Getting to the heart of anatomical diversity and phenotypic plasticity: fish hearts are an optimal organ model in need of greater mechanistic study

The heart has long been recognized as an important organ. Egyptian hieroglyphs, for example, show how, prior to embalming, a person's heart was weighed on a scale against the Feather of Truth; a mummy with a ‘heavy’ heart might not make it into the afterlife. Of course, today, the heart is better known as the powerhouse of the vertebrate cardiovascular system, which delivers vital oxygen to tissues and removes toxic CO₂. The heart generates the blood pressure required to pulse blood to the capillaries – the site of diffusive tissue gas exchange. The heart is also the first-functioning organ during ontogeny, and all adult vertebrates need a continuous cardiac output (see Glossary). Consequently, cardiac damage or dysfunction debilitates whole-animal function, and the absence of a heartbeat indicates death. Understandably then, the terms ‘cardio’, ‘heart’ and ‘cardiovascular’ are prominent key words during JEB's first 100 years, appearing in ∼3% of JEB's articles.

This Commentary has three main parts and highlights some of JEB's literature on cardiac function. Part 1 considers the enormous species diversity of cardiac structure and function that is evident among fishes. The ∼30,000 fish species have radiated into almost every water mass on Earth and represent about half of all extant vertebrate species. How cardiac form and function specifically contribute to the success of a species or population in waters varying dramatically in temperature – as well as in salt, oxygen and carbon dioxide content – remains largely unexplored. Indeed, the cardiac chamber responsible for generating central arterial blood pressure (CAP; see Glossary), the ventricle, has very distinct anatomical forms and extensive variation. Of course, only hundreds of the thousands of fish species have been studied so far: new ventricular morphologies could emerge from studies of additional species. Moreover, the emerging field of population-specific diversity suggests that cardiac form and function for certain fish populations may be adapted to local environments, information that is particularly important for conservationists and ecologists, as well as physiologists (see Eliason et al., 2011; Gilbert et al., 2020). Part 2 of this Commentary highlights the remarkable phenotypic plasticity of cardiac form and function of an individual adult fish that can result from acclimation to a new environment, and considers its cellular basis (see Gamperl and Farrell, 2004). Part 3 emphasizes that cardiac function can be reliably measured in many ways, including at the organ level. These techniques allow cardiac form and function to be more easily related to mechanisms and to species diversity than those of other organs. By extension, fish hearts have the potential to become an excellent organ model for providing mechanistic linkages with genetic diversity, assuming that species diversity has some genetic basis.

Basic cardiac pumping dynamics are fundamentally similar in the small number of fish species studied thus far, which range from ‘primitive’ hagfishes to ‘modern’ teleosts (Satchell, 1991; Fig. 1). Specifically, a syncytial contraction of the entire ventricular myocardium (see Glossary) expels blood into the circulatory system with each heartbeat. Yet, there is not a single cardiac design that is common to all fish. Natural selection, instead, has produced a tremendous range of ventricular shapes and sizes, and variety in the specifics of supporting physiological processes. As shown below, much work is still needed to understand this extensive species diversity.

Relative ventricular mass (RVM; see Glossary) represents a simple metric for assessing an adult fish's investment in its myocardium. For most fish species, RVM is between 0.05% and 0.15% of body mass, but it varies by almost 10-fold across all fish species (Farrell, 1991; Farrell and Smith, 2017). RVM is highest (∼0.4%) in highly active tunas and lowest for benthic species (0.04–0.05%) such as plaice, white sturgeon and shovel-nose ray. Yet, swimming prowess is not the only determinant of RVM, because its value is unimpressive in sportfish such as Pacific blue marlin and Pacific tarpon (0.09% and 0.07%, respectively), whereas RVM is exceptionally high in inactive Antarctic icefishes (∼0.4%; see below).

