A matter of tree longevity

See allHide authors and affiliations

Science  13 Jan 2017:
Vol. 355, Issue 6321, pp. 130-131
DOI: 10.1126/science.aal2449

Faster tree growth stimulated by rising carbon dioxide levels does not translate into more long-term carbon storage in forests.


There is much scientific and political interest in using the transfer of carbon from the atmosphere to the biosphere, or carbon sequestration, to help mitigate the greenhouse effect (1). Because plants fix carbon dioxide (CO2) by photosynthesis and store carbon in their body (close to half of plant dry matter is carbon), faster carbon uptake by plants through faster growth is widely held to increase carbon sequestration. Yet, this assumption is supported by neither theory nor evidence. Any gain in carbon storage from faster tree growth will be transitory.

Carbon sequestration is the process of storing carbon that has been collected and removed as CO2 from the atmosphere. Increasing the greenhouse effect by CO2 release may be seen as a debt put on future generations; translocating carbon from the atmosphere back to the biosphere mitigates this debt, provided that the removal and subsequent storage last for as long as society considers an increasing greenhouse effect unwanted. This desired time scale is likely to be on the order of centuries. Shorter-term storage, followed by carbon release within a couple of months or years, would burden the next human generation and has no sustained effect.

It is with these time scales in mind that we need to investigate whether faster tree growth can cause existing forests to store more carbon (2, 3). The problem is analogous to money turnover and capital in the economy. The carbon turnover of an ecosystem and the size of the ecosystem's carbon pool are commonly not related. To determine the size of the carbon pool, we must subtract the net carbon output from the net carbon input and then multiply the remainder by the carbon residence time. Pools (capital) will become larger only if two conditions are met: The carbon input must exceed the carbon output, and the difference in these carbon fluxes must remain stored in the system for hundreds of years.

If trees grow faster as result of growth-stimulating environmental change, they will either arrive more rapidly at harvesting size or pass through their natural life span faster. Should such a ramping up of the carbon pool (by faster growth) take place in synchrony over large areas, the result will be a transitory gain in carbon storage—often called “buying time”—followed by a slowing of growth and finally by a carbon release wave when these trees get older (2). This appears to be happening right now in Amazonia, where a decreasing trend of carbon accumulation combined with accelerated mortality (a shorter carbon residence time) has been documented since the 1990s (4) . In the long run, growth stimulation increases carbon turnover, but not the carbon residence time (and hence storage).

Over large areas and over long periods of time, the birth and growth of trees can be considered to balance their death and decay (2, 57). For the carbon pool to grow in a given forest, tree demography must shift toward a greater abundance of older age classes—that is, toward higher mean tree longevity. Once that demographic shift has occurred and a forest has reached its new, higher storage capacity, there is no further net gain in carbon, irrespective of the rate of carbon turnover. The best practice to retain the benefit of such stores is to protect carbon-rich old-growth forests (8).

The critical role of longevity becomes obvious when the tree plantation industry is considered. There, the goal is fast production of timber and pulp, which means short carbon residence times and less carbon storage per unit forest area than in slower-growing, old-growth forests. The use of wood in lasting products must not be mistaken as carbon sequestration unless the overall carbon pool of wood products rises. Just as in forests, what matters is the difference between input and output and the mean residence time (currently, a mean of about 20 years for wood products in Europe).

Although it sounds contradictory, forest productivity is commonly negatively correlated with the carbon capital of a forest (7, 9) under a given set of environmental conditions. In tropical forests, the fastest-growing trees also store less carbon per unit land area as a result of short life span and low wood density.

Thus, unless the residence time of carbon (tree longevity) is maintained or enlarged, faster growth does not mean there is more carbon sequestration. And unlike in the economy, the options for extending the growth of carbon capital in trees are naturally limited. This insight had nourished another idea: to increase carbon sequestration in soils. Microorganisms convert plant debris to soil organic matter (SOM), and carbon can be stored in that form for thousands of years. The overall size of the soil organic carbon pool is about three times that of the global biomass carbon. However, SOM stores not only carbon, but also many other chemical elements that plants need for growth. Organic carbon sequestration to soils therefore competes with plant growth for essential nutrients. For instance, the ratio of nitrogen to carbon needed to tie up carbon in organic material can be as low as 1:400 in timber, and as high as 1:10 in SOM (10). The addition of nitrogen to a forest may facilitate SOM formation (11), but other plant nutrients in addition to nitrogen also become locked up in SOM.

Correspondingly, the results of CO2 enrichment experiments with forest trees cannot be translated into future CO2 effects on carbon sequestration. These experiments test whether—and if so, by how much—carbon uptake can be accelerated by higher CO2 supply to photosynthesis. They were not, nor could they have been, designed to identify the mean residence time of carbon. Tree stands that did not show a sustained growth acceleration in response to higher CO2 supply (12, 13) (because chemical elements other than carbon limit their growth) might still slowly build a larger carbon capital by delayed recycling of biomass carbon (delayed mortality), whereas those that showed growth stimulation might reduce the landscape-wide carbon capital in the long run because of a faster life cycle (2, 7, 14). The value of these CO2 fertilization experiments lies in advancing our understanding of the forest carbon cycle, and helping to make model parameterizations more realistic.

Whether forests will grow faster in a changing environment (irrespective of their carbon storage) is of societal interest because forest products can substitute fossil carbon-based products or provide bioenergy. Nevertheless, changes in productivity do not scale with changes in carbon sequestration, which is inevitably tied to residence time (tree turnover, tree mortality) (7, 9). Similarly, earlier bud break and a longer season in temperate trees have little to do with a change in the size of the forest carbon capital, as the phenology research community often assumes. A longer season may contribute to higher annual tree growth to the extent that the internal clock of trees permits it (15), but not to more long-term carbon storage.

The most effective way to enhance forest carbon storage is to prevent logging old-growth forests and to extend the forested land area (8). Once these new forests reach their storage capacity, they will not sequester additional carbon, irrespective of how fast trees grow and turn over carbon.


  1. Acknowledgments: I thank E. Hiltbrunner and G. Hoch for helpful comments.
View Abstract

Stay Connected to Science

Navigate This Article