Multi-species interactions in Northeastern forest ecosystems: Decomposition dynamics in red oak and sugar maple leaf mixtures.

摘要

In forested ecosystems, decomposition of leaf litter is the primary pathway for nutrient return to the soil. Though leaves from many species intermix on the forest floor, only recently have potential interactions during decomposition among leaves of different species been considered. Here I review literature on decay dynamics in litter mixtures, present data from two novel experiments examining decay of litters from co-occurring tree species red oak and sugar maple, and develop a new Bayesian modeling structure appropriate for analysis of data from mixed-species litter decomposition experiments.; Published papers that directly examine decay in leaf mixtures as well as in all component species decaying alone, show that decomposition patterns are not always predictable from single-species dynamics (Ch. 1). Several mechanisms have been proposed for why litters might interact during decomposition, including chemical transfer, changes in microclimate, or decomposer associations. To begin to elucidate the relative importance of these mechanisms, I examine decay in mixtures of maple and oak leaves using reciprocal transplants of litterbags at four sites in northwestern Connecticut (Ch. 2). Overall mass loss and nitrogen accumulation observed in mixtures were less than amounts predicted from observed dynamics in single-species litterbags. Deviations from predicted in the mixed litter were significantly influenced by leaf origin but not strongly by site of decay, suggesting that interactions among litters are more influenced by physical and chemical characteristics of the litter itself rather than characteristics of the surroundings.; Mixing of leaf litters is also influenced by the timing of leaf drop, and the resulting leaf layering on the forest floor could also influence nutrient cycling in ecosystems. The potential effect of leaf stratification was examined in compartmentalized litterbags and microcosms (Ch 3). Within each species, we found significantly greater mass loss and N accumulation in upper litter layers, and these results suggest that leaf mixture experiments could be influenced by a hidden layer effect if there are litters in the mixture that are particularly sensitive to depth.; Finally, I detail a Bayesian model appropriate for analyzing these experiments (Ch 4), providing a strikingly different and complementary method for analyzing and interpreting litter-mix results.