Field-Based Estimates of Thermal Tolerance Limits for Trout: Incorporating Exposure Time and Temperature Fluctuation KEVIN E. WEHRLY* AND LIZHU WANG Institute for Fisheries Research, Michigan Department of Natural Resources, and University of Michigan, 212 Museums Annex, Ann Arbor, Michigan 48109-1084, USA MATTHEW MITRO Wisconsin Department of Natural Resources, Science Operation Center, 2801 Progress Road, Madison, Wisconsin 53716-3339, USA Abstract.—We used temperature and fish data from streams across Michigan and Wisconsin to estimate upper thermal tolerance limits for brook trout Salvelinus fontinalis and brown trout Salmo trutta. Tolerance limits were estimated for the maximum daily mean temperature (MEANT), maximum daily maximum temperature (MAXT), and maximum daily temperature range (RNGT) at exposure lengths of 1, 3, 7, 14, 21, 28, 35, 42, 49, 56, and 63 d. We found no difference in the upper thermal tolerance limit for brook and brown trout. For time periods of 1–14 d, the upper temperatures tolerated by trout decreased rapidly from 25.38C to 22.58C for MEANT and from 27.68C to 24.68C for MAXT. For time periods from 21 to 63 d, the upper temperatures tolerated by trout declined more gradually from 22.18C to 21.08C for MEANT and from 24.28C to 22.98C for MAXT. The 7-d upper tolerance limit was 23.38C for MEANT and 25.48C for MAXT. The maximum RNGT tolerated by trout varied as a function of mean temperature and length of exposure. Our findings suggest that chronic temperature effects as well as temperature fluctuation play an important role in limiting salmonid distributions and therefore should be considered when developing management objectives and water quality standards. Water temperature is an important factor shaping the distribution and abundance patterns of stream fishes, especially salmonids (Binns and Eiserman 1979; Bowlby and Roff 1986; Stoneman and Jones 2000; Wehrly et al. 2003). Temperature affects fishes directly by controlling rates of metabolism, feeding, and growth (Brett 1971; Elliott 1981) and indirectly by affecting prey availability (Hinz and Wiley 1998) and mediating competitive interactions (De Staso and Rahel 1994; Taniguchi et al. 1998; Reese and Harvey 2002). Because temperature is a major determinant of habitat suitability, estimating the thermal tolerance of fishes is of considerable value to resource managers. Thermal tolerance limits can be used to prioritize restoration efforts, to assess risks associated with human-induced changes in water temperature (Eaton and Scheller 1996; Keleher and Rahel 1996), and to develop water- quality criteria to protect fishes from elevated temper- atures (U.S. EPA 1976). Thermal tolerance of fishes traditionally has been estimated by using laboratory methods to determine values such as the critical thermal maximum (CTM) or the upper incipient lethal temperature (UILT). Al- though these methods have been used extensively to develop temperature protection standards, the applica- bility of laboratory-based thermal tolerance limits to fishes in natural settings has been an issue of growing concern (Dickerson and Vinyard 1999; Selong et al. 2001; Johnstone and Rahel 2003; Schrank et al. 2003; Meeuwig et al. 2004). This concern stems, in part, from the use of unnatural thermal regimes such as rapidly increasing (CTM) or constant temperatures (UILT) in the determination of upper lethal limits. Because streams typically exhibit diel temperature fluctuations, laboratory-based tolerance values may not reflect the thermal stress experienced by fishes in nature. An alternative approach for estimating thermal tolerance is to use field observations to identify temperatures that correspond to the limits of a species’ distribution (Eaton et al. 1995; Welsh et al. 2001; Huff et al. 2005). For example, Eaton et al. (1995) used a national dataset of temperature and fish presence to estimate the thermal tolerance of 30 species by determining the 95th percentile of the maximum weekly mean temperature where each species was found. Field-based approaches are attractive because they rely on the thermal regimes experienced by fishes in nature. In addition, because these approaches are based on a species’ realized thermal niche (Magnuson et al. 1979), they account for variation in other factors (e.g., prey availability, behavioral thermoregulation, * Corresponding author: wehrlyke@michigan.gov Received July 7, 2006; accepted November 20, 2006 Published online March 8, 2007 365 Transactions of the American Fisheries Society 136:365–374, 2007 Ó Copyright by the American Fisheries Society 2007 DOI: 10.1577/T06-163.1 [Article]