Biodiversity and plant litter: experimental evidence which does not support the view that enhanced species richness improves ecosystem function

There has been a rapidly increasing recent interest in the effects of biological diversity on ecosystem properties, and while some studies have recently concluded that biodiversity improves ecosystem function, these views are based almost entirely on experiments in which species richness of live plants has been varied over all the species diversity treatments. However, most net ecosystem primary productivity eventually enters the decomposition subsystem as plant litter where it has important "afterlife effects". We conducted a field experiment in which litter from 32 plant species (i.e. eight species of each of four plant "functional groups" with contrasting litter quality) was collected and placed into litter-bags so that each litter-bag contained between one and eight species; the species which were included in the multiple (≥2) species litter-bags were randomly selected. This litter diversity gradient was created within each functional group and across some functional groups. We found large non-additive effects of mixing litter from different species on litter decomposition rates, litter nitrogen contents, rates of nitrogen release from litter and the active microbial biomass present on the litter. The patterns and directions of these non-additive effects were dependent upon both plant functional group and time of harvest, and these effects could be predicted in some instances by the initial litter nitrogen content and the degree of variability of nitrogen content of the component species in the litter-bag. There was no relationship between litter-bag species richness and any of the response variables that we considered, at least between two and eight species. Within plant functional groups our results provide some support for the species redundancy and idiosyncratic hypotheses about how biodiversity alters ecosystem function, but no support for the ecosystem rivet hypothesis or the view that species richness of plant litter is important for ecosystem function. We suggest that increased species diversity of plant litter is less important than that of live plants for determining ecosystem properties (and provide possible reasons for this) and conclude that perceived relationships between biodiversity and ecosystem function may be of diminished significance when the ecological importance of plant litter is fully appreciated.