The capacity to detect change stream fish communities characteristics at the site-level in the Lake Ontario basin.

We investigate natural inter-annual variability of fish community measures within streams of the Lake Ontario basin. Given this variability, we examined coefficients of variation (CV) among the community measures and three scenarios pertaining to the capacity of biologists to detect changes in the fish community at the stream site level. Results indicate that Ontario's stream fish communities are highly variable in time. Young-of-the-year rainbow trout growth was the least variable whereas biomass density scored the highest CV of 0.50 among streams (range 0.22-0.99). Given the CVs and relatively equal sample sizes, our measures of the fish community can be ranked from least to most powerful: biomass, density, richness, diversity, and growth of young-of-the-year rainbow trout. Only large changes in measures can typically be detected. For instance, it would take 4-6 years of monitoring before and after a pulse perturbation to detect a 50 % change in species richness or diversity. We suggest that monitoring abundance is unlikely to result in the detection of small impacts within a short period of time and that large effects can be masked by low statistical power. This evidence voices the need for more research into better sampling methods, experimental designs, and choice of indicators to support monitoring programs for flowing waters.