Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/oxford-university-press/does-aerobic-scope-influence-geographical-distribution-of-teleost-uLa0w4EUgC
- Publisher
- Oxford University Press
- Copyright
- © The Author(s) 2023. Published by Oxford University Press and the Society for Experimental Biology.
- eISSN
- 2051-1434
- DOI
- 10.1093/conphys/coad012
- Publisher site
- See Article on Publisher Site
Abstract
Volume 11 • 2023 10.1093/conphys/coad012 Research article Does aerobic scope influence geographical distribution of teleost fishes? 1, 2 1 1 1 Julie J.H. Nati , Lewis G. Halsey , Paul C.D. Johnson , Jan Lindström and Shaun S. Killen School of Biodiversity, One Health and Veterinary, Graham Kerr Building, Glasgow G12 8QQ, UK Department of Life Sciences, University of Roehampton, Holybourne Avenue, London SW15 4JD, UK *Corresponding author: School of Biodiversity, One Health and Veterinary, Graham Kerr Building, Glasgow G12 8QQ, UK. Email: julienati3@gmail.com .......................................................................................................................................................... Many abiotic and biotic factors are known to shape species’ distributions, but we lack understanding of how innate physio- logical traits, such as aerobic scope (AS), may influence the latitudinal range of species. Based on theoretical assumptions, a positive link between AS and distribution range has been proposed, but there has been no broad comparative study across species to test this hypothesis. We collected metabolic rate data from the literature and performed a phylogenetically informed analysis to investigate the influence of AS on the current geographical distributions of 111 teleost fish species. Contrary to expectations, we found a negative relationship between absolute latitude range and thermal peak AS in temperate fishes. We found no evidence for an association between thermal range of AS and the range of latitudes occupied for 32 species. Our main results therefore contradict the prevailing theory of a positive link between AS and distribution range in fish. Key words: teleost fish species, geographical distribution, ecophysiology, comparative physiology, Aerobic scope Editor: Dr. Steven Cooke Received 14 January 2021; Revised 20 January 2023; Editorial Decision 1 March 2023; Accepted 16 March 2023 Cite as: Nati JJH, Halsey LG, Johnson PCD, Lindström J, Killen SS (2023) Does aerobic scope influence geographical distribution of teleost fishes? . Conserv Physiol 11(1): coad012; doi:10.1093/conphys/coad012. .......................................................................................................................................................... animal to perform oxygen-fueled processes (such as growth, Introduction locomotion and reproduction) above those required for The geographical distributions of species are shaped by basic maintenance (Fry, 1971; Clark et al., 2013). AS in a multitude of interacting factors including temperature ectotherms often increases with rising temperature, peaks (Payne et al., 2018), intraspecific and interspecific interactions at an optimum temperature and then decreases with further (Hansson, 1984) and dispersal capacity (Gaston, 2009; warming (Fry, 1971; Farrell, 2016; cf. Lefevre, 2016). An Kubisch et al., 2014). Another suite of influential traits ambient temperature that differs substantially from an are species’ physiological responses to their environment animal’s AS thermal optimum could therefore lower that (Kearney and Porter, 2009). We can therefore assume that animal’s capacity to deliver oxygen to tissues, which in turn biogeographical trends will covary with species’ physiology may impose constraints on its ability to perform physical (Somero, 2005; Bozinovic et al., 2011; Somero, 2011; activity or invest energy in reproduction or growth (Pörtner, Huey et al., 2012; Naya and Bozinovic, 2012; Seebacher 2001). et al., 2015). For ectotherms, aerobic scope (AS) has been Recently, it has been proposed that a relatively high AS proposed as a physiological constraint on species’ geographic may accommodate a high capacity for behavioural and physi- ranges (Pörtner, 2001; Pörtner and Farrell, 2008). AS ological plasticity (Biro et al., 2018). We therefore hypothesise represents the cardiovascular and respiratory capacity of an .......................................................................................................................................................... © The Author(s) 2023. Published by Oxford University Press and the Society for Experimental Biology. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. Research article Conservation Physiology • Volume 11 2023 .......................................................................................................................................................... a high latitude had a greater AS than populations at a lower latitude. Over a broader geographical scale, across 38 fish species, a positive correlation was demonstrated between latitude range and metabolic scope (Naya and Bozinovic, 2012). This analysis, however, did not account for phylogenetic relatedness and defined metabolic scope as being the difference between routine metabolic rate (RMR) and standard metabolic rate (SMR). This definition does not correspond to AS but, depending on how RMR is defined for each study, may approximate daily routine energy expenditure. Payne et al. (2016) found a positive relationship between the optimum temperature for AS and the highest temperature encountered in the geographical range of 12 tropical and temperate marine fishes, suggesting that AS might set the upper thermal limit for geographical regions that are habitable by these species. Therefore, AS may well be relevant for determining the upper distribution boundaries in Figure 1: Theoretical relationship between aerobic scope and teleost fishes. It is in part due to these findings that there is ambient temperature range, denoting the temperatures across an acceptance in the literature that aerobic performance in which sufficient oxygen can be allocated to all necessary aerobic ectotherms, such as fish, is positively related to their capacity processes for the body to function effectively (‘high aerobic scope’; to occupy a wider range of