Summary A large body of research has revealed (often) positive biodiversity–ecosystem functioning (B–EF) relationships in manipulative experiments. The vast majority of such studies have focused on either micro- or macro-organisms, and none we are aware of have manipulated the diversity of both simultaneously under controlled laboratory conditions. We performed a microcosm experiment in which we manipulated species richness of aquatic fungi and invertebrates, two taxonomically distant sets of consumers that contribute to the same key ecosystem process in freshwaters, the decomposition of terrestrial leaf litter. We used a novel statistical design to maximize parsimony and analytical power in an experiment with three levels of species richness (seven mono-culture, 21 di-culture, and seven tri-culture treatments). Litter decomposition was measured as both leaf mass loss and the production of fine particulate organic matter (FPOM). We tested whether species richness affected these two processes or whether polycultures performed as predicted from their component mono-cultures. Further, we calculated assemblage metabolism in each microcosm to test whether the processes were driven by the metabolic demands of fungi and invertebrates. In general, across the 35 treatments, most species combinations performed in an additive fashion and we found no effect of species richness on either process. There was evidence of assemblage identity effects (i.e. certain species combinations not performing as expected), with instances of significant differences for species combinations that contained both caddis larvae and fungi. These assemblages performed worse than expected, which might have been due to dual vertical and horizontal interactions, with the possibility that although both consumed litter directly the former may also have grazed on the latter. Apart from these particular species combinations, overall performance of a species in polyculture was effectively the same as in mono-culture and reflected its metabolic demands. This suggests that even taxonomically distant consumers might exhibit a degree of functional redundancy for certain processes provided the remaining species can attain sufficient population biomass (and hence metabolic capacity) to compensate for the loss of other species, although whether such compensatory mechanisms operate in the field remains unknown. Further species contribute to a multitude of ecosystem processes and progressively more species are needed to sustain the sum of them. Our experiment highlights how, by taking metabolic demands into account, future B–EF studies could help to disentangle how species contribute to ecosystem processes both separately and in combination, and to help partition the effects of taxonomic and functional diversity.