is just as important in controlling populations (Edmondson, 1993). Properties other than mortality (such as those given above) also vary with toxicity. Neoplastic lesions in fish have been correlated with pollutants (Russell and Kotin, 1957; Tyler et al., 1991). Chironomid head capsule morphology (Warwick, 1988) and diatom shell structure (W. T. Edmondson, personal communication, 1994) can be distorted when the organisms live in conditions containing sublethal toxic material. In Lake Orta in north Italy, copper-rich waste discharged from a rayon factory for many years caused elimination of the entire biota. The sedimentary record for the lake shows an initially normal diatom community; at progressively shallower levels, there are increasing proportions of distorted diatoms, gradual elimination of the various species, and finally no biota. Whether other metals would have the same effect as copper or whether different species of diatoms respond differently to different metals is not known.
Population guilds and communities generally have proven more satisfactory than individual indicator organisms for the assessment of water quality. For example, paleolimnologists working on acid rain have been able to determine the pH of lakes on the basis of groups of diatoms found in bottom sediments. In a slightly different approach, planktonic-diatom biomass estimates were examined for the sediments of three lakes with contrasting types of lake disturbance: acidification (Gaffeln, southwestern Sweden), point-source eutrophication (Lough Augher, northern Ireland), and catchment agriculture (Akassjon, Northern Sweden) (Anderson, 1994). At Gaffeln, biomass declined steadily with acidification until extinction of the planktonic diatoms occurred. In Akassjon, maximum biomass coincided with the maximum areal extent of arable land. At Augher, nutrient input from a creamery resulted in steady biomass increases. Multivariate statistical analyses of species groupings have proven to be more useful than individual indicator organisms in a number of studies, particularly those involving pollution (e.g., Clarke, 1993).
Various methods are employed to obtain samples of the biota for use in water quality assessment. These vary with the type of habitat (e.g., lotic or lentic), location (e.g., open water or benthic), size (microscopic or macroscopic), type of substratum (e.g., mud or rock), and degree of mobility of the organisms being sampled. Application of these methods requires skill and training. Planktonic forms may be trapped, pumped, enclosed, or netted from specific or composite locations within the water column (Lind, 1985; Wetzel and Likens, 1991). Attached microscopic forms can be scraped or brushed to free them from the substratum and then removed by suction for placement in a sample container or collection onto a filter (Lind, 1985; Porter et al., 1993). Benthic macroinvertebrates are collected by a variety of dredges, nets, corers, baskets, and other devices depending on water depth, substratum type, and current velocity (Lind, 1985; Wetzel and Likens, 1991; Cuffney et al., 1993a,b). Fish commonly are collected