different thermal environments dashed line). A species with a high peak AS (blue line) may be able to (e.g. Pörtner, 2001; Pörtner and Farrell, 2008; Farrell, 2016), function effectively over a larger thermal range (blue arrows) (termed a greater thermal breadth of AS) than a species with a low peak AS but the current evidence for this is far from compelling (orange line and arrows). Figure modified from Nati et al., 2016. (Nati et al., 2016). The climatic variability hypothesis (CVH) states that animals at higher latitudes experience greater thermal fluctu- that species with a high thermal peak AS, or that exhibit ations and therefore display broader thermal performance a high AS over a large range of temperatures, have the breadths and high physiological plasticity (Janzen, 1967; plasticity required to survive over a broad range of thermal Stevens, 1989; Bozinovic et al., 2011). This suggests that environments. Indeed, often, such species may present both a temperate species have broader thermal tolerance ranges than high peak AS and a great thermal breadth of high AS. This is tropical species. Although the CVH has been examined in because there appears to be no functional trade-off between several ectothermic taxa (e.g. Drosophila spp., Overgaard peak AS and breadth of high AS across temperatures, at least et al., 2011; insects, Addo-Bediako et al., 2000; lizards, for teleost fishes (Nati et al., 2016), and thus species with Deutsch et al., 2008), it seems to apply only to terrestrial a high peak AS may have an AS that remains high across species (Bozinovic et al., 2011). In addition, Rapoport’s rule a broad temperature range (Fig. 1). We therefore predict a states that the range size of species decreases towards the positive correlation between peak AS and thermal breadth of equator (Stevens, 1989), suggesting that tropical species high AS with latitudinal distribution. would have narrower distribution ranges than temperate Several studies on endotherms and terrestrial ectotherms species. However, it is unclear whether this rule applies to have demonstrated positive relationships between metabolic teleost fishes (Rohde et al., 1993) and how the implications of traits and geographic distributions (Rezende et al., 2001, Rapoport’s rule interact with species’ metabolic physiology. A 2004; Dillon et al., 2010; Naya et al., 2012). For example, greater understanding of how temperate and tropical species a positive correlation between AS and latitude was reported may have evolved different thermal breadths for physiological across 48 rodent species (Naya et al., 2012). For ectotherms, performance or physiological plasticity would provide insight such as fish, physiology is more dependent on ambient into how each of these two thermal groups may be able to environmental conditions and therefore could strengthen respond to a changing global climate. the correlation between AS and geographical distribution. To date, however, no broad comparative study in an Teleost fishes are a diverse ectothermic taxon found in vari- ectothermic taxon, such as teleost fishes, has investigated able aquatic habitats. They have vastly differing geographical the links between AS and species distribution. Most studies ranges and lifestyles, with some species being much more investigating relationships between metabolic traits and active than others. Furthermore, there is now a considerable distribution in fishes have focused on a few species and amount of data available on fish metabolic traits. Therefore, over relatively narrow latitudinal gradients (Gardiner et al., teleost fishes are a suitable study model to test our main 2010; Rummer et al., 2014; Payne et al., 2016). For example, hypothesis (Fig. 1). We propose that a bigger/higher thermal in a four-species comparison of coral reef fish, Gardiner peak of AS and wider thermal breadth of AS in fishes will et al. (2010) showed that populations of each species at associate with them having larger geographical ranges. In the .......................................................................................................................................................... 2 Conservation Physiology • Volume 11 2023 Research article .......................................................................................................................................................... ◦ −1 Table 1: Summary of the PGLS model testing for the eeff cts on absolute latitude range (0 –90 ) of peak aerobic scope (log AS (mg O h )), 10 2 mass (log mass in g), acclimation temperature ( C), lifestyle (benthic, benthopelagic or pelagic), behaviour (migratory or sedentary), thermal group (tropical or temperate), habitat (marine, freshwater or combination marine/freshwater) and the interaction term between thermal group and log AS. R = 0.363, F = 4.215, p < 0.0001, n = 111 species, λ = 0.00, d.f. = 96. For lifestyle categorisation, the reference category is 10 13, 96 ‘benthic’; for behaviour it is ‘migratory’; for thermal group it is ‘tropical’; for the interaction it is ‘tropical’; for habitat it is ‘freshwater’ Term Estimate s.e. t p Intercept 30.87 4.31 7.33 < 0.001 log AS −1.43 4.06 −0.35 0.725 log mass 3.802 3.61 1.05 0.294 Temperature 0.002 0.24 0.008 0.99 Lifestyle Benthopelagic 0.46 2.69 0.171 0.87 Pelagic 4.951 4.11 1.20 0.232 Reef-associated −8.962 4.63 −1.94 0.056 Behaviour Sedentary −3.03 2.74 −1.106 0.272 Thermal group Temperate species −3.22 3.54 −0.909 0.366 Polar species −5.325 5.77 −0.92 0.36 Habitat Marine species 14.07 3.22 4.37 <0.001 Marine/freshwater 10.26 3.32 3.09 0.003 log AS:thermal group Temperate species −5.48 2.26 −2.42 0.02 Polar species 3.52 7.85 0.448 0.655 −1 current study, we focused on using interspecific, phylogeneti- absolute aerobic scope (AS; mg O h ) was calculated as the cally informed analyses to examine relationships between AS difference between MMR and SMR. To collate these data, we and latitudinal distribution range of 111 teleost fish species. searched Web of Knowledge and Google Scholar using key- We investigated whether latitudinal distribution range of fish words ‘standard metabolic rate’ or ‘maximum metabolic rate’ species is associated with, and therefore potentially deter- or ‘aerobic scope’ and ‘fish’. Species were only included when mined by, peak of AS and thermal breadth of