Contrary to expectations, this study highlighted differences in individual response to temperature, where surprisingly individual relative performance was not maintained across temperatures for most performance metrics. Some fish performed relatively better at lower temperatures, others had moderate performance across both temperatures, and others had higher performance the high temperature. This suggests the maintenance of performance potential across temperatures for some variables, while indicating niche performance for other variables, all observed within the confines of a single species.
The individuals that had similar activity and growth rates across both temperatures therefore presented high contextual plasticity, thus a wider thermal optimum and broader tolerance to temperature change. In contrast, the individuals that had markedly greater growth and activity rates when exposed to high temperatures showed low contextual plasticity (Biro et al. 2010). These patterns are obscured in most population studies reporting overall means and variances (Metcalfe et al. 2016). If we had ignored individual variability and only assessed each fish at a particular temperature, the behavioural estimates would have been biased by the temperature itself (Biro et al. 2010). For instance, since fish tend to exhibit bolder behaviours in warmer temperatures, an individual that is usually shy might seem bold, even bolder than a naturally bold fish in colder conditions when observed in warm environments. Considering plasticity, an individual with high contextual plasticity could have appeared relatively bold at cooler temperatures. In contrast, the same individual at a warmer temperature would be identified as shy compared to an individual with low contextual plasticity, where the changes in boldness with temperature are much sharper (Biro et al. 2010).
In contrast with our finding that performance was mostly not preserved across treatments, studies that assessed the Amazon molly (Poecilia formosa) and the green swordtail (Xiphophorus helleri) dominance showed that individuals maintained their losing/winning performance over time (Laskowski et al. 2016). Individuals that only experienced winning interactions in the early life stages held the same trend later in life, remaining top-ranked individuals in the hierarchy while losers remained at the bottom rank and neutral individuals were in the middle (Dugatkin and Druen 2004; Laskowski et al. 2016). Additionally, a study on the spotted catshark (Scyliorhinus canicula) demonstrated how shark social network position was consistent and maintained across different habitats (Jacoby et al. 2014). This study highlighted how previous social experiences could have substantial and long-lasting consequences on the adult population's social behaviour and structure (Laskowski et al. 2016).
Across all species, growth rate and change in total length at 16°C were not significant predictors of relative performance at 20°C. The magnitude of individual variability is surprising where not all individuals followed the population pattern with an increase in performance with temperature, but some moved from low to high performance in cold temperatures while others maintained average performance at both temperatures. This highlights the strong individual differences in thermal optima (Angilletta Jr et al. 2002). As aforementioned, when below the thermal optimum, an increase in temperature usually provides larger aerobic scope and, thus, more energy available for growth which would explain the higher growth rates at higher temperatures experienced by most individuals (Neuheimer et al. 2011). Colder waters lower the energetic costs due to activity that could allow the fish to invest in growth, if a sufficient aerobic scope is available (Killen 2014). Thus, this could justify the higher growth rates in some fish at 16°C.
Differences in growth rates with temperature have important implications for fish shoals, as fish that grow faster become larger, affecting social hierarchy (Cutts et al. 1998). Larger fish are usually more aggressive and dominant members that outcompete the suburbanites in access to food and mates (Symons 1968; Cutts et al. 1998). This social hierarchy can impact the social dynamics within a group, such as reproductive success and therefore affect future generation genetics (Wong et al. 2008). Additionally, this increases the homogenisation of reaction responses while lowering population resilience as the genetic pool shrinks (Morrongiello et al. 2019). This study highlights that these social hierarchies could be modified by a temperature increase as the individuals that grow the most (and thus are likely to be dominant) at 16°C would not be the same ones that grow the most at 20°C. A study on the African cichlid fish species, A. burtoni found that size differences that were once thought to be negligible (< 10% body length) provided a substantial advantage to dominant and larger individuals (Alcazar et al. 2014).
However, despite the fact that faster-growth-rate and larger individuals benefit from this attribute in the social hierarchy, the size-selection that operates at a fisheries level acts against a rapid growth rate, where larger individuals are more likely to be selected and removed (Biro and Post 2008). Additionally, warm adapted-faster-growing genotypes might be at a disadvantage in a winter situation, because the more robust appetite and, thus, higher feeding rates that are required to sustain the higher growth rates might induce more risk-taking and bolder behaviours to forage and thus increase the risk of encounter fishing gear and vulnerability to it (Biro and Post 2008). A lake experiment showed that genotypes that grow faster were captured at three times the rate of the slow-growing ones (Biro and Post 2008).
G.subfasciatus aerobic scope at 16°C was a good predictor of aerobic scope at 20°C, which was expected as these temperatures are within the fish thermal range, and a slight increase in temperature below thermal optimum is expected to boost metabolism (Norin et al. 2014; Coleman et al. 2019). This might result from the consistency in sub-cellular and cellular components and processes within individuals (Biro and Stamps 2010). Additionally, it has been shown that resting metabolic rate is consistently different (repeatable) across individuals in birds, mammals and some fish (Álvarez and Nicieza 2005; Rønning et al. 2005; Nespolo and Franco 2007; Boratyński and Koteja 2009; Larivee et al. 2010; Barnes et al. 2019) and thus we would expect such consistency as environmental conditions are modified. Additionally, relationships between metabolic rates at different temperatures could be influenced by the difference in individual responses to stress that fish with a shy or bold phenotype experience during the actual metabolic test. For instance, shy individuals have been shown to increase their oxygen consumption during respirometry due to their higher sensitivity to confinement and handling stress (Martins et al. 2011).
