Magnitude and Causes of Smolt Mortality in Rotary Screw Traps: an Atlantic Salmon Case Study PAUL A. MUSIC,* JAMES P. HAWKES, AND MICHAEL S. COOPERMAN National Marine Fisheries Service, Northeast Fisheries Science Center, Maine Field Station, 17 Godfrey Drive, Suite 1, Orono, Maine 04473, USA Abstract.—Rotary screw traps (RSTs) are commonly used for collecting and holding fish swimming downstream. We used these traps to collect Atlantic salmon Salmo salar smolts from multiple rivers draining to the Gulf of Maine from 1996 to 2008. Limited prior evaluations suggest that RSTs cause only minimal mortality, but detailed evaluation of the potential for negative impacts specific to Atlantic salmon and other smolts has not occurred. There are concerns that RSTs contribute to smolt mortality via physical injury during collection or via exposure to stressful holding conditions within RST live-cars. We evaluated the magnitude and causes of mortality associated with our use of RSTs and recorded the water temperatures and velocities present within the live-cars at the times smolts were moving downstream. Of the 24,675 Atlantic salmon smolts collected, 48 (0.2%) were classified as trap-caused mortalities. The most frequent causes of death were physical injuries associated with the clogging of traps with debris and overcrowding of the live-cars. Water velocities within the live-cars ranged from 0.09 to 0.57 m/s, and mean velocities never exceeded the published tolerance limits for Atlantic salmon smolts. The seasonal water temperatures within the live-cars (range, 4.2– 20.88C) were not significantly different from those of the water columns (4.3–21.08C) adjacent to the traps. Our analysis suggests that, under typical operations, RSTs represent a minimal threat to Atlantic salmon smolts, and we present methods to further reduce the risk. Rotary screw traps (RSTs; E.G. Solutions, Corvallis, Oregon) are passive-capture devices used for trapping fish moving downstream. The traps consist of a collection cone oriented to capture oncoming flow, an Archimedes screw that dampens the power of the water flow and filters intercepted fishes from the water stream into a live-car (live-box). The live-car retains the fish until they are processed. The rear wall of the live-car contains a cleaning drum that slowly rotates to remove small debris (Figure 1; Chaput and Jones 2004; USFWS 2008). Rotary screw traps have been widely used in studies of juvenile salmonids, including population monitoring and abundance estimates (The- dinga et al. 1994; Bryant et al. 1999; Greenwald et al. 2003), mark–recapture studies (Miller and Sadro 2003), and evaluating the performance of migratory fish (Roper and Scarnecchia 1996; Koed et al. 2006). Over the last decade, our research program has used RSTs as a platform for monitoring Atlantic salmon Salmo salar in the Gulf of Maine (USASAC 2008) and for studying the ecology of smolt emigration (Kocik et al. 2009). We elected to use these traps in 1996 because of their reputation as a safe and effective smolt sampling device (USASAC 1997). Because Atlantic salmon in the Gulf of Maine are now endangered and abundance of naturally produced smolts is extremely low (Fay et al. 2006), research impacts on the population must be minimized. In their review of the status of the U.S. Atlantic salmon population, the National Research Council (NRC) raised concern about the risks associated with smolt trapping with RSTs (NRC 2004). Specifically, the NRC identified the potential for extended periods of exposure to strong currents or high water temperatures in RST live-cars as cause for concern (NRC 2004). We are not aware of any studies or literature that has directly addressed the consequences of RSTs on survival of Atlantic salmon smolts. Chaput and Jones (2004) used RSTs on five New Brunswick (Canada) rivers, all of which have Atlantic salmon populations and reported minimal trapping-related mortality across the fish community. Although these authors document- ed some dead Atlantic salmon smolts (between 0.2% and 0.6% smolt mortality), they did not specifically address the total Atlantic salmon smolt mortality rate or the conditions associated with smolt death. Studies using RSTs to capture Pacific salmon smolts found varying levels of mortality. For example, Pearsons et al. (1998) reported a 0.3% mortality rate (1 dead out of 260) for spring Chinook salmon Oncorhynchus tshawytscha smolts captured in a RST, and Greenwald et al. (2003) estimated 2.2% mortality (17,430 mortalities from a total catch of 811,303) for juvenile Chinook salmon. Knapp et al. (1998) reported RST mortality rates of 2.3–8.6% for freshwater-stage hatchery summer steelhead O. mykiss and age-0 * Corresponding author: paul.music@noaa.gov Received October 30, 2009; accepted February 25, 2010 Published online May 20, 2010 713 North American Journal of Fisheries Management 30:713–722, 2010 Ó Copyright by the American Fisheries Society 2010 DOI: 10.1577/M09-181.1 [Article]