AS. The results both SMR and MMR were measured in the same study. For presented here provide insight into the degree of influence of SMR, we only considered studies which measured metabolic AS on current species distributions. rates as follows: 1) by extrapolating oxygen consumption values measured at different activity levels to zero activity level; or 2) by direct recording of oxygen rates of a post larval, starved and resting fish over a consistent period of Methods time; for MMR: 1) by recording critical peak oxygen rates during forced swimming in a swim tunnel, or 2) by measuring Data collection oxygen consumption immediately after exhausting exercise in a swim tunnel or 3) after a chasing protocol (Clark et al., (a) Metabolic rate data and calculating peak aerobic 2013). These methods are known to provide similar values scope of MMR at the species level (Killen et al., 2017). Where we Data on the standard and maximum metabolic rates of 116 found several studies for the same species, we selected one species of fish were collated from the literature (see suppl. data set per species according to the following criteria. 1) −1 data set). SMR (mg O h ) is defined as the minimum For studies in which differing size classes were measured, energy required to sustain life, and maximum metabolic rate we used the study which collected data using fish with the −1 (MMR; mg O h ) represents the maximum rate of aerobic greatest body mass to select for juvenile or adult stages. These metabolism that an animal is able to perform. Whole animal life stage categories are the less thermally sensitive stages in .......................................................................................................................................................... 3 Research article Conservation Physiology • Volume 11 2023 .......................................................................................................................................................... ◦ −1 Figure 2: Relationship between absolute latitude range occupied by each species (0–90 )and log peak aerobic scope (log AS mg O h ) 10 10 2 centred for 111 fish species, categorised as temperate (n = 66), tropical (n = 37) or polar (n = 8). Each data point represents a distinct species. fishes (Pörtner and Farrell, 2008). Our mass range was from not thermally comparable. Absolute latitude range of each 0.05 to 6450 g. 2) For studies in which multiple temperatures species, calculated as the highest latitude minus the lowest were examined, we used the study that measured metabolic latitude at which a species is found, was used in this analysis rates at the greatest number of acclimation temperatures. Our as an indicator of the current distribution range, referring to ◦ ◦ ◦ ◦ acclimation temperature ranged from 3 Cto35 C. Besides a range from 0 to 90 . For any species with a distribution ◦ ◦ collecting metabolic rates (MO ) from each study, we also range overlapping the equator (e.g. 64 N—18 S; n = 39 recorded the mean body mass (g) of the fish and the accli- species), we considered the absolute latitude range expanding mation temperature used in each study. In cases where MO from the equator up to the maximum latitudinal point was measured at more than one temperature (studies with one of the species’ range in the northern hemisphere (in the acclimation temperature, n = 58; more than one acclimation example, 64 ). Tropical, temperate and polar species are temperature, n = 53), we used the data at the acclimation known to cover different distribution ranges according to temperature for which AS was highest. However, as most the CVH and Rapoport’s rule. Therefore, the mid-latitude studies measured AS at one temperature, we cannot presume position for each of 111 fish species was calculated from that this measurement was indeed a peak AS. In this case, the average of the distribution range (considering northern no optimum temperature for AS could be determined. We hemisphere values as positive and southern hemisphere values prioritised studies, which acclimated their fish. We avoided as negative) and based on this, each species was categorised ◦ ◦ taking AS where fish were acutely exposed to a range of as either tropical (0–20 ), temperate (20–60 ) or polar temperatures. (> 60 ). We also collected information from FishBase on lifestyle (benthic, benthopelagic, pelagic or reef-associated), habitat (marine, freshwater or marine/freshwater anadro- (b) Latitudinal and life history data mous and diadromous species) and behaviour (migratory or Latitudinal range data were available in FishBase (Froese and sedentary species) for all species included in the analyses, to Pauly, 2014) and metabolic rate data for 116 fish species account for these factors. We considered migratory species were taken from the literature. Only five species for which to be those that undergo active migrations for reproductive data were available, Pagothenia borchgrevinki, Sillaginodes and/or feeding purposes. punctatus, Colossoma macropomumin, Alcolapia graham and Bidyanus bidyanus were not included in the analyses, (c) Calculating thermal breadth aerobic scope leaving 111 species. They were excluded from the analysis because they were designated as southern hemisphere species, From our data set of 111 species, we selected species for which and this hemisphere is thermally less variable than the AS data were tested at three or more different temperatures northern hemisphere (Sunday et al., 2011, see Fig. 1b), and the AS thermal reaction norm could be fit to a Gaussian thus species distributions within the two hemispheres are model with three parameters (thermal peak AS, T and b; opt .......................................................................................................................................................... 