Biological Carbon Sequestration: From Deep History to the Present Day

1. Introduction

One of the most important biogeochemical processes on Earth is the global carbon cycle ([1,2]. Carbon compounds are the basis of all life on the planet and are also important components of most non-living systems including rocks, oceans and the atmosphere. In the global carbon cycle, various forms of carbon flow between these systems in response to ‘natural’ external drivers including biological activities, geological forces, climate, and planetary orbital variations. Over recent decades a new factor has emerged which has significantly perturbed the global carbon cycle, namely anthropogenically related carbon emissions to the atmosphere that are not balanced by carbon sequestration into land sinks [3,4]. An estimate of the current status of the carbon cycle, also known as the global carbon budget, is shown in Figure 1. Most of the ~1,200 teratonnes (10¹² tonnes) of terrestrial carbon is locked up as mineral deposits in the earth’s crust. A further ~38 teratonnes are in the oceans, ~2.4 teratonnes in the biosphere and ~0.9 teratonnes as atmospheric CO₂ [5].

Carbon sequestration is defined by the IPCC as ‘the process of increasing the carbon content of a pool other than the atmosphere’ [6,7]. Unfortunately the term has frequently been misused, particularly in the context of climate change mitigation studies [8]. For example, unless a temporal dimension is included it can be unclear whether it applies to long-term processes, such as laying down fossil deposits over millions of years, or short-term processes, such as temporary sequestration, e.g. when CO₂ is fixed to form a plant leaf that is then shed and the carbon re-emitted within a few weeks.

The negative balance between anthropogenically related greenhouse gas emissions and carbon sequestration by plants has led to a >50% increase in atmospheric CO₂ levels from about 280 ppm in the late 19^th century to 424 ppm in 2024. Comparably high levels of atmospheric CO₂ have not been seen for over 16 million years, at a time when the Arctic was ice-free and Iceland had a humid subtropical climate [9]. It is widely acknowledged that elevated anthropogenic CO₂ concentrations are contributing to increasing average global temperatures and related climatic changes, such as droughts, that are adversely affecting human societies and ecosystems around the world. One of the main strategies to tackle this so-called ‘climate emergency’ is to decrease emissions of CO₂ and other greenhouse gases, such as methane (CH₄), for example by ‘de-carbonising’ many transport and industrial processes.

However, somewhat less attention has been paid to the other side of the carbon balance, namely the role of biological carbon sequestration by photosynthetic organisms that act as natural carbon sinks, for example the perennial vegetation in tropical agroforestry systems and coastal wetlands [10]. Ever since life first evolved on Earth there has been a dynamic interplay between geological, climatic and biological processes that has resulted in fluctuations in the balance between atmospheric carbon sources and sinks. To a great extent these processes have driven the course of biological evolution, for example by creating an oxygen-rich atmosphere that enabled the development of complex eukaryotic organisms, such as plants and animals, and their subsequent colonisation of the land surface.

More recently, a new factor has emerged in global carbon dynamics, namely a range of large-scale anthropogenic activities leading to greatly accelerated releases of gaseous carbon compounds. It is only in the past few decades that the potential scale of this anthropogenic carbon release has been fully recognised [11]. Even if true carbon-neutrality is achieved later in the 21^st century (which appears increasingly unlikely), atmospheric CO₂ levels are predicted to continue rising well into the next century, possibly to as high as 1,500 ppm [12]. These high CO₂ levels are likely to persist for millennia thereafter unless robust measures are implemented to reset the carbon balance in favour of improved rates of carbon sequestration, preferably via the biological process of photosynthesis. In this review, the contribution of biological carbon sequestration to atmospheric CO₂ levels over geological time until today will be examined.

