Muskrat-River Otter Interactions in and Adjacent to Mammoth Cave National Park, Kentucky

Anecdotal observations have suggested that muskrat (Ondatra zibethicus) populations were dramatically reduced in streams where the North American river otters (Lontra canadensis) were reintroduced. Muskrats predate upon freshwater mussels and it was speculated that river otter reintroduction could result in increased mussel numbers. My objectives were to evaluate the ecological relationship between otter, muskrat, and mussels on the Green and Nolin rivers in Mammoth Cave National Park (MCNP). Seventeen river otters were captured in or relocated to MCNP from January to May 2007. The augmentation was only marginally successful with 3 male river otters establishing home ranges within the park; of the remaining 13 animals, 10 individuals dispersed >35 km outside of the study area and 3 died shortly after release. Despite the relatively low success of the augmentation, scent-station surveys, trap-site visitation, and scat collection indicated that otter numbers had significantly increased on the Green River since Asmus‟ (2004) study, probably as a result of natural immigration and range expansion. Although spotlight surveys indicated that there was a concomitant decline in the muskrat population along the Green River from 2002 to 2008 (F_{1, 73} = 36.56, P < 0.0001), muskrat hair was only found in 1 of 48 (2%) river otter scats examined. That evidence, coupled with a relatively high number of both otters and muskrats on the Nolin River, did not indicate that the relationship between muskrats and otters in MCNP was causal.

On a more extensive scale, I collected data on muskrat and otter presence at 95 randomly selected bridge crossings across Kentucky but centered on MCNP. I used a 2-species co-occurrence model in Program PRESENCE to determine if the presence of river otters was related to the presence of muskrats. This occupancy model indicated that muskrats occurred independently of river otters (φ = 1.02). Observer, water level, and substrate were important determinants of otter detection, whereas straight-line distance from original river otter release sites was an important occupancy covariate for river otters. Therefore, both my intensive and extensive data analyses do not support the notion of a negative interaction occurs between river otters and muskrats.