The Use of Cluster Analysis in the Assessment of Spills of Hazardous Materials

Abstract

The macrobenthic community of Clinch River, near Carbo, Virginia, has twice been subjected to acute stress caused by major industrial spills from a power plant. The first spill, which resulted in a high pH shock, was from a fly-ash retaining pond in 1967. The second was an acid spill in 1970 with consequent low pH shock. Stream surveys were made in 1969, 1970 and 1971. This paper reports the results of Q-mode cluster analysis of presence-absence data on the total aquatic insect fauna, several orders of insects considered separately, and Gastropoda from those surveys. Recovery from the effects of the fly-ash spill by all elements of the fauna studied except the Gastropoda was well under way by the summer of 1969. Nevertheless, the insect fauna in samples from the area affected by the spill was still different from that in unpolluted reaches of the stream, although it was not possible to discriminate between remnant effects of the spills and chronic stress due to the day-to-day operation of the power plant, The spill of acid in 1970 eliminated many elements of the fauna from about 30 km of the river. Again, by the end of the summer recovery was well under way for all groups except the Gastropoda. Cluster analysis was particularly useful in determining the effects of type of substrate, time of sampling, longitudinal succession, and flooding on the composition of the macrobenthic community. It is suggested that one effect of flooding may be to make the fauna more homogeneous so there is a more nearly equal distribution of macrobenthic organisms among the stations from which samples were collected. INTRODUCTION Since 1967, the Clinch River in southwestern Virginia has been subjected to two major industrial spills and several episodes of flooding that have had serious deleterious effects on the fish population and on the community of macrobenthic invertebrates (Table 1). These spills and floods and their effects on density and diversity of the aquatic community have been discussed previously (Cairns et al., 1971; Cairns et al., 1972; Crossman et al., 1973). The paper by Crossman et al. also presented cluster analysis of some of the data on aquatic insects. The purpose of this presentation is to make more extensive use of I Present address: Teledyne Brown Engineering, Huntsville, Alabama. 2 Work was done while author was Visiting Professor in Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg 24061.