. In addition, the 99-L tanks used for metabolic measurements of warm acclimated fish were fitted with glass aquarium heaters to maintain constant acclimation temperatures while cold 99-L tanks were submerged within an 850-L sea table with continuous flow of ambient seawater to maintain lower temperatures. Fish were placed in respirometry chambers with flush pumpsrunningfor10–12 hpriortodeterminationofoxygenconsump-tion rates. Oxygen consumption measurements were collected over a 20 min interval followed by a 5 min flush cycle to re-oxygenate therespirometry chamber. Following the adjustment period to the respi-rometrychamber,respirationrates(ṀO2) weremeasuredcontinuouslyoverathree-hourperiodatthesametimeeachday.Oncenodiscernible chamber effect could be observed, as indicated by no signifi cant changes in oxygen consumption over a 1 h period and r2 values of >0.95 for the slope describing the rate of oxygen consumption, mean ṀO2 values were calculated by averaging five sequential measure- ments. Additionally, values whose slope deviated from an r2 of >0.95 were excluded, as they are likely indication of fish activity in the cham- ber durin g that particular measurement period. Mass specific oxygen consumption rates were standardized to a 100-g fish ( Steffensen, 2005; Robinson and Davison, 2008a,b ) using a mass exponent of −0.25 ( Schmidt-Nielsen, 1984 ). As we did not have enough individuals to experimentally determine the mass-exponent for our study species,we chose to be conservative with our measurements and utilize the more general model of − 0.25 as a mass exponent. A study performed by Clarke and Johnston (1999) indicated that a wide range of teleost fish, whether polar or temperate, scale similarly.