Although there is increasing interest in the effects of habitat disturbance on community attributes and the potential consequences for ecosystem functioning, objective approaches linking biodiversity loss to functional loss are uncommon. The objectives of this study were to implement simultaneous assessment of community attributes (richness, abundance and biomass, each calculated for total-beetle assemblages as well as small- and large-beetle assemblages) and three ecological functions of dung beetles (dung removal, soil perturbation and secondary seed dispersal), to compare the effects of habitat disturbance on both sets of response variables, and their relations. We studied dung beetle community attributes and functions in five land-use systems representing a disturbance gradient in the Brazilian Amazon: primary forest, secondary forest, agroforestry, agriculture and pasture. All response variables were affected negatively by the intensification of habitat disturbance regimes, but community attributes and ecological functions did not follow the same pattern of decline. A hierarchical partitioning analysis showed that, although all community attributes had a significant effect on the three ecological functions (except the abundance of small beetles on all three ecological functions and the biomass of small beetles on secondary dispersal of large seed mimics), species richness and abundance of large beetles were the community attributes with the highest explanatory value. Our results show the importance of measuring ecological function empirically instead of deducing it from community metrics.

1. The role of several factors that affect the composition of the dung beetle assemblages in an Amazonian rainforest was quantified, together with the effect of these factors on the role that dung beetles play as secondary seed dispersers. 2. A total of 61 dung beetle species was captured during 3360 h of trapping. During nocturnal trapping periods, more dung beetles, of larger mean size, and more species were captured per trap than during diurnal trapping periods. 3. During the rainy season, more dung beetle species were captured per trap than during the dry season, but the number of individuals and their mean size did not vary between seasons. 4. Bait size had a significant effect on the mean number of beetles and mean number of species but not on mean beetle size. As bait size increased from 5, 10, 25, to 50 g, more beetles and more species were captured per trap. 5. Between 6 and 73% of plastic beads, used as seed mimics, were buried by dung beetles at depths that ranged from 0.5 to 7 cm. Both the proportion of beads buried and burial depth decreased with increasing bead size, and increased with increasing amounts of dung surrounding each bead (5, 10, and 25 g). 6. The proportion of buried seeds for three species varying in size between 5 and 27 mm, increased with increasing dung beetle size, using beetles of seven sizes, varying between 10 and 25 mm. 7. Seeds surrounded by dung were buried more often and more deeply when placed on the forest floor during the late afternoon than when placed during the early morning. Seeds were also buried more often when placed on the forest floor during the rainy season than when placed during the dry season, but season had no effect on burial depth. 8. Forests in Central Amazonia hold a rich dung beetle community that plays an active role in secondary seed dispersal, and consequently in plant regeneration. The interaction between seeds and beetles is complex because it is affected by many factors.