Soil mite biodiversity: Its relationship to grass species and influence on decomposition in the Konza tallgrass prairie

Species are disappearing faster than we can record them and before we determine their role in ecosystems. Of terrestrial systems, we are least certain about the diversity and function of soil organisms. Hypotheses that soil mite species diversity is greater beneath grasses in dicultures compared to monocultures, beneath grasses of higher resource quality compared to lower resource quality, and beneath heterogeneous mixes of grass resources compared to homogeneous mixes were tested using natural occurrences of grass species at the Konza Prairie Biological Station (KPBS). Increased grass richness supported a more species and phylogenetically rich mite fauna at depth but not in the upper soil horizon. Soil mite richness increased non-linearly with grass species richness. The proportion of mite size classes in dicultures was considerably different than those for monocultures. There was no difference in soil mite richness between grass combinations of differing resource quality or heterogeneity.; Mites sampled beneath six native and one alien-invasive species of grass were similarly abundant, species rich, diverse, and taxonomically distinct. There was no evidence that mite community composition was specific to grass species or that a significant number of mite species had affinities for different grass species. The mite community was weakly related to soil environmental conditions. Only oribatid mites were related, marginally, to the species of grass present. An alien invasive grass species had no influence on mite community structure.; Rates of cotton strip decomposition, and mite abundance and species richness were measured at high and low fire frequency sites of the KPBS. Likelihood-based and information theoretic approaches were used to examine strength of evidence in data for models of decomposition representing the Null, Rivet and Redundant hypotheses of biodiversity and ecosystem function. The Null model including temperature, moisture and saturating effects in the total abundance of predatory mites (Mesostigmata) had more support in the data than any other models. A significant trend was observed in the models' residuals from low fire frequency sites; trends not observed in high fire frequency sites. Implications of these findings for predicting soil mite diversity and its role in ecosystem functioning are discussed.