Sandy soils in forests of the upper Great Lakes region store ~80 Mg C ha^-1 and function together with forest vegetation as part of the Northern Hemisphere carbon sink. Our study is focused on the effects of exotic earthworm communities on carbon cycling in sandy soils of northern temperate forests and on whether they alter the amount, distribution, and stability of carbon as has been shown for finer textured forest soils. We measured earthworm species diversity, population density, and stable isotope contents (¹³C and ¹⁵N) across an aspen-dominated forest at the University of Michigan Biological Station in northern lower Michigan.  Surveys were conducted using an electro-shocking technique in a 33 hectare treated area of the FASET (Forest Accelerated Succession ExperimenT) manipulation, where aspen and birch trees were girdled in May 2008, and in a reference (untreated) stand.  We characterized spatial and seasonal patterns of earthworm abundance and community diversity using counts of individuals in electro-shocked plots, and differences in resource use using stable isotope abundances of earthworm species. To investigate possible community-specific effects on soil carbon dynamics, a mesocosm experiment was designed using eight earthworm community treatments. Rates of soil CO₂ efflux were measured and the effects of earthworm communities relative to reference mesocosms (containing no earthworms) were quantified.

Results/Conclusions

Earthworm communities were dominated by five species of European-origin: Dendrobaena octaedra, Apporectodea trapezoides, A. caliginosa, Lumbricus rubellus, and L. terrestris. Earthworm densities range from <20 individuals m^-2 to >150 individuals m^-2. Both densities and relative abundances of species and developmental classes varied seasonally. The relationship between earthworm densities and soil moisture was strongest in summer, followed by fall.   No relationship was found in the spring when juvenile densities are highest and soils are moist. The isotopic signatures varied according to species ecological groups and spatially across plots. Ecological groups were separated by ¹⁵N signatures, while ¹³C signatures separated earthworm communities across horizontal space.  While some earthworm communities increase rates of soil CO₂ efflux (e.g. L. rubellus/A. caliginosa community increased CO₂ efflux by up to 31%), others decreased the rate of soil CO₂ efflux (e.g. A. caliginosa/L. terrestris community decreased CO₂ efflux by up to 20%). These data provide insight into the structure of earthworm communities in sandy temperate forest soils. Further, preliminary data suggest that understanding relationships between community composition and resource loading may enhance our ability to predict future impacts on soil carbon dynamics in northern temperate forests.