Deer browsing correlates with songbird community change: Evidence from point count surveys at study sites with contrasting land use
White-tailed deer (Odocoileus virginianus) numbers in the eastern United States have increased dramatically over the past few decades. Human-induced forest fragmentation, extirpation of natural predators, and setting aside of expansive areas in which hunting does not occur have contributed to this population boom. While deer browsing impacts on understory vegetation are predictable, few large-scale studies have examined the influence of deer on forest bird communities. The objective of our study was to examine the effects of deer use on bird density collected over four years (n = 92 sites, 2010-2013) at a forest-suburban study site in Virginia. Additionally, we included 2012 data (n = 99 sites) from a forest-agriculture study site to assess land use effects. We predicted that densities of avian species nesting and foraging in low forest foliage will negatively correlate with deer-pellet density. To quantify deer habitat use and bird abundance, we used point counts to survey birds and fecal pellet counts to estimate relative deer habitat use. We used four 8-min surveys and one 120m transect at each site per year to estimate bird and pellet densities, respectively. Final density estimates were adjusted based on survey-specific covariates.
We detected significant correlations between bird and pellet density estimates both at the guild and species levels. At the forest-suburban site, preliminary results indicate negative correlation between pellet estimates and the understory nesting guild (rho = -0.31, P=0.0004), as well as the understory nesting and feeding guild (rho = -0.29, P=0.0011). Interestingly, the canopy nesting and feeding guild correlated positively with deer pellets (rho = 0.27, P=0.002). At the species-level, five birds correlated negatively (Mniotilta varia, Pipilo erythrophthalmus, Setophaga citrina, Setophaga discolor, Vireo griseus) and nine positively (Empidonax virescens, Cyanocitta cristata, Poecile carolinensis, Thryothorus ludovicianus, Contopus virens, Dryocopus pileatus, Melanerpes carolinus, Piranga rubra, Sitta carolinensis) with deer pellet densities. We found deer pellet density to be higher in the forest-suburban landscape than at the forest-agriculture site where we obtained only a weak negative correlation for the understory nesting and feeding guild (rho = -0.20, P=0.049), while detection number restriction did not allow for species-level analysis. Our results suggest that habitat modification by deer browsing promotes measurable changes in bird community composition, with conservation implications for declining forest songbirds.