Abstract
Brown trout ( Salmo trutta L.) are ecologically and economically important, but like many salmonids their populations are declining. Trout exhibit a high level of phenotypic plasticity, ranging from freshwater residency to anadromy. This complexity causes problems for assessment and constrains application of adult catch‐based assessment methods that are commonly applied to other species such as Atlantic salmon. Juvenile methods provide one of the most promising approaches for the assessment of trout that can quantify the status of populations, identify pressures and guide evidence‐based management actions. This study collated and analysed ad‐hoc multi‐pass electrofishing data from across Scotland to model relationships between fish density and habitat, having accounted for the effects of variable capture probability. Given a lack of consistent data collection and recording, the habitat was characterised by landscape proxies derived from spatial data. The density model was used to derive expectations for healthy trout populations (i.e., a benchmark) that could be used for assessments across spatial scales. Capture probability exhibited a negative gradient between the south‐west and north‐east of Scotland, was generally higher for parr (>0+) than fry (0+), and higher on the first than subsequent passes. Capture probability increased over the summer for fry before levelling off in the autumn. The positive day of the year (DoY) effect was smaller and more linear for parr. Capture probability decreased with upstream catchment area (UCA), with the negative response being stronger for parr than fry. Altitude had a small negative (logistic)‐linear effect, which was common to both Lifestages. Trout densities (both fry and parr) were generally lower in the west and north of Scotland than in the east. Densities decreased non‐linearly with UCA, with stronger responses for fry than parr. Trout fry densities also declined strongly with altitude. However, trout parr densities exhibited a weak positive relationship with altitude and decreased with river distance to sea (RDS). There was no effect of RDS on fry densities. Fry and Parr densities both decreased with DoY. Two possible benchmarks were produced for each Lifestage. The first included all fixed effects. The second excluded regional effects on the assumption that the underlying abundance data contained spatial biases reflecting uncharacterised pressures. When combined with a suitable formal survey design, these benchmarks can support assessment of the status of juvenile trout populations in Scotland for the first time.