The influence of abiotic incubation conditions on the winter mortality of wild salmonid embryos

Abstract Embryos of many valued salmonid species incubate in the hyporheic zone of boreal streams over winter. Influence of individual winter‐related environmental variables on salmonid embryo success has been previously investigated. However, how multiple variables act together to influence embryo incubation remains poorly understood. Using a naturally spawning population of Atlantic salmon ( Salmo salar ) in the Miramichi River basin (New Brunswick, Canada), we related variation in the abiotic embryo incubation habitat in different streams (spatial) and over the course of two winters (temporal) to embryo mortality between fertilisation and hatch. Over two years (2013–2014 and 2014–2015), we introduced fertilised eggs to six simulated salmon redds in each of three riffles in each of five active spawning reaches ( n redds = 90) with a range of hyporheic conditions. Embryo mortality was quantified at an early sampling event (March; pre‐freshet and during late embryonic development) and a late sampling event (May; post‐freshet and post‐hatch). We extracted 22 abiotic predictor variables for statistical analyses from continuous records of hyporheic environmental conditions, collected for the duration of the incubation period in each study reach. Through partial least squares regression analyses, 37.6% of the total variation in mortality was explained by the predictor variables. Each group of predictor variables explained similar proportions of variation (water temperature: 8.4%, water level: 7.4%, dissolved oxygen: 7.1%, ice conditions: 7.2%, and substrate characteristics: 7.5%), which suggests that mortality is influenced by multiple interacting abiotic conditions, rather than a single variable in isolation, and that the factors contributing to ideal salmonid incubation habitats are complex and interconnected. Our research highlights the value of a multi‐faceted research perspective and provides a baseline from which future changes in threatened salmonid populations can be measured and compared in an effort to identify relevant species‐ or population‐specific differences.