| Abstract: | SUMMARY. 1. Blackfly larvae were collected from twenty-one stations in five lake outlets in Southern Quebec. Tiles (total area=500cm2) were introduced in early March, and collected 4 weeks later: randomly selected rocks (30–500cm) from the surrounding area were collected at the same time. 2. Larval densities on tiles were significantly less variable than on rocks. The variance of density estimates on tiles averaged 36% of the observed variance on natural rocks, or 67% when variance on rocks was corrected for average rock size. 3. Tiles significantly overestimated densities on rocks in some streams, and significantly underestimated them in others. These differences could not be explained by microhabitat differences (distance from the lake, depth, current velocity) between rock and tile samples. The bias that tiles introduce in density estimates precludes their use in comparisons among sites. |