2. Mechanisms of Carbon Sequestration

Carbon sequestration can be defined as the net absorption of carbonaceous greenhouse gases, principally CO₂ and CH₄, from the atmosphere and oceans. The contemporary atmosphere is principally made up of gaseous nitrogen (78%) and oxygen (21%), but it also contains about 424 ppm (0.042%) CO₂ and 1.9 ppm methane. These two greenhouse gases can be released into the environment via a variety of anthropogenic and non-anthropogenic (‘natural’) processes. Examples of anthropogenic processes include fossil fuel combustion for heating, cement production, transport and many industrial processes, crop and livestock rearing, removal of coastal wetlands, drainage of peatlands and deliberate burning of vegetation. Examples of non-anthropogenic processes include vulcanism and outgassing from terrestrial and oceanic sources plus breakdown of organic matter by saprotrophs and detritivores [13,14].

Over the past century, and particularly after 1950, anthropogenic processes were responsible for a huge acceleration in the rates of carbonaceous greenhouse gas emissions [11,15,16]. (Ito, 2019; IPCC, 2023; Jones et al, 2023). This was principally due to combustion of previously sequestered carbon in the form of coal, petroleum and natural gas. In 2023, estimated emissions of fossil-derived CO₂ totalled over 35 gigatonnes (10⁹ tonnes), respectively from coal (44%), petroleum (34%), and natural gas (22%) deposits [17]. In contrast to carbonaceous greenhouse gas emissions, nearly all of which are of anthropogenic origin, the vast majority of carbon sequestration occurs via non-anthropogenic ‘natural’ processes, the vast majority from photosynthesis [4].

There are numerous research programmes into anthropogenic strategies for non-biological carbon sequestration. Most of these approaches involve various geological technologies using largely experimental processes that have only been studied or implemented on a relatively small scale. Examples include mineral carbonation [18], graphene production from CO₂ [19], Carbon Capture and Storage (CCS) [20,21], and use of engineered molecules such as MXenes (ternary carbides, nitrides, or carbonitrides), MBenes (transition-metal borides) [22] and tetrahedral organic cages [23]. In the future such approaches might become scalable to global significance, but for the time being they are of only minor practical use for the urgent reduction of the ever-increasing levels of atmospheric CO₂ and methane [24].

3. Biological Carbon Sequestration

Biological carbon sequestration is a non-anthropogenic process that has been modulating global CO₂ levels for billions of years. It largely relies on biological processes, and most notably on CO₂ fixation via terrestrial and aquatic photosynthesis. Photosynthesis is an ancient biological process that originally arose in bacteria, possibly as much as 3.8 billion years ago, but is now mostly carried out by plants and algae [25]. It is estimated that approximately 40 gigatonnes of CO₂ is emitted annually, of which 46% remains in the atmosphere, 30% is sequestered via terrestrial photosynthesis and 24% via aquatic photosynthesis [26,27]. Hence biological carbon sequestration is already removing 22 gigatonnes/yr of CO₂ from the atmosphere.

Estimates of the precise amount of biological carbon sequestration predicted over the coming decades are subject to many uncertainties. For example, ongoing climatic changes could affect the future rates at which biologically sequestered carbon is either formed or broken down. Hence, it is possible that higher levels of ambient CO₂ could lead to elevated rates of photosynthesis, giving rise to enhanced carbon fixation and sequestration. This is the so-called ‘fertilisation effect’ [28,29,30,31]. In one Earth system model that focused on high latitude regions, the net photosynthetic carbon fixation in such regions was about 60% higher than previous estimates [32]. In contrast, in warmer regions elevated temperatures could lead to more rapid rates of vegetation decomposition, mainly via fungi and bacteria, resulting in accelerated release of both CO₂ and CH₄ into the atmosphere [33,34,35].

In addition to climate-related changes, rates of biological carbon sequestration are affected by changing land use patterns whereby strong carbon sinks, such as forests, are being replaced by less effective sinks via deforestation and woodland degradation [26,36,37,38,39,40,41,42,43]. The carbon sequestration effectiveness of new land cover after deforestation varies greatly according to the vegetation type. For example, large perennial crops such as oil palm or rubber are more effective carbon sinks than relatively small-sized annual crops such as soybean or wheat. The scale of deforestation is shown by the loss of a net of 130 million hectares from 2010 to 2015 [44,45]. Not only did this generate an annual 3.7 gigatonnes of additional CO₂ emissions [46], it also drastically reduced any future capacity for carbon sequestration.

