Relation of Physical and Chemical Characteristics of Streams to Fish Communities in the Red River of the North Basin, Minnesota and North Dakota, 1993-95

Document
Description
Fish community composition was determined at 33 reaches (average length 150 meters) at 22 sites in the Red River of the North Basin during 1994. Sites were selected to represent a range of stream sizes and ecoregions within the basin. Physical and chemical characteristics (classified in data sets of instream habitat, terrestrial habitat, hydrology, and water quality) were determined for various sites for periods ranging from two to 48 years. Instream habitat measurements were made from 1993 through 1995 for 31 reaches at 19 sites. Terrestrial habitat measures of land use/land cover, soils, and riparian zones were determined from a geographical information system coverage for 23 reaches at 14 sites. The geographical information system coverage used data from aerial photographs taken from 1990 and 1991, National Wetlands Inventory data, soils maps from the Natural Resource Conservation Service, and data from the U.S. Department of Agriculture soils data base. Water chemistry data were collected from 14 sites in the basin from 1993 through 1994. Hydrologic variability was determined from U.S. Geological Survey gaging records. Correlation analysis, cluster analysis, and principal components analysis were used to determine representative variables which accounted for the most variation in each data set. The representative variables and the fish community data were analyzed with canonical correspondence analysis to determine the relative effect of each source of environmental influence on fish community composition. Instream habitat, terrestrial habitat, and hydrologic variability were analyzed together. Water chemistry data were analyzed separately due to a lack of corresponding sites. Within the instream habitat data set, measures of habitat volume (channel width and depth) and habitat diversity were most significant in explaining the variability of the fish communities. The amount of nonagricultural land and riparian zone integrity from the terrestrial habitat data set were also useful in explaining fish community composition. Variability of mean monthly discharge and the frequency of high and low discharge events during the three years prior to fish sampling were the most influential of the hydrologic variables.The first two axes of the canonical correspondence analysis accounted for 43.3 percent of the variation in the fish community and 52.5 percent of the variation in the environmental-species relation. Water-quality indicators such as the percent of fine material in suspended sediment, minimum dissolved oxygen concentrations, minimum concentrations of dissolved organic carbon, and the range of concentrations of major ions and nutrients were the variables that were most important in the canonical correspondence analysis of water-quality data with fish. No single environmental variable or data set appeared to be more important than another in explaining variation in the fish community. The environmental factors affecting the fish communities of the Red River of the North are interrelated. For the most part, instream environmental conditions (instream habitat, hydrology, and water chemistry) appear to be more important in explaining variability in fish community composition than factors related to the agricultural nature of the basin.
Date Issued
1996
Number of Pages
65
Decade
Publisher
U.S. Geological Survey
Rights Holder
Minnesota Water Research Digital Library
Rights Management
Creative Commons