Nested patterns of community assembly in the colonisation of artificial canopy habitats by oribatid mites

An observed species-area relationship (SAR) in assemblages of oribatid mites inhabiting natural canopy habitats (suspended soils) led to an experimental investigation of how patch size, height in canopy and moisture influence the species richness, abundance and community composition of arboreal oribatid mites. Colonisation by oribatid mites on 90 artificial canopy habitats (ACHs) of three sizes placed at each of three heights on the trunks of ten western redcedar trees was recorded over a 1-year period. Fifty-nine oribatid mite species colonised the ACHs, and richness increased with the moisture content and size of the habitat patch. Oribatid mite species richness and abundance, and ACH moisture content decreased with increasing ACH height in the canopy. Patterns in the species richness and community composition of ACHs were non-random and demonstrated a significant nested pattern. Correlations of patch size, canopy height and moisture content with community nestedness suggest that species-specific environmental tolerances combined with the differential dispersal abilities of species contributed to the non-random patterns of composition in these habitats. In line with the prediction that niche-selection filters out species from the regional pool that cannot tolerate environmental harshness, moisture-stressed ACHs in the high canopy had lower community variability than ACHs in the lower canopy. Colonising source pools to ACHs were almost exclusively naturally-occurring canopy sources, but low levels of colonisation from the forest floor were apparent at low heights within the ACH system. We conclude that stochastic dispersal dynamics within the canopy are crucial to understanding oribatid mite community structure in suspended soils, but that the relative importance of stochastic dispersal assembly may be dependent on a strong deterministic element to the environmental tolerances of individual species which drives non-random patterns of community assembly.