Landscape Ecology of North American River Otter (Lontra canadensis) in the Upper Peninsula of Michigan

As loss of habitat, fragmentation, and climate change continue to alter natural habitats, connectivity of the landscape becomes necessary for species conservation. My dissertation covers several of the factors that affect connectivity for North American river otter (Lontra canadensis) populations in the Upper Peninsula of Michigan. First, we developed a new non-invasive method that captures DNA from snow tracks to identify individual otters. We were successful in identifying 66 individual otters from 87 putative otter samples. This allowed for estimation of population density and genetic diversity for connectivity analyses. Next we conducted a systematic review of the literature and meta-analyses to determine habitat variables that otters select. At the latrine scale, otters were positively associated with forested areas, a high percentage of overhead cover, and complex shorelines. Otters avoided areas with high percentage of herbaceous and shrub cover. At the river segment scale, otters avoided human disturbance and were found in areas with a high percentage of forest, higher number of ponds, closer to lakes, and deeper water than random segments. The significant variables from the meta-analyses were applied to resource selection functions (RSF). The Meta-analysis RSF model was compared with a use/random RSF model, use/road-stream crossing RSF model, and a null model. The Meta-Analysis RSF model did not predict use as well as the use/road-stream crossing RSF model, indicating that meta-analyses may be helpful in determining important habitat variables, however, the coefficients may not be transferrable across the otter’s range. Toxoplasma gondii is a parasite that may be affecting the connectivity of otter populations. We found that prevalence was 28% in sampled otters and 69% percent of T. gondii positives were Type #4 clones. When prevalence was modeled with other factors, the presence of Sarcocystis, the percent area of exotic vegetation, the percent area of agriculture, and sex explained 78% of the variation. Understanding the connectivity of the landscape is dependent on multiple variables that interact on different spatial and temporal scales. However, maintaining connectivity for wildlife populations is necessary for protecting biodiversity in a changing world