Data from: Importance of biotic niches versus drift in a plant-inhabiting arthropod community depends on rarity and trophic group

Communities are mostly composed of rare species; yet, the factors that determine their patterns of occurrence remain obscure. Theory predicts that, in contrast with common species, the occurrence of rare species will be poorly correlated with environmental variables (niches) and more affected by stochasticity (ecological drift), but how this pattern varies across different trophic groups is still poorly understood. Here, we compared the ability of environmental variables (bottom-up biotic niches) to predict the occurrence of plant-dwelling arthropods across different abundance classes in the Cape Floristic Region of South Africa. We compared three trophic groups, including 104 herbivorous hemipteran, 171 parasitoid wasp and 84 spider species, totalling 4511 individuals in 48 quadrats. To quantify bottom-up biotic niches, we studied the influences of species composition of plants on hemipterans, and of plants and hemipterans on spiders and wasps. We compared the observed strength of the correlation between rare species and their niches with expectations that were generated by repeatedly rarefying abundant species. A large proportion of arthropod species were very rare, i.e. with only 1 or 2 individuals (49-55%). Although rarefying abundant species greatly decreased the correlation with bottom-up biotic niches, bottom-up biotic niches generally better predicted the occurrence of rarefied abundant species than very rare ones, suggesting a greater influence of drift on very rare arthropods. That is, (very) rare arthropods are distributed more randomly than rarefied abundant species. Nevertheless, trophic groups differed in the details of their response to bottom-up biotic niches. Plant species composition was a better predictor of rarefied abundant than truly rare hemipterans. In contrast, the importance of bottom-up biotic niches among abundance classes varied less visibly in spiders and wasps. Our study thus suggests that the importance of niches in structuring arthropod communities depends on species rarity and trophic group.HemipteranAbundances_Dec2014Number of individuals per quadrat for hemipteran morpho-species collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).SpiderAbundances_Dec2014Number of individuals per quadrat for spider morpho-species collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).WaspAbundances_Dec2014Number of individuals per quadrat for parasitoid wasp morpho-species collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).PlantCovers_m2_Dec2014Cover (in m2) per quadrat for plant morpho-species sampled at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).HemipteranTaxonomy_Dec2014Taxonomy of hemipterans collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).SpiderTaxonomy_Dec2014Taxonomy of spiders collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).WaspTaxonomy_Dec2014Taxonomy of parasitoid wasps collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).PlantTaxonomy_Dec2014Taxonomy of plants collected on the vegetation at Jonkershoek on December 2014 (six year old vegetation (post-fire age)).PlantTaxo_Dec2014.cs