Sample Analysis
All samples collected from Crater Lake were immediately preserved in 3 percent formalin solution. In the laboratory an aliquot was taken from a thoroughly mixed sample and the organisms counted under a dissecting scope. A graduated bulb pipette was used in extracting the aliquot from the sample.
The relationship between aliquot volume to be counted and sample volume was determined by weight. By dividing the weight of the aliquot into the weight of the sample and multiplying the quotient times the number of organisms counted, an estimate of the number of organisms in the entire sample could be obtained. This relationship is represented as follows:
The organisms were enumerated by species and two age classes (adults and juveniles) to observe any differences in their horizontal and vertical distributions. Identification of juveniles was determined by size. All forms having no eggs and being no larger than one fourth the size of the adults were classified as juveniles.
Two aliquots per sample were counted for the vertical tows in 1967 and the horizontal tows in 1968. Three aliquots per sample were counted for the horizontal tows in August 1967, and the vertical tows in 1968.
Since the T. S. flow-meter is capable of measuring the volume of water sampled, samples taken with nets equipped with a flow meter were computed in organisms per cubic meter of water sampled. Horizontal calibration of the flow meter was done in the Men’s Pool at Oregon State University. Calibration of the vertically towed flow meter was done a t Crater Lake in 1968. Vertical tows at 100 m were made with both standard and closing nets (No. 20 mesh) with and without a flow meter. Averages of flow meter revolutions per 100 m vertical tow were used for tows made without a flow meter. The standard tow net with a No. 6 mesh was always equipped with a flow meter.
All calculations were programmed for computerization. Calculations were made of sample means and the number of organisms/cubic meter of water sampled (these data are included in the Appendix). The horizontal distribution data were analyzed in a multiple analysis of variance by comparing samples taken in the same stations during both 1967 and 1968. Station 21, sampled in 1968, was compared with station 22, sampled in 1967, because of their relatively similar positions on the lake (Figure 1).
Since sampling error from within station variance might bias the results of the previous analysis, the effect of sampling error was determined by analyzing common stations sampled during August 26, 1967, and August 28, 1968, when two samples per station were taken.
***previous*** — ***next***
