance of forbs, whereas reproduction appeared to be more closely correlated to abundance of forbs. Monthly survival and persistence of young generally were greater in tallgrass, where cover was dense and availability of forbs was low, than in alfalfa, where cover was less dense and availability of forbs was high. Proportions of adult females that were reproductively active showed the opposite response, being greater in alfalfa than in tallgrass. Comparison of seasonal differences supports the importance of cover as a factor affecting survival and forb availability as a factor involved in reproduction of M. pennsylvanicus. During winter, persistence of young born in alfalfa, where cover was sparse and availability of forbs was relatively high, was less than in tallgrass, where cover was dense and availability of forbs was lower. In contrast, the proportion of reproductively active females was greater during winter in alfalfa than in tallgrass. Body mass of males, an indirect indicator of overall quality of individuals in the population (Norrdahl and Korpimäki 2002), was greater in alfalfa than in tallgrass during winter, providing evidence for the important influence of food on reproduction, but not on survival of M. pennsylvanicus. On 1st evaluation, survival of M. ochrogaster appeared to be related more to food availability than to vegetative cover. Survival rates were greater in alfalfa than in bluegrass and tallgrass, whereas proportions of reproductively active females did not differ between the 3 habitats. Body masses of males were significantly greater in alfalfa than in the other 2 habitats, suggesting better physiological condition in response to higher forb availability. Better physiological condition, in turn, could result in increased survival, thus suggesting an importance of forb availability to survival, but not to reproduction. However, availability of forbs was greater in bluegrass than in tallgrass, yet no difference was found in survival between these 2 habitats. Although availability of high-quality food may play a prominent role in population densities of M. ochrogaster, we suggest that effects on survival are not the primary reason. Populations of M. ochrogaster are organized into territorial communal groups (Getz et al. 1992, 1993). Habitats with high availability of preferred food, such as alfalfa, can support a larger number of communal groups than can those where availability of preferred food is lower, such as bluegrass and tallgrass (Getz et al. 1992). Because survival is either the same or greater where cover is sparse than where cover is dense, more individuals survive in alfalfa to form more communal groups; thus, M. ochrogaster may achieve higher population densities in alfalfa, which provides abundant high-quality food, than in bluegrass or tallgrass, which do not. The low population densities of M. ochrogaster in tallgrass, the original habitat of this species in east-central Illinois (Getz 1985), were consistent with low densities of the species in tallgrass prairies elsewhere (Bixler and Kaufman 1995; Meserve 1971). Our finding that the species achieved highest densities in alfalfa, a habitat with artificially high availability of preferred food, further demonstrates the importance of food to population densities of M. ochrogaster. The positive response of M. pennsylvanicus to availability of high-quality food in alfalfa (i.e., greater proportion females reproductive) appeared to be insufficient to offset low survival

June 2005 GETZ ET AL. HABITAT-SPECIFIC DEMOGRAPHY OF VOLES 567 because of sparse vegetative cover, thus resulting in low population densities in this habitat. Although bluegrass and tallgrass contained fewer of the preferred forbs, food may not be a limiting factor for this species. Lindroth and Batzli (1984) have shown that M. pennsylvanicus can reproduce and sustain populations on a diet of grasses, whereas M. ochrogaster cannot (Cole and Batzli 1979). Even though the proportion of reproductive females was lower in tallgrass than in alfalfa, greater survival and persistence of young in the former habitat resulted in higher population densities than in alfalfa. Lin and Batzli (2001) concluded that the primary reason for lower population densities of M. pennsylvanicus than of M. ochrogaster in areas with sparse cover was greater susceptibility of M. pennsylvanicus to predation. Lin and Batzli (1995) listed 21 species of predators known to be present in our study sites: 8 raptors, 5 large carnivores, 3 small carnivores, and 5 snakes. Raptors and large carnivores forage above the surface vegetation, whereas small carnivores and snakes hunt below the vegetation at the surface. All would be effective in habitats with sparse vegetative cover, such as alfalfa (Getz 1961). However, in habitats with dense vegetative cover, such as bluegrass and tallgrass, small carnivores and snakes would be more effective than would predators that hunt above the surface vegetation. We suggest that M. pennsylvanicus may be more