Recovery of carbon benefits by overharvested baleen whale populations is threatened by climate change

Abstract

Despite the importance of marine megafauna on ecosystem functioning, their contribution to the oceanic carbon cycle is still poorly known. Here, we explored the role of baleen whales in the biological carbon pump across the southern hemisphere based on the historical and forecasted abundance of five baleen whale species. We modelled whale-mediated carbon sequestration through the sinking of their carcasses after natural death. We provide the first temporal dynamics of this carbon pump from 1890 to 2100, considering both the effects of exploitation and climate change on whale populations. We reveal that at their pre-exploitation abundance, the five species of southern whales could sequester 4.0 × 10⁵ tonnes of carbon per year (tC yr⁻¹). This estimate dropped to 0.6 × 10⁵ tC yr⁻¹ by 1972 following commercial whaling. However, with the projected restoration of whale populations under a RCP8.5 climate scenario, the sequestration would reach 1.7 × 10⁵ tC yr⁻¹ by 2100, while without climate change, recovered whale populations could sequester nearly twice as much (3.2 × 10⁵ tC yr⁻¹) by 2100. This highlights the persistence of whaling damages on whale populations and associated services as well as the predicted harmful impacts of climate change on whale ecosystem services.

1. Introduction

The concentration of atmospheric carbon dioxide has increased dramatically since the beginning of the industrial era, from about 277 ppm in 1750 to over 412 ppm today [1]. The rise of this greenhouse gas in the atmosphere is changing the climate with a range of damaging consequences for ecosystems and human societies [2]. Since the Paris Agreement in 2015, the international community set the objective of containing the global warming below +2C° compared to pre-industrial levels [3].

The ocean is a primary regulator of the climate system, absorbing about 23% of anthropogenic CO₂ emissions [1] and maintaining a primary production equivalent to that of terrestrial ecosystems [4]. In that context, particular attention has been given to understand the drivers of the biological carbon pump and its sequestration potential. Indeed, the open ocean's capacity to sequester carbon mainly relies on the export of organic carbon particles (dead organic matter and faeces) at depth where they can be sequestered for decades, centuries or even longer before returning to surface waters [5].

While the contribution of lower trophic levels (essentially phytoplankton and zooplankton) to the biological pump has been well explored for decades [6], the role of large vertebrates (fish and mammals) on ocean biogeochemistry and their ability to sequester carbon has been garnering increasing research attention [7–9]. Indeed, some of these species, such as whales, are the largest of all time, thus storing vast amounts of carbon throughout their lifetime which can be exported in the deep ocean after their natural death [10]. When reaching the seafloor, whales' carcasses can be consumed by the abyssal fauna and sustain local food webs or be buried in the sediments, being trapped for centuries to millennia [11,12] (figure 1). Therefore, whales generate carbon sequestration i.e. long-term (greater than or equal to 100 years) carbon removal from the atmosphere.

Figure 1.

Conceptual diagram of the carbon sequestration mediated by baleen whales. (1) After natural death, the carcasses start sinking. (2) They can be partially consumed by scavengers like killer whales and sharks as well as degraded by microorganisms. Once a carcass has reached the seafloor, the organic carbon in tissues can be (3a) consumed and respired at depth by local abyssal fauna (like Zoarcidae) and microorganisms. Once respired, the inorganic carbon can be brought back to the surface water and outgassed into the atmosphere on a timescale from centuries to thousands of years depending on the depth and the water circulation [5]. Or (3b), it can be buried in the sediments where it can be trapped on longer timescales, up to millennia [12]. (Online version in colour.)

