550 FORD, FOREMAN, GRUBBS, KROLL, AND WATTS 



lated in the experiments. Data on the chemical composition of 

 the water in these two systems were obtained seasonally and at 

 representative times of the tidal cycle in studies conducted concur- 

 rently. Comparative measurements were made of the trace metals Cu, 

 Zn, Cd, Pb, Co, Cr, and As (Ford et al., 1976; Dorband et al., 1977); 

 dissolved oxygen, phosphate, ammonia, nitrate, and nitrite (Ford 

 et al., 1976); and Cu, Cr, pH, oil and grease, and chlorine 

 (unpublished monitoring data from generating-station records). 



On the basis of the results of t-tests comparing sample means, 

 there were no significant differences (P > 0.05) in the concentrations 

 of these chemical constituents in the two cooling-water systems. This 

 indicates that the chemical composition of the effluent and intake 

 water supplied to the laboratory was essentially the same as that of 

 the cooling-water system simulated. 



Collection, Stocking, and Maintenance of Animals 



Animals used in experiments at the generating-station laboratory 

 were collected in the vicinity of the station at locations outside the 

 limits of the thermal plume. The Scripps control animals were 

 collected in shallow-water areas off Point La Jolla, Calif., which had 

 temperature regimes similar to those of the collecting sites near the 

 generating station. 



Animals were placed in the 1100-liter tanks and held for 3 to 5 

 weeks before stocking. This quarantine procedure served several 

 functions. It allowed accumulation of large numbers of individuals so 

 that representative size groups could be selected and sorted before 

 the tanks were stocked, provided adequate time for individuals to 

 acclimate to laboratory holding conditions at ambient temperatures, 

 and enabled investigators to observe and eliminate individuals injured 

 by collecting and handling. 



In each unit of 20 tanks used for a given experimental treatment 

 or control group at the generating-station laboratory, five tanks 

 stocked with the same combinations of species, size groups, and 

 densities of experimental animals were used as replicates. The Scripps 

 controls used only three replicates because space limitations allowed 

 use of only 12 tanks. 



Animals of each test species used in the experiments were 

 separated into three size categories representing the size ranges of 

 juveniles and adults commonly encountered in the study area. Five 

 individuals of each size category were randomly assigned to each of 

 the five (or three) replicate tanks in a series. 



Natural foods were introduced daily and were maintained in the 

 tanks at all times in slight excess to promote optimum growth. Both 



