Metabolic rate and cortisol measurements
We measured basal metabolic rate (oxygen consumption) of 55 adults (14 parental enriched; 13 parental poor, eight offspring from enriched parents reared in enriched environment; ten offspring from enriched parents reared in poor environment; five offspring from poor parents reared in poor environment; five offspring from poor parents reared in enriched environment), 30 of which were also analysed for epigenetic variation (Table S1). Basal metabolic rates were measured at eight months of age using four identical close respirometers consisting of 120 ml sealed dark chambers filled with oxygen saturated autoclaved water. Two blank trials were carried out to confirm no leakage of oxygen. Oxygen levels were calibrated in trials using saturated oxygenated water (100% dissolved oxygen) and anoxic water (2% dissolved oxygen). Fish were fastened for 48 hrs prior to acclimation for 20h. Oxygen consumption was measured once for each fish for 40 minutes after acclimation, with oxygen levels always above 60%. Chambers were drained and cleaned between runs. Basal metabolic rate was calculated taking into account the rate of oxygen decrease in the chamber, mass of the individual, volume of water and time of measurement (mg O2 g-1 min-1). Averaged background respiration levels across runs was 12.34% (sd=±9.71).
We used ultrasensitive graphene immunosensors (Barton et al. 2018) for measuring waterborne cortisol non-invasively from parents reared in both enriched and poor environments. For this, 120ml of water were taken from the respirometer after each individual measurement of metabolic rate and kept at -80 °C until the analysis. A total of 10ml were centrifuged at 1000rpm for 5 minutes, and 10ul of the supernatant were pipetted onto the modified sensor surface. Electrochemical measurements were conducted with a potentiostat/ galvanostat (Autolab), controlled with NOVA software as in (Barton et al.2018).