Yet, whale populations have been dramatically reduced by historical whaling. The majority of this defaunation occurred in the twentieth century with some populations reduced by greater than 99%, like the southern right (Eubalaena australis) and Antarctic blue whales (Balaenoptera musculus intermedia) [13]. Since 1986, they have been protected from commercial whaling by an international moratorium of the International Whaling Commission (IWC). However, their recovery is now imperilled by climate change via the reduction of prey abundance, increased extreme climate events (El Niños and heatwaves) and disease spread [14–16]. While the dynamics of these populations have already been modelled [14,17], the influence of whaling and then of protection on the carbon sequestration mediated by these marine mammals remains poorly understood and unquantified. Thus, the long-term dynamics of whale-mediated carbon sequestration in the southern hemisphere need to be assessed accounting for both past whaling and various future climate scenarios. Here, we quantified the role of whales in carbon sequestration through carcasses sinking using historical and forecasted whale abundance estimates under two climate scenarios.

2. Materials and methods

(a) . Modelling of whale population dynamics

We used population dynamics for five southern baleen whales that have been recently modelled [14,17] as inputs for our carbon sequestration assessment. This is a Southern Hemisphere spatial ‘Model of Intermediate Complexity for Ecosystem Assessments’ (MICE) that estimates whale population sizes from 1890 to 2100. Whale abundances were estimated based on both the dynamics of their prey (Antarctic krill Euphausia superba and copepods; bottom-up process) and catches (whaling; top-down pressure). To account for the effects of climate change, we used two MICE model outputs. In the first version, whale population dynamics were not coupled to changing climate conditions which were assumed to remain constant from 1890 to 2100. The second model integrates the impacts of climate change according to the RCP 8.5 (business as usual) scenario, on krill and copepod prey availability. See electronic supplementary material for more information on the population model assumptions and outputs.

(b) . The age structure of the population

As carbon sequestration depends on whales' body mass, we derived the age structure of each species population from the outputs of the MICE model. The age structure of the population in a given year, i.e. the number of individuals in a given age class, was constructed using demographic parameters for each species derived from the MICE models (electronic supplementary material, table S1). Whale population abundances provided by the MICE models correspond to mature females only. To obtain the total number, the number of males (E_male) was estimated from the number of females (E_female) and the sex ratio (q):

$$\begin{array}{rl}{E}_{\mathrm{tot}}=& \hspace{0.17em}{E}_{\mathrm{female}}+E_{\mathrm{male}}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\\ \hspace{0.17em}{E}_{\mathrm{female}}=& q\cdot E_{\mathrm{tot}}\\ E_{\mathrm{male}}=& (1-q)\cdot E_{\mathrm{tot}}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\underset{\hspace{0.17em}}{\Rightarrow }\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}{E}_{\mathrm{male}}={\displaystyle \frac{(1-q)}{q}\cdot E_{\mathrm{female}}}\end{array}\}$$ 2.1

For each sex, the number of individuals (N) in each age class (a) between birth and maximum age (z) was then calculated from the number (E) of adult individuals between the age of maturity (T) and maximum age (z) and the survival rate (S or S_juv). A distinction was made between the survival rate of individuals under one-year old (S_juv) and that of individuals over one-year old (S). These rates were species-specific (electronic supplementary material, table S1), considered constant and identical for males and females. The number of individuals of a given year class (individuals aged a) is thus written as a function of the number of individuals aged T:

$$\begin{array}{rl}& \mathrm{\forall }a\hspace{0.17em}\in \hspace{0.17em}]0,z]\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}{N}_{a+1}=S\cdot N_a\hspace{0.17em}\hspace{0.17em}\underset{\hspace{0.17em}}{\Rightarrow }\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}{N}_a=S^{a-T}\cdot \hspace{0.17em}{N}_T\hspace{0.17em}\mathrm{for}\\ & \hspace{0.17em}a=0,N_0=S_{\mathrm{juv}}^{-1}\cdot S^{1-T}\cdot N_T\end{array}$$ 2.2

