An assessment of metrics to determine a rapid, benign method of classifying the metabolic physiology phenotype of individual fish, using dusky kob (Argyrosomus japonicus) as a case study

Abstract

Climate change is a major driving factor in the redistribution of species across the globe. The overall global trend of ocean warming has already driven fish distributions to higher latitudes and deeper depths. This shift in fish distribution can have several negative implications on coastal communities who rely on these fish as a source of food and income. Similarly short-term thermal variability in the coastal ocean, largely driven by increases in the frequency and intensity of upwelling events and marine heatwaves, may push fishes beyond their thermal limits. As ectotherms, fish are required to respond rapidly to such changes in coastal temperatures. For many species, behavioural adjustments, such as movements to more suitable habitats represent the first response to changing thermal conditions. Advanced techniques, such as biotelemetry have allowed scientists to track fish in their natural environments providing important information on movement ecology, and the response to changes in environmental conditions. While this technique provides valuable information on movement patterns it has limited utility in understanding the mechanisms driving these patterns. To understand the mechanisms driving fish behaviour, it is necessary to link behaviour in the wild with the physiological attributes of individual animals. Ideally, the physiological phenotype of the species should be categorised prior to tagging and tracking their movements in the wild. However, clear methodologies to do this have been elusive. This study aimed to fill this knowledge gap by investigating an appropriate laboratory technique for rapidly assessing the physiological attributes of individual fish without compromising their health prior to an acoustic telemetry study. Importantly, the physiological technique should provide appropriate information with which to categorize the physiological phenotype of an individual before it is tagged and released into the wild. The dusky kob, Argyrosomus japonicus (Sciaenidae), was selected as a candidate species, as they have shown extensive diversity in their behaviour in the wild which has been attributed to individual thermal tolerance. They are also currently listed as endangered on the IUCN Red List of threatened species with research gaps on the impacts of climate change and overexploitation on the sustainability of future populations; and hatchery-reared fish were readily available from Kingfish Enterprise in East London, South Africa, which allowed for the physiological phenotypic assessments without compromising wild-caught populations. A total of 40 juvenile A. japonicus were obtained from Kingfish Enterprises and transported back to a laboratory in the NRF SAIAB’s Aquatic Ecophysiology Research Platform (AERP) at the Department of Ichthyology and Fisheries Science (DIFS), Rhodes University. Thirty fish were randomly selected from the original 40 and randomly split into two groups (Batch A and Batch B) and housed separately in two 5900 L tanks. For the purpose of repeated measures, each fish was tagged with a passive integrated transponder (PIT) for individual identification throughout the experiments, and was given a colour coded (either red, blue, or white) spaghetti tag for visual identification. The dynamic method, characterized by the critical thermal minimum (CTmin), and critical thermal maximum (CTmax) was conducted using a repeated measures approach to assess the thermal tolerance of individual fish. Temperature was decreased or increased at a rate of 1 °C per hour to simulate marine upwellings or marine heatwaves until fish reached their critical thermal end points. Thermal breadth was calculated as the difference between CTmin and CTmax. To assess thermal stress prior to reaching critical thermal limits, opercula beats (OB), as a proxy for ventilation rate, were counted for each fish at every degree change using GoPro cameras. A subsample of eight individuals were selected and intermittent flow respirometry was used to determine their standard metabolic rate (SMR), maximum metabolic rate (MMR) and aerobic scope (AS) at three test temperatures (12, 18, and 24 °C) using a repeated measures approach. The results of the thermal tolerance trials indicated that on average hatchery-reared A. japonicus had a broad thermal tolerance, with an average CTmax of 32.7 °C (SD = 0.59), CTmin of 8.2 °C (SD = 0.28) and an average thermal breadth (TB) (CTmax - CTmin) of 24.5 °C (SD = 0.69). Substantial individual variability was also observed in the critical thermal endpoints and TB of A. japonicus where CTmin ranged from 7.5 to 8.6 °C, CTmax ranged from 31.7 to 33.7 °C, and TB ranged from 23.3 to 26.0 °C. Categorisation of fish into physiological phenotypes based on their CTmax, CTmin and TB estimates revealed that the sampled population exhibited a high degree of phenotypic variability. Using the percentile ranking method, individuals were categorised as high performers (broad TB, low CTmin, high CTmax), low performers (narrow TB, high CTmin, low CTmax) and intermediate performers (those individuals that ranked as neither high nor low performers). Individuals were also classified as either “broadly-adapted” (low CTmin, high CTmax), “narrowly-adapted” (low CTmax, high CTmin), “cold-adapted” (low CTmin, low CTmax) and “warm-adapted” (high CTmin, high CTmax).The piecewise linear breakpoint analyses conducted on the ventilation rate (used as a proxy for thermal stress) data indicated the onset of thermal stress at temperatures below 12 °C and above 30 °C. Given that these temperatures are occasionally observed in estuarine and in the nearshore marine environment, these breakpoints may be useful to test hypotheses on thermal drivers using telemetry data. Respirometry data showed a positive relationship between metabolic rates (SMR, MMR) and temperature. The high performance (AS) at all the test temperatures confirmed the eurythermic nature of the species. As with the thermal tolerance data, there was considerable intraspecific variability in SMR, MMR and AS and the percentile rankings were used to identify high, low and intermediate performers across the three temperatures. Overall, the diversity in thermal tolerance and performance of these hatchery A. japonicus was surprising because they were from a limited adult gene pool and were reared in a relatively warm and stable thermal environment. However, the diversity among individuals does suggest that the natural population of this species may have considerable adaptive potential, but a similar study on wild A. japonicus will be required to confirm this. The rankings for the different physiological metrics were used to obtain an overall sum rank which was assumed to provide a numerical representation of the physiological phenotype. Of the various metrics measured, the AS across all test temperatures followed by the CTmin and CTmax showed the greatest alignment with the sum rank. When the handling stress, confinement stress, thermal stress, and time taken to complete the experiment were considered, the most appropriate metrics for assessing the physiological phenotype before a telemetry experiment were CTmin and CTmax. In conclusion this study contributed to the development of a rapid, benign method to assess the physiology of fishes prior to a telemetry experiment. Such experiments will allow researchers to explore correlations between physiological performance and fish behaviour in the wild, furthermore improving our understanding of their adaptive capacity to impacts of climate change.