The previous post, “New Global Carbon Data Revives the 'Missing Sink' Problem,” generated considerable interest and I would anyway like to discuss this topic further. As mentioned in the comments, it is a true detective story—one that is intellectually rewarding. To recap, we examined how the study by Bar-On et al. (2025) revealed that a substantial share of CO₂, which could have accumulated in the atmosphere due to fossil fuel burning, is stored not in live tree biomass, as previously believed, but elsewhere (the so-called "missing sink"). I argued that this is not merely a technical or conceptually inconsequential finding—it pertains to the very fiber of our understanding of what life is.

Why the missing sink could not be predicted by ecological science

The dominant civilization of today was built by humans who lived outside their natural ecological niche—in cold, inhospitable climates with low biological productivity. Limited by environmental scarcity, to which they had continuously adapted, humans applied this same principle to understanding the rest of life.

They viewed life as governed by limitation and adaptation. These two principles are closely linked: if your functioning is constrained by something beyond your control, you must adapt to that limitation. In this view, life—constrained by external environmental conditions—must constantly adapt to them. (Alternatively, if you are in control of your environment, you will face fewer changes to which you would need to adapt.)

Struggling to improve their living standards, humans sought ways to increase biological productivity, enabling them to work less. To achieve this, they extracted a few particularly promising plant species from their natural environments and began experimenting with ways to enhance their productivity. These experiments led to the formulation of what is now known as Liebig’s Law:

Growth is dictated not by total resources available, but by the scarcest resource (the limiting factor).

According to Liebig’s Law, plant growth can be enhanced by adding to the soil the nutrient that is in shortest supply relative to the others. Such nutrients are called fertilizers. As we all know, commonly used fertilizers contain nitrogen and phosphorus.

Since nitrogen and phosphorus do enhance plant growth, Liebig’s Law implies that growth is not limited by carbon and, therefore, cannot be increased by simply raising atmospheric CO₂ levels. This was the first reason why a biotic carbon sink was neither expected nor predicted by ecological science. (There are no links between Liebig’s Law and the CO₂ fertilization effect on Wikipedia—indeed, the contents of these pages are in logical contradiction.)

The second, and probably more important, reason is that there are also heterotrophs that consume what plants produce. These heterotrophs — bacteria, fungi, and animals — are, according to the same view of life, limited by food. This view presumes that, even if plants were to synthesize more food as CO₂ increased, heterotrophs would consume the excess, preventing the formation of a net sink of organic carbon.

Thus, the conventional paradigm of life being limited by external factors effectively doubles the reason for a biotic carbon sink being unlikely. (Imagine how comfortable it would be if, instead of a missing carbon sink, the law of matter conservation pointed to a missing carbon source? In that case, the ecological science would readily offer an explanation: the respiration of heterotrophs is limited by temperature. As global warming progresses, this temperature rises, causing heterotrophs to respire more soil carbon, further contributing to fossil emissions…)

Why it matters that the missing carbon is not stored in live trees

In an effort to explain the missing sink, there have been attempts to circumvent the fact that terrestrial vegetation is not carbon-limited. For example, exposing plants to elevated CO₂ can lead to short-term growth enhancement at the expense of depleting nitrogen and/or phosphorus. These efforts are comprehensively reviewed in the paper “Integrating the evidence for a terrestrial carbon sink caused by increasing atmospheric CO₂” by Walker et al. 2021 in New Phytologist. This paper is in free access, and I highly recommend it. Its informative abstract accurately reflects the contents, and it concludes that

Global carbon budgeting, atmospheric data, and forest inventories indicate a historical carbon sink, and these apparent iCO₂ responses [i.e., responses to increasing CO₂] are high in comparison to experiments and predictions from theory. Plant mortality and soil carbon iCO₂ responses are highly uncertain.

But let us assume, for the sake of argument, that trees do indeed synthesize significantly more organic carbon in response to increasing atmospheric CO₂. Even then, this does not solve the missing sink problem in the conventional paradigm. We still need to explain how it is that other organisms in the biosphere did not consume all the added organic carbon. One possible explanation is that all the added carbon is currently stored in the form of live tree biomass, and thus is actively protected by the trees from being consumed by other organisms.

The ecological science, hesitant to take on the scientific responsibility of explaining the missing sink but left with no other option, seized this opportunity—one that could at least be somewhat reconciled with the principle of limitation. Accordingly, as I mentioned earlier, sophisticated Dynamic Vegetation Growth Models were developed to describe how the extra synthesized carbon is mostly allocated to live tree biomass. (My observation, from different fields of science I’ve been working in, is that models will simulate anything, provided sufficient resources are invested.)

