Whereas litter chemistry, climate, and decomposers have been recognized as controls of litter decomposition, our understanding of the role of soil microbial community composition in linking decomposition rate to litter chemistry is limited.  Variation in litter chemistry measures that are highly correlated with decomposition rate (C:N, lignin:N, defensive compounds) are likely to select for particular microbial functional groups.  As chemical composition of litter changes over the course of decomposition, a succession of microbial functional groups should occur as litter is transformed from a labile to more recalcitrant substrate.  We would predict, then, that the relative proportion these microbial functional groups vary among plant species in relation to litter chemistry and the microbial community should experience a predictable succession.  We created litter bags for a total of eleven plant species that differ widely in C:N, lignin:N, carbon fractions, and leaf toughness and morphology.   To control for potential colonizing microbial community, we placed the litter bags in a “common garden” in a mixed-hardwood forest in north-central Florida.  We collected the litter bags at 28, 58, and 238 days and subjected them to litter chemistry and phospholipid fatty acid (PLFA) analysis for determination of relative abundance of functional groups.

Results/Conclusions

In principle components analysis of PLFA profiles of litter microbial communities, communities clustered according to litter quality and collection date.   Fungal:bacterial biomass increased with decreasing litter quality (high C:N and lignin:N) both among species at a given stage of decomposition and through time as litter quality decreased as decomposition proceeded.  Microbial communities from poor quality litters tended to change less over decomposition than those from high quality litters (time*species interaction, p < 0.0001).  The composition of microbial communities was highly correlated with a leaf chemistry axis (an ordination of litter chemistry variables) (r² = 0.65, p < 0.001).  The importance of the links between microbial composition and function was highlighted as decomposition rate (k) was better predicted  by microbial community composition(r² = 0.76, p < 0.0001) than any measure of litter chemistry.  These results suggest litter quality does drive microbial community composition, but that their function in decomposition is an emergent property that cannot be predicted solely upon litter quality measures alone.