Fish collection, husbandry and experimental design
Between August 21st and 28th 2018, we sampled perch from the lake Erken (59°50’09.6”N, 18°37’52.3”E) in Central Sweden by angling from littoral and pelagic habitats. We caught additional young-of-the year juveniles by beach seining. Using cooled and aerated boxes, we brought the fish to the aquarium facility at Uppsala University.
Fish were anesthetized using 60 mg L-1 benzocaine and weighted to the nearest 0.01g. We grouped the fish according to their habitat of origin and the body weight at the beginning of the experiment into weight classes: <20 g (juveniles of approximately 4 g), 20-30 g, 30-40 g, and 40-50 g. Perch of the weight class 20-30 g were caught in both littoral and pelagic habitats. Due to potential differences, we analyzed this weight class habitat-specifically. 3-9 perch individuals were placed together in 30l –aquaria (50 x 25 x 25cm) with the bottom covered by a 3cm thick layer of sand. Room temperature of the facility was set at 15 C°. Aquaria were equipped with a flow-through system of fresh tap water. As the tap water became colder during the winter months (November to March), we noticed a drop in the tank water temperatures from 18 C° to 14 °C. We fed the fish daily with commercially available Chironomids, with an amount corresponding to approximately 15% of the individual wet weight –day. Unfortunately, we noticed strong variation in isotopic signatures of Chironomids and were forced to switch suppliers to obtain more stable signatures in perch diet (final food: δ13C: -30.6 ‰ ± 0.5, δ15N: 14.0 ‰ ± 3.7; average ± SD, respectively). Using the new food, we started the feeding experiment to estimate TDFs in perch on October 10th 2018 (day 0). We ended the trials between July 12th (day 275) and 25th 2019 (day 288), i.e. the experiment ran slightly longer than 9 months. At this time, perch had approximately doubled their original body mass on average (132.7 % weight gain ± 89.4, Table 1). This is an approximate value as we did not track the weight increase individually, but set initial weights as the average values of the respective weight class. We assumed perch to be in isotopic equilibrium with the Chironomid diet based on the equations for isotopic half-life of vertebrate muscle tissue and liver reported in Vander Zanden et al. (2015) and the predictive turnover equations presented by Thomas and Crowther (2015). For both predictions, we assumed isotopic equilibrium in 4-5 times the half-live (Thomas and Crowther 2015). Before the fish were sacrificed at the end of the experiment with an overdose of benzocaine, we conducted metabolic trials to obtain individual SMR (see section below). We weighed the sacrificed fish to the nearest 0.01g and measured total length to the nearest mm. We dissected a sample of the dorsal muscle tissue and the entire liver for stable isotope analyses and dried the tissue in a drying oven at 60°C for 48 hours.
Over the course of the experiment, we sacrificed a subset of 5 individuals every 6-10 weeks to observe the development of the TEFs over time. In total, we analyzed the Δ13C and Δ15N in muscle and liver for 48 and 47 individuals respectively. In addition to the fish sacrificed over the course of the experiment, 28 individuals were maintained for the entire 9-month period and SMR measurements were taken for 26 of them.
The study was approved by the Uppsala Animal Ethic Committee with permit number: C59/15.