Cluster Analysis of Data from Limnological Surveys of the Upper Potomac River

Abstract

In a study of data from limnological surveys of a portion of the upper Potomac River, cluster analysis and correlation coefficients have shown that redundancy exists in the data. Aquatic insects and, to a lesser extent, diatoms have distributions that most nearly represent the total biota. It is suggested that the study of these two groups and the exclusion of other algae, protozoans, other invertebrates and fish would significantly reduce the cost of such surveys without appreciably affecting the information they provide. INTRODUCTION Limnological surveys of the kind reported by Patrick, Cairns and Roback (1967) of the Savannah River between South Carolina and Georgia have typically involved study by a team of specialists of almost the entire biota, including diatoms and other algae, protozoans, aquatic insects, other invertebrates and fish, from several stations for each of several years. Such surveys are usually used to monitor a source of potential pollution, such as thermal pollution from the outfall of a power station. Because the results of limnological surveys may be used as evidence in the courts, it is imperative that standard methods be used both for the surveys themselves and for their interpretation (Patrick, 1949). Nevertheless, it is clear that a great deal of redundancy may be introduced by the study of the entire biota, so that the cost of limnological surveys could be reduced if only those groups of organisms were selected for study that have distributions that are highly correlated with other major groups. The purpose of this research was to investigate the hypothesis that data from limnological surveys typically include a great deal of redundancy. In this paper we show that this isi true, we present the results of Q-mode cluster analysis using several groups of organisms, and we indicate which groups of organisms are likely to best summarize the distribution of the total biota. We have selected for study data from limnological surveys of the upper Potomac River directed by Dr. Ruth Patrick. Cairns, Kaesler and several co-authors have studied distribution of the several groups of organisms and have described in detail the sampling procedures and methods of analysis used (Cairns and Kaesler, 1969; Roback, Cairns and Kaesler, 1969; Cairns, Kaesler and Patrick, 1970; Cairns and Kaesler, 1971; and Kaesler, Cairns and Bates,, 1971). Data from the surveys are on file as follows: protozoans, National Auxiliary Publications Service No. 00362 and No. 89021; insects, 1National Auxiliary Publications Service, 22 West 34th Street, New York, N.Y. 10001, U.S.A. Microfiche copies $1.00 per set; 8 x 11 inch photo copies $3.00 per set.