The number of individuals in T years can be found with the parameters of the model:

$$\sum _{a=T}^z{N}_a=E\hspace{0.17em}\underset{\hspace{0.17em}}{\Rightarrow }\hspace{0.17em}{N}_T\hspace{0.17em}\cdot \sum _{a=T}^z{S}^{a-T}=E\underset{\hspace{0.17em}}{\Rightarrow }\hspace{0.17em}\hspace{0.17em}{N}_T={\displaystyle \frac{E}{{\sum }_{a=T}^z{S}^{A-T}}}$$ 2.3

Thus, the age structure of the population can be written as a function of N_T:

$$\begin{array}{rl}{N}_0& =S_{\mathrm{juv}}^{-1}\cdot S^{1-T}\cdot N_T;\hspace{0.17em}{N}_1=S^{1-T}\cdot N_T\cdots \\ \hspace{0.17em}\hspace{0.17em}{N}_{T-1}& =S^{-1}\cdot N_T;\hspace{0.17em}\hspace{0.17em}{N}_T;\hspace{0.17em}{N}_{T+1}=S\cdot N_T\hspace{0.17em}\hspace{0.17em}\begin{array}{ccc}\cdots & N_z=S^{z-T}\cdot N_T& \hspace{0.17em}\end{array}\end{array}$$ 2.4

(c) . The whale mortality and biomass

We applied this age structure every year from 1890 to 2100 in order to calculate the number of dying individuals and their biomass within each age class. We assumed that age structure did not vary over time. Indeed, the body mass of individuals is almost constant in adulthood so we considered that no adult year class is more impacted by whaling than another. Among juveniles, very few catches have been reported for the whale species of this study [18].

First, in order to calculate the number of individuals dying naturally each year, the natural mortality rate (1-S or 1-S_juv) was applied to each age class (electronic supplementary material, table S1).

Second, to obtain the biomass of an age class, the number of individuals in this class was multiplied by the corresponding individual body mass. The biomass of individuals at each age follows the Von Bertalanffy equation, whose parameters, depend on both species and sex [10] (electronic supplementary material, table S1). Let a the age, m_inf the maximum size of individuals, k the growth rate and a₀ the theoretical age at which the mass is zero, the mass m of an individual aged a is

$$m(a)=m_{\inf }(1-{\mathrm{e}}^{-k(a-a_0)}).$$ 2.5

Then, the total population biomass (B_tot) was calculated from the biomass of each age class (B_a) as follows:

$$\begin{array}{rl}& B_{\mathrm{tot}}=\sum _{a=0}^z{B}_a\hspace{0.17em}\\ & \mathrm{with}\hspace{0.17em}\hspace{0.17em}{B}_a=\hspace{0.17em}{m}_{\mathrm{female}}(a)\cdot N_{\mathrm{female},a}+m_{\mathrm{male}}(a)\cdot N_{\mathrm{male},a}.\end{array}$$ 2.6

Finally, to assess the amount of carbon sequestered by sinking whales’ carcasses, the biomass was converted into carbon mass. A previous study assumed a total body carbon content for dolphins of 25% (wet weight) [19] based on estimates on humans [20]. A lower estimation of 15% (wet weight) was found assuming 70% of water content and 50% of carbon in the dry weight [21]. We used these 15%, 25% and 20% as lower, upper and mean estimate of the carbon content in the whale biomass, respectively. We assumed that the carbon content is identical for all individuals without distinction of species, sex or age.

(d) . Carbon sequestration

To estimate the amount of carbon sequestered via whale carcasses sinking into the deep ocean (i.e. deadfall carbon), we estimated the proportion of carcasses that reaches the deep sea. Indeed, the fate of the carcasses depends on several factors. We assume that most adult individuals die of senescence or disease since attacks by predators such as killer whales (Orcinus orca) are thought to be rarely lethal, especially on adults [22,23]. However, carcasses do not sink in their entirety to the ocean floor because they are partly consumed by scavengers like sharks [24] or killer whales [25] or degraded by microorganisms [26]. We used a conservative estimate of 50% [27] for all species except for southern right whales, for which we considered that only 10% of carcass biomass reaches the deep sea because of their high proportion of blubber making them float after death [10]. We considered that whales' dead tissues reaching the seafloor before degradation (remineralization) are sequestered [5].

