Canada lynx restoration at Isle Royale National Park: A feasibility study

Licht, Daniel; Moen, Ron; Romanski, Mark; Brown, Paul

Isle Royale is a 535-km2 island located in western Lake Superior, about 22 km from the Minnesota/Ontario shoreline. The island and surrounding islets and waters comprise Isle Royale National Park. The Canada lynx (Lynx canadensis) is a 7-15 kg wildcat indigenous to the boreal forests of northern North America. The species is listed as threatened in the contiguous 48 states per the Endangered Species Act. It is widely believed that Canada lynx historically occupied Isle Royale, but have been essentially absent since the 1930s. The island supports a prey base—specifically, snowshoe hare (Lepus americanus)—that might support a lynx population. As directed by National Park Service policies, we evaluated the ecological feasibility of reintroducing Canada lynx to Isle Royale. Lynx were present on Isle Royale, apparently in great abundance, when explorers and settlers of European descent first reached the island. A harvest of lynx was underway by the 1890s. Reported harvests include a minimum of 48 lynx taken in the winter of 1903-04 and 67 lynx in the winter of 1916-17. We conducted a Population Viability Analysis (PVA) of the reported and assumed harvest rates from the 1890s-1930s. Although there are several possible reasons for the extirpation of lynx in the 1930s, we conclude that over-harvest was likely the primary cause. There were periodic reports of lynx on Isle Royale from the 1960-80s; however, those animals were likely immigrants from the mainland that did not persist. There is no evidence of a resident population since the 1930s. Based on known snowshoe hare densities at Isle Royale, reported lynx densities in Minnesota and Ontario, and other scientific information we concluded that Isle Royale can support about 30 lynx on average. However, we acknowledge that the mean carrying capacity might be substantially more or less than 30 lynx due in part to the unique characteristics of Isle Royale, so we evaluated long-term viability across a range of population sizes. Furthermore, we varied the carrying capacity over time as Canada lynx populations in northern regions often go through a 10-year cycle in abundance. For example, our baseline carrying capacity of 30 lynx fluctuated from 15 to 45 animals every decade. However, some southern lynx populations show little if any cyclicity so we also evaluated non-cyclical scenarios. We conducted a PVA of a potential lynx reintroduction to assess the persistence of the population. When we parameterized the PVA with demographic rates reported from mainland studies our cyclical model indicated a 0.36 chance of population persistence for 100 years. A non-cyclical model indicated a 0.73 probability of 100-year persistence. A sensitivity analysis of the model indicated that assumed levels of inbreeding depression had a significant impact on long-term viability. For example, when we disabled inbreeding depression the probability of persistence increased to 0.96 and 1.00 for the cyclical and non-cyclical scenarios, respectively. When we modeled the addition of 2 unrelated lynx into the population every 10 years the probability of persistence increased to ≥ 0.98 for both models, even when inbreeding depression was enabled. In historical times lynx might have immigrated to Isle Royale every 10 years or so via an ice-bridge connecting the mainland to the island. However, we found that the probability of ice-bridge formation in a winter has declined from 0.8 in 1959 to 0.1 in 2013, apparently due to climate change. As a result, to mitigate for the loss of natural ice-bridge connectivity managers might have to periodically introduce new lynx to the island to avoid inbreeding depression.

Canada lynx restoration at Isle Royale National Park: A feasibility study

Series

Collections