Re-examining density-habitat relationships across thirty years of change in forest management and boreal bird communities

Background/Question/Methods

Habitat loss and degradation are primary drivers of global biodiversity loss, implicated in significant recent population declines of North American avifauna. However, the effects of habitat change are species-specific, and geographically and temporally variable. Deepening our understanding of these dynamics at different scales will help to clarify the fundamental relationship between population density and habitat change, and provide critical information to guide avian conservation strategies. In this light, our study explores the multi-scale habitat associations of a boreal bird assemblage within the Maine North Woods, a dynamic 4-million-ha mosaic of private, commercial, public, and tribal forest parcels with varying structural characteristics and management histories. Following the methods of a study conducted by our group members on the same landscape (1992-1996), we are conducting standardized avian point counts and stand-level vegetation surveys across nine age-type forest classes within this mosaic. We aim to: 1) classify current and prior avian density-habitat associations using boosted regression trees, 2) test alternate hypotheses about the relationship between density and gradients of habitat type and structure, 3) evaluate these changes in the context of continental-scale bird declines. Here we report on data from 2021, where we surveyed 120 points across all forest classes.

Results/Conclusions

We recorded 77 species across all counts, 59 of which were detected within the 50-m radius used for point-level density estimates. Eight new species were detected since 1996, and six were no longer detected, suggesting community change may be occurring at a regional scale. Respective landcover of each forest class has also changed over time. In addition, provisionally, a majority of the species detected in both studies (30 of 51) changed habitat affiliations, by raw assessments of peak densities, since 1996. This was surprising, and could reflect a number of processes, including response to shifting management regimes in the region, broader continental shifts in range core, or niche plasticity. Density estimates and habitat affiliations are being refined by correcting for detectability and using boosted regression trees. Interestingly, our peak species-habitat densities differ from those reported in other boreal regions. We are particularly interested in being able to describe the shapes of species-specific relationships to variation in forest age, composition, and structure. Eventually, we hope to be able to recommend landscape-level silvicultural practices that could improve avian conservation while meeting foresters’ goals over the coming century.