Despite the common view that conditions in winter strongly influence survival and population size of fish, the ecology of salmonids has not been as extensively studied in winter as in other seasons. In this paper, we review the latest studies on salmonid winter survival, habitat use, movement and biotic interactions as they relate to the prevailing physical and habitat conditions in rivers and streams. The majority of research conducted on the winter ecology of salmonids has been carried out in small rivers and streams, where temperatures are above zero and where there is no ice. Investigations in large rivers, regulated and dredged rivers, and under conditions of different ice formations are almost totally lacking, presumably related to sampling difficulties with these systems. The studies-at-hand indicate that a multitude of physical and biological factors affect the survival, behavior, and habitat use of salmonids in winter. The general concept that winter functions as a critical period for the survival of young salmonids is not well supported by the literature. Instead, overwinter survival of juvenile fish appears to be context-dependent, related to specific habitat characteristics and ice regimes of streams. In general, over wintering salmonids prefer sheltered, low velocity microhabitats, are mainly nocturnal, and interact relatively little with conspecifics or interspecifics. Specific descriptions of microhabitat preferences of salmonids are difficult to make due to highly disparate results from the literature. We suggest that future research should be directed towards (1) being able to predict the dynamics of freezing and ice processes at different scales, especially at the local scale, (2) studying fish behavior, habitat use and preference under partial and full ice cover, (3) evaluating the impacts of man-induced environmental modifications (e. g. flow regulation, land-use activities) on the ecology of salmonids in winter, and (4) identifying methods to model and assess winter habitat conditions for salmonids.

Summer water temperatures are rising in many river systems in North America, and this warming trend is projected to intensify in the coming decades. Cold-water fish may alleviate thermal stress in summer by aggregating in discrete cold-water plumes that provide thermal refuge from high ambient river temperatures. Reliance on cold-water thermal refugia is expected to increase in a warming climate, and many river reaches already lack suitable thermal refugia as a result of an absence of thermal diversity. A comprehensive fish management strategy could proactively address this imminent threat to cold-water fish populations across North America by preserving existing thermal refugia, augmenting thermal anomalies to improve performance as refugia, and creating new thermal refugia in uniformly warm river reaches. We provide practical recommendations on how these measures can be accomplished based on insight derived from recent research focused on the Miramichi River, New Brunswick. Opportunities include limiting land use change, construction aggregate extraction (e.g. sand and gravel pits), and groundwater pumping/consumption. Existing thermal anomalies can be enhanced by controlling advective thermal mixing between cold-water tributaries and the river mainstem flow, installing riparian shading, and adding temporary structures for protection from avian predators. New refugia can be created by temporarily pumping groundwater to discrete points within the river during periods of thermal stress. These concepts are discussed in the context of a comprehensive thermal refugia management strategy. Copyright

Anthropogenic influences, including climate change, are increasing river temperatures in northern and temperate regions and threatening the thermal habitats of native salmonids. When river temperatures exceed the tolerance levels of brook trout and Atlantic salmon, individuals exhibit behavioural thermoregulation by seeking out cold‐water refugia – often created by tributaries and groundwater discharge. Thermal infrared (TIR) imagery was used to map cold‐water anomalies along a 53 km reach of the Cains River, New Brunswick. Trout and salmon parr did not use all identified thermal anomalies as refugia during higher river temperature periods (>21°C). Most small‐bodied trout (8–30 cm) were observed in 80% of the thermal anomalies sampled. Large‐bodied trout (>35 cm) required a more specific set of physical habitat conditions for suitable refugia, that is, 100% of observed large trout used 30% of the anomalies sampled and required water depths >65 cm within or adjacent to the anomaly. Densities of trout were significantly higher within anomalies compared with areas of ambient river temperature. Salmon parr were less aligned with thermal anomalies at the observed temperatures, that is, 59% were found in 65% of the sampled anomalies; and densities were not significantly different within/ outside anomalies. Salmon parr appeared to aggregate at 27°C, and after several events over 27°C variability in aggregation behaviour was observed – some fish aggregated at 25°C, others did not. We stipulate this is due to variances of thermal fatigue. Habitat suitability curves were developed for velocity, temperature, depth, substrate, and deep water availability to characterize conditions preferred by fish during high‐temperature events. These findings are useful for managers as our climate warms, and can potentially be used as a tool to help conserve and enhance thermal refugia for brook trout and Atlantic salmon in similar systems.

