Niche‐oriented species–abundance models: different approaches of their application to larval chironomid (Diptera) assemblages in a large river

Abstract

Summary Species–abundance patterns of larval chironomid assemblages in a large river were tested against seven niche‐oriented models, using data on abundances measured monthly at four sites along a cross‐section of the River Danube in Austria. Expected abundance patterns generated by computer simulations were tested against two categories of assemblages: all species found and the dominant species only. Within each category, the whole river cross‐section was compared with the four single sites, and three functional feeding groups were analysed separately. For all assemblages, species‐ and process‐oriented approaches concerning the correspondence between species identity and abundance ranking were used. When all 81 species found within the river section were included, none of the models agreed with the observed process‐oriented patterns, whereas the distributions of collector–gatherers, filterers and (at a single site) grazers fitted the Random Fraction model. This model indicates that the species divide the niche space in a random fashion, with no hierarchy of dominance. None of the models fitted the species‐rich assemblages using the species‐oriented approach. When only the most abundant species were considered, several of the models fitted the observed assemblages. The Random Assortment model provided best‐fits most frequently, suggesting that relative abundances of the species are completely independent. However, the Random Fraction model agreed with observed patterns of abundance in about 50% of cases when the Random Assortment model provided a good fit. Only 25% of all species‐oriented patterns of the most dominant species were described by the seven models, mostly by the Random Assortment model. However, species classed as filterers showed a tendency towards more even distributions according to the MacArthur Fraction model and the Dominance Decay model. Results suggest that the Random Fraction model corresponds best with species‐rich assemblages whereas this model combined with the Random Assortment model may provide a good explanation for patterns formed by the (relatively few) most dominant species. The species‐oriented approach might be a valuable additional form of analysis for uncovering tendencies in abundance patterns which could be overlooked when focusing solely on the process‐oriented approach that is usually used.