| Habitat loss is the leading cause of declining wildlife populations around the world. Predicting species responses to habitat loss therefore represents a fundamental challenge in conservation biology. Effective management strategy necessitates a thorough understanding of (1) the quality of different habitat types for a given species, (2) the response of the individuals immediately impacted by habitat loss, and (3) the larger-scale repercussions of habitat loss on populations in the surrounding landscape. I used a combination of natural history surveys, manipulative field experiments, and matrix population modeling to address these issues for the white-footed mouse, Peromyscus leucopus.;First, I evaluated habitat quality for mice living in mosaic landscapes using four years of field data collected in northeastern Connecticut. I parameterized spatially explicit matrix population models with data collected for each of four mosaic study sites, revealing strong differences in habitat quality. Habitat-specific growth rates, as predicted by these models, were in all cases substantially lower than the growth rates projected for the mosaic sites as a whole, indicating the importance of multiple habitats in maintaining persistent regional populations. Next, using a manipulative field experiment, I showed that territorial interactions are an important mechanism determining the successful relocation of mice that are evicted from their native home ranges.;In addition, I used the matrix models to simulate the impact of habitat loss on the population dynamics across the mosaic sites. To test the predictions of these models, I altered the highest quality habitat at two of the four study sites, one with a prescribed burn and the other with clear-cut logging. Follow-up monitoring and modeling exercises revealed negative impacts of habitat loss across both sites, as predicted. The matrix models and the field experiments illustrated the importance of multiple habitats for ensuring the long term persistence of animals across mosaic landscapes. Combining behavioral and demographic information, and validating model predictions with landscape-scale manipulation experiments is an innovative and effective method with which to assess the impact of habitat loss. |
