An Investigation of Disturbance, Sampling Methods, the River Continuum Concept, and Hydroseres in Hardwood Creek, a Ditched Stream in Minnesota, USA

Abstract

Hardwood Creek is located northeast of the Twin Cities Metropolitan Area near Hugo, Minnesota. Upper portions of the stream have a long history of ditching that has substantially modified the flow regime and physical structure of the stream channel. Ditching has occurred at different intervals along segments of the stream, resulting in a mosaic of segments at varied stages of regression back to more normal channel structure in terms of sinuosity, habitat heterogeneity, and profile. Ditching is a physical disturbance to streams and is expected to modify stream habitats. Sites which have never been ditched, or have not been ditched recently, are expected to have greater habitat heterogeneity when compared to recently ditched sites. As a result, traditional sampling methods may not provide an accurate estimate of taxa richness. Dipnet (DN) and chironomid surface-floating pupal exuviae (SFPE) methods were employed to capture both chironomid and non-chironomid macroinvertebrates. The SFPE method detected more taxa over the entire project and on a site-to-site basis for both chironomid and non-chironomid taxa. For chironomids collected in June, the DN method was most effective across all sites, but at disturbed sites there was no difference between methods. In addition, the SFPE method exclusively collected 1.2 to 8.8 times as many non-chironomid macroinvertebrates, on a site-to-site basis, as the DN method. A separate analysis found the SFPE method exclusively collected twice as many chironomid genera as the DN method. To obtain a comprehensive estimate of taxa richness in ditches, all macroinvertebrates collected in SFPE samples need to be evaluated in conjunction with DN samples. The a priori test of the IDH to account for the longitudinal patterns of macroinvertebrate biodiversity, as a function of ditching-related disturbance, was also examined. The IDH was tested at the community, order, and family level to determine whether different levels of taxonomic scale would reveal differing levels of congruence with the IDH. I found the model does not account for the patterns of species richness and diversity, with the exception of one order and three families, the IDH was rejected. As a consequence of the IDH being rejected, a posteriori models were employed in an attempt to explain community composition in Hardwood Creek. The River Continuum Concept (RCC) was tested by analyzing the macroinvertebrate community using three similarity indices to test eight hypotheses. P-values ranged from 0 .488 to 0.957 leading to the acceptance of all eight null hypotheses. The similarities of communities of adjacent sites did not conform to the prediction of the RCC that sites adjacent to each other should have the most similar community compositions. The failure to explain community composition patterns in Hardwood Creek led to the testing of more a posteriori models, hydroseres and management categories. Eight taxa were associated with a single hydrosere. Similarity indices generally supported the validity of the hydrosere classification based upon time since the last dredging event. However, the hydrosere models failed to explain patterns of taxa richness for most taxa. An alternative classification based on 303d listing considerations and Total Maximum Daily Load (TMDL) management strategies was also investigated. Based on this classification, the number of taxa at upstream and downstream sites was not significantly different. However, similarity indices generally supported the upstream and downstream management reaches. Poor water quality, substrate stability, and differences in aquatic vegetation among sample sites are probably the most important factors confounded with ditching in Hardwood C reek. The confounding factors likely played a role in the limited success of IDH, RCC, hydrosere, and management models to explain patterns of macroinvertebrate community composition in Hardwood Creek.