Despite the above statistics, however, decreases in global land use-related emissions of about 2 gigatonnes of CO₂ since 2000 were reported [17]. This may be largely due to increasing crop yields since 1961 that substantially reduced new land use requirements, possibly by as much as 1.7 billion hectares [47]. In some cases, degraded forest can be subjected to landscape restoration, including re-afforestation, to restore functions such as increased biodiversity and mitigation of negative climatic effects, for example by maximising carbon sequestration. Data from a large metanalysis indicated that the most effective strategy for maximising carbon sequestration in restored habitats was the replanting of forests or woodlots, which could remove up to 41.7 tonnes CO₂/ha/yr over a 20 year period [48]. Interestingly, natural regeneration removal rates were between 9.2 and 18.8 tonnes CO₂/ha/yr while agroforestry rates were between 10.8 and 15.6 tonnes CO₂/ha/yr [48]. While the effectiveness of native tropical forests in biological carbon sequestration is clear, there is some uncertainty about the extent to which this may be offset in contemporary forested regions, for example where anthropogenic activities have already degraded some of their ecosystem functions [39].

Another possible source of biological carbon sequestration is by perennial tree crops, either as part of agroforestry or in single-crop plantations. Much of the carbon fixed by these trees is converted into woody biomass with a lifespan that typically extends to several decades or longer. It is estimated that biological carbon sequestration currently captures about two gigatonnes CO₂ per year [49]. To put this figure in perspective, in order to get back to the pre-1900 atmospheric CO₂ levels of under 300 ppm it would be necessary to remove more than 900 gigatonnes CO₂ via carbon sequestration. However, if the aim is simply to maintain a value of about 400 ppm atmospheric CO₂, new measures to increase biological carbon sequestration rates beyond two gigatonnes CO₂ per year might be useful to implement on a long term basis. In the next three sections we will discuss the dynamics of carbon sequestration over the course of Earth’s 4.54 billion year history.

4. Photosynthesis and Carbon Sequestration in the Early Earth

This section is concerned with the development of oxygenic photosynthesis and its pivotal role in carbon sequestration. During the first four billion years of Earth’s existence there were huge fluctuations in the trajectory of biological carbon sequestration across the globe, none of which involved humans, who had yet to evolve. The vast majority of biological carbon sequestration occurs via photosynthesis, which first arose between three and four billion years ago [25]. Photosynthesis is a biochemical process involving the use of light radiation (typically in the wavelength range of 400-700 nm) to reduce CO₂ into complex organic biomolecules such as carbohydrates. During oxygenic photosynthesis by plants, algae and cyanobacteria, light energy splits water molecules releasing electrons that reduce atmospheric or aquatic CO₂ to carbohydrates ([CH₂O]_n) with the evolution of dioxygen (O₂).

Within a few million years of the Earth’s formation about 4.54 billion years ago, its atmosphere had stabilised to comprise mainly CO₂ and other partially oxidized gases such as CO, N₂, plus a mixture of reduced gases such as H₂, CH₄, NH₃, and H₂S. Due to the presence of highly reducing minerals, there was no free oxygen on the early Earth. However, by 4.5 billion years ago the planet had cooled sufficiently to enable large amounts of free water to accumulate, creating extensive oceans and releasing large amounts of water vapour into the atmosphere. These conditions facilitated the evolution of anaerobic life and it is likely that photosynthesis arose in bacteria soon after the start of the Archean Eon, possibly as early as 3.8 billion years ago [25,50].