3. Results

(a) . Carbon sequestration prior to commercial whaling

We first investigated the ability of the five baleen whale species to sequester carbon when they were at their pre-exploitation abundance. Dead whales represented a biomass of almost 4 million tonnes each year. The sinking of their carcasses generated a flux of 4.0 × 10⁵ (range: 3.0 × 10⁵–5.0 × 10⁵) tonnes of carbon per year (tC.yr⁻¹) towards the deep sea (figure 2). However, all species did not contribute equally to this total carbon flux. Fin and blue whales contributed to 48% and 34%, respectively (figure 2). The other three species had a marginal contribution, particularly the southern right whales, which accounted for only 0.7% of the total carbon flux.

Figure 2.

Amount of carbon sequestered annually by each whale species and for all the five species together (total) at their pre-exploitation levels through the sinking of the carcasses. On the top right, the relative contribution of each species. Errors bars represent high and low estimations for carbon sequestration given parameter uncertainties. (Online version in colour.)

(b) . Carbon sequestration dynamics from 1890 to 2100

We then predicted carbon sequestration dynamics from 1890 to 2100 under two climate change scenarios, both accounting for historical whaling. A stable phase from 1890 to 1912 was followed by a sharp drop in the amount of carbon sequestered over the exploitation period (figure 3a,b). All species experienced a severe population decline, particularly the main sequestration contributors, fin and blue whales, which were reduced to approximately 3% and 0.5% of their pre-exploitation abundance, respectively (figure 3c,d). As a result, carbon sequestration dropped to a minimum of 0.6 × 10⁵ tC yr⁻¹ (range: 0.4 × 10⁵ and 0.7 × 10⁵ tC yr⁻¹) in 1972, so only 15% of the pre-exploitation level.

Figure 3.

Dynamics of carbon sequestration mediated by the five baleen whale species between 1890 and 2100 without climate change (a,c) and with climate change (b,d); (a,b) represent the total sequestration by the five whale species; (c,d) represent the sequestration dynamic for each species. Shaded areas represent the high and low estimations of carbon sequestration given parameter uncertainties. (Online version in colour.)

In the model without climate change, carbon sequestration would reach 3.2 × 10⁵ tC yr⁻¹ (range: 2.4 × 10⁵ and 4.0 × 10⁵ tC yr⁻¹) in 2100, so 80% of the sequestration potential before the whaling period. This recovery is mainly due to the predicted increase of Antarctic minke whales and the recovery, albeit slower, of all other species (electronic supplementary material, figure S3). In the model including the effects of climate change, only the Antarctic minke whale would reach their pre-exploitation population size and their sequestration potential before the end of the twenty-first century (electronic supplementary material, figure S3). Under this scenario, the whale-mediated carbon sequestration via carcasses sinking would reach 1.7 × 10⁵ tC yr⁻¹ (range: 1.3 × 10⁵–2.1 × 10⁵ tC yr⁻¹) in 2100, so less than half of the pre-exploitation level, with Antarctic minke whales accounting for 47% of the total flux, against 5.6% before whaling. In both climate scenarios, fin whales were the major contributors until 1961. While large whales (mainly fin whales and blue whales) recover in the scenario without climate change and overcome the contribution of minke whales, the latter remain the main contributor until the end of the century in the scenario with climate change.

Finally, we showed that harmful consequences of whaling persist for many years after its end. By dramatically reducing population levels of the largest species, over-exploitation has created a carbon sequestration deficit. It represents the amount of carbon that has not been sequestered because of whaling compared to what would have been sequestered with whales’ populations at their pre-exploitation levels. We calculated that from 1890 to 2100, the carbon sequestration deficit caused by whaling is on average 41.9 × 10⁶ tC and 45.2 × 10⁶ tC without and with climate change, respectively (figure 4a,b).

