PhD Dissertation, Joseph LaManna (2009-2015)

Vegetation provides the habitat that underlies animal distributions. Yet mechanisms by which changes in vegetation affect animal fitness, distributions, and communities remain unclear. For example, animal richness and species composition often change with decreased forest structural complexity associated with anthropogenic disturbance, but differences in latitude and vegetation effects on reproductive success may influence species responses to vegetation changes. A global meta-analysis of logging effects on bird communities revealed substantial species loss in tropical but not temperate forests. This suggested greater habitat specialization across tropical birds compared to temperate relatives. Furthermore, changes in species composition in north-temperate forests were greater where logging impacted nest success more.
This meta-analysis suggested that changes in reproductive success can influence the response of animal species to vegetation change. Therefore, habitat use and nest predation rates were examined for 16 bird species breeding along a deciduous to coniferous vegetation gradient and with experimental conifer removal from aspen stands. For most bird species, decreasing abundance was associated with increasing predation risk along both natural and experimentally modified vegetation gradients. This landscape-scale approach strongly supports the idea that vegetation-mediated effects of predation are associated with animal distributions and species turnover.

While direct predation mortality clearly had effects on bird distributions, the risk of predation alone may have equally large effects on reproduction and, ultimately, fitness. Yet the severity and generality of such demographic ‘costs of fear’ is unknown across species. We tested phenotypic responses to risk and associated demographic costs for 10 songbird species breeding along natural nest predation gradients and by experimentally increasing risk for four species. Parents decreased offspring development periods and feeding rates to reduce time-dependent nest mortality with natural and experimental increases in risk. Reproductive output from nests that escaped predation generally declined along risk gradients, but the severity of this cost varied across species. Ultimately, demographic costs of fear strongly reduced fitness across bird species, but not as much as direct predation mortality. These landscape and experimental tests suggest that vegetation affects direct predation mortality and predation risk, and that both these demographic costs strongly influence avian behaviors, fitness, distributions, and communities.