The form of photosynthesis that evolved in cyanobacteria involved the use of sunlight to split water which allowed oceanic, and later atmospheric CO₂, to be reduced into carbohydrates and more complex biomolecules, such as proteins and nucleic acids, releasing oxygen as a byproduct. This was the first form of biological carbon sequestration whereby CO₂ was gradually removed from the oceans and atmosphere. Not all the photosynthetically fixed CO₂ is truly sequestered, i.e. removed from circulation as stored organic carbon. As much as half of recently fixed carbon is soon lost due to aerobic respiration, which results in the re-release of some fixed CO₂ into the oceans or atmosphere. Respiration is essential for cellular metabolism and therefore occurs constantly in all photosynthetic organisms [51]. After a rather slow start, both photosynthetic carbon sequestration and oxygen evolution eventually had far-reaching global geological and biological impacts as shown in Figure 2.

During the Archean Eon (2.4 – 4.0 billion years ago), oxygenic photosynthesis was confined to relatively slow-growing oceanic cyanobacteria. During this time, increasing rates of bacterial carbon sequestration resulted in a 100-fold decrease in atmospheric CO₂ concentrations from about 10% to 0.1% (the current value is 0.043% and rapidly rising). In contrast, oxygen levels remained very low because almost all the oxygen evolved via photosynthesis was immediately removed by mineral iron deposits. The extremely high levels of atmospheric CO₂ (and CH₄) in much of the Archean would have exerted a powerful greenhouse effect. This enabled liquid water to be present on Earth, despite the fact that the early solar output was only about 70% of current values [52,53]. The gradual sequestration of atmospheric CO₂ during and after the Archean therefore involved both biological and geological processes [53,54].

Soon after the start of the Proterozoic Eon (0.8 – 2.4 billion years ago), photosynthetic oxygen release by cyanobacteria exceeded the storage capacity of iron deposits and vast amounts of oxygen gas were released into the oceans and atmosphere. This ‘Great Oxidation Event’ (GOE) irreversibly altered the trajectory of biological evolution and created the oxygen-rich planet we live in today. Oxygen was toxic to the anaerobes that comprised the bulk of early life and GOE resulted in probably the greatest mass-extinction event during Earth’s history with the very few surviving anaerobes being relegated to marginal habitats such as volcanic vents. However, aerobic cyanobacteria thrived and the plentiful oxygen that they produced enabled much larger aerobic eukaryotic organisms to thrive. In one momentous episode a heterotrophic eukaryotic cell captured a cyanobacterium to create the first photosynthetic algal cell. This green alga was much larger and faster growing than cyanobacteria and was also the precursor of all of our land plants [25].

Over the next two billion years, as photosynthetic algae thrived and diversified, they sequestered ever more carbon (some of which later accumulated as natural gas and oil deposits as discussed below). These algae also released more oxygen in a second GOE that further supported the evolution of large multicellular life forms including the earliest animals. By the end of the Proterozoic, about 800 million years ago, the increased rate of biological carbon sequestration, mainly by algae, had lowered atmospheric CO₂ levels to about 1,000 ppm. This value is not far from the some of the climatic model estimates for likely CO₂ levels at the end of the current century [55,56]. The massive proliferation of algae and cyanobacteria during Proterozoic led to the formation of the Earth’s earliest coal deposits, some of which may be 1-2 billion years old [57].

5. Photosynthesis and Carbon Sequestration by Land Plants from 500 Million Years Ago

The next landmark in the history of biological carbon sequestration came around 500 million years ago with the establishment of a protective stratospheric ozone (O₃) layer that greatly reduced the amount of damaging high-energy radiation reaching the ground. This enabled some algae to colonise the land surface and evolve into plants. Within 100 million years, increasingly large and complex vascular plants, such as club mosses and ferns, had evolved and the first forests appeared around 400 million years ago [25,58,59]. In general, plants are faster growing and more effective at carbon sequestration than algae, with perennial woody species, such as trees, being by far the most efficient organisms in terms of the removal of atmospheric CO₂ and its long-term sequestration as organic carbon deposits. This is because trees convert much of their fixed CO₂ into lignified tissues such as wood, which first appeared during the Devonian Period about 400 million years ago [60,61].

