Data for: Impact of invasive Typha and wetland interspersion on muskrat declines in North America

Muskrat (Ondatra zibethicus) populations are declining in North America. The exact cause of these declines is largely unknown. Along a similar timeframe, wetlands have been experiencing an invasion of cattail (Typha) throughout the continent. Specifically, T. x glauca, a hybrid of native T. latifolia and non-native T. angustifolia, has been increasing in range and abundance. This hybrid is associated with many negative impacts on wetland ecosystems, including reductions in biodiversity, open water habitat, and interspersion of water and emergent vegetation, the latter of which is an important habitat feature for muskrats. We sought to determine the impact of invasive T. x glauca on muskrat populations. We sampled 39 Typha-dominated marshes in southern Ontario, Canada to test the hypotheses that muskrats are declining in North America due to: (1) the increased relative abundance of T. x glauca in marshes, and (2) reduced wetland interspersion, which is associated with T. x glauca invasions. We estimated muskrat population density using house counts, sampled Typha communities to determine the relative abundance of T. x glauca, and measured interspersion using remote sensing techniques. We found that muskrat population density was positively associated with interspersion, but not associated with the relative abundance of T. x glauca. However, most sites were highly dominated by T. x glauca, limiting our inference. Our findings suggest that changing wetland structure may be contributing to muskrat population declines in North America, but more research is needed to determine the full impact of T. x glauca invasions on muskrat population declines. 

Methods
Muskrat house density was estimated by collecting aerial imagery at study sites in March 2021 and February 2022 and using the imagery to identify muskrat houses at each site. We used a random-sampling approach whereby a grid of 1-hectare cells was overlaid on each study site and cells that did not represent suitable muskrat habitat were eliminated; from the remaining cells, we randomly selected 10 cells to sample (only a few sites had less than 10 suitable cells, in which case all suitable cells were sampled). Muskrat houses appearing in these random cells were included in the calculation of mean house density in houses per hectare.

Typha x glauca relative abundance was estimated by sampling the Typha community in the field along predetermined transects at each site. Transects were accessed from the water, or in a few cases, on foot. Number of sampled transects ranged from 3 to 7 per site. Each transect was approximately 6 meters in length and was roughly perpendicular to the water's edge. At the start of the transect, the first ramet within reach with visible spikes (inflorescences) was sampled. Sampling consisted of obtaining a sprinkling of pollen when present, followed by a clipping of half of the pistillate (female) spike. Whenever possible, three ramets were sampled along each transect, with the first being collected at the water's edge, and the subsequent samples at 3 m increments along the transect. Pistillate spikes were placed under a dissecting microscope and used to identify the plant based on a botanical key. In a few cases, pollen was placed under a compound micriscope and was instead used to identify the plant using the same key.

Mean interspersion and mean proportion of open water were both measured using geospatial processing. This consisted of conducting land cover classifications of each site, identifying areas of Typha-dominated marsh, other vegetation, and open water. From the cells that represented suitable muskrat habitat, the proportion of open water was calculated for each 1-hectare cell and averaged across those cells. In each of those same cells, the total length of water polygon edges was calculated and averaged across those cells, This effectively represented emergent-water edge density in meters per hectare, which we used as a measure of interspersion.