Figure 4.

Cumulative carbon sequestration deficit from 1890 to 2100 without climate change (a) and with climate change (b). For each year, the total amount of non-sequestered carbon is compared to that corresponding to the pre-exploitation levels of whale populations. Shaded areas represent the high and low estimations for carbon deficit given parameter uncertainties. (Online version in colour.)

4. Discussion

(a) . The role of whales in the carbon cycle and carbon sequestration

Although our estimates are accompanied by a wide range of uncertainty, we showed that whales—as massive and fast-sinking organisms—efficiently sequester carbon after natural death. Through this process, they trap carbon away from the atmosphere for centuries to millennia, helping to mitigate climate change. However, the annual sequestration capacity of whales in the southern hemisphere is one or two orders of magnitude smaller than sequestration carbon flux estimated on blue carbon ecosystems (BCE) like mangroves, tidal marshes or seagrasses, sequestering annually 31.2–34.4 × 10⁶ tC yr⁻¹, 4.8–87.2 × 10⁶ tC yr⁻¹ and 41.4–82.8 × 10⁶ tC yr⁻¹ on a global scale, respectively [28].

Nevertheless, whales could play other important roles in the carbon cycle that are less explored. First, thanks to their longevity and high weight, they store large amounts of organic carbon in their body mass throughout their lifetime, that can be up to one century for some species like blue whales (electronic supplementary material, figure S1). Second, whales have been shown to be efficiently recycle nutrients and boost primary production [29–32]. Indeed, the Southern Ocean is largely considered to be a high-nutrient low-chlorophyll zone, i.e. a zone where macronutrients (nitrogen and phosphorus) concentrations are high but primary productivity is low [33]. Phytoplankton growth is limited by the availability of trace elements (Fe, Cu, Zn, Co and Cd), especially iron [33] as confirmed by many short-term iron-addition experiments [34,35]. In that context, southern whales play a critical functional role since they feed mostly on krill which is an iron accumulator [31,36]. Thus, they can alleviate the growth limitation of phytoplankton through the supply of iron-rich faeces. Indeed, whales' faeces are highly concentrated in iron (the iron concentration of whales’ faeces was estimated to be 145.9 ± 133.7 mg kg⁻¹, being approximately 10 million times that of Antarctic seawater [31]). Plus, they are likely highly bioavailable as they are liquid and buoyant, remaining in the euphotic zone where whales defaecate, and iron is also excreted with other nutrients, preventing phytoplankton growth from co-limitations and successive limitations. On the other hand, although krill can also recycle iron, the pool of iron in krill is probably poorly available for phytoplankton, as krill release most of its iron in dense and fast-sinking pellets [36]. Therefore, whales play a unique role in fertilizer production in the Southern Ocean. But even though they help maintain high levels of primary productivity, whales do not enrich i.e. bring new iron in the system themselves. Thus, they mainly generate ‘regenerated’ primary production, as opposed to ‘new’ primary production generated by the addition of new nutrients in the system.

On the other hand, krill is known to feed a lot in the benthos [37,38], bringing new iron in the system. Thus, whales and krill could maintain highly productive systems through these processes, potentially boosting the biological carbon pump and sequestering carbon.

As only the new primary production leads to additional carbon fixation by the phytoplankton and carbon sequestration in the deep ocean (through particles flux toward the depth), it remains difficult the quantify the fraction of new and regenerated primary production and subsequent carbon sequestration mediated by whales. Nevertheless, research has shown that krill was more abundant in the Southern Ocean before the defaunation of whales [39], indicating that whales and krill were highly mutualistic. Since krill can export carbon at depth very efficiently through various processes (fast sinking of faecal pellets, sinking of exuviates [40–42]), we suggest that the pre-whaling ocean sequestered much more carbon than a megafauna-depauperated ocean. The role of whales in the carbon cycle, mainly through their indirect role of nutrient cycling represents an avenue for future research, especially regarding climate change mitigation and potential identification of new natural climate solutions (NCS). In order to have an exhaustive overview of the impact of whales on the carbon cycle and their potential as an NCS, we need comprehensive studies that compare the carbon cycling of the entire food web in the presence or absence of whales.