susceptible than M. ochrogaster to predation by raptors and large carnivores. M. ochrogaster, on the other hand, may be more susceptible than M. pennsylvanicus to snakes and small carnivores. Species differences in the detection and avoidance of specific types of predators may be involved (Derting and Cranford 1989; Glickman and Morrison 1969; Müller- Schwarze and Müller-Schwarze 1971; Spiegel et al. 1974). Such differences in predation risks would be consistent with observed differences in survival and demography of the 2 species in the 3 habitats in our study. ACKNOWLEDGMENTS The study was supported in part by National Science Foundation grant DEB 78-25864, National Institutes of Health grant 09328, and by the University of Illinois School of Life Sciences and Graduate College Research Board. We thank the following individuals for their assistance with the fieldwork: L. Verner, R. Cole, B. Klatt, R. Lindroth, D. Tazik, P. Mankin, T. Pizzuto, M. Snarski, S. Buck, K. Gubista, S. Vanthernout, M. Schmierbach, D. Avalos, L. Schiller, J. Edgington, B. Frase, and the 1,063 undergraduate mouseketeers without whose extra hands in the field the study would not have been possible. C. Haun, M. Thompson, and M. Snarski entered the data sets into the computer. LITERATURE CITED ADLER, G. H., AND M. L. WILSON. 1989. Demography of the meadow vole along a simple habitat gradient. Canadian Journal of Zoology 67:772 774. ANIMAL CARE AND USE COMMITTEE. 1998. Guidelines for the capture, handling, and care of mammals as approved by the American Society of Mammalogists. Journal of Mammalogy 79:1416 1431. BATZLI, G. O. 1992. Dynamics of small mammal populations: a review. Pp. 831 850 in Wildlife 2001: populations (D. R. McCullough and R. H. Barrett, eds.). Elsevier Applied Science, New York. BATZLI, G. O. 1996. Population cycles revisited. Trends in Ecology and Evolution 11:448 449. BIRNEY, E. C., W. F. GRANT, AND D. D. BAIRD. 1976. Importance of vegetative cover to Microtus cycles. Ecology 57:1043 1051. BIXLER, S. H., AND D. W. KAUFMAN. 1995. Prairie voles at low density in ungrazed tallgrass prairie in the Flint Hills region of eastern Kansas. Prairie Naturalist 27:33 40. BURT, W. H. 1940. Territorial behavior and populations of some small mammals in southern Michigan. Miscellaneous Publications of the Museum of Zoology, University of Michigan 45:1 58. COLE, F. R., AND G. O. BATZLI. 1978. Influence of supplemental feeding on a vole population. Journal of Mammalogy 59:809 819. COLE, F. R., AND G. O. BATZLI. 1979. Nutrition and populations of the prairie vole, Microtus ochrogaster, in central Illinois. Journal of Animal Ecology 48:455 470. DERTING, T., AND J. A. CRANFORD. 1989. Physical and behavioral correlates of prey vulnerability to barn owl (Tyto alba) predation. American Midland Naturalist 121:11 20. DESY, E. A., AND G. O. BATZLI. 1989. Effects of food availability and predation on prairie vole demography: a field experiment. Ecology 70:411 421. DUESER, R. D., M. L. WILSON, AND R. K. ROSE. 1981. Attributes of dispersing meadow voles in open-grid populations. Acta Theriologica 26:139 162. GAINES, M. S., AND L. R. MCCLENAGHAN, JR. 1980. Dispersal in small mammals. Annual Review of Ecology and Systematics 11:163 196. GETZ, L. L. 1961. Hunting areas of the long-eared owl. Wilson Bulletin 73:79 82. GETZ, L. L. 1968. A method for measuring light intensity under dense vegetation. Ecology 49:1168 1169. GETZ, L. L. 1970a. Influence of vegetation on the local distribution of meadow vole in southern Wisconsin. Occasional Papers in Biology, University of Connecticut 1:213 241. GETZ, L. L. 1970b. Habitat of the meadow vole during a population low. American Midland Naturalist 83:455 46l. GETZ, L. L. 1985. Habitats. Pp. 286 309 in Biology of New World Microtus (R. H. Tamarin, ed.). Special Publication, The American Society of Mammalogists 8:1 893. GETZ, L. L., F. R. COLE, L.VERNER, J.E.HOFMANN, AND D. AVALOS. 1979. Comparisons of population demography of Microtus ochrogaster and M. pennsylvanicus. Acta Theriologica 24: 319 349. GETZ, L. L., D. GUDERMUTH, AND S. BENSON. 1992. Pattern of nest occupancy of the prairie vole, Microtus ochrogaster, in different habitats. American Midland Naturalist 128:197 202. GETZ, L. L., AND J. E. HOFMANN. 1999. Diversity and stability of small mammals in tallgrass prairie habitat in central Illinois, USA. Oikos 85:356 363. GETZ, L. L., J. E. HOFMANN, B.J.KLATT, L.VERNER, F.R.COLE, AND R. L. LINDROTH. 1987. Fourteen years of population fluctuations of Microtus ochrogaster and M. pennsylvanicus in east central Illinois. Canadian Journal of Zoology 65:1317 1325. GETZ, L. L., J. E. HOFMANN, B.MCGUIRE, AND T. W. DOLAN III. 2001. Twenty-five years of population fluctuations of Microtus ochrogaster and M. pennsylvanicus in three habitats in east-central Illinois. Journal of Mammalogy 82:22 34. GETZ, L. L., B. MCGUIRE, T.PIZZUTO, J.E.HOFMANN, AND B. FRASE. 1993. Social organization of the prairie vole (Microtus ochrogaster). Journal of Mammalogy 74:44 58.

