Assessing the Sustainability of Fish Communities in Glacial Lakes: Habitat Inventories and Relationships Between Lake Attributes and Fish Communities

Broad-scale aquatic habitat inventories have rarely been done in South Dakota waters. Even less has been done to understand the relationships between habitat variables and the fish communities in large lakes. The objectives of my research were to: 1) quantify the available aquatic habitats in 15 South Dakota large lakes, 2) investigate the use of Geographical Information System technology to create aquatic habitat layers, and 3) determine the relation of abiotic and biotic habitat characteristics to fish growth, size structure, body condition, and relative abundance in large South Dakota lakes. Aquatic habitat types were surveyed during the sununer months of 2000, 2001, and 2002 and found to be highly variable among large South Dakota lakes. Some of the lakes were recently expanded (due to above normal precipitation) and others have been relatively stable. The percent coverage of hard substrates (e.g., sand, gravel, or rock) ranged from 2.4 to 42. l. Submergent macrophyte coverage ranged from 0.5% to 31.1 %. The range of maximum depths was 2.8 m to 4.9 m. Shoreline development indices ranged from 1.4 to 4.5. There was between 15.9% and 66.7% cropland in Hydrologic Units associated with the study lakes. Total phosphorous levels ranged from 0.01 ppm to 0.46 ppm. Trophic State Indices based on chlorophyll-a ranged from 47.9 to 73.7. Secchi visibilities ranged between 0.5 m and 2.5 m. These characteristics result largely from the geology and land-use practices found in this region. The accuracies of Geographical Information System raster maps created from vector (point) data were assessed. Lake substrate data sets consisting of individual sample points were used to create interpolated substrate coverages. The accuracies of the coverages were analyzed based on sample point density. Analysis suggested that substrate sample point densities should be at least 1 point / 8 hectares to achieve an accuracy of 65% on interpolated maps of lakes having relatively homogenous substrates (e.g., Lake Madison). Lakes having substantial substrate heterogeneity (e.g., Pickerel Lake) should be sampled at a density of at least 1 point / 4 hectares to achieve the same accuracy. Correlation analysis suggested that a number of relationships existed between habitat and fish population characteristics. Submergent vegetation was negatively correlated with yellow perch Perca flavescens relative weight (Wr), yellow perch proportional stock density (PSD), and northern pike Esox lucius PSD. Lake basin ruggedness was negatively correlated with black crappie Pomoxis nigromaculatus PSD, northern pike Wr, and walleye Stizostedion vitreum Wr. Lake surface area was positively correlated with walleye Wr, walleye length at age 3, and yellow perch Wr. Nitrate levels were positively correlated with black bullhead Ameiurus me/as CPUE and northern pike CPUE, but negatively correlated with northern pike PSD and walleye Wr. Total phosphorous levels were positively correlated with black bullhead Wr and black crappie Wr. Chlorophyll-a levels were positively correlated with black crappie CPUE. Principal components analysis of the physical and biological habitat data produced three habitat component groupings. Maximum depth, basin slope, and two measures of basin complexity loaded highly on the Lake Basin a negative correlation between this component and northern pike Wr. Surface area, component. There was maximum fetch, and watershed area loaded highly on the External component. Yellow perch Wr was positively correlated to this component. For the Landscape component, submergent macrophyte coverage (positive), shoreline development index (positive), and the percent cropland in the Hydrologic Unit (negative) loaded highly. Black crappie PSD and black bullhead relative abundance were negatively correlated with the Landscape component. This study should serve as a foundation for future studies. The habitat sampling and GIS methodology should be duplicated on additional lakes for further comparisons. Habitat data can be used to monitor changes over time or implemented into fisheries management activities and research efforts. Correlations between habitat attributes and fish population characteristics may be considered when planning further studies, fish stockings, or habitat improvement projects