From the outset, we would like to thank Farrell, Pörtner and Giomi (Farrell, 2013; Pörtner and Giomi, 2013) for their thoughts on our previous publication (Clark et al., 2013), and we acknowledge their valuable contributions to science as well as their desire to find unifying principles to help understand the evolutionary pathways of animals. Of course, we share their desire to bring simplicity to complex biological systems and processes.
The oxygen- and capacity-limited thermal tolerance (OCLTT) hypothesis explicitly states that thermal tolerance, physiological performance, climate change sensitivity and field distribution of ectothermic animals are causally determined by oxygen transport capacity, concisely summarised by Pörtner and Farrell (Pörtner and Farrell, 2008): ‘Direct effects of climatic warming can be understood through fatal decrements in an organism's performance in growth, reproduction, foraging, immune competence, behaviors and competitiveness. Performance in animals is supported by aerobic scope, the increase in oxygen consumption rate from resting to maximal. Performance falls below its optimum during cooling and warming. At both upper and lower pejus temperatures, performance decrements result as the limiting capacity for oxygen supply causes hypoxemia’. While we agree that increasing temperatures negatively affect the mentioned variables, we disagree on the generality of the proposed mechanism (tissue hypoxaemia at pejus temperatures leading to secondary losses of performance). Furthermore, we propose that different processes (‘performances’) can be thermally limited by factors unrelated to oxygen transport, by known and unknown mechanisms, and that the way forward is to test these contrasting perspectives experimentally. We fail to see how this could be viewed as taking ‘misguided sideswipes’, and instead we see it as encouraging scientific progress.
It is apparent from recent literature that the OCLTT hypothesis is treated as a confirmed theory by many scientists. Indeed, if we were new to the field of thermal physiology and read the literature on OCLTT, we would probably form the conclusion that the physiological mechanism underlying thermal performance and tolerance in ectotherms has been discovered in the form of OCLTT. This is worrying, as it implies that aerobic scope measurements over a relevant temperature range will provide a thorough understanding of the thermal ecology and climate-related responses of ectotherms, and it may lead researchers to downplay and overlook important results that do not fit the OCLTT hypothesis. We also raise the point that the OCLTT literature often extrapolates acute thermal exposures to global warming, despite the fact that we know even relatively short thermal acclimation can have profound effects on aerobic scope (Norin et al., 2013). Different rates of heating, from minutes in the lab to decades on a global scale, should not be assumed to affect animal physiology by identical mechanisms.
Our original paper was not intended to provide a thorough evaluation of the ecological and evolutionary framework for OCLTT, as we feel this topic has been addressed thoroughly in a string of papers by Pörtner and colleagues. Moreover, we were cognisant of the fact that a discussion in this direction would necessarily rely on speculation and hypothetical models as empirical data are lacking. Indeed, Pörtner and Giomi question our focus on aerobic scope and ToptAS (the temperature eliciting maximum aerobic scope) while they promote OCLTT as a powerful framework to universally explain broad-scale evolutionary constraints and climate-related effects on aquatic ectotherms [and terrestrial ectotherms (Pörtner, 2002)]. In doing this, they essentially transition away from the fundamental principles of the hypothesis (that oxygen transport limitations induce tissue hypoxaemia at pejus temperatures) and move towards a complex hypothetical framework that does not easily lend itself to experimental testing. The fact remains that the underlying assumptions of OCLTT rely on aerobic scope and its thermal dependence. We argue that individual performance in vivo is a fundamental level at which to test oxygen limitation. As these measurements have been most accessible and relevant to the broader scientific community in the context of OCLTT, our paper was targeted in this direction.
We do not agree with the assertion by Farrell that we have an ‘obvious error’ in fig. 1B (see Clark et al., 2013). We suspect that this criticism has arisen because some scientists misleadingly use the term Topt (without any qualifier) to describe the temperature where aerobic scope is maximal. We strongly recommend that the term Topt should be reserved as an abbreviation for the optimal temperature of the animal, which will be context dependent and governed by interacting optimal temperatures of different physiological and ecological performance metrics. Any use of the term Topt in the context of a particular performance metric should include a qualifier. We use the term ToptAS to describe the temperature that is optimal for maximising aerobic scope. Topt, not ToptAS, is what we have annotated in fig. 1B (see Clark et al., 2013), highlighting that aerobic scope may not set the optimal temperature of the animal. Indeed, we propose that the ‘onset of loss of performance’ of the animal can occur at the peak of the Fry curve where aerobic scope is maximal. In other words, performance metrics like growth and reproduction can be markedly reduced at ToptAS, as highlighted in the study on killifish cited by Farrell (Healy and Schulte, 2012), which we discussed in our original paper.
