Recent studies considering the contribution of biodiversity to ecosystem functioning have emphasised the functional importance of individual species and, in so doing, have rekindled the use of categorical descriptors that group species according to their relative contribution to ecosystem processes or functioning. Such functional effect groupings, however, tend to be based on specific traits or contributory roles that are assumed to adequately characterise the functional importance of a species, rather than being based on direct measures of ecosystem processes and functions. This decoupling of organism−environment interaction is difficult to reconcile and, when applied widely, distorts understanding of the mediating role that species play in natural ecosystems. In this study, we begin to address this problem by characterising the functional contributions of 7 benthic invertebrate species for 2 ecosystem processes (particle reworking and bioirrigation) linked to 4 ecosystem functions (changing concentrations of NH4-N, NOx-N, PO4-P and SiO2-Si) and use these data to derive functional effect groupings. We show that whilst it is possible to categorise species according to how they influence ecosystem properties, the membership and number of functional effect groups depends on which ecosystem property is considered. Furthermore, we demonstrate that categorisations based on functional effects are not synonymous with species taxonomy and that they cannot be applied generically even when considering closely linked biogeochemical processes. Collectively, our findings call for a rethink of how functional effect groups are defined and emphasise the need to interrogate presumed links between species and ecosystem properties across a range of biodiversity−environment contexts.