Gas chromatography and mass spectrometry
Whole brains and samples of dorsal muscle tissue were removed immediately upon death from the observers of the remaining triads (i.e. 30 trout), also representing all combinations of treatment group, flash frozen in liquid nitrogen, then freeze-dried and stored at −80 °C to limit lipolytic degradation. The immediacy of this method of preservation was not compatible with the brain-measuring procedure described above. Samples of food sources were included with tissue samples in subsequent fatty acid analyses.
Lipid extraction and esterification from freeze-dried samples followed the protocol described by Pilecky et al. (2023). Briefly, whole brains (ca.  10 mg) and ca.  30 mg of muscle and food source samples were weighed to the nearest 0.01 mg, homogenised and stored in chloroform (2 mL) under N2 gas overnight at −80 °C. With the addition of 1 mL MeOH and 750 μL of 0.9 % NaCl solution, samples were repeatedly sonicated, vortexed and centrifuged to remove non-lipid materials. Solvent was fully evaporated from extracted lipids and 2 mL chloroform was added under N2. Gravimetry of aliquots was performed as a measure of total lipids for each sample.
Neutral and polar lipids were separated using BondElut Ultra Inert GC columns (Agilent Technologies, Inc., Santa Clara, CA, USA). Column equilibration was conducted by passing hexane through each column; then, 10 mg of lipids-chloroform solution were loaded. Neutral lipids (NL) were isolated with 4 mL of 2:1 chloroform:isopropanol and evaporated fully. Then, to remove free fatty acids, 4 mL of 2 % acetic acid in diethyl ether was run through the columns and discarded. Polar lipids (PL) subsequently eluted with 4 mL of methanol and evaporated fully.
Fatty acid methylated esters (FAME) of the extracted NL and PL were formed by incubation with 1 % H2SO4 in MeOH for 16 h at 50 °C, followed by the addition of 2 mL of 2 % KHO3 and 2 mL hexane, following Pilecky et al. (2023). Samples were mixed and centrifuged, and the upper organic layer of each sample was collected and concentrated under N2gas.
Gas chromatography (TRACE GC 1310, Thermo, Waltham, MA, USA) of FAME followed the protocol established by Pilecky et al. (2023), using external standards for calibration; concentrations were reported as mg g-1 dry weight. Data involving candidate n‑3 and n‑6 FAME were carried forward for further investigation.
For fatty-acid-specific stable isotope ratio mass spectrometry (DELTA V Advantage, ThermoFisher Scientific, Waltham, MA, USA), the gas chromatograph was coupled via CONFLO IV (Thermo, Waltham, MA, USA). Samples were run against certified Me-C20:0 stable isotope reference material (USGS70: δ 2H = −183.9 ‰, USGS71:δ 2H = −4.9 ‰ and USGS72:δ 2H = +348.3 ‰) and corrected for methylation, as described elsewhere (Pilecky et al., 2023). Food sources were used further to correct the δ 2H signature of each FAME in each sample:
Δδ 2HFAME = sampleδ 2HFAME – mean(food sourceδ 2HFAME).
Depletion of Δδ 2H in successive FAME in the biosynthesis pathways indicates bioconversion of shorter chain PUFA rather than a dietary source of LC-PUFA (Pilecky et al., 2022).