Ecology of the burrowing amphipod, Pontoporeia affinis, in Lake Michigan / by Wayne Paul Alley.

88

mental error or inefficient sampling techniques. The Wilcoxon sign
rank test indicated there was no significant difference in the abundance
of amphipods between the stations of the two sampling dates.
The result of the rank test justified the combining of the two data
sets for each station to obtain an estimate of sample variation. The coefficient of variation, which measures the relative variation of amphipod
abundance, was calculated for these combined data (Fig. 43). There is
no particular pattern to the relative variation with respect to depth or
position along the transect. The average relative variation for the seven combined stations was 9. 3 percent.
The amphipod counts for all 35 stations of the long-term study
area were pooled for each month of the three sampling seasons (Fig. 44).
A Kruskal-Wallis sign rank test indicated there was no significant difference between the average standing crop of the three sampling seasons.
Table 8 shows the grand geometric mean number of amphipods of all
depths combined for each of the sampling seasons and the results of the
Kruskal-Wallis test. The average number of amphipods for 1964 is 53
while the means for 1965 and 1966 are 55 and 54 respectively.
The Pontoporeia collected at the 35 stations were also combined
into four depth zones: 10-35 m, 36-65 m, 66-105 m, and greater than
105 m for the monthly sampling periods (Fig. 44). A Kruskal-Wallis
test was calculated to determine if the average density of amphipods
was significantly different for the four depth zones. A significant differ