Fish sampling and data collection
In January 2019 brown trout were captured in Scotland, UK, at Loch Sloy (56.2632175°N, 4.7707667°W; N = 14; FL = 261±52 mm mean±SD) and Carron Valley Reservoir (56.0338314°N, 4.1057406°W; N = 14; FL = 227±37 mm) using 30 m × 1.5 m single-mesh (38 mm) benthic gill nets. Three nets were deployed in late afternoon and retrieved the following morning. Loch Sloy has surface area of 1.33 km2 and mean depth 25.5 m. Carron Valley Reservoir has surface area 3.76 km2 and mean depth 9.6 m. Fish community in Loch Sloy is composed of brown trout, European whitefish Coregonus lavaretus , and European eel Anguilla Anguilla. Carron Valley Reservoir contains the same fish species as Loch Sloy complemented by rainbow trout Oncorhynchus mykiss , three-spined stickleback, and perch Perca fluviatilis . In October 2019 brown trout were captured by electrofishing (e-fish, UK) in small tributaries of Loch Lomond, Scotland, UK (56.0470272°N, 4.5504428°W): Ross Burn (N = 12, FL = 153±16 mm) and Wood Burn (N = 3, FL = 165±42 mm). Ross Burn has a mean discharge 0.3 m3.s-1 and length 2.3 km and Wood Burn has a mean discharge 1.1 m3.s-1 and length 3.1 km. Fish communities in Ross Burn and Wood Burn are the same, composed of brown trout, brook lamprey Lampetra planeri , European eel, and occasional Atlantic salmon Salmo salar and European minnowPhoxinus phoxinus . Finally, in January 2020 hatchery young-of-year brown trout (AE Fishery, Moffat, UK) were transported to the Scottish Centre for Ecology and the Natural Environment (SCENE). Hatchery trout were held in unadorned 120 L cylindrical tanks at a density of 20 fish per tank. Tanks were fed individually with water on a flow-through system directly from Loch Lomond at natural temperatures (low of ~3° C in winter to high of ~18° C in summer). Inflows of ~100 L per hour were angled to provide a current and an air stone was added to each tank. Lighting simulated ambient sunlight at the latitude of the facility (~56° N). Hatchery trout were fed daily to satiation on commercial salmon pellets containing 31.3 % of fish oil (Ewos Ltd., UK). These commercial pellets are a standard diet of hatchery trout and contain high amount of n-3 LC-PUFA (Heissenberger et al. , 2010). Hatchery fish were held under these conditions until June 2020, when brain samples of randomly selected individuals were extracted (N = 15, FL = 178±22 mm). All fish were killed with an overdose of benzocaine. Heads of fish were removed and fixed in 4% buffered (pH 6.9) paraformaldehyde solution. Brains were then dissected out by opening the skull along the anteroposterior axis and removing muscle tissue and bones around the brain until the brain could be lifted from the skull. Dissected brains were stored in 4 % buffered paraformaldehyde until further procedures were conducted. Brains were photographed with a Canon EOS 1300D DSLR camera with an EF-S18-55 III lens (Canon) and 13- and 31-mm extension tubes designed for Canon DSLRs (Xit Inc.). For each dissected brain sample, an image was taken from dorsal, left lateral and ventral views. Each brain was measured to calculate total volume and the volumes of the cerebellum, optic tectum, telencephalon, olfactory bulb, and hypothalamus. Measurements were completed using ImageJ 1.48 (Schneider et al. 2012) and used to calculate volume with the formulas outlined by Pollen et al. (2007).
We performed geometric morphometric analysis of body shape to confirm the predominant habitat use of wild caught individuals (see Appendix). This analysis indicated clear differences in body shape of individuals caught in lake and stream habitat (Procrustes ANOVA: F2,64 = 11.6, p = 0.001), which suggests that the presence of individuals in the sampled habitat (i.e. , lake or stream) was not coincidental, and it corresponds to their long-term habitat preference.