The potential of water spray to attract Atlantic salmon kelt (Salmo salar) in a regulated river

Sammanfattning

This study investigates the downstream migration challenges faced by Atlantic salmon (Salmo salar) during the post-spawning kelt life stage at hydropower dams. Surviving kelts can live on to spawn multiple times, and those repeat spawners can be of great importance because they can increase recruitment and population resilience. However, kelt migrating downstream are unlikely to survive passage through the turbines of a hydropower dam. In this study, sonar and acoustic telemetry were employed above the Stornorrfors hydropower dam to evaluate the effectiveness of water spray as a potential attractant for kelt. If successful, such a method could potentially be used to facilitate methods to either capture or guide kelt towards safer passages. The findings of our study only partially support the hypothesis that surface spraying would attract salmon kelt. Sonar data, which do not allow differentiation between species, revealed that larger fish exhibited longer residence times near the active spray. While this indicates that water spray might hold the attention of larger fish, further results were more nuanced. Specifically, a significant decrease in fish count when the spray was on during twilight hours suggested a possible repellent effect. While such a decrease was not observed during daylight, the daylight data did not indicate a significant positive effect of water spray on fish count. Telemetry data revealed a non-directional increase in swimming speeds when the spray was active, which could potentially also be linked to a repellent effect. Notably, the pump feeding the water spray was found to produce a great amount of noise, potentially triggering a flight response. Furthermore, the mean delay time from first to last detection was 19.2 minutes, which is considerably shorter than reported by previous studies (> 20 h), questioning the impact of overwintering conditions, physiological factors, and environmental conditions. These short delay times in effect lead to a decreased sample size, limiting the generalisability of the results. However, our results serve to increase the understanding of kelt behaviour in the intake channel, highlighting locations where future measures could be best implemented, and suggesting how experimental setups could be improved in the future.