Laserfiche WebLink
and Rotifera were counted using a dissecting microscope. All benthic and planktonic <br />organisms from the three treatments were counted for weeks 1, 3, and 5. Benthic and <br />planktonic organisms from 10 randomly chosen benthic samples were analyzed from <br />weeks 2, 7, and 10. Chironomidae, Copepoda, and Cladocera were identified from at <br />least ten benthic samples from each of the sample series. Cladocerans and adult <br />copepods from all sorted plankton samples were identified. Chironomidae were <br />mounted on slides in Hoyer's solution and identified to genus using keys by Merritt and <br />Cummins (1984), Wiederholm (1983), and Mason (1968). Cladocerans and adult <br />copepods were identified to species using. a keys by Yeatman (1959), and Pennak <br />(1989). <br />Statistical analysis <br />Data from weeks 1, 3, and 5 were examined with analysis of variance (ANOVA) <br />tests. During week 7 we noted a significant reduction in permeability of the enclosure <br />cages caused by growth of algae on the screens. This introduced a potential bias <br />when comparing these areas with the open control sites for weeks 7 and 10 and for <br />that reason the closed and perforated treatments were excluded from the analysis. <br />We performed a weighted ANOVA on the treatment means (of sample counts) for <br />each group of organisms. The weighting was based on the number of samples <br />processed for the mean. The ANOVA model was: <br />Y;;k=N+B;+D;+BD;j+Tk+BT;k+DT,k+BDT;;k <br />where Y;jk is the count mean of the it' backwater on the jt' date from the kt' treatment, <br />B; is the backwater (block) effect, Dj is the date effect, and Tk is the treatment. F- <br />6