Comparisons of Fish Species Traits From Small Streams to Large Rivers

Robert M. Goldstein¹ and Michael R. Meador²

¹U.S. Geological Survey, 196 Whitten Road, Augusta, Maine, USA

²U.S. Geological Survey, 12201 Sunrise Valley Drive, MS 413, Reston, Virginia, USA

ABSTRACT.-To examine relations between fish community function and stream size, we classified 429 lotic freshwater fish species based on multiple categories within each of six species traits: (1) substrate preference, (2) geomorphic preference, (3) trophic ecology, (4) locomotion morphology, (5) reproductive strategy, and (6) stream size preference. Stream size categories included small streams, small rivers, medium rivers, large rivers, and no size preference. Frequencies of each species trait category were determined for each stream size category based on life history information from the literature. Cluster analysis revealed the presence of co-varying groups of species trait categories. One cluster (RUN) included traits of planktivore and herbivore feeding ecology, migratory reproductive behavior and broadcast spawning, preferences for main channel habitats, and a lack of preferences for substrate types. Frequencies of classifications for the RUN cluster varied significantly across stream size categories (P = 0.009), and were greater for large rivers compared to small streams and rivers. Another cluster (RIFFLE) included traits of invertivore feeding ecology, simple nester reproductive behavior, preferences for riffles, and preferences for bedrock, boulder, and cobble/rubble substrates. No significant differences in the frequency of classifications among stream size categories were detected for the RIFFLE cluster (P = 0.328). Our results suggest that fish community function, based on trophic, reproductive, geomorphic, and substrate traits, may be structured by large-scale differences in habitat and is different for large rivers compared to small streams and rivers. Our findings support theoretical predictions of variation in species traits among stream reaches based on ecological frameworks such as landscape filters, habitat templates, and the river continuum concept. We believe that the species trait classifications presented provide an opportunity for further examination of fish species relations to physical, chemical, and biological factors in small streams to large rivers.