The primary purpose of this project is to acquire long-term data series ontemperature of selected lakes to support management of nursery habitat of lakerearingjuvenile sockeye salmon (Oncorhynchus nerka) in relation to climatechange. We adopted protocol developed by the National Park Service (NPS) toestablish moored all-season vertical temperature monitoring arrays in eight lakesof Kodiak, Togiak, and Alaska Peninsula/Becharof National Wildlife Refuges(NWR) in summer and fall 2011. We recorded lake temperature at a resolution of0.02°C on an hourly basis at various depth strata between lake surfaces and lakebottoms. Monitoring sites were visited annually or biannially to extract data andto service monitoring equipment. A custom database utility was developed tofacilitate data organization, basic analysis, and archiving. Initial evaluation oftemperature data indicated that lake study sites conformed to a dimictic pattern ofstratification and isothermy at Kodiak NWR and Togiak NWR. Although studysites exhibited similar overall patterns of interannual temperature variation, theydiffered in magnitude of variation during the summer-fall stratification period,which we attributed mainly to differences in lake volume. At Kodiak Island,variation in monthly mean air temperature was related to variation in monthlymean lake temperature at the 5 m depth strata. Annual mean air temperature was -1.8°C below the long-term (1981-2010) mean (i.e. normal) between Oct 2011 andSep 2012 but near normal (0.3°C) between Oct 2012 and Sep 2013. However,seasonal differences were more pronounced. Prolonged ice cover of study siteswas facilitated by below normal mean air temperature (-2.5°C) during Dec-Mar2011-12. Conversely, above normal mean air temperature (1.3°C) during Jun-Seplikely triggered periodic exceedance of the Alaska Department of EnvironmentalConservation’s (ADEC) 15°C water temperature standard at1, 5, and10 m depthstrata between late-Jul and early Aug 2013. During this same period, we observeda rate of temperature change of up to 1°C/m between 5 and 10 m depth strata ofRed Lake, which possibly indicated thermocline development. We conclude withrecommendations for protocol refinement, measures to increase understanding oftemporal and spatial variation in temperature estimates, and an interagencyprocess to facilitate cooperative analysis of long-term series of lake temperaturedata for nursery lakes of sockeye salmon in the southwest Alaska region.
