Summary 1.Much of the literature on the relationship between species richness or functional group richness and measures of ecosystem function focuses on a restricted set of ecosystem function measures and taxonomic groups. Few such studies have been carried out under realistic levels of diversity in the field, particularly in high diversity ecosystems such as tropical forests. 2. We used exclusion experiments to study the effects of dung beetle functional group richness and composition on two interlinked and functionally important ecological processes, dung removal and secondary seed dispersal, in evergreen tropical forest in Sabah, Malaysian Borneo. 3. Overall, both dung and seed removal increased with dung beetle functional group richness. However, levels of ecosystem functioning were idiosyncratic depending on the identity of the functional groups present, indicating an important role for functional group composition. There was no evidence for interference or competition among functional groups. 4. We found strong evidence for overyielding and transgressive overyielding, suggesting complementarity or facilitation among functional groups. Not all mixtures showed transgressive overyielding, so that complementarity was restricted to particular functional group combinations. 5. Beetles in a single functional group (large nocturnal tunnellers) had a disproportionate influence on measures of ecosystem function: in their absence dung removal is reduced by approximately 75%. However, a full complement of functional groups is required to maximize ecosystem functioning. 6. This study highlights the importance of both functional group identity and species composition in determining the ecosystem consequences of extinctions or altered patterns in the relative abundance of species.

No abstract

Summary1. Many studies document high levels of functional redundancy in ecosystems, suggesting that species extinctions will not be detrimental to ecosystem functions and services. However, apparently redundant taxa may prove critical for sustaining ecosystem functions and services in the context of environmental perturbations. 2. Dung beetles (Coleoptera:Scarabaeidae) provide a valuable ecosystem service in temperate agro-ecosystems by increasing rates of dung decomposition and nutrient cycling. However, there is concern that these services may be negatively affected by changes in species richness and composition due to changes in pasture management and negative effects of anthelmintics used to control livestock parasites. 3. We used a mesocosm experiment to investigate the functional importance of dung beetle species richness in a system perturbed by the anthelmintic, ivermectin. We varied dung beetle species richness within three functional groups in factorial combination with ivermectin treatment. 4. In the short term (1-4 weeks), multi-species dung beetle assemblages achieved higher decomposition rates than monocultures, but only in ivermectin-treated dung. Varying speciesspecific sensitivities to ivermectin meant that species-rich assemblages sustained ecosystem functioning in the context of this anthropogenic perturbation. 5. Over the longer term (36 weeks), there was a significant, positive effect of species richness on dung decomposition in both ivermectin-treated and untreated dung, underlining the functional importance of maintaining a species-rich dung processing community even in the absence of perturbations to the system. 6. Synthesis and applications. The interacting effects of dung beetle species richness and ivermectin highlight the importance of maintaining diverse assemblages in the face of anthropogenic perturbations and suggest that apparent functional redundancy of species in agroecosystems should be interpreted cautiously. Furthermore, different farm management practices (e.g. pesticide use and fragmentation of habitats) may have consequences for ecosystem functions and services that exceed the effects of each when considered in isolation.

Agricultural expansion and intensification are major threats to global biodiversity, ecological functions, and ecosystem services. The rapid expansion of oil palm in forested tropical landscapes is of particular concern given their high biodiversity. Identifying management approaches that maintain native species and associated ecological processes within oil palm plantations is therefore a priority. Riparian reserves are strips of forest retained alongside rivers in cultivated areas, primarily for their positive hydrological impact. However, they can also support a range of forest-dependent species or ecosystem services. We surveyed communities of dung beetles and measured dung removal activity in an oil palm-dominated landscape in Sabah, Malaysian Borneo. The species richness, diversity, and functional group richness of dung beetles in riparian reserves were significantly higher than in oil palm, but lower than in adjacent logged forests. The community composition of the riparian reserves was more similar to logged forest than oil palm. Despite the pronounced differences in biodiversity, we did not find significant differences in dung removal rates among land uses. We also found no evidence that riparian reserves enhance dung removal rates within surrounding oil palm. These results contrast previous studies showing positive relationships between dung beetle species richness and dung removal in tropical forests. We found weak but significant positive relationships between riparian reserve width and dung beetle diversity, and between reserve vegetation complexity and dung beetle abundance, suggesting that these features may increase the conservation value of riparian reserves. Synthesis and applications: The similarity between riparian reserves and logged forest demonstrates that retaining riparian reserves increases biodiversity within oil palm landscapes. However, the lack of correlation between dung beetle community characteristics and dung removal highlights the need for further research into spatial variation in biodiversity–ecosystem function relationships and how the results of such studies are affected by methodological choices.