4 Conservation Physiology • Volume 11 2023 Research article .......................................................................................................................................................... Life’ (see Appendix 1, Supplementary Fig. S1.1)(Hinchliff et al., 2015) using the ‘rotl’ package (Michonneau et al., 2016). A measure of phylogenetic correlation, λ, was estimated by fitting PGLS models with different values of λ to find the value that maximises the log likelihood. Lambda is the degree to which trait evolution deviates from a ‘Brownian motion’ model (traits evolving by the accumulation of small, random changes over time), and thus provides a measure of the degree of phylogenetic correlation in the data (Freckleton et al., 2002). Lambda = 1 retains the Brownian motion model, indicating that the trait covariance between any two species is directly proportional to their shared evolutionary history, while lambda = 0 indicates phylogenetic independence (the trait values across species are entirely unrelated to the phy- logeny of those species). Intermediate lambda values indicate that trait evolution is phylogenetically correlated but less than expected under the Brownian motion model (for more details, see appendix A of Halsey et al., 2006). All variables in the model were centered prior to analysis. (a) Absolute latitudinal range model for peak AS The following PGLS model was fitted to 111 species: ALR = β + β log (AS ) + β log (mass ) i 0 1 i 2 i 10 10 + β temperature + β I lifestyle = benthopelagic 3 4 i i + β I lifestyle = pelagic + β I lifestyle = reef 5 6 i i + β I habitat = marine 7 i + β I habitat = marine&freshwater 8 i Figure 3: Absolute latitudinal range categorised by (a) lifestyle (benthic = 47, benthopelagic = 32 and pelagic = 11, + β I behaviour = sedentary 9 i reef-associated = 21 species), (b) habitat type (freshwater = 36, marine = 55, marine/freshwater = 20 species) and (c) behaviour + β I thermalgroup = temperate (migratory = 67, sedentary = 44 species). Each data point represents a + β I thermalgroup = polar distinct species. 11 + β log (AS ) I thermalgroup = temperate 12 i 10 i + β log (AS ) I thermalgroup = polar + ε 13 i i 10 i n = 32 species). For each species, this model was a bell-shaped curve with T (optimal temperature) taken as the temper- opt where the response variable was the absolute latitude ature at which AS peaks. We then calculated the range of ◦ ◦ range from 0 to 90 of the ith species (ALR ). The temperatures at which AS remains above the 90% threshold continuous predictor variables were log (AS in mg O 10 2 (Anttila et al., 2014; Farrell, 2016). As the thermal range −1 ◦ h ), log (mass in g) and the acclimation temperature ( C) within which AS is above 90% of peak AS is arbitrary, we at which AS was measured, and the categorical predictor also calculated the range of temperatures at which AS remains variables were lifestyle (benthic, benthopelagic, pelagic, reef above 80, 75 and 60% of peak AS. associated), habitat (freshwater, marine, freshwater/marine), behaviour (migratory, sedentary) and thermal group (tropical, temperate, polar). AS and mass were log -transformed Data analysis to homogenize variance in absolute latitudinal range Statistical models of absolute latitudinal range were gener- with respect to these variables. Log -transformed mass ated with the phylogenetic generalised least squares (PGLS) was included as a covariate in the model because it has method (Grafen, 1989; Garland and Ives, 2000) using the scaling effects on metabolic rate data, such as AS (Gillooly caper package (Orme, 2013) in R (version 3.3.0 R Foundation et al., 2001). Including log-transformed mass in an additive for Statistical Computing). The statistical significance level regression model containing log (AS) has the effect, on the of all tests was set at p = 0.05. The phylogenies for the fish exponential scale, of flexibly adjusting the AS regression species in the analyses were generated from the ‘Open Tree of coefficient for the expected multiplicative effects of mass .......................................................................................................................................................... 5 Research article Conservation Physiology • Volume 11 2023 .......................................................................................................................................................... on AS, in addition to other potential effects of mass on latitudinal range not mediated through AS. Temperature was included in the model because it influences metabolic rate. All the following variables can impact latitudinal range sizes and were therefore also included in our PGLS model: lifestyle (benthic, benthopelagic or pelagic), habitat (freshwater, marine or marine/freshwater combination), behaviour (migratory or sedentary) and thermal group. Different lifestyles, habitats (Rohde et al., 1993) and behavioural strategies can either constrain or favour the dispersal capacity of a species and therefore limit or increase its ability to expand its distribution range. The thermal group of a species can affect its distribution range according to Rapoport’s rule (Stevens, 1989), which states that a decrease Figure 4: There is no discernible relationship between absolute in latitudinal distribution range can be observed towards latitudinal range and thermal breadth of 90% aerobic scope for 32 fish species. Each data point represents a distinct species. The shaded the equator, implying that tropical species tend to have area represents 95% confidence intervals around the line of best fit. smaller distribution ranges (Stevens, 1989). Furthermore, daily, seasonal and annual temperature variations differ in magnitude depending on latitude of occurrence, with high Results thermal fluctuations at temperate latitudes and low thermal variations at tropical and polar latitudes (Sunday et al., 2011). (a) Is latitudinal range predicted by peak Thus, thermal group was included in the model to account AS? for thermal variability across latitudes. Thermal history experienced by a species affects all physiological processes Overall, our model explained 36.3% of the observed variation such as AS, and depends on thermal group; consequently, we in absolute latitudinal range for 111 fish species, but log peak allowed for the possibility that the effect of AS might vary AS was a non-significant main effect (Table 1;t= −0.35, by thermal group by including an interaction term between p = 0.725). The model indicated that lifestyle strategies these two variables. In total, 14 fixed effects were estimated, do not influence range distribution (Table 1, Fig. 3a). including the intercept, β , 11 main effects, β , and two 0 1−11 Furthermore, the model showed that species spending at interaction