Comparisons of swimming performance in rainbow trout using constant acceleration and critical swimming speed tests

Maximum swimming performance of seasonally acclimated rainbow trout Oncorhynchus mykiss was compared among short-duration constant acceleration tests ( U _max) and with the well established, but longer duration critical swimming speed ( U _crit) test. The present results show that U _max was insensitive to a range of acceleration rates that differed by more than three-fold. Thus, test duration could be reduced from 58 to 18 min without affecting the estimate of U _max. The value of U _max, however, was up to 57% higher than U _crit. Only the slowest acceleration rate tested (an increase of 1 cm s^?1 every min) had a significantly lower U _max, and this was up to 19% higher than U _crit. Even so, the potential saving in the test duration was small (70 v. 90 min) when compared with a ramp- U _crit test (a standard U _crit test but with the water velocity initially ramped to c . 50% of the estimated U _crit). Therefore, swim tests that are appreciably shorter in duration than a ramp- U _crit test result in U _max being appreciably greater than U _crit. An additional discovery was that the ramp- U _crit performance of cold-acclimated rainbow trout was independent of the recovery period between tests. These results may prove useful in making comparisons among different swim test protocols and in designing swim tests that assess fish health and toxicological impacts.