The semi-aquatic habitats of terrestial Coleoptera in a lowland river floodplain

Abstract

281 species of terrestrial ground-living beetles were recorded from 69 riparian and wetland sites in the floodplain of the lowland River Soar, England. Differences in species composition between pitfall trapped and timed hand-collected samples were smaller than those attributable to environmental and seasonal factors. Detrended Correspondence Analysis consistently ranked all sites against seasonal variations between April and June and floodplain sites against annual variations. DCA axis 1 scores were slightly better correlated with important environmental variables at the ecohabitat (<5Om) scale rather than the microhabitat scale. Canonical Correspondence Analysis detected assemblage responses to flooding disturbance and grazing pressure along the main channel as well as to water level stability in the floodplain. A conceptual model of floodplain land-use and river management postulated a dynamic equilibrium between flooding disturbances and vegetational succession, producing geomorphic and vegetational structures which serve as semi-aquatic habitats for terrestrial beetle assemblages with appropriate species traits. Impoundment for navigation affects assemblages by modifying the severity of flooding disturbance. The effects of grazing pressure resemble flooding disturbance. The short-term (<5yr) impact of bank regrading was explained by differences in severity, predictability and frequency compared to the beetles' generation length. Evenness and species richness were affected only by flooding and grazing disturbance. This response was not predicted by the intermediate disturbance hypothesis because the frequencies of flooding and grazing disturbances in the Soar valley are not appropriate to the hypothesis, which more closely relates to disturbance by bank regrading. In comparison to diversity indices, a rarity index was much less sensitive to environmental factors than species diversity indices and more robust against seasonal and yearly fluctuations. Consequently, it has more potential for use in site quality assessment.