Climate change is expected to increase the frequency and magnitude of extreme thermal events in rivers. The Little Southwest Miramichi River (LSWM) and the Ouelle River (OR) are two Atlantic salmon (Salmo salar) rivers located in eastern Canada, where in recent years, water temperatures have exceeded known thermal limits (~23°C). Once temperature surpasses this threshold, juvenile salmon exploit thermal heterogeneity to behaviourally thermoregulate, forming aggregations in coolwater refuges. This study aimed to determine whether the behavioural thermoregulation response is universal across rivers, arising from common thermal cues. We detailed the temperature and discharge patterns of two geographically distinct rivers from 2010 to 2012 and compared these with aggregation onset temperature. PIT telemetry and snorkelling were used to confirm the presence of aggregations. Mean daily maximum temperature in 2010 was significantly greater in the OR versus the LSWM (p = 0.005), but not in other years (p = 0.090–0.353). Aggregations occurred on 14 and 9 occasions in the OR and LSWM respectively. Temperature at onset of aggregation was significantly greater in the OR (Tonset = 28.3°C) than in the LSWM (Tonset = 27.3°C; p = 0.049). Logistic regression models varied by river and were able to predict the probability of aggregation based on the preceding number of hours >23°C (R2 = 0.61 & 0.65; P50 = 27.4°C & 28.9°C; in the OR and LSWM respectively). These results imply the preceding local thermal regime may influence behaviour and indicate a degree of phenotypic plasticity, illustrating a need for localised management strategies.

River temperatures often surpass the thermal limits of juvenile Atlantic salmon (Salmo salar). Using thermal monitoring data to replicate a natural heat event, we investigated how cooler nighttime temperatures would affect how juvenile salmon cope with several days of heat cycling.

Chromosomal inversions have been implicated in facilitating adaptation in the face of high levels of gene flow, but whether chromosomal fusions also have similar potential remains poorly understood. Atlantic salmon are usually characterized by population structure at multiple spatial scales; however, this is not the case for tributaries of the Miramichi River in North America. To resolve genetic relationships between populations in this system and the potential for known chromosomal fusions to contribute to adaptation we genotyped 728 juvenile salmon using a 50K SNP array. Consistent with previous work, we report extremely weak overall population structuring (Global FST = 0.004) and failed to support hierarchical structure between the river's two main branches. We provide the first genomic characterization of a previously described polymorphic fusion between chromosomes 8 and 29. Fusion genomic characteristics included high LD, reduced heterozygosity in the fused homokaryotes, and strong divergence between the fused and the unfused rearrangement. Population structure based on fusion karyotype was five times stronger than neutral variation (FST = 0.019) and the frequency of the fusion was associated with summer precipitation supporting a hypothesis that this rearrangement may contribute local adaptation despite weak neutral differentiation. Additionally, both outlier variation among populations and a polygenic framework for characterizing adaptive variation in relation to climate identified a 250 Kb region of chromosome 9, including the gene six6 that has previously been linked to age-at-maturity and run-timing for this species. Overall our results indicate that adaptive processes, independent of major river branching, are more important than neutral processes for structuring these populations..

For the past 22 years, we have monitored hydro‐meteorological conditions and fish population dynamics in Catamaran Brook, a 52 km2 catchment in the Miramichi River system of New Brunswick, Canada. Given the long‐term nature of the multidisciplinary dataset, we are able to provide an overview of the complex interaction between streamflow and fish population dynamics drawing on previously published material as well as new data analyses. For autumn‐spawning fishes like the Atlantic salmon, access to headwater reaches was directly related to streamflow during late October (when they ascend spawning tributaries), as well as the propensity of beaver dams in the stream (also a function of streamflow). Winter streamflow was positively correlated with egg survival, except when rain‐on‐snow conditions induced severe ice break‐up events that likely caused the highest mortalities on record for salmon and other fishes. Juvenile recruitment was significantly influenced by density‐dependent processes of growth and competition but further mediated by density‐independent factors like winter flow. In spring, fry emergence was largely temperature‐driven, although peaks in fry drift were sometimes synchronized with secondary discharge peaks and temperature. Tributaries like Catamaran Brook provide thermal refugia for coolwater fishes like salmon and trout during the summer when wide, shallow main‐stem Atlantic rivers experience low discharge and high water temperatures that induce physiological and behavioural stress (i.e. > 23 °C). These phenomena are discussed in detail, especially in terms of how they may be compromised by future changes in hydrologic conditions resulting from predicted climate change scenarios. Copyright © 2012 John Wiley & Sons, Ltd.

Active tracking of passive integrated transponder (PIT)-tags using portable antennae is becoming an increasingly common technique in fish habitat studies in shallow rivers. We carried out ''blind testing'' to test the efficacy (% tags found) and accuracy (distance between predicted and true tag location) of a portable antenna system (Texas Instruments) in winter conditions using 23-mm PIT-tags. Up to 90 cm reading range was achieved and signals penetrated ice, rock, wood and water. In the ''blind test'' trials, a majority of the hidden tags (N = 12-30) were found indicating high tracking efficacy. PIT-tags that were oriented with their cylindrical axis parallel to the plane of the antenna coil inductor loop resulted in a bimodal detection field that had low detection range in the centre of the loop. The utilization of this bimodal detection field proved to be a very accurate method for identifying tag position (mean ± SE distance from predicted to true tag location 10.9 ± 1.4 cm) and thus well suited for microhabitat and activity studies in winter conditions. Aggregations of tags (multiple tags within 1 m 2 ) and obstacles for the antenna maneuvering (e.g., boulders, logjams) reduced the pinpointing accuracy (mean ± SE 13.3 ± 1.8 cm), but the reduction in accuracy was statistically nonsignificant between the single and aggregated tags.