effects, β and β . Finally, ε represents the 12 13 i least part of their life cycle in marine habitats have wider residual error of the ith species. The residual errors were absolute latitudinal distributions than do entirely freshwater assumed to be drawn from a normal distribution with mean species (t = 4.37, P < 0.001, t = 3.09, marine marine/freshwater zero and a variance–covariance matrix with a structure that P = 0.003, Fig. 3b), while sedentary species tend to have allows phylogenetically close species to covary to an extent smaller absolute latitudinal ranges than do migratory determined by λ. species (t = −1.106, P = 0.272, Fig. 3c.). There was sedentary a significant interaction between log peak AS and thermal group: temperate species showed a negative relationship (b) Absolute latitudinal range model for thermal between AS and absolute latitudinal range (t = −2.42, P = 0.02), while tropical and polar species exhibited no breadth AS relationship (Fig. 2). The slope value for temperate species PGLS analysis was also used to explore the relationship was −5.48 (95% CI, −10.32 to −0.64). Thus, temperate between absolute latitudinal range and 90% of thermal species exhibit an estimated 5.48 decrease in absolute breadth of AS in 32 species (see Appendix 2, Fig. S2.1, latitudinal range distribution for each 10-fold increase in Fig. S3.1-S3.3). The model was: peak AS. Lambda for the model was negligible (λ = 0.00), suggesting no phylogenetic inertia in trait covariance. ALR = β + β TBAS + β log (mass ) + ε i 0 1 i 2 i i (b) Is latitudinal range predicted by thermal breadth AS? where the absolute latitudinal range of the ith species (ALR ) was modelled as the sum of an intercept, β , the effects of The PGLS model including the thermal breadth of AS thermal breadth of AS (TBAS) and log (mass in g), modelled explained just 5.9% of the variation observed in the lati- by β and β , respectively. Due to the smaller sample size of tudinal range of 32 species. Lambda was negligible (λ = 0.00), 1 2 32 species, to avoid overfitting, only two predictor variables suggesting no phylogenetic inertia in trait covariance. There were included in the model (Harrell Jr, 2015). The residual was no relationship between 90% of peak AS thermal breadth errors, ε , were modelled as in the model for peak AS. We on latitudinal range (t = −0.83, P = 0.41, Fig. 4, Table 2). This performed models for other thresholds of thermal breadth AS was also the case for the calculated thresholds 80% and 60% (80%, 75% and 60%, see suppl.material). AS thermal breadths (see Tables S3.2 and S3.4, Fig. S3.1 and .......................................................................................................................................................... 6 Conservation Physiology • Volume 11 2023 Research article .......................................................................................................................................................... Table 2: Summary of the PGLS model testing for the eeff cts on absolute latitude range (0 –90 ) of thermal breadth of aerobic scope (thermal 90% −1 2 AS (mg O h )) and mass (log g). R = 0.059 F = 0.905, p = 0.416, n = 32 species, λ = 0.00, d.f. = 29 2 10 2,29 Term Estimate s.e. t p Intercept 30.47 4.93 6.18 <0.001 Thermal 90% AS −0.42 0.5 −0.83 0.41 Log mass 2.63 2.23 1.18 0.25 S3.3). However, at the 75% threshold, there was a negative It is possible that, within a species, local adaptation in relationship between AS breadth and absolute latitude range populations across a broad geographical range could result (Table S3.3, Fig. S3.2). in specialisation to cope with regional thermal conditions. A species might also cope with thermal extremes across its range by decreasing its activity and feeding, particularly during sea- sonal thermal shifts, allowing it to occupy a wide thermal and geographical range despite having a relatively low AS. This is Discussion a strategy of many temperate freshwater species and at least Current theory implies a positive correlation between aerobic some temperate marine species during overwintering (Ultsch, capacity in ectotherms, such as fishes, and their geographic 1989). Reducing activity during thermal extremes may be less distribution (Naya and Bozinovic, 2012). Contrary to these viable for tropical species, however, because they are exposed expectations, however, in teleost fish, we observed no rela- to relatively high temperatures even during the coolest parts of tionship between thermal breadth aerobic scope (AS) and the year, and so must possess a large AS for activity and other latitudinal range in teleost fish, and a negative relationship physiological processes year-round (Nati et al., 2016). This between peak AS with interaction of latitudinal midpoint and could be one reason why the negative association between latitudinal range, in temperate species (Fig. 2). Thus, while AS and geographical range is attenuated in tropical species. peak AS did not explain latitudinal range in either polar or Species with a narrower distribution may evolve greater spe- tropical species, in temperate species those with a high AS cialisation to living within a defined range of environmental showed a lower-latitude range than did those with a low AS. parameters with fewer functional tradeoffs, perhaps permit- ting a larger AS. In 92 fish species, for example, increased The data here indicate that a greater aerobic capacity AS is generally accompanied by an increase in energetic (both in terms of AS peak and the breadth of temperatures maintenance costs and energetic demand even at rest (Killen across which AS is high) does not convey an advantage et al., 2016). Species that are specialised for relatively constant for a fish species to inhabit a wider geographical range thermal regimes may be able to circumvent this fundamental (Fig. 1). This contradicts the CVH, which states that high- trade-off to some extent therefore reducing the costs of an latitude species have broader thermal tolerance and greater increased AS. We note, however, that we were able to find suit- physiological plasticity due to the thermal fluctuations they able data on far fewer tropical species than temperate species, experience. However, this hypothesis has mainly been studied which may have influenced our results. Additionally, there