COS 107-7
The synergic trio of exotic mammals at the southern end of the Americas: Evidence of an invasional meltdown among beavers, muskrats, and mink on Navarino Island (55°S), Chile

Thursday, August 13, 2015: 10:10 AM
338, Baltimore Convention Center
Ramiro D. Crego, Department of Biological Sciences, University of North Texas, Denton, TX
Jaime E. Jiménez, Sub-Antarctic Biocultural Conservation Program, IEB-UMAG-UNT, Chile
Simon Castillo, Institute of Ecology and Biodiversity, Santiago, Chile
Matías Barceló, Institute of Ecology and Biodiversity, Santiago, Chile
Nicolás G. Carro, Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata, La Plata, Argentina
Ricardo Rozzi, Sub-Antarctic Biocultural Conservation Program, IEB-UMAG-UNT, Denton, TX

Invasive species are a major driver of biodiversity loss, even in remote regions of the world. Negative impacts on invaded ecosystems can be more severe when two or more invasive species act synergistically (invasional meltdown). At the southern end of the Americas, three species that naturally interact in their native range in North America ─American beavers (Castor canadensis), muskrats (Ondatra zibethica), and American mink (Neovison vison)─  were independently introduced into Tierra del Fuego and spread to Navarino Island, Chile. We investigated how muskrat presence relates with beaver-modified habitats in stream systems of forest-dominated areas of Navarino Island, and if there are trophic interactions of the mink with muskrats and beavers. Between February and April 2014 and 2015 we estimated presence/absence of muskrats by sampling their signs of activity along 200 m transects, separated > 100 m. We completed 75 transects in four types of systems: active beaver dams with a pond, inactive beaver dams that still preserve a pond, inactive beaver dams where the stream recovered its flow, and lotic streams without beaver activity. We recorded altitude, distance to coast, and habitat type: forest or meadow. We fitted generalized linear models with binomial distribution, using Akaike’s Information Criterion for model selection. Additionally, we analyzed 202 mink feces, 96 collected on the coast, and 106 collected on beaver dams and rivers. 


Our results show a positive association between muskrat presence and beaver activity. We found two best-supported models that best explained the presence of muskrats. These two models included type of system as a explanatory variable. Model average coefficients showed that muskrats preferred lentic-beaver modified systems, compared to lotic ones. The probability (±SE) of finding muskrats was higher for inactive beaver dams with a pond (0.57 ± 0.13), followed by active beaver dams (0.68 ± 0.1), low for beaver-modified habitats where the stream recovered the flow (0.39 ± 0.11), and almost null in rivers without beavers (0.05 ± 0.05). Overall, mammals represented the main prey item for mink followed by fish. Muskrats represented almost 50% of the bulk of mink diet in inland waters. Our results suggest that beaver activity creates suitable habitat that allows muskrat establishment. In turn, muskrat presence provides a stable prey base for mink on inland territories. Today, beavers, muskrats, and mink seem to synergistically interact to invade and impact the pristine ecosystems, affecting local biodiversity and habitats at the southernmost archipelagos of the Americas.