Living wood can exist for decades or even centuries with some (admittedly exceptional) tree species having lifespans of over 5,000 years. Moreover, although dead wood is normally subject to biological decay within a few years, in some cases it can remain intact for many millennia. For example, recently discovered wood deposits at a waterlogged site in Zambia are over 476,000 years old [62]. In some cases the decay of dead wood and other plant organic matter, such as periderm, is delayed or even halted as these materials accumulate in anaerobic wetlands where biodegradation or oxidation may not occur. This can lead to accumulation of vast quantities of sequestered organic carbon in the form of peat, which can be stable for many millennia or longer. Depending on geological conditions, peat deposits can sometimes be further compressed and mineralised to produce carbonaceous materials, such as coal, oil or gas, which may be stable for hundreds of millions of years.

The immense power of land plants in terms of long term carbon sequestration is dramatically illustrated by the accumulation of many teratonnes of plant-derived coal deposits during the late Carboniferous and early Permian Periods, around 323-252 million years ago [63,64]. The exact causes of this event, during which 90% of all of the Earth’s coal was deposited, are still uncertain, but are likely to involve a combination of climatic, geological and biological factors. The massive increase in biological carbon sequestration around 300 million years ago was initially due to accelerated rates of photosynthesis by the prevailing vegetation at a time when the climate was especially favourable with plentiful water and relatively high atmospheric CO₂ levels. The vegetation mainly consisted of a mixture of large, tree-like vascular plants related to contemporary club mosses, ferns, horsetails and seed plants [59].

These arborescent species formed extensive so-called ‘coal forests’ in coastal and inland wetlands throughout the world. When they died their organic remains were buried in sediments where they were transformed over millennia into peat, lignite, and, eventually, coal [63]. The huge increases in photosynthesis in the proliferating coal forests, coupled with a much reduced rate of organic decomposition, radically altered the Earth’s carbon balance as the vast amounts of CO₂ removed from the atmosphere (about 100 gigatonnes/year) were not balanced by releases due to decomposition. Most estimates show that atmospheric CO₂ was relatively high at about 600 ppm at the beginning of the event but had fallen to a low of 100 ppm by 297 million years ago [64]. The removal of such a vast quantity of CO₂ from the atmosphere greatly reduced its greenhouse function and brought it dangerously close to the threshold of 50 ppm, below which global glaciation to create a ‘snowball Earth’ could take place [64].

The consequences of the late Carboniferous explosion in both photosynthetic activity and long term carbon sequestration are very much still with us today in the form of the vast coal deposits laid down about 300 million years ago and still largely present in the ground. Estimates of the total Carboniferous coal deposits vary considerably but they probably run into several teratonnes of which 400 gigatonnes have been mined to date and the carbon returned to the atmosphere as CO₂. An estimated 968 gigatonnes of recoverable coal reserves remain in the ground, which is sufficient for a further 130 years of extraction at current rates [65]. Despite environmental concerns, global coal extraction continues to increase at an estimated compound annual rate of 5.35% projected for 2024-2028 with 8.6 gigatonnes due to be extracted during 2024 [66].

In contrast with coal, most petroleum and natural gas deposits were formed from the remains of marine organisms, such as algae, at various times between 65 and 250 million years ago [67]. The remains of these organisms were compressed and heated by geological processes converting them into a solid insoluble organic material called kerogen that is found in sedimentary rocks [68]. Although kerogen is the ultimate source of petroleum and natural gas deposits (including shale oils), the vast majority of this form of biologically-derived sequestered carbon remains inaccessible via available extractive technologies [69]. It is estimated that kerogen deposits contain as much as 10¹⁶ tonnes of carbon, 90% of which is unlikely ever to be commercially available for extraction [69]. However, the amount of commercially recoverable petroleum and natural gas deposits continues to increase as more deposits are discovered, with a 2021 estimate of 244 megatonnes for petroleum reserves alone [70].

