Laserfiche WebLink
8 <br />squeezed out onto a glass slide. One or two drops of water were added to <br />make a paste and then a 22 mm square cover slip was placed over the <br />material so that the paste distribution was uniform. Using a microscope <br />with a 10 X occular micrometer and 10 X objective, seston particles were <br />identified and counted in 5 random fields. The number of seston per gut <br />was equal to the average number from 5 fields X 207 (number of field/ <br />cover slip) X gut volume/2. Biomass of diatoms, plant fragments, and <br />green and blue green algae was estimated by the cell volume method (APHA <br />1976). <br />Statistics <br />Data analysis was carried out by the use of Statistical Analysis <br />System (SAS 1979) and hand calculations where appropriate. <br />The following variables were recorded for each fish: stratum, loca- <br />tion, number (to identify fish from collection), species, age, length and <br />numbers of items and weights for each of the 53 food items. Initial <br />analyses indicated the necessity of grouping the data where possible so <br />the 53 food items were grouped into 15 categories as indicated in Figure <br />2. Fish species-age group categories were formed by combining the species <br />and age variables and were used throughout the analysis. <br />Frequencies of occurrence were obtained by using SAS PROC RECODE and <br />PROC FREQ. Index variables were assigned by PROC RECODE; zero for zero <br />numbers, one for numbers greater than zero. The output for PROC FREQ <br />includes a column percent appropriate for frequency of occurrence when <br />used with recoded variables.