Effects of land use, stream habitat, and water quality on biological communities of wadeable streams in the Illinois River Basin of Arkansas, 2011 and 2012

Abstract

The Illinois River Basin includes an area of diverse land use in northwestern Arkansas. Land-use data collected in 2006 indicate that most of the land in the basin is agricultural. The agricultural land is used primarily for production of poultry and cattle. Eighteen sites were selected from the list of candidate sites based on drainage area, land use, presence or absence of an upstream wastewater-treatment plant, water quality, and other information gathered during the reconnaissance. An important consideration in the process was to select sites along gradients of forest to urban land use and forest to agricultural land use. Water-quality samples were collected for analysis of nutrients, and a multiparameter field meter was used to measure water temperature, specific conductance, pH, and dissolved oxygen. Streamflow was measured immediately following the water-quality sampling. Macroalgae coverage was estimated and periphyton, macroinvertebrate, and fish communities were sampled at each site. Stream habitat also was assessed. Many types of land-use, water-quality, and habitat factors affected one or more aspects of the biological communities. Several macroinvertebrate and fish metrics changed in response to changes in percent forest; sites that would be considered most disturbed, based on these metrics, are sites with the highest percentages of urban land use in their associated basins. The presence of large mats of macroalgae was one of the most noticeable biological characteristics in several streams within the Illinois River Basin. The highest macroalgae percent cover values were recorded at four sites downstream from wastewater-treatment plants. Macroalgae percent cover was strongly correlated only with bed substrate size, canopy closure, and specific conductance. Periphyton metrics were most often and most strongly correlated with riparian shading, specific conductance, substrate turbidity, percent agriculture, poultry house density, and unpaved road density; some of these factors were strongly correlated with percent forest, percent urban, or percent agriculture. Total biovolume of periphyton was not strongly correlated with any of the land use, habitat, or water-quality factors assessed in the present study. Although algal growth typically increases with higher nutrient concentrations and less shading, the standing crop of periphyton on rocks can be reduced by herbivorous macroinvertebrates and fish, which may explain why total biovolume in Ozark streams was not strongly affected by water-quality (or other habitat) factors. A macroinvertebrate index and several macroinvertebrate metrics were adversely affected by increasing urban and agricultural land use and associated environmental factors. Factors most commonly affecting the index and metrics included factors associated with water quality, stream geometry, sediment, land-use percentages, and road density. In general, the macroinvertebrate index was higher (indicative of least disturbance) at sites with greater percentages of forest in their basins, lower percentages of urban land in their basins, and lower paved road density. Upstream wastewater-treatment plants affected several metrics. For example, three of the five lowest macroinvertebrate index scores, two of the five lowest percent predator values, and two of the five highest percent gatherer-collector values were at sites downstream from wastewater-treatment plants. The Ozark Highlands fish index of biotic integrity and several fish metrics were adversely affected by increasing urban and agricultural land use and associated factors. Factors affecting these metrics included factors associated with nutrients, sediment, and shading. In general, the fish index of biotic integrity was higher at sites with higher percentages of forest in their basins, lower percentages of urban land in their basins, higher unpaved road density, and lower paved and total road density. Upstream wastewater-treatment plants seemed to affect some fish community metrics substantially but had little effect on other metrics. For example, three of the five lowest relative abundances of lithophilic spawner minus stonerollers and four of the five highest stoneroller abundances were at sites downstream from wastewater-treatment plants. Interpretations of the results of the study described in this report are limited by a number of factors. These factors individually and collectively add to uncertainty and variability in the responses to various environmental stresses. Notwithstanding the limiting factors, the biological responses of macroalgae cover and periphyton, macroinvertebrate, and fish metrics to environmental variables provide multiple lines of evidence that biological communities of these streams are affected by recent and ongoing land-use practices. For several biological metrics there appears to be a threshold of about 40 to 50 percent forest where values of these metrics change in magnitude. However, the four sites with more than 50 percent forest in their basins were the four sites sampled in late May–early June of 2012 (rather than July–August of 2011). The relative influence of season and forest percentage on the biological communities at these sites is unknown.