Recent research has enabled more precise estimates of the variability of atmospheric CO₂ levels during the more recent geological periods from 55 to 5 million years ago. In some of the examples discussed above, such as the late Carboniferous events, biological factors probably played an important role in such fluctuations. However, non-biotic processes, for example geological and/or climatic changes variations in the Earth’s orbit, have also been crucial drivers of CO₂ dynamics [56,71]. In general there has been a gradual, albeit punctuated, decline in atmospheric CO₂ levels between the mid-to-late Eocene epoch (33.5 – 55 million years ago) and the 20^th century CE [72].

Hence, during the Eocene atmospheric CO₂ levels typically ranged between 1,000 to 1,600 ppm. During the Oligocene epoch (23.5 – 33.5 million years ago), CO₂ levels ranged from 500 ppm to 1,000 ppm and more recently in the Miocene epoch (5 – 23.5 million years ago), they had dropped to around 400 to 500 ppm. Subsequent geochemical processes then led to the transport of much carbon to the oceans where it remained. The overall effect of these processes was the non-biotic removal of large amounts of CO₂ from the atmosphere and its eventual long term sequestration in marine systems. As shown by the NASA data in Figure 3, this trend continued into the more recent Pleistocene epoch when atmospheric CO₂ levels averaging about 220 ppm, with a range between 170 to 300 ppm, were the norm for almost one million years.

Modern humans, Homo sapiens, first arose in Africa about 200,000 years ago. However, they remained as relatively small, isolated populations of hunter gatherers until the development of agriculture and other technologies during the Holocene epoch [25,73]. The only significant, albeit regional, ecological impact of pre-Holocene humans seems to have been a series of localised megafaunal extinctions after about 120,000 years ago, although none of these events involved photosynthesis or carbon sequestration [74]. As we will see in the next section, during the almost the entirety of the current Holocene epoch, CO2 levels continued at low values of below 300 ppm until the massive increases following the more recent Anthropocene events around the 1950s CE.

For the ~4.54 billion years prior to the onset of the Holocene, global carbon sequestration was an exclusively ‘natural’ (non-anthropogenic) process driven primarily by biological and geochemical mechanisms. This began to change about 12 millennia ago when the favourable climatic interlude of the Holocene created conditions whereby human activities gradually started to perturb the carbon cycle. Firstly, the development of agriculture drastically changed floral and faunal biota in ways that diminished biological carbon sequestration. Secondly, the later development of extractive technologies enabled the release of vast quantities of fossil carbon deposits and their combustion to provide energy. This led to the depletion of carbon deposits with the release of carbonaceous greenhouse gases, principally CO₂ and CH₄ in a process that was greatly accelerated after the year 1950. As a result, there has now been a decisive shift in the global carbon cycle in favour of CO₂ emissions and away from carbon sequestration. This shift is predicted to have significant climatic effects that could threaten food production and social stability. In the following two sections, the disruption of the carbon cycle during the Holocene and the current status of biological carbon sequestration will be discussed.

7. The Post-1950 Inflection Point or ‘Anthropocene Event’

The inflection point in net CO₂ and methane emissions around the year 1950 (see Figure 6) has led to the designation by many researchers of a new stage in the Earth’s evolution called the ‘Anthropocene’. There is a vigorous scientific debate around the validity of the concept of the Anthropocene as a bona fide geological stage, for example as a new human-dominated epoch that follows the Holocene [86,87,88,89,90,91,92]. Rather than trying to fix the Anthropocene as a time-constrained geological time period, it might be more productive to characterise it as what Walker et al. have recently called “an ongoing, intensifying, diachronous event” [93,94]. In the context of this review, the term ‘Anthropocene’ can be usefully employed to describe the post-1950 events, with the caveat that these events are an integral part of a series of earlier human-related processes that date back at least to the beginning of the Holocene, and possibly earlier. The Holocene remains as the current geological epoch with the Anthropocene simply marking new global processes that started around the year 1950.

