Occurrence and activity of American martens Martes americana in relation to roads and other routes

Abstract

In order to clarify the relationship between American martens Martes americana (Turton, 1806), and roads in boreal forests, we tested the effect of distance from road on marten habitat use by comparing the frequency of occurrence, and the number of, marten snow tracks in 27 pairs of 300 and 400 m transect segments perpendicular to access roads, and corresponding segments 800 or 1000 m away from the road, for a total of four spatial analyses. The number of black spruce Picea mariana, stems was generally lower near roads (-94.0 < T(s) < -63.0, p < 0.05), but otherwise forest composition generally did not differ significantly between road and forest segments. Depending on the spatial scale, 11 - 14 road and 15 - 18 forest (out of 27) segments had marten tracks, but the difference was not statistically significant (Fisher's exact test, p > 0.05). In all analyses, marten tracks were significantly fewer near roads (Wilcoxon test: -74.0 < T(s) < -62.0, p < 0.01) than away from roads. In segments with tracks, track density was significantly lower near roads than away from roads in three of four spatial analyses (-3.3 < T(s) < -2.9, p < 0.01). The findings suggest that snow track samples near roads can detect marten occurrence, but do not reflect the clumping pattern observed away from roads. Habitat selection studies using tracks should consider the possible effect of roads on marten movement patterns.