Emergent Rarity Properties in Carabid Communities From Chinese Steppes With Different Climatic Conditions

Species abundance distributions (SADs) are increasingly used to investigate how species community structure changes in response environmental variations. SAD models depict the relative of recorded a and express fundamental aspects structure, namely patterns commonness rarity. However, influence differences conditions on characteristics is still poorly understood. In this study we carabid beetles (Coleoptera: Carabidae) three grassland ecosystems (desert, typical, meadow steppes) China. These characterized by different aridity conditions, thus offering an opportunity SADs influenced (mainly vegetation cover, hence productivity). We various models, including meta-community zero sum multinomial (mZSM), lognormal (PLN) Fisher’s logseries (LS), uni- multimodal gambin models. Analyses were done at level steppe type (coarse scale) for sectors within same (fine scale). found that mZSM model provided, general, best fit both analysis scales. Model parameters scale analysis. Moreover, LS was desert SAD. If abundances rarefied smallest sample, results similar those without rarefaction, but estimates become more evident. Gambin unimodal provided with lowest ?-value observed higher values typical steppes, which strongly affected use rarefaction. Our indicate all investigated communities adequately modeled two distributions, LS, scales analyses. This indicates (1) relatively small number species, most rare, (2) meta-communities large maintain basic shape local communities. The produced exaggerated ?-values, they overfit simple Overall, ?, ?, ?-values substantially lower implies decreasing harshness (aridity) from steppes.