However, individual P.sexlineatus' aerobic scope at the lower temperature was not a good predictor of its aerobic scope with a temperature increase (Norin et al. 2016). The processes behind intraspecific variation in metabolic rate are uncertain (Metcalfe et al. 2016). Intraspecific variation in metabolic rate could be the physiology of individuals, such as the leakiness of mitochondrial membranes or the protein turnover rates (Killen et al. 2011). Some studies showed how variation in body-size correlated with metabolic rate, revealing a positive relationship between metabolic rate and the size of metabolically costly organs such as the brain and heart (Konarzewski and Książek 2013). In Salmo trutta, intraspecific changes in metabolic rate were positively associated with the activity of some mitochondrial enzymes (Norin and Malte 2012), while changes in erythrocyte size were found to be negatively correlated with metabolic rate in the loach Cobitis taenia (Maciak et al. 2011). These species differences in aerobic scope associations across temperatures highlight how performance maintenance, such as aerobic scope, is mainly species and individual-specific. Therefore different individuals and species will be affected differently by climate change (O'Connor and Booth 2021).
Activity such as boldness was previously identified as an intrinsic trait across individuals (Careau et al. 2008). Members of the same species frequently behave differently, some being more aggressive or bold and others being more docile or shy (Killen et al. 2011), with some individuals having an innate propensity to be more active all the time (Careau et al. 2008). As such, this might justify the relationship between boldness at 16°C and 20°C in P.sexlineatus and C.australis as fish adapted to the changes in temperature to maintain the same relative performance.
P. sexlineatus exhibited a significant relationship across temperature treatments for bite rate, boldness, and shelter response. Under natural conditions, an individual's food intake rates and, thus, growth can be heightened through a combination of boldness and activity (Biro et al. 2010). And this process could therefore increase the positive relationships in these metrics across temperatures where a consistent increase in behavioural activity such as boldness can allow for higher foraging efficiency thus, higher food intakes (Biro and Stamps 2008).
Skeletal muscle activity affects individual performance in essential activities such as locomotion, prey capture and predator avoidance (James and Tallis 2019). The lack of relationship in escape response across temperatures detected in all species could be motivated by differences in cellular physiology that act at the base of responsiveness to stimuli such as predators (Ioannou et al. 2008).
The difference in swimming performance across individuals influences their spatial positioning within a school. Individuals with a relatively high aerobic scope and gait transition speed might locate themselves in anterior positions within the school (Metcalfe et al. 2016). These higher swimming speeds would allow these individuals to be found at the front of the school while performing other tasks such as feeding or digestion (Metcalfe et al. 2016). Consequently, these intrinsic differences in swimming capabilities may be associated with differences in intra-species migration success and probability of evading predator attacks (Walker et al. 2005; Eliason et al. 2011). Individual differences in boldness positively influence migratory inclination, the possibility of becoming dominant, foraging success, and reproductive performance. Bold actions, however, also have adverse fitness effects, such as greater predation risk, which can lower long-term survival (Forsatkar et al. 2016).
Differences in shelter usage across individuals might be linked to trade-offs between energy usage and foraging success (Conallin et al. 2012). For instance, sheltering might lower fish energy expenditure linked to performing mechanical tasks (e.g., swimming) (Hafs et al. 2014). Sheltering can also lower the energy expenditure of non-mechanical tasks such as thermoregulation (Beck and Jennings 2003), or high-energy activities such as camouflage, alertness and vigilance (Millidine et al. 2006). Therefore individual differences in shelter use might have significant impacts on individual survival and growth (Conallin et al. 2012).
In addition to physiological traits such as metabolism, different activity types can induce fish to choose different strategies to cope with stressors such as temperature changes. This is performed by selecting and migrating to cold habitats that lower energy expenditure (less active fish) or warmer habitats that promote a more active lifestyle (Killen 2014).
The processes mentioned above highlight behavioural traits' influence on the fish stock's social hierarchy, and where metrics are related to social dominance for example, it may be a mechanism by which the dynamics of estuarine fish populations could differ under climate change. Thus, the lack of relationship between many of these performance metrics at the lower vs higher temperature indicates that stock structures will likely be modified if temperatures increases, lowering fisheries resilience (Morrongiello et al. 2019). Individuals that currently present a physiological advantage at lower temperatures might drop to lower subordinate levels in favour of the individuals that are better adapted to these new warmer environments (Biro and Post 2008). As individuals present different performance optima, these individuals performance niches may serve as a buffer to population-level responses, initially resulting in a reshuffling of individual performance ranks as temperature increases, without altering the population average. After this reshuffling is finalized, a detectable population response may emerge as the temperature continues to rise. Thus, individual performance niches have the potential to mitigate population responses, or at the very least, impede our ability to perceive them.
To conclude, individual fish exhibited substantial differences in performance across temperatures. This highlights how climate change will likely favour a niche of adapted individuals who will thrive in warmer conditions. In contrast, the subset of individuals that perform best at lower temperatures might encounter higher competition as their niche temperature habitat retreats, which could pose a risk for them to transfer their 'cold acclimated' hereditary traits (Metcalfe et al. 2016). Thus, future communities will likely present different genetic compositions, physiological and behavioural characteristics than the current ones. The underlying shift in genetic composition, to genotypes that perform better under warmer temperatures, may actually mask a climate change response from studies that continue to focus on population averages when testing performance; changes in relative individual performance might initially counterbalance a population-level response, buffering climate change responses and potentially hindering our ability to detect them. Detecting individual changes in performance may provide an early warning for future population-level consequences of warming. To further enhance the knowledge of individual fish performance related to climate change, future studies could include broader performance metrics (such as courtship, shoaling aggression, and impact of temperature on neurological systems) (Domenici and Hale 2019; Little et al. 2020) and more fish species (Astles and McLeod 2018).