Validity of winter sampling for estimation of salmonid abundance by electrofishing

Electrofishing removal sampling is a common method for abundance estimation of juvenile salmonids in streams and is usually performed during autumn. European guidelines specify that sampling should be conducted at water temperatures ≥ 5 °C, but studies investigating the validity of sampling below this threshold are lacking. We conducted electrofishing experiments during autumn and winter in three Norwegian rivers, comparing abundance estimation accuracy, rate of fish emigration from electrofishing sites, and mortality for juvenile Atlantic salmon Salmo salar and brown trout Salmo trutta. Abundance estimates were significantly more accurate during autumn than during winter, less accurate for age-0 fish than for older parr, and more accurate when the number of electrofishing passes was increased. By digging through the substrate after electrofishing, we found that a higher proportion of fish were uncatchable during winter than during autumn, and that age-0 fish were more likely than older fish to be uncatchable. For salmon older than age-0, both autumn and winter sampling consistently yielded relatively accurate abundance estimates. However, the rate of fish emigration from electrofishing sites was significantly higher during autumn than during winter and if block nets are not used, emigration during autumn sampling could cause severe underestimation of Atlantic salmon abundance. Fish mortality did not differ between autumn and winter sampling and was low for both species and all age groups. We recommend the use of block nets and conducting five electrofishing passes to reduce the likelihood of severe estimation error.