(b) . Climate change and whaling footprint on whale-mediated carbon sequestration

Although we showed that whales have the ability to sequester non-negligible quantities of carbon, their sequestration capacities have been largely impaired by commercial whaling, with consequences extending well beyond the exploitation period. Indeed, due to their long life cycles, the recovery of whale populations after over-exploitation is a very slow process. Because of over-exploitation, their carbon sequestration capacity is currently limited in 2022 to 1.2 × 10⁵ tC yr⁻¹ (0.9 × 10⁵–1.4 × 10⁵ tC.yr⁻¹), i.e. 30% of the pre-exploitation level. This has created a sequestration deficit reaching between 41.9 × 10⁶ tC and 45.2 × 10⁶ tC on average in 2100, depending on the climate scenario. In addition to past whaling, the sequestration capacity of whales is now reduced by climate change. Indeed, the recovery of whale populations and of the carbon pump may be delayed and weakened by climate change as carbon sequestration barely reaches half of its historical value in the model with climate change. This is primarily explained by changes in the abundance and distribution of krill due to changing primary productivity patterns in the Southern Ocean [14]. Furthermore, the distribution of krill is expected to contract southward due to increasing temperature and reduced sea-ice extent [43]. This is particularly deleterious for whale species predominantly feeding in mid-latitudes areas (humpback whales, fin whales and southern right whales). Minke whales and blue whales could benefit from the ice-extent reduction in the Southern Ocean because of their ice dependency, assuming they can shift their distribution southwards to follow the krill. However, the higher abundance of minke whales, which have been less exploited, may reduce the prey availability for other species. Consequently, the MICE model predicts slower recovery for blue whales, decline for humpback, fin and southern right whales, but population increase for Antarctic minke whales during the twenty-first century in the scenario with climate change [14]. As a result, despite a predicted increase of minke whale populations, the total carbon flux would not return to its pre-exploitation level due to the negative impact of climate change on other species. A negative feedback loop between climate change and whale populations could therefore occur in the southern hemisphere.

On the other hand, thanks to their rapid increase, Antarctic minke whales could maintain, at least partially, the carbon sink and limit the carbon sequestration deficit due to over-exploitation. Their predicted increase throughout the twenty-first century would be a key resilience factor since they would allow a faster recovery of carbon flux towards its historical value. The asynchrony between the different population dynamics enables the overall carbon sequestration of the whale community to be more resilient, thus exhibiting a positive relationship between diversity and stability under the portfolio effect [44].

(c) . Restoration of whales and the associated carbon pump

Even though whales are vulnerable to climate change, primarily through the decrease in krill prey density and southward contraction of geographical distributions [43], their restoration could be promoted by mitigating current threats. In addition to the moratorium established in 1986 by the IWC that protects whales from commercial exploitation, restriction on maritime routes, fishing zones and authorized speed of boats can reduce the mortality and low reproduction rates associated with ship strikes and noise generated by boats. Promoting sustainable exploitation of krill, the whales' primary prey in the Southern Ocean, would also preserve whale populations and their associated carbon sequestration. Thus, the protection of both krill and whales, in particular under the legislation of the CCAMLR and the IWC, would help maintain whale populations and associated carbon sequestration.

Although our study does not account for the spatial distribution of whales, and thus, cannot map the associated carbon flux toward the depth, conservation strategies should be determined regarding their feeding, breeding grounds and migratory routes. Coupling tracking studies and assessment of the whale-mediated services like carbon sequestration could help identify areas of ecological importance and inform policies in the design of protected areas [45]. This would be especially relevant for fin whales and blue whales that were the main contributors to the carbon sequestration in the past, and that are currently categorized as ‘vulnerable’ and ‘threatened’ by the IUCN (see electronic supplementary material, figure S1).

