Influence of long-term dormant-season burning and fire exclusion on ground-dwelling arthropod populations in longleaf pine flatwoods ecosystems

Abstract

Frequent dormant-season prescribed burns were applied at 1-, 2- and 4-year intervals to longleaf pine stands, Pinus palustris, for over 40 years on the Osceola National Forest in Baker County, Florida. Control plots were unburned for the same period of time. Pitfall traps were operated from November 1994 to October 1999 to measure the short- and long-term effects of prescribed burning frequency on the relative abundance and diversity of ground-dwelling macroarthropods. We also measured dead and live plant biomass to determine how long-term frequent fires affected the structure of the forest floor. The average total dead plus live plant biomass was significantly higher on plots where fire had been excluded. Annual and biennial burning resulted in about the same amount of total plant biomass (dead and live plant material combined) which was significantly less than the quadrennially burned plots. Shannon diversity ( H′) and evenness of ground-dwelling arthropods were reduced by burning. Annually burned plots had the lowest diversity and evenness while biennially and quadrennially burned plots also were significantly lower than unburned control plots. Dormant-season burning did not increase the number of rare genera regardless of frequency. Percent similarity of arthropod communities was highest for comparisons between plots that had been burned (60–68%) and lowest for the comparison of annually burned plots to unburned controls (37%). Examination of diversity and similarity through time showed that changes were due to short-term effects caused by the application of fire and not long-term changes in the ground-dwelling arthropod community. Burning significantly reduced the numbers of predators regardless of fire frequency and resulted in an increased number of detritivores. A total of 42 genera were reduced by prescribed burning; 32 genera were captured in greater numbers on annually burned plots, and 11 genera had higher numbers in one or both of the intermediate burn frequencies (biennial or quadrennial). Twenty-six genera were captured in equal numbers on quadrennially and annually burned plots, but in significantly lower numbers than on unburned plots, demonstrating that 4 years was insufficient time for their populations to recover from mild dormant-season fires. Arthropod response to burning appeared to be species specific so attempts to generalize how arthropods will respond based on a few species or groups should be avoided. The slow recovery rate of so many species suggests that management oriented toward conservation of biodiversity in longleaf pine flatwoods should include areas of fire exclusion.