There has been some debate about the extent to which the late Holocene increases in CO₂ and CH₄ emissions are indeed anthropogenic, or whether other factors are partially or even principally to blame [85,95]. However, the current scientific consensus decisively supports the anthropogenic case [11,96]. The linkage between accelerated CO₂ emissions after 1950 and an increase in global surface temperatures is also clear, with an inflection point in the mid-1960s after which average surface temperatures rose much more rapidly than in the previous century [11,15]. The IPCC estimates that the majority of the global greenhouse gas emissions (i.e. 56%) are produced by the energy sector and industry, plus 15% by transport and 6% by buildings. Interestingly, and despite its obvious importance as a carbon sink (due to photosynthesis), agriculture was also deemed responsible for a surprisingly high proportion of emissions at 22% of the total, primarily due to deforestation [11].

The world is currently experiencing record levels of anthropogenic greenhouse gas emissions, coupled with a reduced capacity for carbon sequestration, which mostly occurs via photosynthesis. In 2023, it was estimated that the global figure for annual carbon sequestration by land-based plants was 6.6 gigatonnes of CO₂ equivalents, but this was massively exceeded by annual anthropogenic emissions totalling 59 gigatonnes of CO₂ equivalents [11]. The rate of marine photosynthesis, mainly by algae, is estimated to be approximately equal to that of land plants [97], but this would only increase the overall biological carbon sequestration figure to about 13 gigatonnes of CO₂ equivalents, which is a mere 22% of the anthropogenic greenhouse gas emissions.

Also in 2023, a new record atmospheric CO₂ concentration of 424 ppm was recorded by the NOAA Global Monitoring Lab in Hawaii [98]. This figure was equivalent to total CO₂ emissions of 40 gigatonnes during the previous 12 months. To put this figure into a historical context, ice core data show that average atmospheric CO₂ concentrations over the past 800,000 years have been in the range of 180 to 220 ppm [99]. Due to its potency as a greenhouse gas, the unprecedentedly high CO₂ concentrations would be expected to contribute to an increase in average global temperatures, as predicted by Arrhenius well over a century ago [100]. A more recent analysis of CO₂ levels over the past 66 million years is even more compelling [101]. Hence, at about 50 million years ago atmospheric CO₂ concentrations were estimated at ~900 ppm, which is close to the projected 2100 level of 800 ppm at current rates of increase [102]. At 50 million years ago the mean global surface temperature was 12°C higher than now and Antarctica and Greenland were ice-free.

The last time that CO₂ concentrations were consistently higher than today was during the Middle Miocene Climatic Optimum (MMCO) (about 16 MYA) when dense rainforests covered much of northern Europe, Greenland was unglaciated, and the sea level was some 50 m higher than the present day. Due to the similarity between the elevated Miocene CO₂ concentrations and the projected values later in the 21^st century, the Miocene conditions have been used as possible models for a future scenario for Earth [9]. It did not come as a surprise, therefore, when in April 2024 the EU Copernicus Global Climate report highlighted that average global temperatures for the previous ten months were the hottest on record, exceeding 1.5°C higher than pre-industrial levels, sometimes surpassing values in excess of 2.0°C higher [103].

The current prognosis for anything approaching a reversal of the 1950 inflection point in global carbon emissions and temperature rises during the 21^st century appears remote. While some public and private actors are gradually implementing strategies to decrease anthropogenic emissions these initiatives are nowhere near likely to be at the scale required to halt the inexorable rise of both carbon emissions and the related adverse climatic consequences. An important question for humanity as a whole is: does this climatic uncertainty really matter as long as we can still produce enough food for ten billion people in 2050? The short answer is that we cannot yet foresee the consequences of the current climatic trends on downstream factors such as food production and social cohesion at a global scale but if the Holocene climate is disrupted the consequences could be far reaching.