(d) . Limits and uncertainties

While our estimations are subject to several uncertainties, we adopted a conservative approach. First, our study is restricted to five baleen species in the southern hemisphere, whereas there are 15 species of baleen whales globally. We considered here only baleen whales that were included in the MICE models [14,17], i.e. species commercially exploited in Antarctic waters, in most cases feeding predominantly on Antarctic krill and for which enough survey data were available. Therefore, our study may significantly underestimate the importance of carbon sequestration mediated by whales at global scales by excluding other southern species (Bryde's whale, pygmy right whale and dwarf Antarctic minke whale), northern species (bowhead whale, grey whale, Omura's whale and northern right whale) and toothed whales.

We used the most updated population dynamics and converted the number of individuals of each species accounting for sex and age variability in the body mass. However, we ignored seasonal variations in body mass. Indeed, these migratory species experience significant weight variations during the year [46]: they may gain several tonnes during the summer and be considerably thinner at the end of the breeding season. The amount of carbon sequestered therefore depends on the seasonality of natural mortality, which is not taken into account in our study.

Additionally, a main limitation in our study is the lack of empirical data on several processes. Indeed, the carbon content in whale tissues and the proportion of biomass reaching the deep ocean have not been experimentally measured on large whales. First, we used the same carbon content in whale tissues for each species, ignoring the inter-specific variability. Second, it was assumed that the biomass–carbon conversion does not change with carcass degradation. However, not all tissues have the same carbon concentration [47] and some (fat tissue, muscle) may be consumed primarily by scavengers [48]. In order to gain precision, it seems essential to determine the carbon level in the different types of tissue (bone, muscle, blubber and viscera) for each species. Finally, the proportion of biomass reaching the deep ocean before being consumed or remineralized is uncertain and probably highly variable (depending on the presence of scavengers or currents, for example). The estimates we used are conservative, thus we provide a lower bound of the potential whale-mediated carbon sequestration. Data collection on these crucial processes is therefore needed to refine our estimates.

5. Conclusion

We showed that despite efficient carbon sequestration capacities at their pre-exploitation levels, whale-mediated carbon sequestration has dropped dramatically because of commercial whaling and will be far from complete recovery by the end of the century as a result of climate change. Therefore, our results call for protection and restoration of whale populations as a potential BCE providing opportunities for climate change mitigation [49–51]. Indeed, further research should refine the quantification of their carbon sequestration potential and explore other sequestration pathways, especially those mediated by bottom-up processes like nutrient cycling. Further identifying and quantifying the carbon sequestration capacities of marine vertebrates in general would provide additional evidence to support the protection of 30% of the oceans by 2030, a new target proposed to the United Nations [52]. In the case of whales, protection should be designed as a meaningful assemblage including feeding and breeding grounds and migratory routes [45]. In this way, restoring marine vertebrates could contribute to achieve our climate objectives while generating other services beneficial to the functioning of the biosphere and the well-being of human societies [53,54].

Data accessibility

We use existing outputs from already published models: https://doi.org/10.1111/gcb.14573 [14]. The underlying codes of this paper are provided as electronic supplementary material (.py files and .ipynb file) [55].

Authors' contributions

A.D.: formal analysis, methodology, visualization and writing—original draft; G.M.: conceptualization, methodology, supervision, validation and writing—review and editing; V.T.: resources and writing—review and editing; M.S.S.: methodology, resources and writing—review and editing; M.T.: conceptualization, methodology, supervision, validation and writing—review and editing; D.M.: conceptualization, methodology, supervision, validation and writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

We received funding from the Institut Universitaire de France for this study. M.S.S. was supported by MAC3 Impact Philanthropies.