In theory, a greater investment in ventricular muscle mass should produce a higher CAP, which also varies by around 10-fold across fish species. That CAP and RVM are associated with each other is exemplified by tunas, which have the highest CAP and RVM across fish species (Brill and Bushnell, 1991; Farrell and Jones, 1992). Yet, the lowest CAP of all fishes (in hagfishes) is associated with an intermediate RVM (0.1%; Cox et al., 2010). Thus, other factors (such as cardiac output; see discussion of Antarctic icefishes, below) should be considered in order to explain species variability in RVM. The specifics of ventricular architecture may be another factor, but a systematic comparison of muscle tension development for different ventricular shapes and muscle fibre arrangements is needed to understand the biomechanical consequences of different myocardial thicknesses and different muscle fibre orientations of the ventricle (see Sanchez-Quintina and Hurle, 1987; Farrell, 1991). I suspect that mechanical efficiency will vary across ventricular designs and that the Law of Laplace (which relates cardiac muscle tension to overall spherical radius; see Glossary) applies only to certain ventricular designs. Whereas a pyramidal ventricular design is associated with the highest CAP in fishes, the benefits of nearly circular and elongate ventricular shapes are unknown. Domestication of salmon, for example, has produced a more rounded and less pyramidal ventricle. Likewise, greater knowledge regarding myocardial oxygen consumption will yield insights into mechanical efficiency and the overall cost of cardiac pumping expressed as a proportion of a fish's overall metabolic rate (estimated at ∼1%; Farrell and Steffensen, 1987).

Working cardiomyocytes (see Glossary) normally require a continuous and reliable oxygen supply for continuous cardiac pumping. This oxygen need is directly related to work performed by the heart, its cardiac power output (see Glossary; Farrell and Jones, 1992). Fishes evolved two different supply routes for this vital ventricular muscle oxygen supply. However, the specifics of these supply routes vary considerably with ventricular architecture. An updated classification system summarizes these routes (Table 1; for earlier classifications, see Tota et al., 1983; Davie and Farrell, 1991; Farrell et al., 2012; see also Graham, 1997, for the additional cardiac diversity of air-breathing fishes).

The most common oxygen supply route (found in ∼60% of teleost species) is associated with a ventricle with only a trabecular architecture (also called spongy myocardium). This Type 1 ventricular design (Fig. 2A) has an elaborate arrangement of muscular trabeculae that span segments of the ventricular wall with varying lengths, diameters and branching patterns across species. The Type 2 ventricle also has a trabeculated myocardium, but this spongy layer is completely encased by a layer of compacted cardiomyocytes (Fig. 2B,C) and it has a dedicated coronary circulation that provides a well-oxygenated arterial blood supply directly from the gills.

The sole myocardial oxygen supply to the trabeculae of a Type 1 ventricle is the venous blood inside the ventricular chamber. This supply route is likely the oldest among chordates (it is found in hagfishes) and remains a successful evolutionary solution for the majority of teleosts. Its success relies in part on the narrow and numerous trabeculae of spongy myocardium (Fig. 2C,D) providing an enormous surface area and a short distance for oxygen to diffuse from venous blood. The Type 1 design, therefore, does not require a dedicated coronary circulation (see Glossary), thus saving on building and maintenance costs. Compaction in the Type 2 design, however, reduces access to venous blood and slows the rate of oxygen diffusion, hence the need for a coronary circulation. During early ontogeny, however, all vertebrates (including mammals) have a temporary trabecular design before the dedicated coronary circulation of the Type 2 ventricle develops.

Two coronary arrangements exist in Type 2 fish ventricles, Types 2c and 2v (Table 1). In Type 2c, the coronary vessels are confined to the compact myocardium (the trabeculae receive oxygen from venous blood as for Type 1). In Type 2v, however, the trabeculae and the compact myocardium are highly vascularized. Vascularized trabeculae can have a much larger diameter than those of Type 2c hearts (well beyond 1 mm for several shark species; Cox et al., 2016). The percentage of compact myocardium in the Type 2 ventricles is another source of species diversity, regardless of the subtype. The percentage varies by over 10-fold across fish species. It is highest in tunas (73%) and lowest in chimaeras (3%), the most primitive extant vertebrate. The reason for such huge species variability is unknown, as are the mechanisms that limit compaction to some predictable adult percentage for a fish species. Addressing this unknown would not be a hard task because compact myocardium can be easily quantified by ‘peeling’ it from the trabeculae, much like peeling an orange (Farrell et al., 1988b, 2007; Simonot and Farrell, 2007), and expressing its mass as a percentage of the entire ventricle mass.