Farrell refers to a previous study on pink salmon (Clark et al., 2011), of which the take-home messages are (1) the breadth of acute thermal tolerance of pink salmon is exceptional, (2) aerobic scope across a broad temperature range is higher than has been reported for other Pacific salmon, and (3) the ToptAS of 21°C is a temperature that causes death within days in captivity (Jeffries et al., 2012) and one that most individuals in the population would never experience in their lifetime (i.e. a clear mismatch between Topt and ToptAS). This provides evidence that aerobic scope is not tailored to the mean river migration temperature encountered by the population (16–17°C), and instead aerobic scope simply increases with temperature up to (and often beyond) the highest temperature historically encountered by the species.
The Fry curve is repeatedly presented as a bell-shaped curve in most studies concerning OCLTT (e.g. Pörtner and Farrell, 2008; Farrell, 2009). That shape gives the false impression that aerobic scope declines gradually as temperature is increased beyond ToptAS, giving plenty of room for oxygen limitation to progressively affect other physiological systems. In reality, most species that have been investigated using appropriate techniques show high aerobic scope all the way up to lethal temperatures. If there is a general consensus that the Fry curve is not bell-shaped as implied by Farrell, then we see no reason why it should be continually portrayed this way in the literature.
Pörtner and Giomi ask whether there are any alternative concepts and whether OCLTT must be replaced with these alternatives in different ectothermic species. They also wonder how any species can be exempt from the hypothetical principles of OCLTT. Of course, aerobic scope is vital as it provides the capacity for increasing aerobic metabolism, but is ToptAS important? Must animals have access to maximum attainable aerobic scope in order to perform optimally? As portrayed in fig. 7B of our original paper, we believe that a temperature eliciting sub-maximal aerobic scope can be beneficial to maximise fitness as it may enhance other important performance metrics (e.g. growth rate, food conversion efficiency, reproduction). While a greater aerobic scope may be obtained at higher temperatures, the costs involved to other physiological and biochemical systems may outweigh any potential benefit. If aerobic scope increases with temperature throughout the ecologically relevant range (e.g. Claireaux et al., 2000; Clark et al., 2011), then ToptAS has limited power to predict the optimal temperature of the organism. If thermal acclimation causes aerobic scope to become temperature independent across most of the ecologically relevant temperature range (e.g. Norin et al., 2013), then any perceived benefit of ToptAS is lost and OCLTT loses predictive power.
The OCLTT hypothesis has given the broader scientific community the perception that they can answer many of their ecological and climate-related questions by measuring aerobic scope across a relevant temperature range. In this context, Pörtner and Giomi criticise the use of enforced exercise to gain an understanding of maximum aerobic metabolism, yet one wonders how this variable (and thereby aerobic scope) can be quantified in a ‘stress-free situation’. It was the goal of our original paper to stimulate critical thinking around the claimed generality of OCLTT, to find where it may make useful predictions, and to determine where its limitations may lie. We encourage scientists to critically assess their data and the published literature and ask some relevant questions: (1) is there a statistically significant ToptAS, does it align with data on other performance metrics (e.g. growth efficiency, reproduction) and, if so, how can the direction of causality be established?; (2) does the reported ToptAS make sense in light of the thermal history, evolution and ecology of the animal?; (3) does aerobic scope progressively approach zero as temperature exceeds ToptAS towards the upper critical temperature?; (4) did the animals survive the experimental protocols at all temperatures or could any observed reductions in aerobic scope be a consequence (not a cause) of poor health and impending death?; and (5) how does thermal acclimation influence aerobic scope and ToptAS? These questions may help to guide more thorough testing of the general relevance of the OCLTT hypothesis across species.
In their concluding paragraph, Pörtner and Giomi seem to suggest that we should stop debating and instead throw our support behind OCLTT to provide a unified voice from the experimental community about climate-related physiology. If we blindly endorse one view without rigorous testing, then we fail in our role as scientists and our perception in society would rightfully suffer. We believe that the OCLTT hypothesis has some major shortcomings in assuming that ToptAS has significant ecological and evolutionary relevance. Our desire to have a unifying principle should not trump empirical evidence. An increasing number of studies have shown that organisms maintain an excellent capacity for oxygen transport at the temperatures predicted to occur in their environments over the next 100 years (Brett, 1972; Clark et al., 2011; Overgaard et al., 2012; Norin et al., 2013; Gräns et al., in press), yet there are clearly other physiological and/or biochemical processes that may deteriorate, compromise fitness and survival, and explain the shape and position of performance curves on the temperature scale [see ‘Conclusions’ of our original paper (Clark et al., 2013)].
Our formalisation of the idea of ‘multiple performances – multiple optima’ [see fig. 7B of our original paper (Clark et al., 2013)] highlights that different performances can have different thermal optima, and the interrelations between these performances govern the thermal preference and fitness of animals. Of course, the challenge for scientists is to pinpoint important performance metrics and determine their thermal dependence in the context of animal fitness. This is likely to differ between organisms, life stages and rates of heating. It is therefore unfortunate that Farrell sees our idea as poorly supported and misleading when our intention is to advance the field by encouraging scientists to test the assertion that thermal performance and resilience to climate warming must be governed by oxygen transport capacity. Our common goal should be to find the causal mechanisms for deteriorations in animal performance and use them to understand the responses of animal populations to environmental perturbations including climate change.