Local forest structure, climate, and human land cover determine broad-scale gradients of boreal bird species richness

Background/Question/Methods

Little information exists about how local habitat structure, climate, and human disturbance combine to determine broad-scale richness gradients. Here, we separated how various measures of local forest structure, abiotic factors, and human land cover constrain species richness of bird guilds in the boreal forest of Western Canada. Using a spatially fine-grained (<1 km resolution) dataset covering >200 sites across a large geographic extent (ca. 1000 km), we modeled bird guild species richness in relation to forest structure (woody plant richness, forest productivity, number of vegetation layers, canopy openness), abiotic environment (temperature, precipitation, elevation), and percent area of human land cover. We classified bird species into different guilds based on dietary preference, habitat specialization, and migratory status, and used structural equation modeling to quantify effect strengths of predictor variables.

Results/Conclusions

We found that temperature, low levels of human land cover, woody plant richness and number of vegetation layers had strong (and positive) effects on overall bird species richness in the boreal forest. However, local forest structure showed a pronounced variation in its effect on species richness of different guilds. Insectivores, old-growth forest specialists, forest generalists, long-distance migrants and winter residents showed strong positive responses to woody plant richness whereas old-growth forest specialists and winter residents were also strongly affected by forest productivity. The number of vegetation layers showed a positive effect on all guilds (except carnivores) whereas the response to canopy openness was most pronounced for old-growth forest specialists (being negatively affected). Overall, in addition to climate and human land cover, local forest structure is an important determinant of broad-scale richness gradients. However, the strength and direction (positive/negative) of determinants is guild-specific suggesting that community structure is likely to change under future climate and land use change.