568 JOURNAL OF MAMMALOGY Vol. 86, No. 3 GLICKMAN, S. E., AND B. J. MORRISON. 1969. Some behavioral and neural correlates of predation susceptibility in mice. Communications in Behavioral Biology 4:261 267. HAKEN, A. E., AND G. O. BATZLI. 1996. Effects of availability of food and interspecific competition on diets of prairie voles (Microtus ochrogaster). Journal of Mammalogy 77:315 324. KLATT, B. J., AND L. L. GETZ. 1987. Vegetation characteristics of Microtus ochrogaster and M. pennsylvanicus habitats in east-central Illinois. Journal of Mammalogy 68:569 577. KLEMOLA, T., E. KORPIMÄKI, AND K. NORRDAHL. 1998. Does avian predation risk depress reproduction of voles? Oecologia 115: 149 153. KORPIMÄKI, E., K. NORRDAHL, T.KLEMOLA, T.PETTERSEN, AND N. C. STENSETH. 2002. Dynamic effects of predators on cyclic voles: field experimentation and model extrapolation. Proceedings of the Royal Society of London, B. Biological Sciences 269:991 997. KORPIMÄKI, E., K. NORRDAHL, AND J. VALKAMA. 1994. Reproductive investment under fluctuating predation risk microtine rodents and small mustelids. Evolutionary Ecology 8:57 368. KREBS, C. J. 1996. Population cycles revisited. Journal of Mammalogy 77:8 24. KREBS, C. J., AND J. H. MYERS. 1974. Population cycles in small mammals. Advances in Ecological Research 8:267 399. LIN, Y. K., AND G. O. BATZLI. 1995. Predation on voles: an experimental approach. Journal of Mammalogy 76:1003 1012. LIN, Y. K., AND G. O. BATZLI. 2001. The influence of habitat quality on dispersal, demography and population dynamics of voles. Ecological Monographs 71:245 275. LINDROTH, R. L., AND G. O. BATZLI. 1984. Food habits of the meadow vole (Microtus pennsylvanicus) in bluegrass and prairie habitats. Journal of Mammalogy 65:600 606. MESERVE, P. L. 1971. Population ecology of the prairie vole, Microtus ochrogaster, in the western mixed prairie of Nebraska. American Midland Naturalist 86:417 433. MÜLLER-SCHWARZE, D., AND C. MÜLLER-SCHWARZE. 1971. Responses of chipmunks to models of aerial predators. Journal of Mammalogy 52:456 458. NORRDAHL, K., AND E. KORPIMÄKI. 2000. The impact of predation risk from small mustelids on prey populations. Mammal Review 30: 147 156. NORRDAHL, K., AND E. KORPIMÄKI. 2002. Changes in individual quality during a 3-year population cycle of voles. Oecologia 139: 239 249. OSTERMEYER, M. C. 1983. Maternal aggression. Pp. 151 179 in Paternal behavior of rodents (R. W. Elwood, ed.). John Wiley & Sons, Chichester, United Kingdom. PELES, J. D., AND G. W. BARRETT. 1996. Effects of vegetative cover on the population dynamics of meadow voles. Journal of Mammalogy 77:857 869. SAS INSTITUTE INC. 1999. SAS/STAT user s guide. Volumes 1 3. SAS Institute Inc., Cary, North Carolina. SLADE, N. A., L. A. RUSSELL, AND T. J. DOONAN. 1997. The impact of supplemental food on movements of prairie voles (Microtus ochrogaster). Journal of Mammalogy 78:1149 1155. SPIEGEL, R., E. PRICE, AND U. W. HUCK. 1974. Differential vulnerability of wild, domestic and hybrid Norway rats to predation by great horned owls. Journal of Mammalogy 55:386 392. SPSS, INC. 2001. SPSS 10.0.7 for Macintosh. SPSS, Inc., Chicago, Illinois. THOMPSON, D. Q. 1965. Food preferences of the meadow vole (Microtus pennsylvanicus) in relation to habitat affinities. American Midland Naturalist 74:76 86. VERNER, L., AND L. L. GETZ. 1985. Significance of dispersal in fluctuating populations of Microtus ochrogaster and M. pennsylvanicus. Journal of Mammalogy 66:338 347. ZIMMERMAN, E. G. 1965. A comparison of habitat and food species of two species of Microtus. Journal of Mammalogy 46:605 612. Submitted 24 August 2004. Accepted 13 October 2004. Associate Editor was Douglas A. Kelt.