106 
David C. Culver and Timothy J. Ehlinger 
ficial stream. Since individuals that washed out sometimes died, this in- 
troduces a potential bias if there are differences in washout propensity 
among individuals of the same size class. This would result in higher 
washout rates in earlier runs of the same experiment. There was no 
evidence that this happened. 
There is a very clear tendency for small C. holsingeri to have higher 
washout rates in large gravels than in small gravels, and for large C. can- 
nulus to have higher washout rates in small gravels than in large gravels 
(Table 2). The washout frequency of C. holsingeri is significantly higher 
(P < 0.01) in large gravels (x = 0.75) than in small gravels (x = 0.15). 
Similarly, the washout frequency of C. cannulus is significantly higher (P 
<0.05) in small gravels (x = 0.55) than in large gravels (x = 0.24). These 
data suggest that there should be a “match” between gravel and isopod 
size distributions in natural streams. 
Table 2. Washout rates of small C. holsingeri ( <C.4.5 mm) and large C. cannulus 
( ^ 8 mm) in small gravels (.2 cm diam. 1 cm) and large gravels 
(2.2 cm <: diam ^3.8 cm). Statistical analysis used washout frequency 
transformed to sin"' VT, where x is the washout frequency. 
X 
sim' V\ = y 
Species 
Gravel Size 
N 
x 
S.E. 
y 
S.E. 
C holsingeri 
small 
4 
0.15 
0.05 
20.0 
6.7 
C. holsingeri 
large 
4 
0.75 
0.05 
60.2 
3.2 
C. cannulus 
small 
4 
0.55 
0.05 
47.9 
2.4 
C. cannulus 
large 
5 
0.24 
0.12 
23.3 
10.2 
t P 
5.46 0.01 
2.07 -=^0.05 
There was also evidence that size differences among isopods reduced 
the washout rate in an artificial stream with a variety of gravel sizes 
(Table 3). The washout frequency of large C. cannulus is significantly 
lower when small C. holsingeri are present (x = 0.08) than when the same 
total density of isopods, all of them large, are present (x = 0.44). The 
washout frequency of small isopods was slightly lower when large 
isopods were present (0.36 compared to 0.42), but the difference was not 
significant (Table 3). 
The difference in washout frequency shown in Table 3 could be due 
either to size differences per se or to species differences per se. Washout 
experiments with intermediate-size isopods (Culver and Ehlinger, in 
preparation) indicate that size per se is more important. In these experi- 
ments, five small C. holsingeri (or five large C. cannulus) were put in the 