The evolutionary roots of the coronary circulation also need greater study. Type 2v is found in all extant elasmobranchs studied to date and might have been the original vertebrate trait that appeared after the cyclostomes (Type 1a). Teleost species would then have largely abandoned the Type 2v morphology, replacing it with either Type 1 or Type 2c morphologies (see Farrell et al., 2012). Coronary vascularity studies with ‘primitive’ fishes such as polypterus and garfish which, like elasmobranchs, have a coronary supply to cardiac muscle surrounding their conus arteriosus, might be informative. The vascularity of in trabeculae the Type 2v tuna ventricle (Tota, 1978) should be quantified too.

Although Type 1 and Type 2c morphologies dominate among teleost species, the myocardial oxygen supply route has been described as ‘precarious’ (Farrell, 2016) because the last major tissue to receive oxygen from the circulatory system is a part or all of the ventricle. Moreover, venous partial pressure of oxygen (Pv_O2), which drives diffusion of oxygen into the trabeculae of Type 1 and 2c ventricles, is low and very variable. Pv_O2 decreases appreciably, for example, with intense swimming, aquatic hypoxia or a supraoptimal water temperature (Farrell and Clutterham, 2003; Steinhausen et al., 2008; Seymour et al., 2007; Clark et al., 2008; Ekstrom et al., 2016). However, a mechanism to boost Pv_O2 was recently discovered inside the Type 2c ventricle of rainbow trout. The mechanism involves the expression of plasma-accessible carbonic anhydrase (paCA) on the trabecular endothelium (Alderman et al., 2016). This enzyme is already known to substantially elevate the P_O2 in skeletal muscle capillaries in fish under acidifying conditions (Rummer et al., 2013; Randall et al., 2014; Harter and Brauner, 2017; Harter et al., 2019; McKenzie et al., 2004), and it seems to be important for CO₂ excretion in Antarctic icefishes (Harter et al., 2018). The cardiac expression of paCA among the many thousands of other fish species is unknown, however. One suggestion is that paCA first appeared around the Permian era and led to the very successful radiation of teleost fishes (Harter et al., 2019). If this is true, its appearance may have coincided with an evolutionary ‘reappearance’ and success of the Type 1 and 2c ventricular morphologies in teleosts.

A trabecular morphology has advantages. Foremost, it promotes efficient emptying of venous blood with each ventricular contraction. The anatomical dead space of the ventricular chamber at end-systole (end-systolic volume) is just 5% of end-diastolic volume (Franklin and Davie, 1992), which contrasts with a much larger ∼50% for the human ventricle (Shave et al., 2019). Efficient emptying means that the venous myocardial oxygen supply is almost completely refreshed with every cardiac contraction. Trabeculae may have another advantage (according to the Law of Laplace) that needs experimental testing. The trabeculae may create ‘miniature hearts’ with very small radii that generate greater muscle tension when compared with the larger radius of the entire ventricle (Johansen, 1965).

The Antarctic Ocean is a harsh but biologically productive environment that is home to specialized and abundant biota. Its frigid water temperature (as low as −1.9°C) has been stable for at least 30,000 years, ample time for genetic species adaptations to occur. Indeed, some of the Antarctic notothenioids (Sidell and O'Brien, 2006) lack functional genes to manufacture haemoglobin and myoglobin, two crucial respiratory pigments (O'Brien et al., 2021). Consequently, their blood appears a ghostly white (Fig. 3), as do their heart and skeletal muscles (they were once called ghost fish, but are now called icefishes). Distinguishing blood vessels from the surrounding tissues during surgical dissections of these fish is tricky.

