Biological Significance of Thermal Refugia for Juvenile Atlantic Salmon in a Changing Climate

Atlantic salmon (Salmo salar L.) parr show a behavioural and physiological stress response when water temperature exceeds 23 ^oC. Various climate change scenarios for eastern Canada predict that such water temperature thresholds will be exceeded more often. In the current study, we examined i) the threshold conditions for initiating behavioural thermoregulation of salmon parr in situ and  ii) how the incidence of temperature stress events and proximity to thermal refugia affect the distribution and abundance of wild, juvenile salmon in a large (>60m wide) river. To answer these questions, the movements and behaviour of 635 individually Passive Integrated Transponder (PIT) tagged 1+ and 2+ Atlantic salmon parr was monitored during two field seasons (summers 2009 and 2010) in the Little Southwest Miramichi River, NB, Canada. The monitoring of movements was carried out in a 11 km river segment where various cold water refugia (seeps, tributary plumes) were identified (from airborne thermal remote sensing) and in relation to a river reach with no available cold water refugia present. In 2009, no salmon parr aggregations were observed in cold water refugia despite maximum daytime water temperatures exceeding 24 ^oC for seven consecutive days in mid-August (max. 26.1 ^oC); night-time temperatures during the corresponding period were < 23 ^oC. In 2010, large salmon parr aggregations were observed (involving 46 % of the PIT tagged fish) in early July when water temperature exceeded 27 ^oC in daytime (max. 31 ^oC) and remained >23 ^oC at night. Concurrently, large scale mortality was apparent in all salmon age classes. Salmon parr aggregations remained in place for three consecutive days when water temperature consistently exceeded ~23  ^oC. Aggregations of 0+ Atlantic salmon were also observed when the daytime water temperature exceeded 30 ^oC. Most PIT tagged parr found in refugia were marked within the same river reach (within 1 km); the longest movements in response to high water temperatures were in excess of 10 km. River reaches with no cold water refugia were void of tagged parr during the high temperature events and tagged fish from these areas were observed ~ 5 km up- and downstream. The dataset will be subsequently anylysed with GIS tools to identify locations where artificial cold water refugia should be situated to best mitigate the effects of warming climate on juvenile Atlantic salmon.