We found no significant differences in either sockeye or sculpin
numbers between sites–depths within sampling locations and strong
positive correlations between sockeye and sculpin numbers over time
within each site (results not presented). As such, we present only the
mean number per sampling location, pooled over the two depths.