Delineating Local Lynx Distributions
Squires, J. R., K. S. McKelvey, and L. F. Ruggiero. 2004. A snow-tracking protocol used to delineate local lynx, Lynx canadensis, distributions. Canadian Field-Naturalist 118:583-589.
Delineating the current distribution of lynx is critical to the conservation of the species in the contiguous United States. We developed, tested, and published a peer-reviewed survey method for conducting extensive and representative lynx surveys based on winter snow-tracking (Squires et al. 2004).
Lynx have many attributes that make them ideal candidates for snow-track surveys such as large home ranges (approx. 150 km2 for males and 70 km2 for females; Aubry et al. 2000) and high daily travel rates (Ward and Krebs 1985). In the past, detection protocols that depended on snow-tracking often suffered from track misidentification and non-representative surveys. A misidentified Endangered or Threatened species can have major management ramifications.
We initiated a grid-based survey that is spatially representative, and that formally incorporates genetic sampling to prevent track misidentifications (Squires et al. 2004, McKelvey et al. 2006). In other words, we view the track as a collection device for obtaining genetic samples that are obtained in a spatially representative manner.
This method uses an 8 x 8 km grid overlaid over the survey area of interest. Each cell of this grid is a single survey unit, and 10 km of survey routes via snowmobile or snowshoe are surveyed within each cell. After completing a survey of all cells within the grid, the researcher has a distribution map delineating which survey units are occupied by lynx (Figure 19). We have conducted snow-track surveys over an extensive area of western Montana and Wyoming, and they have provided an effective
means of delineating local lynx distributions (Figure 20).
During winter 2006, we field-tested this survey method by conducting snow-track surveys within the home ranges of collared lynx on 2 study areas. We surveyed a total of 20 survey units, and repeated these surveys twice during the winter. We knew that between 17–90% of each cell was within a known lynx home range. Despite this variability, we detected lynx in 100% of cells where lynx were known to be present. Although the protocol specifies a survey of 10 km per cell, we first detected lynx in each cell after an average of 2.3 km of search effort and traveled an average of 3.4 km per lynx detection (Figure 21). We are still evaluating whether further testing of this protocol is warranted next winter.
next section Predicting Lynx Habitat at the Landscape Scale