Track Counts as Indices to Abundances of Arboreal Rodents
Journal of Mammalogy
Counting tracks to obtain an index of abundance for species difficult to capture offers a promise of efficiency and effectiveness when broad surveys of populations are necessary. Sand plots, smoked kymograph paper, and, recently, smoked aluminum plates have been used to record tracks (Raphael et al., 1986; Taylor and Raphael, 1988) . Findings of studies of carnivores conducted by using smoked-aluminum track-plates suggest the technique might prove useful for study of arboreal rodents,
... northern flying squirrels, Glaucomys sabrinus (Raphael et al., 1986) . Raphael et al. (1986) reported patterns of abundance of flying squirrels based on track-plate surveys and recommended track plates for that purpose. We tried unsuccessfully to calibrate a track index to densities of flying squirrels and Townsend's chipmunks (Tamias townsendii), as determined by live trapping, and abundance of Douglas' squirrels (Tamiasciurus douglasii), as determined from point counts of calls. Although the techniques are not directly comparable, rank orders of abundance derived from live trapping and point counts (direct observations of individual animals) can be contrasted with rank orders derived from track counts (observation of sign). We report those results here. We placed 12 track stations, 100 m apart, in each of 25 stands west and northwest of Roseburg, Douglas Co., Oregon, in the Coast Ranges between 26 June and 28 August 1985. Fifteen stands were old-growth Douglas-fir (Pseudotsuga menziesii), 220-525 years old; five stands were mature, 80-120 years old; and five were young, 40-70 years old. Each track station consisted of two adjoining 0.8-by 0.4-m sheets of 1.0-mmthick aluminum covered with a thin film of soot from a kerosene torch. We used cans of tuna-flavored cat food as bait; a can was placed in the center of each of the 0.8-by 0.8-m stations. After 3 nights, we recorded all tracks for each station. All but two stands were sampled once; one stand was sampled seven times during the 8-week period (two of these samples were ruined by rain), and-one stand was sampled twice. These multiple samples were averaged for contrasts with the live-trapping and point-count estimates. We calculated a trackplate index (the number of plates with ≥1 track divided by the number of plates deployed). The sampling procedure was based on a pilot study in which the expected frequency of tracks was estimated (K. B. Aubry, in litt.). We counted Douglas' squirrels within 50 m of the stations during seven visits to each of 20 stands during April, May, and June 1985; 10 min were spent at each station. We tried to avoid duplicate counts of the same individuals in a visit. We calculated a point-count index (number of stations at which squirrels were recorded ≥ 1 time divided by total number of stations). Douglas' squirrels are vocal and easily observed (Smith, 1968) . They are individually territorial with diameters of territories averaging about 100 m. Our point-count index should have provided a reliable estimate of Douglas' squirrel abundance, because our station spacing was comparable to the average diameter of territories. We trapped northern flying squirrels during October and November 1985, on grids encompassing the track stations in eight stands. Northern flying squirrels are not easily caught during the dry summer; live trapping is effective in the rainy spring and autumn (when track plates would be impractical). Live trapping for 5 years demonstrated that populations of flying squirrels do not fluctuate dramatically between spring and autumn.