Species traits as generalized predictors of forest community response to human disturbance

During the past decade, substantial effort has been aimed at identifying a group of easily measured, widely applicable plant traits that could serve as a metric to predict temporal change in plant communities. Traits that transcend regional differences in species composition and ecological conditions through their consistent response to disturbance would give managers a simple tool for tracking ongoing and future forest change in response to human disturbance. Although a wide range of plant traits has been associated with human disturbance, consistent traits have not emerged in the literature. However, this may due to differences in methodology among studies. Previously collected data from two eastern deciduous forest floras of North America (Iowa and Massachusetts) allowed us to evaluate whether such traits emerged, while controlling for methodological differences. We created a plot × traits matrix for each site and ordinated them used using principal components analysis (PCA) to identify disturbance gradients. We then assessed how well the traits most strongly associated with the disturbance gradients corresponded. The data sets shared only 4 of the 12 traits associated with disturbance and 6 of 12 traits associated with undisturbed sites. We did not find a consistent association between dispersal limitation and undisturbed sites or high dispersal capacity associated with disturbed sites. However, in both data sets degree of habitat specialization was an important variable on both ends of the disturbance gradient. Habitat generalists were associated with disturbance and habitat specialists were associated with more pristine sites in both data sets. These results agree with the findings of a wide range of site-specific studies, and we therefore propose that this variable is a promising candidate trait to provide a signal of forest community response to human disturbance. Our results are should be particularly encouraging for managers because in many regions morphological trait data are not readily available, and compiling such data is a very time-intensive task and unlikely to be feasible for most managers to undertake. With a list of species and a published flora, upwards of several hundred species can quickly be coded for degree of habitat specialization and used to track the impact of current disturbance, to predict future impacts, and to target specific species for reintroduction or restoration.