Biological evaluation of stream restoration efforts in the Cache River basin in southern Illinois.

摘要

Stream restoration practices are becoming increasingly common, but biological assessments of these improvements are still limited. Constructed riffles were implemented in the upper Cache River in southern Illinois, USA, in 2001 and 2003-2004 to control channel incision and protect high quality riparian wetlands as part of an extensive watershed-level restoration. Construction of these riffles provided an opportunity to examine biological responses to a common in-stream restoration technique. I compared macroinvertebrate assemblages on previously constructed riffles and newly constructed riffles to those on snags and scoured clay streambed, the two dominant substrates in the unrestored reaches of the river. Ephemeroptera, Plecoptera, and Trichoptera (EPT) biomass and total insect biomass were significantly higher on constructed riffles than the streambed for most sample periods. Snags supported intermediate EPT and total insect biomass compared to rock weirs and streambed. Nonmetric multidimensional scaling ordinations revealed distinct assemblage groups for constructed riffles, snags, and the streambed. Analysis of similarity supported visual interpretation of NMDS plots. Results indicate positive responses by macroinvertebrate assemblages to in-stream restoration in the Cache River, particularly by insects in general and more sensitive EPT taxa. Construction of another series of riffles in Big Creek, a third order tributary of the lower Cache River, began in summer 2005. I conducted pre-restoration sampling in Big Creek from September 2000--October 2002. Data presented here comprise an extensive pre-restoration dataset. I quantified macroinvertebrate production and organic matter standing crops for riffle/run and pool habitats in two stream reaches along Big Creek. The lower reach was designated as the site for constructed riffles. The upper reach served as a statistical reference for both pre- and post-rehabilitation biological assessment. Habitat-weighted secondary production estimates were 39 and 42 g AFDM m-2 y-1 in the upper and lower reaches, respectively, in year one of the study and 27 and 21 AFDM m -2 y-1 in the second year. Nontanypodine Chironomidae and Oligochaeta combined accounted for ≥65% of production in both reaches. Gammarus contributed an additional 14% of production in the upper reach in year one and 9% in year 2. Estimating production of nontanypodine Chironomidae was problematic because of taxonomic difficulties, short generation times, and overlapping cohorts. Therefore, I developed assemblage-level instantaneous growth models for Big Creek. These temperature-specific models will also allow more accurate estimates of chironomid production in central U.S. streams with thermal regimes and larval midge assemblages similar to those in Big Creek.