Behavioral Responses of Rainbow Smelt to Sensory Deterrent Systems

Since the late 1970s, rainbow smelt Osmerus mordax have been the primary prey fish for top-level predators in Lake Oahe. Historically, water level management and entrainment losses of rainbow smelt at the Lake Oahe dam have likely contributed to the reduced growth rate of walleye Sander vitreus, chinook salmon Oncorhynchus tshawytschu, as well as other piscivores. In 1997, an estimated 439 million rainbow smelt were entrained through Oahe Dam from June through September. With high predator biomass, a single, weak year class coupled with high entrainment can lead to a substantial reduction in rainbow smelt abundance. The objectives of this study were: 1) to test the effectiveness of an underwater sound projector array (SPA) and (2) strobe lights as fish deterrents for avoidance and acclimation behavior of rainbow smelt; 3) to determine physical characteristics that affect light attenuation in Lake Oahe; and 4) to evaluate predator-prey interactions of rainbow smelt and zooplankton in close proximity to strobe lights. For objectives 1 and 2, splitbeam hydroacoustics were used to assess the effectiveness of the SPA and the strobe light system by comparing proximal densities of rainbow smelt before and after a deterrent method was activated. For objective 3, limnological data such as chlorophyll-a, total suspended solids, water temperature, and light intensity were collected throughout the summer months of 2004 and 2005. To evaluate objective 4, experimental gill nets were used to collect rainbow smelt from two different locations for stomach content analysis. Results from the sensory deterrent systems indicated that strobe lights successfully repelled rainbow smelt to a minimum of 21 m at both 1-h and 4-h post-activation time intervals. A comparison of vertical strata 10 m above and below the strobe light showed that rainbow smelt were displaced to approximately 6 m. Conversely, when exposed to underwater sound, rainbow smelt were only deterred approximately 1 m away from the SPA, exhibiting little to no avoidance in the vertical plane. Thus, deterrence of rainbow smelt was more pronounced using strobe lights as compared to underwater sound. Analysis of limnological variables indicated that light attenuation varied little throughout the summer months of 2004 and 2005 in Lake Oahe. Throughout both years chlorophyll-a concentrations were low, which is similar to other oligotrophic systems. Total suspended solids were also low and inorganic solids were more prevalent than organic solids. Stomach content analysis indicated that rainbow smelt collected in close proximity to the effective deterrence distance had a significantly lower percentage of empty stomachs along with more total prey items compared to fish from distant locations. Cyclopoid copepods composed 94% of the entire diet of rainbow smelt collected from both net locations. Length measurements of cyclopoid copepods were similar between locations. In summary, strobe lights were more effective as a deterrence system and may prove useful for reducing rainbow smelt entrainment. A reduction in entrainment should provide a more stable prey community. It is recommended that the importance of entrained rainbow smelt to the Lake Oahe tailrace/Lake Sharpe food web be evaluated prior to installation of strobe lights. Light attenuation appears to fluctuate minimally, indicating that strobe lights should consistently deter rainbow smelt throughout the summer months in Lake Oahe. Diet analysis of rainbow smelt indicated that strobe lights may have increased the feeding efficiency of rainbow smelt. Lastly, further research may be warranted to evaluate the extent at which predator-prey interactions are altered