The Influence of Habitat Quality on Dispersal, Demography, and Population Dynamics of Voles

Abstract

We compared the effects of habitat quality on dispersal, demography, dynamics, and fitness of prairie vole (Microtus ochrogaster) and meadow vole (M. pennsylvanicus) populations by manipulating habitat patches in experimental landscapes. Four habitat patches in each of four replicate landscapes differed in availability of high-quality food and amount of vegetative cover in a 2 × 2 factorial design. High cover had a strong positive effect on basic habitat quality, as reflected by the performance of founders early in the season, but supplemental food had only a small effect. Population growth ceased for prairie voles after week 18 (mid-October) when densities had reached much higher levels in habitats with high cover (260 ± 27 voles/ha in high cover with either high or low food; mean ± 1 se) than in habitats with low cover (115 ± 38 voles/ha in high food, low cover and 60 ± 15 voles/ha in low food, low cover). Population growth had ceased in three habitat types for meadow voles by week 20 at much higher densities than for prairie voles in high cover (636 ± 101 voles/ha with high food, high cover; 556 ± 117 voles/ha with low food, high cover; 110 ± 74 voles/ha with high food, low cover; and 51 ± 16 voles/ha with low food, low cover). Correlates of fitness for prairie voles, particularly the number of young surviving to adulthood per female, indicated greater individual fitness in high cover than in low cover, which suggests ideal despotic habitat selection. The proportion of the total population found in habitat patches with high cover continued to increase until prairie vole populations stabilized late in the season. This result did not agree with expectations of either the ideal free model or the ideal despotic model of habitat selection. Fitness differed little for meadow voles in different habitat types, and the proportions of the populations in different habitats remained constant after the founders had settled and population growth began. Both of these patterns supported the ideal free model of habitat selection for meadow voles. Because positive net recruitment occurred in all habitats for both species of voles, source–sink dynamics could not occur in our experimental system. Two subpopulations of meadow voles in low-food, low-cover habitats did go extinct temporarily, but this habitat type did not appear to be a population sink because the losses occurred primarily from emigration rather than mortality. Emigration rates for both species of voles were inversely related to carrying capacity of the habitat (estimated as the density at which populations stabilized). This relationship and ideal free habitat selection are required by the balanced dispersal model, which produces equal numbers of dispersers between a pair of habitat types. Dispersal of meadow voles was balanced throughout the growing season, but dispersal of prairie voles was unbalanced, with net movement of individuals from low-quality to high-quality habitats until late in the season when populations stabilized. Unbalanced dispersal early in the season may have reflected a delayed response to habitat quality by prairie voles. Populations in high cover approached stable densities largely as a result of density-dependent in situ net recruitment, whereas populations in low cover stabilized more as a result of density-dependent net movement for both species of voles. Per capita emigration did not show a consistent positive relation to population growth in high cover, as would be expected if dispersal contributed to the tendency for prairie voles to display multi-annual population cycles in such habitats. In fact, emigration generally declined as density increased (inversely density dependent), which should have fostered greater population growth as densities increased. Nevertheless, population growth declined late in the season, which implied that emigration (dispersal) did not strongly affect the pattern of population dynamics in high-quality habitats. The overall results for the two species were remarkably similar, although prairie voles did differ from meadow voles in several respects. Meadow voles performed less well in low-cover habitats than did prairie voles, which probably reflected their greater sensitivity to predation risk. Prairie voles tended to move into habitats with supplemental food, and young prairie voles had greater body growth in habitats with supplemental food, whereas meadow voles showed neither response. Finally, although the results for meadow voles consistently indicated ideal free habitat selection throughout the growing season, those for prairie voles did not until late in the season. This difference between species likely reflected a more rapid response to habitat quality by the more vagile meadow voles.