Quantifying avian and forest communities to understand interdependencies of ecological systems and inform forest bird conservation
Bryce T. Adams, PhD
Advisor: Stephen N. Matthews
Forests represent the largest terrestrial biome on Earth, providing a wealth of ecological and social services. Their effective conservation and management under intensifying anthropogenic threats, climate change, and shifting disturbance regimes hinges on an accurate knowledge of ecological process and spatial pattern to address questions related to their dynamics, how they are changing, and what resources they provide to wildlife. Predictive models are currently the main tools used to quantify landscape-level forest parameters and resource use of wildlife communities. Advances in remote sensing technologies and new, innovative ways to characterize these data offer great potential for improved quantification and monitoring of ecological systems.
My overall research seeks to integrate new methodologies for landscape-level quantification of avian and forest communities and to investigate interrelationships that inform forest bird conservation in southeastern Ohio. The study area, positioned within the Central Hardwoods Region, displays a pronounced floristic gradient, recognized as one of the most speciose forested regions in the eastern US. I sampled avian and woody plant assemblages across a spectrum of forest stands with different vegetation composition and structure within six study sites during 2015 and 2016. My objectives include: (1) to determine the relative importance of plant taxonomic composition versus vegetation structure on the species composition of avian assemblages; (2) to examine the effectiveness of multi-sensor fusion of different remote sensing platforms that are sensitive to improved monitoring of forest successional state; (3) to incorporate concepts of community-continua to map floristic assemblages; (4) to develop predictive models of bird species abundance to map potential habitat quality for selected species and examine the importance of various remotely-sensible data attributes; and (5) to develop a community-level model of bird species composition and evaluate its efficacy in providing species-level inference. Collectively, my results help to reinforce (1) the importance of environmental heterogeneity in maintaining bird diversity within managed forests, (2) the many interdependencies among avian and floristic assemblages, and (3) the effectiveness of remote sensing platforms in quantifying ecological process and spatial pattern of forested landscapes. Many of the vegetation factors I examined, considered decisive in determining avian resources, were rarely independent. The strong individual and community-level responses of birds to forest compositional gradients can also be taken as a response to variation in structure among these different forest types, as vegetation structure and floristic composition were intimately related in most cases. But species-specific preferences for certain tree species and the integrative nature of plant composition data helped explain its strong predictive power for avian assemblage composition. Nonetheless, stand management aimed to manipulate structural complexity within and among forest stands will influence plant species composition in ways that meet the fine-scale resource needs of certain species. A diversity of growth stages can provide not only diverse vegetation structures, but also a diversity of tree species. The diversity and composition of understory shrubs should not be overlooked as well, as many forest birds utilize the understory strata for nesting and foraging. Foundational oak species (Quercus spp) will require special attention to ensure restoration of these species assemblages into the future. Policies should aim to maintain a diversity of stand types, in terms of structure, growth stage, and tree composition, to provide a range of habitat conditions for birds and other wildlife. Additionally, climate change is expected to impact forest communities in unpredictable ways. Full-stand taxonomic composition of forest stands of different type and structure should be frequently evaluated in the coming decades to ensure policies are successfully implemented to maintain environmental heterogeneity and desired levels of diversity. The methods I utilized show great potential for remotely inventorying the compositional patterns of forests at landscape extents. Restoration of important tree species can be greatly advanced by these methods to ensure viable species pools remain under shifting vegetation pressures.