Taxonomic resolution and functional traits in the analysis of tropical oribatid mite assemblages.

We analysed species-level datasets representing Oribatida assemblages along a gradient of old-growth primary tropical forests, secondary forests, and plantation forests in Dong Nai Biosphere Reserve, Vietnam. We identified patterns in abundance, species richness and species assemblages of Oribatida, then applied taxonomic sufficiency approach to the datasets. Using three levels of higher-taxon aggregation, we evaluated whether aggregated datasets are useful in identifying ecological patterns, in comparison to species-level data. Species-level data on Oribatida assemblages clearly separated plantation forests from other forest environments; there was no significant separation between primary and secondary forests. Geographical structuring of species-level assemblages was significant, separating sites from two regions of the reserve. There was a significant concordance between multivariate ordination plots produced for species-level and aggregated (families, suborders/superfamilies) datasets, with Oribatida assemblages of plantation forests consistently separated from two other forest types. Mycobatidae (at family level) and Ceratozetoidea (at suborder/superfamily level) were indicators of plantation forests. The coarsest taxonomic resolution dataset with only four aggregated groups produced no separation of Oribatida assemblages by forest type or region. Moderate level of taxonomic aggregation applied to Oribatida community data did not cause great differences in patterns revealed by multivariate analysis, and therefore could be a valid approach to analysing the structure of tropical Oribatida assemblages. The taxonomic level of suborders and Brachypylina superfamilies appears to be the best compromise for ecological information and ease of identification. Two traits-body size and reproductive mode-were recorded for collected Oribatida species. Community-weighted mean trait value, modified Mason's index of functional divergence, and Rao's index of functional diversity were calculated for each trait in each of the sampled Oribatida assemblages. Sexual reproduction was a dominant reproductive mode in soil Oribatida and did not vary across forest types, indicating similar levels of resource limitation for this trait. For body size, lower functional divergence in plantation forests suggests less scope for niche differentiation and higher competition among different body sizes in this forest type. Use of functional traits can enhance and complement the analysis of Oribatida communities, but more data are needed on feeding- and diet-related traits in tropical Oribatida.