Ecological Traits Fail to Consistently Predict Moth Species Persistance in Managed Forest Stands

Species traits have been used as predictors of species extinction and colonization probabilities in fragmented landscapes. Thus far, trait-based analytical frameworks have been less commonly employed as predictive tools for species persistence following a disturbance. I tested whether life history traits, dietary traits, and functional traits were correlated with moth species persistence probabilities in forest stands subjected to varying levels of timber harvest. Three harvest treatments were used: control stands (unharvested since 1960), shelterwood cut stands (15% canopy removed), and patch cut stands (80% standing bole removed). Logistic regression models were built to assess whether species persistence probabilities were a function of species traits; separate models were constructed for each level of timber harvest treatment. Species persistence probabilities were mainly a function of pre-harvest abundances. Species traits had idiosyncratic effects on species persistence depending on the level of timber harvest employed. These results suggest that species traits may indirectly influence how moth species assemblages change as a result of forest management by determining pre-harvest abundance rather than persistence per se. The absence of significant trait effects on persistence probabilities may also reflect prior reduction in species trait space. That is, the range of species trait combinations sampled in this study was much lower than observed in historically unlogged eastern deciduous forest systems. Thus, the lack of significant trait-persistence correlations observed here might indicate historic extinctions of species from prior logging events that have not been offset by post-harvest recovery of original species assemblages