1.3.3. The Future of the Terrestrial Carbon Sink

Table 1-2 illustrates the considerable uncertainty about the relative importance of the present global oceanic and terrestrial sinks in acting as negative feedbacks on the rate of CO2 increase. These sinks may also change in the future as atmospheric CO2 continues to rise. The question therefore arises whether these feedbacks may reach some limit, so that no additional carbon can be stored on land. This possibility would imply that the airborne fraction of emissions would increase more rapidly in the future than at present.

How long the current carbon sink capacity of the terrestrial biosphere is likely to be maintained into the future is a matter of conjecture; several hypotheses proposed as the basis for quantitative explanation are discussed here. Consideration of the nature of the uncertainties for such projections is essential, however. The time scale of our immediate concerns with respect to the biospheric sink as a possible means of reducing the impact of GHG emissions is considered to be several decades, but longer time scales may become relevant if emissions of GHGs continue to rise. Projections of future sources and sinks of carbon in terrestrial ecosystems depend on multiple aspects of the future environment: climate (temperature, precipitation, humidity, and radiation), atmospheric CO2 concentration, nutrient deposition, land use, and ecosystem management (Thompson et al., 1996). Projections of most of these variables into the medium term are highly uncertain, particularly with respect to their future spatial patterns. The regional distribution of ecosystems is an important factor influencing the overall sink capacity. Therefore, improved regional scenarios for human land use are as important for better sink strength projections as they are for better regional climate scenarios.

We may assume that the total capacity for carbon storage in terrestrial ecosystems has an upper limit as a result of mechanical and physiological constraints on the amount of aboveground biomass and physiochemical constraints on the amount of carbon that can be held in soils. These fundamental limits are presently not known in quantitative terms, but it does not seem very likely that they will become of importance within a few decades. Over the past 10,000 years-since the last glaciation-carbon has accumulated in the boreal and north temperate forest ecosystems very largely as transfer from the atmosphere, via vegetation, to the soil organic carbon (SOC) pool. Experimental addition of nutrients has demonstrated the large capacity of these ecosystems for additional tree growth, production of detritus, and SOC. We may therefore conclude that this accumulation of carbon has been constrained since the retreat of the ice by the slow concurrent accumulation of nutrients, particularly nitrogen, in the ecosystem. The present small size of trees in the boreal region and their low nitrogen content suggest that the capacity of the system to store carbon will not by itself limit the transfer of carbon from the atmosphere to the tree and soil pools in the immediate future. Because an accelerated rate of climate change, as well as continuing atmospheric inputs, may increase the availability of nitrogen, it is possible that boreal ecosystems could increase their aboveground and below-ground carbon storage considerably.

By contrast, moist tropical forests in general have large standing stocks of carbon in the trees (Table 1-2). Therefore, there may be limited scope for additional storage of carbon in the trees, although that is less likely to be the case in sparse or secondary tropical forest (Phillips et al., 1998). At present, moist tropical forests currently have high NPP and NEP (Section 1.3.2.2), so it is questionable whether this ongoing flow of carbon from the atmosphere into tropical forests can be maintained. Significant amounts of carbon leave tropical forests in floodwaters as particulate and dissolved organic carbon, some of which is subsequently stored in coastal sediments (Schlesinger, 1997). This and other removals of carbon from tropical forests may provide the opportunity for some increased uptake from the atmosphere and transfer to storage.

