Disturbance increases functional diversity but decreases phylogenetic diversity of an arboreal tropical ant community

Tropical rainforest canopies host a highly diverse arthropod fauna, which contribute to ecosystem function through their functional (FD) and phylogenetic diversity (PD). While a lot of previous research has documented the severe negative impacts of disturbance on the FD and PD of ground invertebrate communities, our understanding of arboreal counterparts is limited. Here, we studied the effects of forest disturbance on an ecologically important invertebrate group, the ants, in a lowland rainforest in New Guinea. We exhaustively sampled 4000 m2 area of a primary and a secondary forest for canopy ants. We report > 2800 occurrences of 128 ant species in 852 trees, one of the most comprehensive arboreal collections to date. To test how ant PD and FD differ between the two forests, we constructed the ant species-level community phylogeny and measured 10 functional traits. Furthermore, we assessed by data exclusion the influence of species which were not nesting in individual trees (visitors) or only nesting (nesters), and of non-native species on FD and PD values. We expected that disturbance would decrease FD and PD in tree dwelling ants. We hypothesized that traits in primary forests would be more overdispersed due to the greater availability of ecological niches, while secondary forests would have stronger trait clustering due to a a stronger habitat filtering caused by more extreme microclimate. Primary forests had higher species richness and PD than secondary forest. Surprisingly, we found higher FD in secondary forest. This pattern was robust even if we decoupled functional and phylogenetic signals or if non-native ant species were excluded from the data. Visitors did not contribute strongly to FD, but they increased PD. Community trait means further corroborate the functional distinctiveness of arboreal ants among secondary and primary forest, with almost all traits being impacted by disturbance and forest succession. We conclude that the most plausible explanation is increased competition among closely related ant species in the secondary forest, which drives trait divergence. In the primary forest, abiotic habitat filters leads to more similar morphology and thus lower FD of phylogenetically more diverse ant assemblages.