The study by Bar-On et al. (2025) has shown that the live biomass of trees has not been increasing, contrary to the predictions of the models. Below is a graph I redrew from Figure 1 of Bar-On et al. (2025) (I'm unsure if I can reproduce their original figure as it is copyrighted by Science). It highlights the mismatch between the models and actual observations. While the uncertainties (not shown) are generally large, the difference in trends is conspicuous.

Now that it turns out the extra carbon must be stored in non-living forms, such as soil or dissolved organic matter—none of which are protected like live biomass—this compensatory response to increasing carbon dioxide falls completely outside the scope of what conventional ecological views can explain. It is a crucial blow to the principle of limitation. It's no wonder that Professor Pacala referred to the missing sink as the “most vexing problem in global change science.”

The global carbon budget re-partitioned

Before we move on to explore what is actually happening, let's pause once more to examine the global carbon budget. The graph below is for the years 2009-2018 following Friedlingstein et al. 2019.

There is a source of 1.5 GtC/year related to “land use change,” and a sink of 3.2 GtC/year referred to as “land uptake.” The source, historically reflected in global carbon studies, largely corresponds to net deforestation. “Net” means it accounts for both deforestation and the re-growth of trees in previously deforested areas. Taken together, these figures imply that the 3.2 GtC/year sink is maintained by ecosystems that have not been disturbed by deforestation.

Bar-On et al. (2025) combined all vegetation without distinguishing between deforested and non-deforested ecosystems, showing that vegetation biomass has remained nearly constant from 1992 to 2019. This does not necessarily mean that there has been no extra growth of tree biomass in undisturbed ecosystems—it could still be happening. However, it does suggest that a significant carbon sink, estimated at 1.7 GtC/year for 2009-2018, occurs in other forms.

Biotic regulation of the environment

If a particular biological species, when isolated from its ecological community, is indeed limited by the need to find resources to survive, an ecosystem as a whole is not. This is because most of the materials life depends on are provided by life itself. We all know that plants synthesize food for us from atmospheric carbon dioxide, but what may be missing from our collective understanding is how short-lived that stock of carbon is.

With its global net primary productivity of 100 GtC/year, global photosynthesis could deplete atmospheric carbon (around 800 GtC) in less than ten years. So, all the CO₂ that plants need is almost immediately provided to them by the rest of the biosphere, specifically by those organisms that consume organic matter synthesized by plants and decompose it back into CO₂.

Furthermore, both oceanic and terrestrial biotas contain large reservoirs of organic carbon that can be decomposed to increase atmospheric carbon levels. These include dissolved organic carbon in the ocean (700 GtC) and the organic carbon in soils (around 2000 GtC). Therefore, it’s not just that the global biota rapidly recycles atmospheric carbon, but it also possesses reserves from which carbon can be added or removed.

According to the biotic regulation concept, life creates and maintains conditions favorable for its own perpetuation. In this framework, if the pre-industrial atmospheric CO₂ concentration had been stabilized by the biosphere, once we began to disturb it, the biosphere started compensating for the disturbance. This is a response at the level of the ecological community as a whole.

Indeed, consider that heterotrophs refrain from decomposing the 2000 GtC of soil carbon or the 700 GtC of dissolved organic carbon in the ocean, thus preventing significant carbon perturbations in the atmosphere. For heterotrophs, there is plenty of food available, but they don’t eat it up. (Why is this the case, considering the persistence of soil matter is an ecosystem property?) When extra carbon was added to soil or oceanic organic pools, heterotrophs also refrained from consuming it, initiating a stabilizing response.

We should not view this ultra-complex community-level response in anthropocentric terms, such as imagining a bacterium that sees an extra sugar molecule floating by and refrains from eating it "thinking" that it is "for the sake of the environment." Or imagining that, under increased CO₂, an earthworm may “feel” apathetic and depressed, unwilling to consume extra food coming its way. However, we need to recognize that this stabilizing response exists, even if we do not fully understand its mechanisms. Its existence provides experimental support for the biotic regulation concept and calls for a re-evaluation of the importance of natural ecosystems for climate stability.

Conclusion

We have been running a costly and life-threatening experiment on the biosphere by polluting the atmosphere with excess carbon dioxide. What are the results of this experiment? Even though our actions have degraded natural ecosystems to a large degree, they remain functionable and work to mitigate the disturbance, reducing the rate of carbon accumulation by a significant amount. They could do much more if we had exploited them less.

The fact that the biotic carbon sink was not predicted by ecological science means that the common understanding of the biosphere, ignoring its capacity to regulate the environment, has been wrong. Learning from this dramatic global experiment is both an urgent imperative and a unique opportunity. Let us shift our mentality. We cannot hope to meaningfully manage natural ecosystems that are far more intricate and complex than our limited understanding allows. We must just give them space to operate. Life will cope and self-recover if we give it a chance. Let us reduce extracting from it—it is not a "renewable resource." Let us stop destroying it. There is no alternative path out of the current crisis.