ELK PELLET-GROUP DECOMPOSITION AND DETECTABILITY IN COASTAL FORESTS OF WASHINGTON

Short count intervals (< 60 days) for pellet-group surveys in Pacific coastal forests, where precipitation is high and winter temperature mo derate, are costly but necessary to reliably estimate elk (Cervus elap has roosevelti) use of winter habitat because of potentially rapid pel let decomposition. We wanted to determine rate of pellet-group decay a nd minimum count interval necessary to detect > 90% of overwinter pell et groups in coastal rain forests on the Olympic Peninsula, Washington . We analyzed effects of elevation and aspect in a randomized block ex perimental design, with forest canopy effects examined as a split-plot effect. Decomposition of experimental pellet groups was rated periodi cally on an ordinal scale from December 1991 through May 1993. Pellet groups persisted longer overwinter than expected from other studies, i ndicating that pellet-group counts can be a reliable technique in many situations. At least 90% of pellet groups were present after the firs t winter or 4 months after deposition. The detection rate of pellet gr oups on forested slopes was > 90% up to 1 year after deposition. Pelle t groups in valley bottoms decayed more rapidly (P = 0.001) than those on mid- and upper slopes. Decomposition rates on north and south aspe cts were not different (P < 0.43). Pellet groups decayed more rapidly in clear-cuts than in forests (P < 0.001), which was opposite the patt ern found in an earlier study. Decomposition rates and detection proba bilities that we estimated are conservative and should be generally ap plicable for designing pellet-count studies in coastal forests of the Pacific Northwest with similar environmental conditions.