Remarkable compensatory cardiovascular adaptations have resulted from the lack of functional red blood cells, which reduces the arterial oxygen content of icefishes to almost an order of magnitude lower than that of other vertebrates. Their RVM, for instance, compares with that of tunas and humans (all are around 0.4% of body mass). But their large ventricle is not particularly powerful; the maximum pressure a perfused icefish heart (that of Chaenocephalus aceratus) can generate is almost half of that a red-blooded Antarctic cousin (Egginton et al., 2019) and a fifth of that of a tuna. Instead, the thin-walled, Type 1 ventricle has a huge internal volume that permits an enormous maximum stroke volume (almost 10 ml kg⁻¹), i.e. about 10 times that of tunas and humans. Thus, cardiomegaly enables a high cardiac output despite the maximum heart rate being just 17 beats min⁻¹ at 0.6°C (Joyce et al., 2018a). The strategy of a thin-walled, large-capacity, slowly pulsing ventricle that trades off pressure generation for an exceptionally high cardiac output is viable because the absence of red blood cells also appreciably lowers blood viscosity and because the arterial vessels of icefish have unusually wide diameters (Sidell and O'Brien, 2006).

Despite its unusual compensatory adaptations, the icefish heart is not operationally limited (O'Brien et al., 2021). Routine cardiac output near 0°C is about a third (Joyce et al., 2018a,b) of their impressive maximum (around 0.1 l min⁻¹ kg⁻¹; Egginton et al., 2019; Acierno et al., 1997; Tota et al., 1991), which means that cardiac output can almost triple as needed, as in other fishes (Farrell and Smith, 2017). The initial reports of exceptionally high routine cardiac output in icefish (Hemmingsen et al., 1972; Holeton, 1970) probably represented a stressed and possibly anaemic state (see below) because visualizing blood loss from an icefish is almost impossible.

Understanding whether or not the icefishes' unique cardiac design would be problematic should the water temperature in Antarctica increase with global climate change is an important research need. Their myocardial oxygen supply is the most precarious of all fishes; Pv_O2 is extremely labile during exercise with no red blood cells to elevate venous oxygen content. Furthermore, cellular myoglobin, which can facilitate the rate of oxygen diffusion within cardiomyocytes (see Bailey and Driedzic, 1986; Bailey et al., 1990), is missing. Although other fishes have adapted to life without cardiac myoglobin (Driedzic and Stewart, 1982), only icefishes lack both cardiac myoglobin and functional red blood cells.

A particularly exciting and emerging field of fish biology is population-specific diversity, whereby a population has tailored its form to local environmental conditions. Although this has traditionally been a field of greatest relevance to evolutionary geneticists and ecologists, many examples of population-specific physiological adaptations are emerging, and these include cardiac adaptations.

Sockeye salmon populations spawning in the Fraser River watershed (BC, Canada) are an excellent example. These populations are genetically distinct and reproductively isolated; a population returns with remarkable fidelity to its natal spawning area. Importantly, their river migration is rapid (20–40 km day⁻¹), which leaves little time for acclimation, and individuals die after spawning. Moreover, during this once-in-a-lifetime spawning river migration, populations can encounter very different river temperature and hydraulic conditions (see Farrell et al., 2008), and migration distances from 100 to 1000 km. Many population-specific adaptations are evident (Crossin et al., 2003, 2004) and these include differences in cardiac morphology and function (Eliason et al., 2011, 2013a). For example, RVM is related to a population's migration rate; percentage compact myocardium is related to both migration distance and migration effort; and absolute aerobic scope (AAS; see Glossary) is related to migration distance (Fig. 4). Of course, the only ‘experience’ of river conditions actually encountered during the river migration is through ancestry. Remarkably, the optimum temperatures for maximum scope for heart rate, cardiac output and AAS are all similar within a population, and each thermal optimum also coincides with the population-specific modal river temperature experienced by previous generations of that population during the river migration (Eliason et al., 2011, 2013b).

Ecological impacts of population-specific diversity in cardiorespiratory physiology are already evident. For instance, the Weaver Creek sockeye population attempted its relatively short ∼120 km spawning migration in 2004 when river temperature was 21°C versus a historical 14–15°C. The ensuing en route mortality was ∼70% – catastrophic for lifetime fitness. But it was not unexpected: the optimum temperature for this population's maximum scope for cardiac output and for oxygen uptake is 14.7°C, but is near zero at 21°C (Farrell et al., 2008). Therefore, population-specific adaptations for other fish species, which await discovery, may have important applications for conservation and fisheries management.

