Snowshoe hare population ecology in lynx-occupied areas of Washington

Paul Otto Jensen

doi:10.7273/000004202

Boreal forest ecosystems are being disrupted by climate change, resulting in alterations to species distribution and abundance. Snowshoe hares (Lepus americanus) are a keystone boreal forest herbivore that has begun to experience range contraction in the continental United States. Despite their importance, we have limited spatio-temporal knowledge of their densities along the southern edge of their range where boreal habitat is fragmented and climatic conditions are marginal. To address this limitation, we explored the accuracy of three non-invasive techniques for estimating hare density including two indices (pellet density and camera trap detection rate) and a camera-trap-based model for direct density estimation (Random Encounter Model - REM) at 13, 10.25 ha sites in northcentral Washington. We compared these values to spatially explicit capture-recapture (SECR) density estimates from live-capture efforts at the same 13 sites. SECR ranged from 0.16-1.58 hares/ha. Pellet densities had a strong linear association to SECR (R2 = 0.91) and camera trap detection rates had a strong exponential relationship with SECR (R2 = 0.81). REM densities using our camera-trap-derived animal speed (2.4 km/day) ranged from 0.16 to 1.10 hares/ha, though estimates were sensitive to animal speed. Average absolute difference between REM- and SECR-based density estimates were 0.19 hares/ha (range = 0.00-0.84, SE = 0.07), with REM-based estimates for 10 of 13 sites falling within SECR confidence intervals. Additionally, using pellet density data maintained by two external sources and remote sensed data derived to describe forest structure and topography, we developed a landscape-scale pellet density model, converted pellets to density, and related hare density to detection rates of a threatened species, Canada lynx (Lynx canadensis) that relies heavily on hare as prey. We found pellet density is best predicted by stand age, horizontal and vertical cover and lynx detection rates are significantly higher when hare density is > 0.5 hares/ha. As hare populations change along their southern range edge, non-invasive techniques will enable the collection of cost-effective large-scale data that can be used to inform management and conservation, and provide the means to better assess trophic relationships within forested communities.

Snowshoe hare population ecology in lynx-occupied areas of Washington

Files and links (1)

Abstract

Metrics

Details