Fish species‐environment and abundance relationships in a Great Plains river system

Abstract

Fish assemblages in the upper Red River system of southwestern Oklahoma (USA) were predictable along measured environmental gradients Conductivity was the most important variable predicting structure of fish assemblages followed by stream size, alkalinity woody debris and water clarity Classification of abundance data identified four groups each of species and sites Species groups were separated on a habitat template indicating similar environmental responses within groups However, site groups showed considerable overlap on the template Correlations among species environmental preferences were significantly associated with correlations of species abundances Likewise, site correlations on the basis of measured environmental variables and on the basis of species abundances were significantly similar We tested abundance and distribution data for agreement with the hierarchical model of Kolasa Several testable predictions of the model described our data well specialist species outnumbered generalist species and were less abundant on average, than generalist species Average abundance of species was highly correlated with their ecological ranges and species were clumped along both ecological range and abundance axes