is a in terrestrial ectotherms (Naya and Bozinovic, 2012) and has lack of physiological data for fish species inhabiting the south- limited evidence for aquatic species. Furthermore, according ern hemisphere (Seebacher et al., 2015). It is also possible that to Rapoport’s rule, range distributions should be smaller in migratory species are able to track their optimal or preferred species that occur at lower latitudes (Stevens, 1989). However, thermal niches during migrations. There are currently insuffi- this rule was originally applied to terrestrial species, with cient data on the migratory patterns of species and the thermal mixed evidence of it applying to fishes (Rohde et al., 1993; conditions they experience to have included this factor in our Rohde and Heap, 1996). Similarly, in the current study, we analyses, but it is an important area for future research. observed no significant effect of thermal group (temperate, tropical or polar) on the latitudinal range of species. Our The fact that peak AS relates to distribution range accord- data thus support the view that Rapoport’s rule (Stevens, ing to thermal group (temperate), and that breadth of AS does 1989) may only apply within specific latitudinal ranges and not influence distribution range, implies that fish adjust AS biogeographical contexts. according to energetic needs and environmental conditions In our study, the AS data came from studies that mostly (Norin et al., 2016). It should be noted that sample size for the measured AS at one acclimation temperature (58 of 106 thermal breadth of AS was relatively low while the variation in the data was considerable. While this may have contributed studies). Due to this, we cannot presume that these AS data to the lack of an observed effect, the parameter estimates correspond to the thermal peak AS. In the wild, species may indicate that any effect of thermal breadth would nonetheless rarely utilise their entire aerobic capacity. Our results might be extremely small. Species or individuals can have vary- display a trend of the influence of AS on distribution ranges ing degrees of plasticity in AS allowing them to cope with in 111 fish species. .......................................................................................................................................................... 7 Research article Conservation Physiology • Volume 11 2023 .......................................................................................................................................................... environmental stressors (Munday et al., 2012, 2013; Norin et Recherche Luxembourg (4005263), the Natural Environment al., 2016). One threshold breadth (75%) had a negative asso- Research Council Advanced Fellowship (NE/J019100/1) and ciation with the absolute latitudinal range. It seems that at this the European Research Council Starting Grant (640004). breadth, species display a trade-off between maintaining this breadth and spreading their range. They might favour nearby food resources rather than expanding their range. The current Conflict of Interest statement study indicates that despite the plasticity of AS, species or indi- The authors declare no competing interests. viduals do not necessarily need to have a high AS to function efficiently over a larger range of environmental temperatures (Nati et al., 2016). Furthermore, other characteristics, such Data availability as lifestyle, body size, trophic level and fecundity, can all interact to influence the distribution of species and possibly Data supporting the results of this study are available in this outweigh any direct effect of AS on distribution. Fishes with manuscript and its supplementary files. different lifestyles (benthic, benthopelagic and pelagic) have different energy requirements and constraints (Killen et al., 2016), dispersal capacities and migration patterns (Eliason Authors’ Contributions et al., 2011; Demer et al., 2012). Benthic fish species are known to have a lower AS than pelagic species, giving them Conception: J.J.H.N., J.L., S.S.K.; data collection: J.J.H.N., less aerobic capacity to direct toward dispersal (Killen et al., S.S.K.; data analysis: J.J.H.N., L.H., P.C.D.J., S.S.K.; manuscript 2016). Furthermore, benthic species might be less constrained writing: J.J.H.N.,. S.S.K.; manuscript reviewing: J.J.H.N., by a reduction in AS due to their less active lifestyle. Here, L.H., PC.D..J., J.L., S.S.K. lifestyle was not a predictor for latitudinal distribution ranges. Further, different life stages might be more vulnerable than others. Larvae, eggs and spawning adults are believed to be the Acknowledgements most affected (Dahlke et al., 2020, Pörtner and Farrell, 2008). In our data set, we had juveniles and adults, the life stages that We are grateful to M. Ryan for assistance with data collection are the more robust in term thermal challenges. Another area and to two anonymous reviewers for valuable comments and for additional targeted research would be the effects of life suggestions. We would like to thank the two reviewers’ com- stage on the interplay between aerobic capacity and distribu- ments and thoughts, it helped us to improve the manuscript. tion. Although body mass appears to be the primary driver of changes in aerobic capacity during ontogeny both within and among species (Killen et al., 2007; Killen et al., 2016), Supplementary Material behavioural differences between life stages (e.g. juvenile ver- Supplementary material is available at Conservation Physiol- sus adult) could affect the proportion of aerobic scope remain- ogy online. ing for species after factors such as activity are accounted for. Additionally, other than AS, fishes, populations and indi- viduals can vary in their thermal limits (CT ). CT set max max References the upper latitude boundaries. We know that tropical not only have the highest CT but also have the lowest intraspecific max Addo-Bediako A, Chown SL, Gaston KJ (2000) Thermal tolerance, cli- variation in their CT (Nati et al., 2021). This makes max matic variability and latitude. Proc R Soc B Biol Sci 267: 739–745. tropical species less resistant to future warming events. https://doi.org/10.1098/rspb.2000.1065. In conclusion, we found evidence that peak AS is nega- Anttila K, Couturier CS, Øverli Ø, Johnsen A, Marthinsen G, Nilsson GE, tively related to the geographical distribution of temperate Farrell AP (2014) Atlantic salmon show capability for cardiac accli- teleost fish, suggesting that greater AS can be a constraint in mation to warm temperatures. Nat Commun 5: 4252. https://doi. this regard. Maintaining their maximum aerobic capacity is org/10.1038/ncomms5252. believed to come with an energetic cost. It has been suggested Biro PA, Garland T, Beckmann C, Ujvari B, Thomas F, Post JR (2018) that fish species distributions may be linked to the thermal Metabolic scope as a proximate constraint on individual behavioral sensitivities and limits of mitochondrial stability and func- variation: eeff cts on personality, plasticity, and predictability. Am Nat tioning of the heart (Iftikar et al., 2014). 