4.1 Reasons for decline
It is unclear whether the apparent declines in muskrat abundance that we have demonstrated are a result of a broad-scale underlying cause or site-specific factors.
In recent studies of muskrat occurrences in Lake Ontario-St. Lawrence River coastal wetlands it was found that water-level regulation of that system had a negative influence on muskrat abundance (Greenhorn et al., 2017; Toner et al., 2010). While the two Great Lakes that our study sites occur on are not subject to such tightly controlled water levels as Lake Ontario, both Lake Erie and Lake Huron have experienced record high and extended low water levels over the duration that our investigation spans (NOAA, 2020), primarily brought on by weather extremes and climate variability over the past few decades (Gronewold and Rood, 2019). The presence of water, either through excessively high water levels or periods of drought, has been considered the greatest selection pressure affecting muskrats (Bellrose and Brown, 1941; Errington, 1963; Virgl and Messier, 1996; Proulx and Gilbert, 1983; Ahlers et al., 2015; Ward and Gorelick, 2018). Essentially, the water level in a marsh is suitable for muskrats when it is deep enough to maintain travel routes and allow access to houses and feeding areas underwater, even in cold winters when ice cover can be very thick, yet shallow enough to permit the growth of emergent aquatic vegetation. It is also important for muskrats that any changes in water level occur gradually, so that houses are neither flooded out nor left high and dry. We have not conducted a detailed analysis of water level patterns at Pelee and Matchedash in relation to muskrat occupancy, therefore this is an avenue that warrants closer examination.
Closely related to water levels is the hypothesis that habitat change has led to dramatic reductions in muskrat numbers. For example, Ward and Gorelick (2018) suggested that the loss of critical wetland habitat due to drying is the primary driver responsible for the decline of muskrats in the Peace-Athabasca Delta over the past half-century and potentially across their entire native range. While the amount of wetland habitat on the landscape available to muskrats is certainly an important factor influencing their population levels, the sizes of the wetlands in our study area have not changed appreciably over the timespan of our investigation. Province-wide, Ontario has experienced significant loss of wetlands over the past two centuries, but most of this loss (70% of original wetland area) occurred between the time of European settlement (circa late 1700s) and 1982 (Ontario Biodiversity Council, 2010). Only a relatively small additional amount of wetland area loss (3.8%) occurred between 1982 and 2014 (Ontario Biodiversity Council, 2015). Thus, it is clear that the vast majority of wetland loss in Ontario occurred prior to the recent period of muskrat decline that we have demonstrated. Therefore, it seems unlikely that a loss of habitat area is a major factor driving the recent declines we have seen in muskrat populations in Ontario over the past 30-50 years.
A more likely factor responsible for muskrat decline in Ontario is a change in the overall quality of muskrat habitat, particularly in the structure and composition of wetlands that were previously occupied in large numbers by muskrats. Proulx and Gilbert (1983) demonstrated that muskrats prefer to occupy areas of marsh with a 1:1 interspersion of emergent vegetation and open water, a structural pattern of habitat patchiness often referred to as a “hemi-marsh”. Yet there is evidence around the Great Lakes basin that the structural diversity of wetlands has declined. The research of Wilcox et al. (2008) has shown that Lake Ontario coastal wetlands have experienced a significant decrease in emergent marsh habitat heterogeneity and the amount of emergent-open water edge in the past 50 years since regional water level control was implemented. At the Point Pelee Marsh, one of our two study sites, Markle et al. (2018) demonstrated that both marsh habitat diversity and open water connectivity have declined since 1931, with the most significant reductions occurring between 1959 and 2015. Similarly, at our other study site, the Matchedash Bay-Gray Marsh, air photo interpretations by Taylor et al. (2015) demonstrate a large increase in the areal coverage of dense emergent vegetation along with a correspondingly large decrease in the extent of aquatic habitat between the years of 1973 and 2008. These reductions in habitat structural quality coincide with the time periods over which we have demonstrated a decline in muskrat abundance at both sites.
The research of Markle et al. (2018), Wilcox et al. (2008) and others (e.g., Farrell et al., 2010, Wilcox et al., 2003) also demonstrate changes to the compositional quality of Great Lakes coastal wetlands over the past 50 years that coincide with the decline in muskrats that we report here. Not only have they demonstrated class-level habitat changes, such as the expansion of dense emergent cattail marshes at the expense of meadow marsh and other shallow mixed marsh communities, but they have documented dramatic increases in the prevalence of invasive emergent marsh species such Phragmites australis subsp. australis, Typha angustifolia and Typha X glauca . In particular, the hybrid species of cattail (Typha X glauca) has increased significantly in extent, displacing both the native broad-leaved cattail (Typha latifolia ) and other aquatic plant species, and is now the dominant emergent species in most of our coastal wetlands. How such changes in both the structure and composition of coastal wetlands may be driving muskrat decline is a question that deserves further exploration.
Several other explanations for the decline in muskrat populations are possible but seem less likely for a variety of reasons. Overharvest by trappers is often postulated as the reason behind furbearer population declines, however the demographic characteristics of muskrats makes them resilient to overharvest (Boutin and Birkenholz, 1987). Even when heavy harvest does occur in a wetland, the muskrat’s high reproductive output and dispersal ability usually enables a population to rebound relatively fast. As well, one of our study sites (Pelee) has been closed to muskrat trapping since 1959, therefore muskrat harvest levels have had no direct influence on the population dynamics there for at least 50 years. At our other study site, local long-time trappers have reported low harvests and great difficulty finding muskrats for many years now (M. Dunlop, personal communication, March 2014).
Predation and disease are important natural aspects of muskrat ecology that can affect their abundance over the short-term, but which typically balance out over the long-term. Errington (1963) suggested that significant mortality of muskrats due to predation usually only occurs when muskrats are already vulnerable as a result of disease or habitat changes such as loss of cover, drought/flooding, or freeze-outs. As well, muskrat populations typically respond to predation mortality by a variety of compensatory processes, such as increased reproduction and survival among the remaining individuals (Errington, 1951). Mink (Neovison vison ) are considered the most important predator of muskrats (McDonnell, 1983), but there is evidence from harvest data of long-term mink population decline in Ontario (Gorman, 2007) and no evidence of local mink population booms at our sites (personal observations). Disease outbreaks among muskrats are typically localized and usually occur when a muskrat population is under some form of stress, such as overcrowding, food shortage or drought (Errington, 1963). Also, as with predation, the effects of disease on muskrat population size are often short-term because of compensatory responses in other sources of mortality or in reproduction. We are unaware of any documented outbreaks of disease in muskrats at either of our study sites or elsewhere in Ontario over the time period that our study covers (Ganoe et al., 2020).