In a paper in Nature, Terrer et al.1 reveal an unexpected trade-off between the effects of rising atmospheric carbon dioxide levels on plant biomass and on stocks of soil carbon. Contrary to the assumptions encoded in most computational models of terrestrial ecosystems, the accrual of soil carbon is not positively related to the amount of carbon taken up by plants for biomass growth when CO2 concentrations increase. Instead, the authors show that carbon accumulates in soils when there is a small boost in plant biomass growth in response to CO2, and declines when the growth of biomass is high. Terrer et al. propose that associations of plants with mycorrhizal soil fungi are a key factor in this relationship between the above- and below-ground responses to elevated CO2 levels.
Read the paper: A trade-off between plant and soil carbon storage under elevated CO2
Rising levels of atmospheric CO2 are thought to have driven an increase in the amount of carbon absorbed globally by land ecosystems over the past few decades, a phenomenon known as the CO2 fertilization effect2. This occurs because, at the scale of leaves, higher CO2 levels enhance photosynthesis and the efficiency with which resources (water, light and nutrients such as nitrogen) are used to assimilate CO2 and support biomass growth3. Evidence supporting the existence of the CO2 fertilization effect has been observed in experiments in which the atmosphere around plants or plant communities is enriched with CO2. But at the level of whole ecosystems, responses to CO2 enrichment are more difficult to track, because the effects are diluted throughout a chain of connected processes. Constraining estimates of the response of the global land carbon sink to rising CO2 levels therefore remains a major challenge (see go.nature.com/3vgvhj).
Changes in soil carbon are inherently difficult to detect, and studies that assess the effects of elevated CO2 levels on soil-carbon stocks have been equivocal4. Terrer and colleagues set out to investigate these effects by carrying out a meta-analysis of 108 CO2-enrichment experiments. The authors estimate that, in these studies, soil-carbon stocks increased in non-forest sites but remained almost unchanged in forests. By evaluating the effects of multiple environmental variables, the authors found that, surprisingly, the best explanation for the observed patterns is that the changes in soil carbon stocks are inversely related to the changes in above-ground plant biomass: high accumulation of carbon in biomass was associated with soil-carbon loss, whereas low biomass accumulation was associated with soil-carbon gain. This relationship was evident only in experiments in which no nutrients had been added to the studied systems, leading the authors to propose that plant nutrient-acquisition strategies are responsible — which, in turn, depend on the mycorrhizal soil fungi associated with the plants.
Soils linked to climate change
A previous study reported5 that only a small increase in above-ground biomass occurs in CO2-enriched plants that associate with a particular family of mycorrhizae (arbuscular mycorrhizae; AM). AM-associated plants benefit from the fungi’s extensive network of hyphae (branching filaments that aid vegetative growth), which support the plants’ uptake of nitrogen from the soil solution. However, AM have only a limited ability to ‘mine’ nitrogen from organic matter in the soil. The availability of soil nitrogen therefore limits the increase of biomass growth of AM-associated plants in response to elevated CO2 levels. By contrast, plant species that associate with a different group of soil fungi (the ectomycorrhizae; ECM) exhibit a greater increase in above-ground biomass in CO2-enrichment studies, because some of their carbon is allocated to ECM that can mine for nitrogen5. Mining for nutrients by ECM is, however, thought to accelerate the decomposition of organic matter in soil.
Terrer et al. now find that AM-associated plants produce a bigger increase in soil-carbon stocks in CO2-enrichment experiments than do ECM-associated plants. The authors suggest that this is because AM-associated plants allocate more carbon to fine roots and to compounds exuded by the roots, resulting in soil-carbon accrual (Fig. 1a). By contrast, nutrient acquisition by ECM-associated plants results in increased turnover — and therefore loss — of soil organic matter (Fig. 1b). Overall, this would lead to an ecosystem-scale trade-off between the amount of carbon sequestered in plants and that sequestered in soil, in a CO2-enriched atmosphere.
Most Earth-system models that account for land carbon-cycling processes assume that rising levels of atmospheric CO2 will increase plant growth, thus producing more plant litter and thereby increasing stocks of soil carbon6. The authors compared the changes in soil carbon and above-ground plant biomass predicted by various models, both in simulations of six open-air CO2-enrichment experiments, and in global simulations of historical and future increases in atmospheric CO2. None of the models reproduced the negative relationship between carbon sequestration by soil and growth in plant biomass that was observed in the current study.
Terrer and co-workers’ findings thus provide another urgent warning that current climate models overestimate the amount of carbon that will be sequestered by land ecosystems as atmospheric CO2 levels increase — not only because the models largely ignore the effects of nutrient limitations, but also because they overestimate the amount of carbon that could be sequestered in soil, particularly in forest ecosystems7. But the new study also reveals that grasslands, shrublands and other ecosystems that already have high soil-carbon stocks have great potential to accumulate more soil carbon as CO2 levels increase. These results thus add weight to previous calls to protect existing soil-carbon stocks to mitigate the effects of climate change8.
Carbon dioxide loss from tropical soils increases on warming
There are some limitations to the set of CO2-enrichment experiments included in Terrer and colleagues’ meta-analysis. The experiments are biased towards temperate systems, and most of the forests studied are associated with ECM, whereas the grasslands are all AM-associated. The authors did not find that the type of ecosystem had a substantial effect on the observed responses to CO2, but it remains to be seen whether the reported trade-off between above- and below-ground carbon sequestration for AM- compared with ECM-associated plants applies to forests alone9. Further experiments, especially in tropical ecosystems, are now needed to address these issues.
Tropical ecosystems are large contributors to the global terrestrial carbon sink10, but they are notoriously under-studied. Field observations are scarce and few manipulation experiments — such as CO2 enrichment or nutrient additions — have been carried out in these ecosystems11,12. Below-ground processes are particularly challenging to assess in the tropics, where the effects of multiple nutrient scarcities often come into play12. Terrer and colleagues’ study provides a promising framework that can be elaborated to describe diverse plant–soil interactions in various terrestrial ecosystems in the future.
CO2-enrichment experiments generally last for just a few years, or just over a decade at most13. Such timescales are unlikely to capture the effects of elevated CO2 levels on plant mortality, plant-species composition and soil-carbon turnover time, all of which can affect the sequestration of carbon by ecosystems in different ways in the longer term. Mechanistic understanding gained from experiments about the coupling between carbon and nutrient cycling can, however, be integrated into computational models. And this will allow us to constrain estimates of the size of the terrestrial carbon sink in the coming decades. The interactions between plants and their associated soil fungi, as well as other crucial below-ground agents and processes such as microbial communities, are already stirring up modelling efforts14,15. Terrer and colleagues’ study now invites researchers to test hypotheses about the processes that drive coordinated above- and below-ground responses to rising CO2 levels. Such studies could be a real step forwards in our understanding of the fate of the terrestrial carbon sink.