Resource-consumer diversity: testing the effects of leaf litter species diversity on stream macroinvertebrate communities

1. Understanding relationships between resource and consumer diversity is essential to predicting how changes in resource diversity might affect several trophic levels and overall ecosystem functioning. 2. We tested for the effects of leaf litter species diversity (i.e. litter mixing) on litter mass remaining and macroinvertebrate communities (taxon diversity, abundance and biomass) during breakdown in a detritus-based headwater stream (North Carolina, U.S.A.). We used full-factorial analyses of single- and mixed-species litter from dominant riparian tree species with distinct leaf chemistries [red maple (Acer rubrum), tulip poplar (Liriodendron tulipifera), chestnut oak (Quercus prinus) and rhododendron (Rhododendron maximum)] to test for additivity (single-species litter presence/absence effects) and non-additivity (emergent effects of litter species interactions). 3. Significant non-additive effects of litter mixing on litter mass remaining were explained by species composition, but not richness, and litter-mixing effects were variable throughout breakdown. Specifically, small differences in observed versus expected litter mass remaining were measured on day 14; whereas observed litter mass remaining in mixed-species leaf packs was significantly higher on day 70 and lower on day 118 than expected from data for single-species leaf packs. 4. Litter mixing had non-additive effects on macroinvertebrate community structure. The number of species in litter mixtures (two to four), but not litter species composition, was a significant predictor of the dominance of particular macroinvertebrates (i.e. indicator taxa) within mixed-species packs. 5. In addition, the presence/absence of high- (L. tulipifera) and low-quality (R. maximum) litter had additive effects on macroinvertebrate taxon richness, abundance and biomass. The presence of L. tulipifera litter had both positive (synergistic) and negative (antagonistic) effects on invertebrate taxon richness, that varied during breakdown but were not related to litter chemistry. In contrast, the presence/absence of L. tulipifera had a negative relationship with total macroinvertebrate biomass (due to low leaf mass remaining when L. tulipifera was present and higher condensed and hydrolysable tannins associated with leaf packs lacking L. tulipifera). Macroinvertebrate abundance was consistently lower when R. maximum was present, which was partially explained by litter chemistry [e.g., high concentrations of lignin, condensed tannins, hydrolysable tannins and total phenolics and high carbon to nutrient (N and P) ratios]. 6. The bottom-up effects of litter species diversity on stream macroinvertebrates and litter breakdown are different, which suggests that structural attributes of macroinvertebrate communities may only partially explain the effects of litter-mixing on organic matter processing in streams. In addition, stream macroinvertebrates colonising decomposing litter are influenced by resource diversity as well as resource availability. Broad-scale shifts in riparian tree species composition will alter litter inputs to streams, and our results suggest that changes in the diversity and availability of terrestrial litter may alter stream food webs and organic matter processing.