More generally, swimming at a supra-optimal temperature can decrease heart rate and cause cardiac dysrhythmias in salmon (Eliason et al., 2013b). This is also a worry for tropical fishes that live close their upper thermal ‘ceiling’ (Rummer et al., 2014). Therefore, determining thermal safety margins (the difference between habitat temperature and the upper critical temperature) is an imperative future task. Also, properly predicting the consequences of a warmer world with an increasing frequency and severity of hypoxic events on global fish distributions will require knowledge on whether or not upper thermal tolerance co-varies with hypoxia tolerance across populations or strains (e.g. Zhang et al., 2018).

The heart of an adult fish is not necessarily a structurally ‘static’ organ; it can show tremendous plasticity. Consequently, it has tremendous value for the study of plasticity of form and function. Indeed, with structural and functional malleability already recognized (see Gamperl and Farrell, 2004), exploring cardiac plasticity in individual fish is fertile ground for future mechanistic studies that circumvent large genetic differences across species and populations. As shown below, only a few mechanistic details have emerged so far for cardiac plasticity; much remains to be understood.

In fish, acclimation to warm temperature can produce a smaller ventricle with an improved oxygen supply, because RVM can decrease by 20–40% and the proportion of compact myocardium can increase by 15–30% (Farrell et al., 1988b; Graham and Farrell, 1989). Such plasticity may facilitate the faster rates of cardiac contraction and relaxation associated with a warmer heart. Connective tissue content, ventricular compliance and myofilament function can also change with temperature acclimation (see Klaiman et al., 2011, 2014). Functional studies are needed to ascertain the benefits of such morphological changes, and more information is needed on the molecular signals for these changes.

The changes in heart rate and cardiac output with warming in a fish are better understood (Farrell, 2002, 2007, 2009). As mentioned above, acute warming always increases heart rate (Sandblom and Axelsson, 2007), but only up to a peak heart rate (Farrell, 2016). Acclimation to a higher temperature can change heart rate in beneficial ways: the peak heart rate can increase; the temperature at which peak heart rate is reached can increase; and the baseline pacemaker rate can be slowed for a given temperature. However, not all three changes necessarily occur within a species (Anttila, et al., 2014; Farrell, 2016). Why species differ in their acclimation potential is unclear and needs study. Adrenergic modulation of heart rate also changes with temperature acclimation (see Ask, 1983), through changes in the density of ventricular β-adrenoceptors (Keen et al., 1993) and in the level of cardiac adrenergic tonus (Axelsson, 2005). β-Adrenergic cardiac stimulation seems to be especially important near upper thermal limits (Gilbert et al., 2019). Elegant electrophysiological studies of single, isolated cardiomyocytes have begun to unravel some of the thermal compensatory mechanisms in fish hearts and their species diversity (e.g. Vornanen, 2016), but more research is needed.

Acclimation to hypoxia can produce beneficial cardiac changes too. Heart rate can be better maintained following acclimation (Petersen and Gamperl, 2010a; Leeuwis et al., 2021) without changing resting or maximum cardiac output (Petersen and Gamperl, 2010a,b; Motyka et al., 2017; Leeuwis et al., 2021). The contractility of hypoxic compact (but not spongy) myocardium can also improve (Petersen and Gamperl, 2010b; Carnevale et al., 2020; Roberts et al., 2021). Contractile impairment of hypoxic spongy myocardium has been viewed, however, as both detrimental (Bolli and Marban, 1999; Overgaard et al., 2004) and adaptive (limiting the work performed by the heart under hypoxia: Syme et al., 2013; Gesser and Rodnick, 2019; Roberts et al., 2021). Mechanistically, hypoxic cardiomyocytes can have improved oxygen utilization for ATP production after hypoxic acclimation (Cook et al., 2013). Acclimation to hypoxia also increases the density and binding affinity of ventricular β-adrenoceptors (Motyka et al., 2017), the sensitivity of cardiac mitochondrial respiration to nitric oxide inhibition (Gerber et al., 2019) and electrophysiological properties (Stecyk et al., 2008). Future studies should consider species and population differences: hypoxic cardiac strips from a hypoxia-tolerant species produce more force than those from hypoxia-intolerant species (Joyce et al., 2016), and a more severe hypoxia is needed for loss of in situ heart function of a hypoxia-tolerant strain of rainbow trout than an intolerant strain (Faust et al., 2004).

