Impact of environmental variables on the occurrence of dolichopodid flies in marshland habitats in Belgium (Diptera: Dolichopodidae)

Publisher Website
Google Scholar
Abstract
Distribution of dolichopodid flies in relation to environmental variables was investigated in 11 marshland habitats. These sites were situated in marshlands along six different creeks in northern Belgium. Dolichopodidae were sampled by means of seven white-water trap units per site from February to October 1990. Seven site and three water-related variables were recorded on one occasion during June. In all 8964 specimens representing 73 species were obtained. The main analysis was based on the 30 most abundant species, which accounted for > 97% of all specimens gathered. Multivariate techniques such as detrended correspondence analysis (DCA), canonical correspondence analysis (CCA) and two-way indicator species analysis (TWINSPAN) were applied to examine the impact of environmental variables on the occurrence of the species in the sampling sites. In a first approach the saltmarsh site was separated distinctly from the other marshlands. After exclusion of this site analysis of the remaining data by DCA and CCA revealed that most of the variation within the species data matrix could be explained by sets of environmental variables. In this regard, light intensity, little layer thickness and soil humidity proved to be the most important, whereas aerial humidity and vegetation height contributed to a much lesser extent. With all sites considered, some species proved to be abundant only along creeks with a chloride content > 1059 ppm. Dolichopodid communities in sites with a similar vegetation structure and composition differed more than those from sites belonging to the same marshland area. True reedmarsh-inhabiting species appeared to inhabit mainly rather dense and moderately humid reedmarsh habitats.