Research has demonstrated that in the short term, increased photosynthesis resulting from the rise in atmospheric CO2 concentration diminishes at higher CO2 concentrations, whereas RA and RH increase exponentially with increasing temperature (e.g., Boone et al., 1998; Rayment and Jarvis, 2000). Thus, Scholes et al. (1998) have hypothesized that as atmospheric CO2 concentration and temperature rise, the overall capacity to take up additional carbon from the atmosphere will progressively diminish so that at some point respiration will exceed photosynthesis and the carbon sink will become a source. Based on these two short-term physiological response functions, Scholes (1999) evaluated the global terrestrial carbon sink under various assumptions and concluded that the global sink would be likely to decline from its current level of approximately 2 Gt C yr^-1 and become a source within a few decades. This approach assumes that acclimation to higher CO2 concentrations and temperatures does not occur, that respiration is independent of photosynthesis, and that there are no feedbacks involving nutrition between the processes. A similar, parallel, hypothesis is that the present sink might be a temporary consequence of CO2 fertilization that eventually will be overtaken by respiratory losses of carbon as temperature rises and respiration "catches up." It is supposed that while photosynthesis increases in response to the increased availability of CO2, respiration will initially lag behind but will eventually catch up as the supply of substrate increases or the temperature lag resulting from the thermal inertia of the oceans declines.

To better understand the range of future terrestrial uptake of carbon, calculations with a suite of several more detailed terrestrial biosphere models have been based on the IPCC IS92a scenario of CO2 increase and the associated climate change simulated by the HadCM2 General Circulation Model (GCM) (Cramer et al., 2000). These models include integrated physiological effects of climate, CO2, and nutrition on plant growth and plant population dynamics but include no consideration of land-use or management effects. Stimulation of carbon uptake as a result of enhanced uptake of soil nitrogen is included in several of the models. For the IPCC IS92a scenario, all of the models confirm that there is a terrestrial CO2 sink of the same order of magnitude as above that explains the present-day overall balance; that this sink might increase in strength with further increases in CO2 concentration; and that when the CO2 concentration increases beyond about 500 ppmv CO2 (i.e., around 2030-2050), the rate of increase of the sink approaches zero as the sink reaches a steady asymptote.

A similar result has been obtained using a well-validated local-scale ecosystem model (G'Day) that treats interactions and feedbacks between the effects of CO2 and temperature on the pools and fluxes of carbon and nitrogen within vegetation and soil compartments of terrestrial ecosystems (Dewar et al., 1999; McMurtrie et al., n.d.). This model was parameterized for a boreal coniferous forest in northern Sweden; temperature and CO2 concentration were increased according to the IPCC IS92a scenario as above. The G'Day model indicates that, after initiation, NPP consistently exceeds RH and the carbon sink rises to an asymptotic maximum after about 60 years and remains there for another 50 years without significant decline (McMurtrie et al., n.d.). The capacity at which the sink stabilizes depends on the availability of nitrogen and the fraction of nitrogen released in decomposition that is subsequently re-immobilized. Thus, with appropriate feedbacks and N:C relationships considered, this model also predicts that the carbon sink will be maintained well into the medium term.

On the other hand, sinks may disappear if climate change or deforestation would lead to widespread tree mortality (Friend et al., 1997; Cao and Woodward, 1998; Walker et al., 1999). Of the suite of models discussed above, one model (HYBRID) (Friend et al., 1997) predicts a drastic decline in NPP that is primarily the result of drought-induced die-back of tropical forest in the Amazon basin. This finding highlights two points: First, changes in vegetation structure can have significant effects on terrestrial carbon sinks and indeed can be as important or more important than the physiological effects discussed above (see also Mooney et al., 1999); second, projections of climate change impacts should be based on a range of predictions of future regional climates. The HadCM2 GCM produces somewhat drier climate change predictions for the Amazon than do some other GCMs, which might well lead to different results.

These model experiments and other analyses (e.g., Walker et al., 1999) suggest that the current global terrestrial sink will be maintained over the short term of the next few decades unless serious mortality occurs but may diminish toward the end of the 21st century. With the present state of knowledge, projections beyond a few decades must be regarded as uncertain; this situation is improving rapidly, however.

This conclusion is based on consideration of the effects of future CO2 and climate change on the present terrestrial sink only; it does not take into account future deforestation or possible actions taken to enhance the terrestrial sink through land use, land-use change, and afforestation or reforestation. Higher CO2 concentration and warmer climate may be expected to modify the effectiveness of such activities to enhance the medium-term terrestrial sink. Improved understanding of these effects would enable these activities to be better adapted to future CO2 and climate conditions.

IPCC Homepage