Fire exclusion and fire return interval affect small mammal populations in longleaf pine forests

Longleaf pine (Pinus palustris) forests are among the most fire-dependent forest types in the world and once dominated much of the southeastern United States. Loss and fragmentation of longleaf pine shrank its distribution to approximately 4% of its historic range; as a result, longleaf pine forests currently receive much conservation and restoration emphasis. In all cases, this emphasis includes prescribed fire. Prescribed fires promote herbaceous understories which maintains habitat for some small mammals, but fires can also result in temporary declines of small mammal populations. Therefore, short return intervals may not allow small mammal populations to recover between fires, and long return intervals may result in declining habitat quality and reduced small mammal abundance. During the first phase of our research, we studied the long-term influence of prescribed fire on small mammals. We annually estimated small mammal abundance and species richness during 2001 - 2014 in eight longleaf pine-dominated plots in southwestern Georgia, USA. Plots were either biennially burned (reference) or had fire excluded during the study. Small mammal species richness was similar in fireexcluded and reference sites. Overall small mammal abundance was greatest in reference plots. Five species were captured with sufficient frequency to justify species-specific analyses. Of these, three (hispid cotton rats, Sigmodon hispidus; oldfield mice, Peromyscus polionons; and house mice, Mus musculus) were more abundant in reference plots. For the second phase of our research, we reintroduced fire in half of the fire excluded plots (burned during 2015, 2016, and 2017) and began sampling small mammals in three additional plots that were burned annually since 2003. We collected data during 2017 and 2018 for this phase of the research to evaluate small mammal response to four prescribed-fire return intervals (annual growing season burns since 2003, biennial burns since 2001 (reference), fire excluded since 2001, and fire excluded since 2001 and then burned during 2015, 2016, and 2017). Small mammal species richness was not affected by burn interval. Cotton rat abundance was greater in reference plots than in other treatments, whereas cotton mice (Peromyscus gossypinus) and oldfield mice were most abundant in the annually burned treatment. Three prescribed fires occurred in fire reintroduced plots prior to sampling, but these fires were insufficient to restore small mammal abundance to that of annually or biennially burned plots. Our work provides further evidence that fire is important for maintaining habitat for some small mammals and that time between fires has a species-specific effect on small mammal abundances. Thus, manipulating fire return interval is a promising tool for managing relative abundance of small mammals and likely other wildlife.