192: 142–154. https://doi.org/10.1086/697963. Bozinovic F, Calosi P, Spicer JI (2011) Physiological correlates of geo- graphic range in animals. AnnuRevEcolEvolSyst 42: 155–179. https:// Funding doi.org/10.1146/annurev-ecolsys-102710-145055. This work was supported by an Aides à la Formation Clark TD, Sandblom E, Jutfelt F (2013) Aerobic scope measurements of Recherche doctoral grant from the Fonds National de la fishes in an era of climate change: Respirometry, relevance and rec- .......................................................................................................................................................... 8 Conservation Physiology • Volume 11 2023 Research article .......................................................................................................................................................... ommendations. J Exp Biol 216: 2771–2782. https://doi.org/10.1242/ Hansson S (1984) Competition as a factor regulating the geographical jeb.084251. distribution of fish species in a Baltic archipelago: a neutral model analysis. JBiogeogr 11: 367. https://doi.org/10.2307/2844802. Dahlke FT, Wohlrab S, Butzin M, Pörtner HO (2020) Thermal bottlenecks Harrell FE Jr (2015) Regression Modeling Strategies: With Applications to in the life cycle define climate vulnerability of fish. Science 369: 65–70. Linear Models, Logistic and Ordinal Regression, and Survival Analysis. https://doi.org/10.1126/science.aaz3658. Springer, pp. 1–582 Demer DA, Zwolinski JP, Byers KA, Cutter GR, Renfree JS, Sessions TS, Hinchliff CE, Smith SA, Allman JF, Burleigh JG, Chaudhary R, Coghill LM, Macewicz BJ (2012) Prediction and confirmation of seasonal migra- Crandall Keith A, Deng J, Drew BT, Gazis R et al. (2015) Synthesis of tion of Pacific sardine (Sardinops sagax) in the California current phylogeny and taxonomy into a comprehensive tree of life. PNAS 112: ecosystem. Fish Bull 110: 52–70. 12764–12769. https://doi.org/10.5061/dryad.8j60q. Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak Huey RB, Kearney MR, Krockenberger A, Holtum JAM, Jess M, Williams SE DC, Martin PR (2008) Impacts of climate warming on terrestrial (2012) Predicting organismal vulnerability to climate warming: roles ectotherms across latitude. PNAS 105: 6668–6672. of behaviour, physiology and adaptation. Philos Trans R Soc B Biol Sci Dillon ME, Wang G, Huey RB (2010) Global metabolic impacts of recent 367: 1665–1679. https://doi.org/10.1098/rstb.2012.0005. climate warming. Nature 467: 704–706. https://doi.org/10.1038/ Iftikar FI, MacDonald JR, Baker DW, Renshaw GMC, Hickey AJR (2014) nature09407. Could thermal sensitivity of mitochondria determine species distri- Eliason EJ, Clark TD, Hague MJ, Hanson LM, Gallagher ZS, Jeffries KM, bution in a changing climate? J Exp Biol 217: 2348–2357. https://doi. Gale MK, Patterson DA, Hinch SG, Farrell AP (2011) Differences in org/10.1242/jeb.098798. thermal tolerance among sockeye salmon populations. Science 332: Janzen DH (1967) Why mountain passes are higher in the tropics. Am Nat 109–112. https://doi.org/10.1126/science.1199158. 101: 233–249. https://doi.org/10.1086/282487. Farrell AP (2016) Pragmatic perspective on aerobic scope: peaking, Kearney M, Porter W (2009) Mechanistic niche modelling: combining plummeting, pejus and apportioning. J Fish Biol 88: 322–343. https:// physiological and spatial data to predict species’ranges. Ecol Lett 12: doi.org/10.1111/jfb.12789. 334–350. https://doi.org/10.1111/j.1461-0248.2008.01277.x. Freckleton RP, Harvey PH, Pagel M (2002) Phylogenetic analysis and Killen SS, Costa I, Brown JA, Gamperl AK (2007) Little left in the tank: comparative data: a test and review of evidence. Am Nat 160: metabolic scaling in marine teleosts and its implications for aerobic 712–726. https://doi.org/10.1086/343873. scope. Proc R Soc B Biol Sci 274: 431–438. https://doi.org/10.1098/ Froese R, Pauly D (2014) FishBase: World Wide Web electronic publica- rspb.2006.3741. tion. http://www.fishbase.org . Killen SS, Glazier DS, Rezende EL, Clark TD, Atkinson D, Willener AST, Halsey LG (2016) Ecological influences and morphological correlates Fry FEJ (1971) The effect of environmental factors on the physiology of of resting and maximal metabolic rates across teleost fish species. Am fish . Fish physiology. Academic Press, London, pp. 1–98, https://doi. Nat 187: 592–606. https://doi.org/10.1086/685893. org/10.1016/S1546-5098(08)60146-6. Killen SS, Norin T, Halsey LG (2017) Do method and species lifestyle Gardiner NM, Munday PL, Nilsson GE (2010) Counter-gradient vari- affect measures of maximum metabolic rate in fishes? J Fish Biol 90: ation in respiratory performance of coral reef fishes at elevated 1037–1046. https://doi.org/10.1111/jfb.13195. temperatures. PloS One 5: e13299. https://doi.org/10.1371/journal. pone.0013299. Kubisch A, Holt RD, Poethke HJ, Fronhofer EA (2014) Where am I and why? Synthesizing range biology and the eco-evolutionary Garland T, Ives AR (2000) Using the past to predict the present: confi- dynamics of dispersal. Oikos 123: 5–22. https://doi.org/10.1111/ dence intervals for regression equations in phylogenetic compara- j.1600-0706.2013.00706.x. tive methods. Am Nat 155: 346–364. Lefevre S (2016) Are global warming and ocean acidification Gaston KJ (2009) Geographic range limits of species. Proc R Soc B Biol Sci conspiring against marine ectotherms? A meta-analysis of the 276: 1391–1393. https://doi.org/10.1098/rspb.2009.0100. respiratory eeff cts of elevated temperature, high CO2 and their Gillooly JF, Brown JH, West GB, Savage VM, Charnov EL (2001) Effects of interaction. Conserv Physiol 4: 1–31. https://doi.org/10.1093/ size and temperature on metabolic rate. Science (80) 293: 2248–2251. conphys/cow009. https://doi.org/10.1126/science.1061967. Michonneau F, Brown JW, Winter DJ (2016) Rotl: an R package to interact Grafen A (1989) The phlyogenetic regression. Philos Trans R Soc London with the open tree of life data. MethodsEcolEvol 7: 1476–1481. 