Acclimation rates, however, are context specific. Cardiac plasticity associated with hypoxic acclimation, for instance, can occur at high but not low temperatures (Carnevale et al., 2020; Motyka et al., 2017). Furthermore, although molecular events that reset the cardiac pacemaker rate can begin within hours at a new temperature (Sutcliffe et al., 2020), organ-level changes more typically take days to several weeks (e.g. Franklin et al., 2007). Therefore, acclimation experiments that simulate real environmental variability are needed. Water temperature and oxygen concentration can vary tidally, daily and seasonally, for example, but not as a simple oscillation between two extremes, which is the typical experimental manipulation. Instead, environmental chambers could be programmed from sensors that have logged real environmental variability, aided perhaps by AI. Ultimately, global fish distribution models, which typically consider only species differences, will have to incorporate acclimation potential and population differences to be truly useful; cardiac plasticity is only one of many mechanisms whereby fish become inherently more tolerant of warming or hypoxia.

A huge unknown is why cardiac phenotypic plasticity is so extensive in fish. A contributing mechanism could be the fact that adult fish cardiomyocytes retain a capacity for hyperplasia (see Glossary) Consequently, fish cardiomyocytes can remain skinny, long and mononuclear (Vornanen et al., 2002a,b) by adding new cells during normal developmental growth; the volume of a single cardiomyocyte increases disproportionately less than overall ventricular mass (Farrell et al., 1988b). Human cardiomyocytes, in contrast, thicken considerably with age. Cardiomyocyte thickness has very important implications for cellular diffusion distances, electrical capacitance and coupling (see Vornanen et al., 2002a,b; Shiels, 2017; Vornanen, 2017), and even for the attachment of the compact layer to the trabeculae layer (see Pieperhoff et al., 2009; Klaiman et al., 2011).

Sexual maturation of male, but not female, salmonids can increase RVM by 20–90%. The mechanisms underlying this sex difference in heart morphology are unexplained, but involve myocardial hyperplasia and hypertrophy (see Glossary), and androgenic steroids (Davie and Thorarensen, 2006). Why maturing male fish would invest more resources than females in their myocardium to potentially improve overall cardiac pumping capacity also needs study. One possibility is the need to support the energetic costs of male courting behaviours. Another is that additional cardiac capacity might be required to support a more meandering upriver spawning migration compared with that of females (Hinch and Rand, 1998).

The heart stands out from other organs because there are dozens of established experimental approaches to measure its function. These established methodologies are broadly grouped as in vivo and in vitro, each of which is discussed in more detail below.

In vivo measurements (e.g. electrocardiogram, cardiac output, blood pressure) can measure many aspects of cardiac function. Interventions of some sort, such as the use of exercise protocols, adjustments to water quality (e.g. hypoxia, temperature, hypercapnia, hyperoxia; e.g. Clark et al., 2007) and pharmacological drugs (e.g. Farrell, 1981), provide the mechanistic insight. Simultaneous respirometry adds holistic insight. All the same, redundant and interrelated organ systems can confound mechanistic interpretations, an issue not unique to in vivo cardiac studies. Why, for instance, should cardiac output change if oxygen uptake does not? Also, a surgical intervention has to consider fish welfare and potential blood loss. Without adequate post-operative recovery to reduce stress, routine metabolic rate, cardiac output and heart rate can be elevated, thereby reducing their functional scopes. Persistent anaemia in a fish, besides elevating heart rate and cardiac output, can even increase RVM as a compensation (Simonot and Farrell, 2007).

