Barred Owl Effects on Spotted Owl Resource Selection: A Meta‐Analysis

Abstract

ABSTRACTWe quantified the effects of barred owls (Strix varia) on fine‐scale (~2 ha), spatially explicit selection of putative foraging habitat by 175 radio‐tagged northern spotted owls (Strix occidentalis caurina) from 1998 to 2012 in western Oregon and northern California, USA. We constructed annual nighttime discrete‐choice resource selection functions (RSFs) for northern spotted owls in 8 study areas. We compared RSFs from pooled data from 4 study areas where barred owls were absent or rare (≤14% of spotted owl territories potentially influenced by barred owls) with RSFs from 4 study areas where barred owls were abundant (≥89% of spotted owl territories potentially influenced by barred owls), and we conducted a meta‐analysis of all 8 study areas. Top‐ranked RSFs for individual study areas indicated that the relative probability a forest patch was selected by northern spotted owls for foraging was most strongly associated with basal area of large coniferous trees (≥66 cm diameter at breast height [dbh]) or quadratic mean diameter (QMD, the diameter of trees of average basal area) of all coniferous trees. The influence of such trees was non‐linear, usually quadratic, in 7 study areas, suggesting that an optimal basal area of large‐diameter trees or QMD may be associated with selection of foraging patches. The relative probability of selection of such trees decreased nonlinearly with increasing distance from nest sites in ≥4 study areas. Basal area of all or specific hardwood tree species was linearly associated with increases in the relative probability of a patch being used for foraging in 7 of 8 study areas. The relative probability of selection was also positively associated with densities of large snags (>50 cm dbh), coarse woody debris, and undergrowth shrubs in most study areas where we estimated those covariates. The relative probability that a patch was selected decreased nonlinearly with increasing distance from streams (7 study areas), increasing elevation (5 study areas), and decreasing slope (5 study areas). Barred owls exerted strong effects on habitat selection by spotted owls, which may have attempted to minimize interference competition via spatial segregation. Spotted owls occupying study areas where barred owls were abundant used foraging patches at greater distances from streams, at greater distances from nest sites, and on steeper and warmer slopes than those occupying areas with few or no barred owls. In addition, northern spotted owls in areas where barred owls were abundant exhibited reduced selection for large‐diameter trees close to nest sites but may have compensated by increasing use of large trees in home‐range locations distant from nest sites. Programs that seek to improve the link between spotted owl demographic performance or occupancy rate and habitat conditions should account for the influences of fine‐scale variation in vegetation composition and structure and physiographic conditions. Forest managers may contribute to spotted owl conservation via silvicultural prescriptions that retain coarse woody debris and large snags, promote undergrowth shrubs, and retain intermediate to high densities of large, old shade‐intolerant early‐seral trees and riparian hardwoods within and beyond core areas surrounding nest sites. © 2019 The Wildlife Society.