Eighteen species of fish including a reference species, were subjected to toxicity bioassay using petroleum refinery effluent as a toxicant. Twenty-four-hour and 96-hour median tolerance limits Were calculated using a straight-line graphical interpolation based on ten specimens per concentration with a replication. Collection, laboratory, and bioassay histories were recorded for each test species and a general suitability statement made for each. Twenty-four-hour and 96-hour adjusted resistances obtained by the “Preadjusted-Abbreviated Doolittle” method were subjected to analysis of variance and to a modification of Duncan's new five percent multiple range test. Six methods were employed to adjust the relative resistance for differences in tests. The tests were ranked according to results obtained by each adjustment from most to least resistant. The “Interval” method was preferred over the other adjusted procedures on the bases of computational ease, ready addition of new data, compatible rankings with the “Preadjusted-Abbreviated Doolittle” method, and a lack of a “number effect.”