326: 119–157. https://doi.org/10.1111/2041-210X.12593. http://cran.r-project.org/ web/packages/rotl/index.html. Halsey LG, Butler PJ, Blackburn TM (2006) A phylogenetic analy- sis of the allometry of diving. Am Nat 167: 276–287. https://doi. Munday PL, McCormick MI, Nilsson GE (2012) Commentary impact of org/10.1086/499439. global warming and rising CO2 levels on coral reef fishes: what hope .......................................................................................................................................................... 9 Research article Conservation Physiology • Volume 11 2023 .......................................................................................................................................................... for the future? J Exp Biol 215: 3865–3873. https://doi.org/10.1242/ Pörtner HO, Farrell AP (2008) Ecology: physiology and climate jeb.074765. change. Science (80) 322: 690–692. https://doi.org/10.1126/ science.1163156. Munday PL, Warner RR, Monro K, Pandolfi JM, Marshall DJ (2013) Predict- ing evolutionary responses to climate change in the sea. Ecol Lett 16: Rezende EL, Bozinovic F, Garland T (2004) Climatic adaptation and the 1488–1500. https://doi.org/10.1111/ele.12185. evolution of basal and maximum rates of metabolism in rodents. Evolution 58: 1361–1374. Nati JJH, Lindström J, Halsey LG, Killen SS (2016) Is there a trade-off between peak performance and performance breadth across tem- Rezende EL, Silva-Dura I, Fernando Novoa F, Rosenmann M (2001) Does peratures for aerobic scope in teleost fishes? Biol Lett 12: 20160191. thermal history aeff ct metabolic plasticity?: a study in three Phyllo- https://doi.org/10.1098/rsbl.2016.0191. tis species along an altitudinal gradient. J Therm Biol 26: 103–108. https://doi.org/10.1016/S0306-4565(00)00029-2. Nati JJH, Svendsen MBS, Marras S, Killen SS, Steffensen JF, McKenzie DJ, Domenici P (2021) Intraspecific variation in thermal tolerance differs Rohde K, Heap M (1996) Latitudinal ranges of teleost fish in the between tropical and temperate fishes. Sci Rep 11: 1–8. https://doi. Atlantic and indo-Pacific oceans. Am Nat 147: 659–665. https://doi. org/10.1038/s41598-021-00695-8. org/10.1086/285873. Naya DE, Bozinovic F (2012) Metabolic scope of fish species increases Rohde K, Heap M, Heapt D (1993) Rapoport’s rule does not apply to with distributional range. Evol Ecol Res 14: 769–777. marine teleosts and cannot explain latitudinal gradients in species richness. Am Nat 142: 1–16. https://doi.org/10.1086/285526. Naya DE, Spangenberg L, Naya H, Bozinovic F (2012) Latitudinal patterns in rodent metabolic flexibility. Am Nat 179: E172–E179. https://doi. Rummer JL, Couturier CS, Stecyk JAW, Gardiner NM, Kinch JP, Nils- org/10.1086/665646. son GE, Munday PL (2014) Life on the edge: thermal optima for aerobic scope of equatorial reef fishes are close to current day tem- Norin T, Malte H, Clark TD (2016) Differential plasticity of metabolic rate peratures. Glob Chang Biol 20: 1055–1066. https://doi.org/10.1111/ phenotypes in a tropical fish facing environmental change. Funct Ecol gcb.12455. 30: 369–378. https://doi.org/10.1111/1365-2435.12503. Seebacher F, White CR, Franklin CE (2015) Physiological plasticity Orme D (2013) The Caper Package: Comparative Analysis of Phyloge- increases resilience of ectothermic animals to climate netics and Evolution in R. http://cran.r-project.org/web/packages/ change. Nat Clim Chang 5: 61–66. https://doi.org/10.1038/ caper/index.html nclimate2457. Overgaard J, Kristensen TN, Mitchell KA, Hoffmann AA (2011) Ther- Somero GN (2005) Linking biogeography to physiology: evolutionary mal tolerance in widespread and tropical drosophila species: does and acclimatory adjustments of thermal limits. Front Zool 2: 1–9. phenotypic plasticity increase with latitude? Am Nat 178: S80–S96. https://doi.org/10.1186/1742-9994-2-1. https://doi.org/10.1086/661780. Somero GN (2011) Comparative physiology: a “crystal ball” for Payne NL, Meyer CG, Smith JA, Houghton JDR, Barnett A, Holmes BJ, predicting consequences of global change. Am J Physiol - Nakamura I, Papastamatiou YP, Royer MA, Coffey DM et al. (2018) Regul Integr Comp Physiol 301: R1–R14. https://doi.org/10.1152/ Combining abundance and performance data reveals how temper- ajpregu.00719.2010. ature regulates coastal occurrences and activity of a roaming apex predator. Glob Chang Biol 24: 1884–1893. https://doi.org/10.1111/ Stevens GC (1989) The latitudinal gradient in geographical range: how gcb.14088. so many species coexist in the tropics. Am Nat 133: 240–256. https:// doi.org/10.1086/284913. Payne NL, Smith JA, van der Meulen DE, Taylor MD, Watanabe YY, Taka- hashi A, Marzullo TA, Gray CA, Cadiou G, Suthers IM (2016) Tempera- Sunday JM, Bates AE, Dulvy NK (2011) Global analysis of thermal toler- ture dependence of fish performance in the wild: links with species ance and latitude in ectotherms. Proc R Soc B Biol Sci 278: 1823–1830. biogeography and physiological thermal tolerance. Funct Ecol 30: https://doi.org/10.1098/rspb.2010.1295. 903–912. https://doi.org/10.1111/1365-2435.12618. Ultsch GR (1989) Ecology and physiology of hibernation and overwinter- Pörtner H (2001) Climate change and temperature-dependent biogeog- ing among freshwater fishes, turtles and snakes. Biol Rev - Cambridge raphy: oxygen limitation of thermal tolerance in animals. Naturwis- Philos Soc 64: 435–515. https://doi.org/10.1111/j.1469-185X.1989. senschaften 88: 137–146. https://doi.org/10.1007/s001140100216. tb00683.x. ..........................................................................................................................................................
Journal
Conservation Physiology – Oxford University Press
Published: Mar 29, 2023
Keywords: teleost fish species; geographical distribution; ecophysiology; comparative physiology; Aerobic scope