2.2.4 Long-term depuration experiments: examining the allocation of released body carbon by labeling fish bodies with14C
To examine the release (/efflux) rate of fish body carbon and the allocation of released fish body carbon as DOC, CO2, and PC, the marine medaka were 14C-labeled by feeding the fish14C-labeled rotifers for 15 days (0.7–1.1 Bq/µg C). Then, three to five fish were collected for 14C measurement in fish at time zero, and other fish were individually transferred into 50-mL feeding beakers containing 25 mL clean FSW for 8-day depuration. During the depuration period, the fish were fed nonradioactive rotifers at two rations of1000 and 2000 ind./fish/day, corresponding to 2.2% and 4.3% of the fish DW, respectively. These rations are within the typical range of daily food consumption for wild mesopelagic fish (Davison et al., 2013). During depuration, fish, feces and water samples were collected every 24 h for the measurement of the14C retained in fish and released into seawater in the 2000 ind./fish/day ration treatment. The samples were collected on days 1, 3, 6, and 8 in 1000 ind./fish/day ration treatment. All the samples were handled and measured by the same methods as described for the short-term depuration experiments.
The release (/efflux) rate of fish body carbon was calculated as the slope of the linear regression of the natural logarithms of the14C retained in fish with the time of depuration. Only the data on the second day and afterward were used for the regression. Based on the mass balance, the apparent body carbon release (B 14C release) was calculated as the difference between the amount of 14C in fish at the beginning and the 14C retained in the fish at the end of depuration. The sum of the collected14CO2, DO14C and P14C during depuration was taken as the measured body carbon release (B 14Crelease). The ratio ofB 14Crelease/B 14C releasewas used to indicate the 14C recovery of the total released body 14C in seawater.
2.3Scenarios for e stimating carbon release from mesopelagic fish in the global open ocean
Carbon release model scenarios were separately established for four groups of mesopelagic fish classified on the basis of their ocean latitudes (40°N–40°S vs. 40–70°N/S) and fish behavior (diel vertically migrating (DVM) vs. nonmigrating (NM)) (Table 1). The carbon release parameters of marine medaka were extrapolated to wild mesopelagic fish. Other parameters, such as the mesopelagic fish biomass and the daily food intake of mesopelagic fish, were derived from the literature.
We assumed that the wild fish would release the same proportions of their daily food carbon to seawater in the forms of DOC, CO2, and PC as marine medaka. The food carbon release parameters measured after 24-h depuration were used in the model scenarios for two reasons. First, the release rates of DOC, CO2, and PC from food decreased to relatively constant low values during the 16–24 h depuration phase, indicating that 24 h or less was enough for the fish to completely allocate the ingested food carbon to assimilation or food carbon release. Second, the feeding behavior of DVM mesopelagic fish usually has a diel rhythm. Therefore, the mean proportions of ingested food allocated to AE (38.9%) and released as DOC (32.7%), CO2 (20.9%), and PC (7.5%) during 24-h depuration were used in the model scenarios (Table 1).
The body carbon release (/turnover) rate (K e) of mesopelagic fish was extrapolated from the carbon release (/efflux) rate of marine medaka by using the power-law function reported in Gillooly et al. (2001) and Davison et al. (2013). That is, K e is proportional to 1924 ×WW 0.75 × e(-5020/K ), where WW is the wet weight of fish (g), and K is the absolute temperature. Based on this formula, theKe s of 0.5-g mesopelagic fish living at different temperatures could be derived from the K e of marine medaka. The K e (0.053 d-1) of marine medaka at the daily food ration of 4.3% fish DW was used as the basis for estimating the K es of mesopelagic fishes (0.5 g in WW) (Table 1). Following the pattern of marine medaka, 40.4%, 42.6% and 16.9% of the released body carbon of wild mesopelagic fish was assumed to be in the forms of DOC, CO2, and PC, respectively. Carbon released through reproduction was not considered in the current estimation.
Variations in the body carbon release rate during different activities, such as swimming, feeding and resting, were not considered in the present estimation. As a simplification, DVM mesopelagic fish were assumed to feed and live in surface waters at night (for 12 h) and inhabit mesopelagic depths during the daytime (for 12 h). NM mesopelagic fishes were assumed to feed and live at mesopelagic depths all day. That is, the DVM mesopelagic fish have two differentK es, depending on the mean temperatures of the surface waters and of the mesopelagic waters. In contrast, the NM mesopelagic fish have only one K e, depending on the mean temperature of the mesopelagic waters.
The biomass of the mesopelagic fish in the open ocean was assumed to be constant over time. The total WW of the mesopelagic fish in the open ocean between 40°N and 40°S was assumed to be 109–1010t, and the WW of the mesopelagic fish in other regions between 40°N–70°N and 40°S–70°S was assumed to be 0.3×109–1010 t (Lam & Pauly, 2005; Irigoien et al., 2014). Furthermore, 30%–50% of the mesopelagic fish were assumed to undergo diel vertical migration (Davison et al., 2015; Klevjer et al., 2016).
For mesopelagic fishes living in the open ocean, the ratio of fish DW to WW was 19.1%, and the carbon content was 43.8% of the fish DW (Childress & Nygaard, 1973), which is similar to that (46.7%) of marine medaka. The mean WW of individual fish was assumed to be 0.5 g (Davison et al., 2013, 2015).
The mean temperatures in the surface and mesopelagic waters in the open ocean between 40°N and 40°S were assumed to be 25°C and 9°C, respectively (Davison et al., 2013; Irigoien et al., 2014), whereas the mean temperatures in the surface and mesopelagic waters at high latitudes, 40°N–70°N and 40°S–70°S, were assumed to be 8°C and 3°C, respectively (Kaeriyama & Ikeda, 2004; Max et al., 2012). The daily food ration for a 0.5-g mesopelagic fish was assumed to be temperature-dependent according to Davison et al. (2013), i.e., the daily food ration was 10%, 5%, 5%, and 4% of the fish WW at 25°C, 9°C, 8°C, and 3°C, respectively. All mesopelagic fishes were assumed to be zooplanktivorous, and the carbon content of zooplankton was assumed to be 0.12 mg C/mg WW (Harris et al., 2000).
The annual rate of carbon release in each form was calculated by multiplying the daily rate by 365.