There is a growing body of research that shows how previous climatic shocks during the Holocene often had dire consequences at a local level, including the extinction of complex agro-urban cultures such as the Maya and Harappans[77]. On the other hand, some cultures were more resilient as they retreated to subsistence food production and abandoned or downsized their urban centres [81,82,84,104]. Farmers were also able to mitigate the effects of climate by changing the suite of crops to suit the prevailing conditions. For example, during the climatic shock of 4,200 years ago, Mesopotamian wheat cultivation declined due to increased aridity and salinity but farmers switched to less palatable but more drought- and salt- tolerant barley crops [77].

However, as we progress through the Anthropocene we will be facing a new series of challenges beyond what has occurred hitherto. Firstly, all of the previous examples of Holocene climatic changes were relatively localised whereas the current predicted climatic changes will be at a global level. Secondly, all of the previous examples of successful human resilience occurred in pre-industrial times at a time when food production was strictly local and populations were a tiny fraction of today’s eight billion. Thirdly, the predicted temperature and rainfall changes would likely dwarf those of previous Holocene episodes and could threaten the viability of some of the most important global staple crops. In this regard we are significantly lacking in resilience in terms of crop diversity.

Indeed, it is a remarkable fact that the vast majority of our food production still relies on the tiny number of crops and livestock species that were domesticated at the beginning of the Holocene. For example, just three cereal crops (rice wheat and maize) provide over half of all food calories consumed by people today [105]. Wheat is particularly vulnerable to drought and breeders are now using methods such as genome editing to produce drought tolerant varieties [106,107,108,109]. The situation is even more marked for livestock with just three species (chickens, pigs and cattle) providing 92% of global meat consumption [110]. The extreme dependence on such a small number of crop and livestock species for global food supplies means that the system is highly vulnerable to perturbations, such as climate shifts or new diseases. These could drastically reduce the availability just one of these foods and one option is to domesticate new crops, especially in the tropics [111].

Finally, the extreme inter-connectedness of our eight billion global population in terms of food production and international trade makes the current system highly vulnerable to a sudden shock that could at best degrade its functionality and at worst lead to its collapse. For these reasons, the preservation of Holocene climatic stability by restoring the global carbon cycle to its pre-industrial status is an urgent priority. The important topic of strategies aimed at enhancing the biological carbon sequestration capacity of the Earth’s ecosystems, both ‘natural’ and anthropogenic, is the topic of a separate study that will be published in the near future.

8. Conclusions

• Over the past ~4.5 billion years the global carbon cycle has fluctuated considerably as geological and biological factors modulated the balance between carbon sequestration from and emission into the atmosphere.
• About 300 million years ago atmospheric CO₂ levels declined considerably following carbon sequestration due to coal formation and global cooling. This was followed by an erratic climate, especially over the past million years when CO₂ levels averaged ~220 ppm.
• The Holocene epoch beginning ~11,700 years ago ushered in a period of unusual climatic stability that enabled humans to develop agro-urban cultures and advanced technologies.
• This resulted in a gradual decline in carbon sequestration as cropland replaced natural vegetation and increased CO₂ emissions due to fossil fuel combustion. By ~1950 an inflection point was reached with anthropogenic CO₂ emissions rising exponentially while carbon sequestration remained static or declined.
• In 2023, plants sequestered 6.6 gigatonnes CO₂ eq, but this was greatly exceeded by anthropogenic emissions of 59 gigatonnes of CO₂ eq. These emissions are a major factor in ongoing climatic changes that could undermine the Holocene climatic stability that underpins food production for billions of people.
• In order to mitigate this situation it will be necessary to restore the global carbon cycle by drastically reducing emissions and increasing carbon sequestration. As part of this strategy the IPCC has set a carbon sequestration target of 10 gigatonnes/yr plus measures to reduce emissions via decarbonisation.
• The most effective carbon sequestration agents include coastal wetlands, tropical forests and tree crops; hence their removal should be avoided and high-yield perennial tropical food crops encouraged.