Many in vitro techniques provide direct mechanistic interpretations. These techniques include the use of isolated cardiomyocytes, isolated cardiac trabeculae and muscle strips. The working whole heart preparation, however, has been especially valuable because the entire organ is either removed or left in situ for perfusion with oxygenated or aerated saline. In vitro perfusion with saline is particularly suited to Type 1 and Type 2c fish hearts (e.g. Driedzic, 1978; Driedzic and Gesser, 1994; Farrell et al., 1985; Egginton et al., 2019) because the saline P_O2 is much higher than Pv_O2 and compensates for there being no red blood cells. Maximum cardiac output in situ can be even greater than that in vivo (Farrell et al., 1990), likely because venous influences on the myocardium (see above) are avoided with fresh perfusate. The perfused heart is also devoid of normal neural and humoral inputs. It beats with a self-generated pacemaker rhythm (or can be electrically paced). Consequently, perfused hearts have provided many valuable mechanistic insights (Farrell and Milligan, 1986; Farrell et al., 1988a; Cousins et al., 1997; Franklin and Davie, 2005; Milligan and Farrell, 1991; Baker et al., 2011; Farrell et al., 1986; Keen et al., 1993; Farrell and Stecyk, 2007; Gillis et al., 2015; Gilbert et al., 2019), including insights into the role of cardiac myoglobin (Bailey and Driedzic, 1986).

Still needed are mechanistic cardiac studies of species and population diversity, as well as phenotypic plasticity in fishes. These studies need not be confined to the lab. The in situ working fish heart preparation has been used in the Arctic, the Antarctic and the Amazon (Farrell et al., 2013; Egginton et al., 2019; Hanson et al., 2009). Moreover, emerging biotelemetry and biologging technologies on free-swimming fishes should be more broadly used to circumvent some of the challenges of ‘hard-wiring’ a fish to expensive, non-waterproof measuring equipment. Perhaps they will permit investigations of cardiac plasticity in an individual fish.

In closing, fishes have invaded almost every water mass on Earth, and the secrets of their success and huge species diversity need to be unlocked. With genome sequencing for less than 1000 USD, such a task will require a ‘blue sky’ research consortium that brings together the best fish physiologists, geneticists, developmental biologists and ecologists. The diverse form and function of the fish heart should provide excellent ground material for mechanistic explorations of genetic diversity given the expectation that some genetic differences will underlie the extensive species diversity in cardiac form and function identified in this Commentary. Although identifying the unique cardiovascular adaptations of Antarctic fishes is well underway, little is known of the cardiac adaptations of deep-sea fishes to their cold, dark, high-pressure and food-limited habitat (e.g. Greer-Walker et al., 1985; Santer, 1985).

Progress is also being made in terms of coupling molecular techniques with functional analyses of phenotypic plasticity in fishes. However, loss-of-function technologies, gene expression and cell culture studies have been more widely used with other organ systems in fish than with hearts (e.g. Kopp et al., 2005; 2007; Malby and Childs, 2010; Johnston and Gillis, 2017; Johnston et al., 2019; Tzaneva and Perry, 2016; but see Zimmer et al., 2019). The Fluidigm system, which measures the expression of many genes, is a very powerful, high-throughput approach for such studies and for discovering cellular response ‘signatures’, but its use with acclimation studies is still rather limited (e.g. Miller et al., 2011a,b; Houde et al., 2019; Sutcliffe et al., 2020).

I suspect that studying the beneficial plasticity of fish hearts might even benefit the biomedical community. For example, knowing the molecular triggers that permit extensive myocyte hyperplasia, or that stop cardiac compaction at a different, species-specific percentage among adult fish, or that provide plasticity to modulate adult cardiac form and function with acclimation could benefit care following a cardiac arrest. In human hearts, compaction and hyperplasia are largely completed during fetal development, and compaction fails in the rare human disorder of endomyocardial morphogenesis. Solving conservation issues is another huge area for potential applications of knowledge on cardiac plasticity (Cooke et al., 2012). The fish heart has been posited as a ‘weak link’ for internal oxygen transport at supra-optimal temperatures (Pörtner and Farrell, 2008; Farrell et al., 2008), and a better understanding of population-specific tailoring of the thermal performance of the heart could impact conservation efforts in this era of global warming and increasing amounts of aquatic hypoxia (Farrell et al., 2009; Hinch et al., 2021).

My entire research career was generously supported by NSERC Canada: a University Research Fellowship, a Canada Research Chair and Discovery, Strategic and Equipment grants. I thank the many trainees and collaborators that spanned all seven continents and who were central to my career successes. A special thanks to Dr Kurt Gamperl who made generous suggestions for the hypoxia section. Lastly, I thank JEB for being my ‘go-to’ journal over the decades.