Spatial Variation of Headwater Fish Assemblages Explained by Hydrologic Variability and Upstream Effects of Impoundment
Hydrologic variability and connectivity influence stream-fish assemblage structure, but other correlated habitat variables often confound attempts to document such relationships. By their nature, dams decrease stream connectivity, but upstream influences on tributary fish assemblages are infrequently documented and poorly understood. We sampled stream fishes seasonally for one year to compare headwater assemblages of the unimpounded East Fork (nine sites) and the impounded West Fork (12 sites) of the San Jacinto River in Texas. Using multiple regression and canonical ordination, we partitioned variation in species' distributions into that explained by instream structural, hydrologic, and physicochemical factors, drainage features, and season. In addition to effects of temporal variation of connectivity (as related to base flow and season) and other environmental factors, spatial patterns of species distribution between tributaries in the East and West Fork indicated upstream effects of impoundment caused by two probable mechanisms. Lentic habitats of impounded lower reaches in the West Fork apparently reduced movement by fluvial specialists among streams and, thus, reduced their opportunity to recolonize dry reaches compared to populations in East Fork streams. Moreover, West Fork assemblages had more macrohabitat-generalist species, which were abundant in the littoral zone of the impoundment and able to recolonize or tolerate environmental conditions in intermittent reaches. We caution that upstream effects of impoundment could be mistakenly attributed to other factors if they are not explicitly considered in species-environment studies.Abstract
Abundance of (top) largemouth bass and (bottom) redfin pickerel across 25-mm groups in tributaries of the East and West Forks of the San Jacinto River
Ordination plot of results from the canonical correspondence analysis showing the relationship on the first two canonical axes between explanatory variables and distribution of fishes in streams of the East and West Forks of the San Jacinto River, Texas. Effects of season were removed by including them as covariables, before ordination. Vectors indicate the direction environmental factors increase in value in relation to species' relative abundances. Vectors also extend in the opposite (negative) direction but for simplicity are not shown. Smaller angles between a vector and an axis indicate higher correlation of the variable with the axis, and longer vectors indicate greater variation accounted for in species' distributions. The approximate center of distribution (highest relative abundance) for a species with respect to an environmental vector is the perpendicular intersection of the line drawn from its centroid the (positive or negative) vector. Open circles are centroids at the center of the distribution of sites within each fork based on the correlation of species' weighted averages and values of the environmental variables. Species abbreviations are the first letter of the genus and first three letters of the species in Appendix 1. The arrows for Nnoc, Psci, and Igag indicate species having ordination scores that lie outside of the axis-one boundary
Ordination plots of the separate canonical correspondence analyses for the first two canonical axes showing the relationship between explanatory variables and distribution of fishes in the East and West Forks of the San Jacinto River. Effect of seasons was first removed by including them as covariables, before analyzing the remaining variation. Instructions for interpretation of vectors for environmental variables and centroids for fork and species are as for Figure 2. Species abbreviations are the first letter of the genus and first three letters of the species in Appendix 1