Premium
A Comparison of Fish‐based Classification Schemes for Reference Streams and Rivers in Nebraska
Author(s) -
Heatherly Thomas,
Bazata Ken,
Schumacher David,
Traylor Elbert
Publication year - 2014
Publication title -
journal of environmental quality
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.888
H-Index - 171
eISSN - 1537-2537
pISSN - 0047-2425
DOI - 10.2134/jeq2013.03.0102
Subject(s) - ordination , multidimensional scaling , streams , interpretability , river ecosystem , abundance (ecology) , multivariate statistics , ecology , relative species abundance , taxonomic rank , canonical correspondence analysis , environmental science , cluster (spacecraft) , ecosystem , statistics , mathematics , biology , computer science , machine learning , computer network , taxon , programming language
Proper assessments of lotic ecosystems depend on our ability to isolate natural differences from anthropogenic disturbances. We examined Nebraska river and stream classification strength, based on fish species, using multiple response permutation procedures for common classification strategies: ecoregions, watersheds, hydrologic‐landscape regions, and assemblage structure from cluster analyses. Next, we tested the ecological interpretability of classification schemes using nonmetric multidimensional scaling ordinations and ANOVAs. Finally, we used nonparametric ANOVA to identify environmental predictors of overall fish assemblage structure. Hydrologic‐landscape regions had the highest classification strength, but cluster groups had the most ecological interpretability based on the discreteness of the groups in ordination space and on the large number of common species that had different abundances across cluster groups. In addition, presence/absence data provided groups with more classification strength and interpretability than abundance data. Temperature, stream size, total phosphorus concentrations, and the percentage of fine substrates were significantly correlated to nonmetric multidimensional scaling ordinations and to overall fish structure in the multivariate ANOVA models. Cluster analyses using presence/absence were therefore the best classification scheme, and we identified the environmental variables that are likely to be useful for determining whether streams should have similar biotic assemblages. This information will be a valuable guide for separating natural variability in biotic assemblages from anthropogenic influences.