The purpose of the investigation presented in chapter 1 was to determine which of the following artificial stream designs would be most logistically simple yet effective in maintaining riffle insects during a 30-d bioassay: 1) static and no current (S-NC); 2) flow-through and no current (FT-NC); 3) static with current (S-C); or 4) flow-through with current (FT-C). Flow-through and current, when provided, were 12 ml min⁻¹ and 30 cm sec⁻¹, respectively. Streams were covered by emergence traps, and daylight equivalent lights provided a natural photoperiod. The four stream designs were evaluated in triplicate based on changes in insect species-abundances after 30 d. Test organisms were transferred to the artificial streams in rock-filled containers previously colonized for 30 d in a third-order mountain stream riffle. Relative to benthic samples taken directly from the source riffle, the artificial substrates selected for collector-filterers and against collector-gatherers. The FT-C and S-C stream designs maintained most taxa at or above initial densities. Emergent adults comprised a large proportion of mayfly and chironomid densities and must be monitored during bioassays with aquatic insects.

The Investigation reported in chapter 2 was conducted to determine if contaminant-induced changes in macroinvertebrate and periphyton communities in laboratory stream microcosms could be used to predict macroinvertebrate and periphyton responses In a natural stream receiving the same contaminant. The microcosms were dosed in quadruplicate with four (0.0, 0.1, 1.0, and 10.0%) concentrations of a complex effluent; these concentrations reflected those in the field. Mayfly densities in the microcosms were significantly (P≤0.05) reduced at 1.0 or 10.0% effluent depending on species. Hydropsychlds were not affected by the effluent, and chironomids and periphyton were stimulated. Overall, the stream microcosms accurately predicted the macroinvertebrate and periphyton response observed in the field.

Chapter 3 compared responses to a complex effluent from microcosms of indigenous macroinvertebrates and protozoans to responses observed in acute tests with Daphnia magna, Ceriodaphnia dubia and Pimephales promelas and chronic survival and reproductive tests with C. dubia The predictive utility of these various tests was then evaluated against observed effects in the receiving stream. The LC₅₀_s (% effluent) from the acute tests were 63.09 for Pimephales promelas, 18.8 to 31.3 for Daphnia magna and 54.7 for Ceriodaphnia dubia. Results from 7-day chronic tests indicated that C. dubia survival was significantly (P≤0.05) affected at 30% effluent and reproduction was affected at concentrations ≥3.0% effluent. In the protozoan microcosms, community composition was significantly (P≤0.05) changed at 1.0%; while protozoan species richness was significantly reduced at 3.0% effluent. The microcosms not only were the most sensitive indicators of effluent toxicity, they also correctly predicted which indigenous organisms would be lost and which would be stimulated at various ambient concentrations of the effluent.

In the fourth chapter canonical discriminant analysis, 2 diversity indices, and 7 community comparison indices were evaluated to determine their utility in quantifying macroinvertebrate response to a complex effluent in laboratory microcosms. A permutation and randomization procedure was used to test the hypothesis of no treatment effect based on the community comparison indices. The Bray-Curtis index provided the most meaningful condensation of the data.