Effects of prescribed fire on breeding forest songbird abundances in central and Southern Illinois

The proliferation of shade-tolerant trees such as maple (Acer spp.) in Midwestern forests is a challenge that forest managers must overcome to retain forests in their oak- and hickory-dominated status. Prescribed fire is one of the tools that forest managers use to meet this challenge. While the forest managers monitor the effects of prescribed fire on the structure and species composition of the trees, it is imperative to also understand how these burns impact wildlife that reside in the forest, including breeding bird species. Monitoring the effects of prescribed fire helps to inform managers’ decisions regarding how often and where to perform prescribed burns. Previous studies have shown that responses of birds to different types of forest management can vary substantially among species, often as a function of their preferences for particular nesting substrates and their foraging behaviors. I used 5 years of point count data collected in historically oak- and hickory-dominated forests in central and southern Illinois to evaluate the influence of prescribed fire (i.e., burn frequency, time-since-fire) on breeding bird abundances and probabilities of occurrence. I also included landscape variables (i.e., proportion forest cover, distance to forest edge) in analyses, as well as data from points where no prescribed fire had occurred. I chose to focus on 10 relatively common species that represented a broad range of nesting and foraging guilds and that were species of interest for conservation efforts. I found that management and landscape variables had additive and interactive effects on breeding birds, though the effects varied depending on the species. Of the ten species analyzed, three were common enough to use abundance values (i.e., actual numbers present per point), whereas the probability of occurring (i.e., presence/absence) was used for the remainder. A prescribed fire treatment effect was present in the top model for nine of the ten species, with burn frequency, rather than time-since-fire, being in the majority of top models. Landscape variables appeared in nine of the ten top models with proportion of forest cover appearing in most models rather than distance to edge. Species such as Acadian flycatcher (Empidonax virescens) and wood thrush (Hylocichla mustelina) had increased abundances in response to frequent fire (more than 2 prescribed fire events in the previous 5 years) and had greater abundances farther from forest edge. Both landscape variables were in the top model for Kentucky warblers (Geothlypis formosa), which were more likely to occur in areas farther from forest edge and in areas that had a moderate (i.e., 33-65%) amount of forest cover within 2 km. Red-eyed vireos (Vireo olivaceus) were more likely to occur in areas of non-managed forest compared to forests that were frequently or infrequently burned. These results suggest that prescribed fire, particularly burn frequency, can increase or decrease breeding bird abundances depending on the species being considered because of the measured response of each guild to the prescribed fire performed in these sites. My results also highlight that landscape variables can add to or interact with effects of prescribed fire and should be included whenever possible when evaluating the effects of forest management on breeding bird abundances. Finally, forest managers can use the information documented here to make more informed decisions about prescribed fire; creating a mosaic of forest units that are burned at various frequencies may allow forest managers to modify the tree composition while providing benefits to the bird species that breed and live there