Fish and rearing environment
Approximately 1000 late-stage ‘eyed’ brown trout (Salmo trutta , Linnæus 1758) eggs were obtained in early February 2021 from a stocking hatchery (AE Fishery, Moffat, UK). The eggs were the offspring of a mixture of riverine and lacustrine parents removed to captivity from the wild three to four generations ago. They were transported to the Scottish Centre for Ecology and the Natural Environment (SCENE) on Loch Lomond, UK and approximately evenly distributed between 12 cylindrical 30 L flow-through tanks held at ca. 4°C. Hatching was complete within two weeks of arrival at the centre.
At first feeding (ca. 10–14 days post hatch), to ensure that differences between individuals in aggression, swimming abilities, etc. did not affect group-level results in later behaviour trials, alevins-turned-fry were randomly split into two diet groups and moved to bare flow-through tanks as above, six replicates per diet group, but without temperature control. Ambient water temperature naturally warmed throughout the growth season from ca. 4–21°C. Two diet types of near-identical nutritional value were specially prepared by Garant-Tiernahrung GmbH (Pöchlarn, Austria): one diet was high in n-3 LC‑PUFA (EPA = 3.69%, DHA = 4.95%); the other diet was deficient in n-3 LC‑PUFA (EPA = 0.9%, DHA = 0.89%) (Appendix, Table A1). Fry were fed to satiation twice daily for ca. four months.
For the final twelve weeks of rearing, trout from each diet treatment were further randomly divided between ‘simple’ and ‘complex’ habitats. Simple habitats consisted of bare flow-through cylindrical 90 L tanks with water supplied directly from Loch Lomond at ambient temperatures; 30–40 individuals per tank approximated conventional hatchery densities, which have been shown to be socially simpler than natural densities by inhibiting the development of competitive behaviours (Brockmark and Johnsson, 2010). Complex habitats used the same tank type but were heavily ornamented, with ornaments changing position every second day (i.e. physical complexity); 6–7 fish per tank and weekly part-exchanges of fish between pairs of replicates fostered social complexity by allowing the continuous formation and reformation of dominance hierarchies (Sloman et al., 2000). All trout remained on their original diets and were, for the final twelve weeks, fed once daily to satiation, giving four distinct treatment groups, from which selections were made for further analysis: simple habitat, high n-3 LC-PUFA diet (four replicates); simple habitat, low n-3 LC-PUFA diet (four replicates); complex habitat, high n-3 LC-PUFA diet (eight replicates organised as four pairs part-exchanging inhabitants); and complex habitat, low n-3 LC-PUFA diet (eight replicates organised as four pairs part-exchanging inhabitants) (Table 1; Fig. 1).
Table 1. Treatment groups and replicates at three stages of experimental rearing period of captive brown trout