Northern Prairie Wildlife Research Center
The survey of carnivore tracks was a modification of the survey of scent stations designed to index coyote abundance (Linhart and Knowlton 1975; Roughton and Sweeny 1982). Rather than using scent tabs to attract predators to small prepared sites where their tracks could be detected, we relied on personnel experienced in identification of tracks to find tracks at available unprepared sites in quarter-section search units. This modification was made possible by the abundance of good tracking media (exposed soft soil) at many sites (e.g., muddy wetland edges, gravel road edges, dirt trails, field edges) in nearly all quarter-sections during at least one search period. Our premise was that if a carnivore species regularly visited a quarter section, its presence was soon revealed by observable tracks at one or more locations, provided sites and conditions were suitable for finding tracks.
The survey consisted of searching for tracks of the coyote, red fox, raccoon, striped skunk, badger, and mink during two periods (April-mid-May, late May-late June) in each quarter section of each study area (only a mid-May-late June search was conducted in the Central Flyway study areas because of staff shortages). The decision to conduct two searches during these periods was based on examination of data from Canada in 1983 when four searches were conducted. Observers were instructed to wait at least 2 days after precipitation that destroyed tracks before conducting searches and to spend less than or equal to 0.5 h in each quarter section examining the most suitable sites for tracks of each species. All-terrain vehicles were used where permitted by landowners to facilitate rapid coverage of study areas. Two observer-days (one observer for 1 day) were required to search each study area once.
Results of track surveys were expressed as the percentage of searched quarter sections in which tracks of each species were found at least once during both searches combined. Data from the Central Flyway study areas, which were searched once, were expanded to estimate results as if they had been searched twice. We used data from study areas that had been searched twice to arrive at regression equations relating percentage of searched quarter sections by study area in which tracks of that species were found in either search to the percentage of searched quarter sections in which tracks were found during only the late search. We applied the regression equations to study areas that had been searched only once to estimate results as if the areas had been searched twice. The correlations between results from the double-search and single-search surveys for individual species ranged from r = 0.75 to r = 0.93. The statistical procedures for the expansions resulted in assigning limited occurrence of a species to a study area even if no tracks of that species were found. However, we classified as undetected each species for which no tracks were found, unless we had evidence substantiating its presence (e.g., sighting).
The survey of tracks differentiated better among low than high populations because presence of a single individual was likely to be detected in more than 1 quarter section. When a species became sufficiently abundant to regularly leave tracks in nearly all quarter sections, higher abundance could not be detected.