Ecological studies of the American marten (Martes americana): Quantifying cryptic processes in an elusive species.

摘要

Habitat fragmentation is a major concern of conservation biologists because its two components, habitat loss and the conversion of contiguous habitats into smaller isolated patches, threaten biodiversity worldwide. Dispersal, the main process connecting animal populations inhabiting isolated habitat patches, is difficult to document and quantify, especially for wide-ranging species. This is particularly true for elusive species, like the American marten (Martes americana). Due to its habitat specialization, large spatial requirements, and longevity, the marten has been identified as an indicator species for mature, structurally complex forests in North America. In particular, martens were chosen as a design species for the Tongass National Forest (TNF). However, we lack information on the response of martens to fragmentation in this region, particularly as it relates to dispersal, so the effectiveness of such reserve design systems is unclear. Further, these designs are based largely on conceptual models developed for small mammals; it has not been determined whether they are applicable to systems featuring forest carnivores, including American martens.;To directly assess the effect of fragmentation on population processes and dispersal of martens within highly fragmented forests of the North Pacific, I quantified individual- and population-level attributes of American martens on Prince of Wales Island---an island that experienced extensive clearcut logging during 1954-1995. Because of the special challenges associated with working with an elusive and wide-ranging carnivore, I developed a series of novel, non-invasive techniques, which are described herein: (1) identifying individuals via microsatellite profiles, and gender with a sex-linked gene (Appendix B); (2) determining the age class of individuals from their telomere lengths (Chapter 4) and; (3) quantifying dispersal rates using isotopic labels within marten hair acquired from consuming isotopically enriched baits (Chapter 3). By developing these approaches, I was able to estimate population size, demographic structure, and independently corroborate rates and distances of dispersal as well as identify individual (age-sex class), population (demographic structure, animal abundance and density) and landscape features that influenced dispersal (Chapter 6).;I found that juvenile martens dispersed from sites featuring a high-degree of forest loss and fragmentation towards ones featuring larger, more contiguous forests. Although most landscape metrics were not strongly related to dispersal rates among study sites, the proximity of high-volume forest patches was positively related connectivity of marten populations. Collectively our findings contradict those predicted by conventional source-sink dynamics. Presumably from extreme temporal fluctuations in population size and relatively high reproductive potential, martens did not exhibit the source-sink dynamics expected from the landscape we studied. These results illustrate the complexities associated with identifying source and sink populations, particularly for species that fluctuate or cycle frequently.