Further Tests of Changes in Fish Escape Behavior Resulting from Sublethal Stresses Associated with Hydroelectric Turbine Passage

Ryon, M.G.

Oak Ridge National Laboratory

关键词: Stresses;    Orientation;    Vulnerability;    Tanks;    Turbines;   

DOI  :  10.2172/885567
RP-ID  :  ORNL/TM-2003/288
RP-ID  :  DE-AC05-00OR22725
RP-ID  :  885567

美国|英语

来源: UNT Digital Library

【 摘 要 】

Fish that pass through a hydroelectric turbine may not be killed directly, but may nonetheless experience sublethal stresses that will increase their susceptibility to predators (indirect mortality). There is a need to develop reliable tests for indirect mortality so that the full consequences of passage through turbines (and other routes around a hydroelectric dam) can be assessed. The most commonly used laboratory technique for assessing susceptibility to predation is the predator preference test. In this report, we evaluate the field application of a new technique that may be valuable for assessing indirect mortality, based on changes in a behavioral response to a startling stimulus (akin to perceiving an approaching predator). The behavioral response is a rapid movement commonly referred to as a startle response, escape response, or C-shape, based on the characteristic body position assumed by the fish. When viewed from above, a startled fish bends into a C-shape, then springs back and swims away in a direction different from its original orientation. This predator avoidance (escape) behavior can be compromised by sublethal stresses that temporarily stun or disorient the fish. Initial studies demonstrated that turbulence created in a small laboratory tank can alter escape behavior. As a next step, we converted our laboratory design to a more portable unit, transported it to Alden Research Laboratory in Holden, Massachusetts, and used it to test fish that passed uninjured through a pilot-scale turbine runner. Rainbow trout were either passed through the turbine or exposed to handling stresses, and their behavior was subsequently evaluated. Groups of five fish were given a startle stimulus (a visual and pressure wave cue) and filmed with a high-speed (500 frames per s) video camera. The reactions of each group of fish to the startle stimulus were filmed at nominally 1-, 5-, and 15-min post-exposure. We compared the behaviors of 70 fish passed through the turbine and another 70 under control conditions (either transferred from the holding tank or injected into the Alden loop downstream of turbine). The resulting image files were analyzed for a variety of behavioral measures including: presence of a startle response, time to first reaction, duration of reaction, time to formation of the maximum C-shape, time to completion of the C-shape, completeness of the C-shape, direction of turn, and degree of turn. The data were evaluated for statistical significance and patterns of response were identified. The most immediate measure of potential changes in fish behavior was whether test and control fish exhibited a startle response. Unlike earlier studies, there was no significant difference among the treatment group and the controls for startle response. The majority of rainbow trout in all groups responded to the startle stimulus. There were however, significant differences in some of the particular aspects of the subsequent escape behavior. The time to first reaction, the duration of the reaction, and the times associated with maximum C-shape formation were all significantly different between the tank controls and the two groups of fish injected into the Alden turbine loop. There were no significant differences in behavioral responses between the trout passed through the turbine runner and those injected downstream of the runner. Other behavioral parameters, such as C-shape completeness ratio, were not significantly affected. The effect of the Alden turbine loop on some aspects of the escape behavior suggest that the process of movement through the system is important, but that the role of the added stress, if any, of passage through the turbine runner is minimal. It may be important that statistically significant differences in timing of phases of the startle response were detected, even though the majority of stressed fish still exhibited the startle response. This is in contrast to earlier studies, where timing of phases of the startle responses were only affected when the overall startle response was impaired. This pattern suggests that the escape behavior response can be a very sensitive tool for detecting potential effects of sublethal stresses associated with turbine passage. Field studies will be needed to determine whether such subtle changes would be translated into increased vulnerability to predation. If a link between changes in escape behavior and indirect mortality can be demonstrated in the field, this technique could become a rapid and inexpensive tool for assessing the total losses of fish at hydropower plants.