Rapid bioassessment of macroinvertebrate communities is suitable for monitoring the impacts of fish farm effluents

Development of the fish farming industry in Iran in an environmentally and economically sustainable manner requires an effective and low-cost means of regularly monitoring receiving environments. Biomonitoring using macroinvertebrates is known to be effective for assessing water quality. The problem, however, is that biomonitoring can be labour intensive and analyses can have a long turnaround time. Rapid bioassessment methods have been developed to overcome these limitations, but it is not known whether they are as sensitive to changes in water quality as are their more time-consuming counterparts. To answer this question, we compared three methods for sampling and measuring macroinvertebrates. We refer to these as the quantitative method, semi-quantitative method, and qualitative method respectively. The quantitative method was a single habitat method with taxonomic identification of macroinvertebrates to genus level that counted all taxa. The semi-quantitative method involved multi-habitat sampling with identification to family level and quantification as relative abundance. The qualitative method was the same as the semi-quantitative method except that incidence (presence/absence) was recorded instead of abundance. The study was carried out at three fish farms in Iran with sampling done once per season for a year from the outfall of each farm as well as from the receiving rivers, with one sample taken upstream of the effluent discharge and two samples downstream. Analysis by permutational multivariate analysis of variance (PERMANOVA) revealed that the effects of three variables of season, farm, and site on macroinvertebrate communities were significant for all three methods. Qualitative sampling was the only method that showed a statistically significant interaction between farm and season as well as a difference among the sites within each farm. Although the results of a BEST (Bio-Bio) analysis showed that different families were responsible for the differences between the sites, all three methods were able to detect the differences between the sites within each farm. However, pairwise comparisons between sites within farms indicated some differences between the three methods. The quantitative method revealed fewer differences than did the other two methods. The qualitative method did not lose any important information and had the added advantage of saving considerable time and effort in sampling and enumerating. These results suggest that rapid bioas