An assessment of habitat quality of heavily- and less-browsed Michigan forests for a shrub -nesting songbird.

Abstract

Our best opportunity for conserving migratory songbirds in northern forests is to include species of concern in landscape-scale management plans. However, evaluating management effects on species requires knowledge of the relative quality of forests that vary in species composition, management history, and browse pressure from white-tailed deer (<italic>Odocoileus virginianus </italic>). I used estimates of relative abundance, demographic data, and spatial analysis of territories to compare the habitat quality of heavily-browsed (HB) and less-browsed (LB) forests for a migratory songbird, the black-throated blue warbler (<italic>Dendroica caerulescens</italic>). I studied blue warblers because they nest and forage in browse-susceptible understory vegetation, and worked in managed forests within northern Michigan's Hiawatha National Forest (1997--2001). Heavily-browsed (HB) sites had short (0.50--0.75 m) hardwoods in the understory (not completely browsed due to snow protection), and patches of browse-resistant balsam fir (<italic>Abies balsamea</italic>), while LB-site understory was primarily hardwoods. Blue warblers were most common in high shrub-density areas, regardless of browse level. Within high shrub-density HB sites, high fir-density areas had significantly more warblers. I conducted demographic and spatial distribution studies at densely-vegetated HB and LB sites. At both site-types, over 90% of pairs reproduced successfully, with 22--26% (LB vs. HB) of pairs fledging two broods. This similarity in reproductive success was in spite of higher nest survival on HB sites. HB sites had higher percentages of older males (78% at HB, 63% at LB), especially among new arrivals, but yearly return rates of adults were similar, with approximately 45% of males and 30% of females returning. Territory sizes (estimated using a bootstrapped minimum convex polygon) typically ranged from 1--4 ha, and tended to be larger on HB sites, although values at all sites were highly variable. Geographic boundary analyses suggested that patches of dense hardwood saplings 1--3 m tall were most often occupied at LB sites, while HB-site territories were concentrated in areas with small firs, or with dense patches of both browsed hardwoods and firs. Overall, habitat quality appeared slightly higher in LB forests, primarily as a function of pair density, with firs providing key refuge habitat within HB forests.