Terrestrial carbon sinks – uncertain

1 Biologist (2002) 49 (4)155 carbon is the basis of life on Earth. It is incorporated intoplants through photosynthesis, absorbed by animals throughtheir food, present in the atmosphere as carbon dioxide,locked into rock as limestone, and pressed into fossil fuelssuch as coal and oil. In the Jurassic age, there was muchmore carbon in the atmosphere than there is now. The carbontaken up by plants on land and in the sea gradually, overmillennia, exceeded the amount released during decay, andthis excess carbon became locked away as fossil fuelsbeneath the surface of the planet. In effect, life on Earth wasacting as a carbon sink . For the last two centuries, this trendhas been drastically reversed as forests have been reducedand fossil fuels have been burnt, meaning that more carbonhas been released into the atmosphere than has beenabsorbed. There is considerable concern that this is leadingdirectly to global warming.

2 carbon sinks are now the focus ofconsiderable the oceans are currently the greatest carbon sink, Terrestrial carbon sinks are also important. The greatestterrestrial carbon sinks occur in young, growing forests,because a hectare of trees holds up to 50 times more carbonthan a hectare of grasses or crops. Older forests and soilsmay also accumulate carbon but the rates are generally lowcompared with the rate for young trees. Nevertheless, lowrates over large areas, including grasslands as well asforests, can yield a sink that is globally large. Because theamount of carbon in soils is large and variable, it is difficult tomeasure an annual carbon sink, even in a small, well-definedarea. To measure a change globally is nearly that Terrestrial ecosystems are accumulating carbon insome regions and releasing it in others, it is very difficult todetermine, directly, the magnitude of the Terrestrial carbonsink.

3 (See Box 1 for a description of the methods used tomeasure or infer the magnitude of Terrestrial sinks .)Direct measurement of changes in the concentrations ofcarbon dioxide and oxygen in the atmosphere provides oneestimate of the global Terrestrial sink (Table 1). Nearly half ofthe carbon released each year from fossil fuel combustionaccumulates in the atmosphere. The oceans and someterrestrial ecosystems take up the rest. According to a recentglobal analysis based on changes in the concentrations ofcarbon dioxide and oxygen (Plattner et al., in review), theworld s Terrestrial ecosystems (vegetation and soils) were asmall net sink during the 1980s and of these estimates of a net Terrestrial sink,changes in land use (including deforestation whether forsettlements or agriculture reforestation, and the manage-ment of forests and wood products) are calculated to havereleased an average of x 1015grams of carbon per year(PgC yr-1), and slightly more during the 1990s (Houghton, inpress).

4 These estimates include the releases of carbon fromthe oxidation of wood products and logging debris, as well asTerrestrial carbon sinks uncertainexplanationsR A HoughtonWoods Hole Research Center, USAThe carbon cycle is fundamental to life on Earth and it determines the amount of carbonTerrestrial carbon sinks uncertainLowland rainforest, Borneo (F Lanting) Minden Pictures/FLPA156 Biologist (2002) 49 (4)the accumulations of carbon in regrowing forests. Strangely,the difference between the net Terrestrial sink and the emis-sions from land-use change suggests that there is a residualterrestrial sink, not well understood, that locked away asmuch as PgC yr-1during the last two decades (Table 1).The exact magnitude, location and cause of this residualterrestrial sink are uncertain , but it is important to understandthe mechanisms responsible for at least two reasons. First,the Kyoto Protocol requires that sinks resulting directly fromhuman activity ( , planting trees) be distinguished fromsinks unrelated to direct human management ( , growthenhancement through CO2fertilisation), so as to avoid undue credits.

5 Second, some potential mechanisms aremore likely to persist into the future than others are. Withouta greater understanding of the global carbon cycle, plans forremedial action will be, at best, reliant on some is even the chance of inadvertently worsening the situ-ation by harming existing mechanisms proposed as being responsible for terres-trial sinks fall into two broad categories (Table 2): Physiological or metabolic factors that affect rates of photo-synthesis, respiration, growth, and decay. These factorsinclude elevated concentrations of atmospheric CO2, increased availability of nutrients, and changes intemperature and rainfall, any of which could increasegrowth rates in forests locking more carbon away fromthe atmosphere. These factors generally result indirectly fromhuman activities. Disturbance and recovery mechanisms, including bothnatural disturbances and the direct effects of changes inland use and management.

6 Disturbances affect the mortal-ity of forest stands and thus the age structure of forests;recovering forests are sinks for carbon . Different landmanagement strategies may also affect the amount ofcarbon stored in non-forest or metabolic mechanismsresponsible for the current sinkThe longest-standing physiological mechanism proposed forthe accumulation of carbon on land is have long known that annual plants respondto higher levels of CO2with increased rates of growth. Theconcentration of CO2in greenhouses is often deliberatelyincreased to make use of this effect. Similarly, experimentshave shown that most C3plants (all trees, most crops, andvegetation from cold regions) respond to elevated concentra-tions of CO2with increased rates of photosynthesis andgrowth. Despite the observed stimulatory effects of CO2on photo-synthesis and plant growth, it is not clear that the effects willresult in an accumulation of carbon (that is, a net sink) inTerrestrial carbon sinks1.

7 Global budgets based on atmospheric data and modelsThe CO2concentration of air is measured weekly at anetwork of nearly 100 sites around the world. Using modelsof atmospheric transport, scientists calculate the surfacesources and sinks of carbon that explain the observed varia-tion in CO2concentrations. This inverse modelling approachcan also be used to distinguish Terrestrial vs. oceanic carbonsinks when the atmospheric measurements include tempo-ral variations in O2and in the isotopic composition of CO2( , 13CO2; Table 5).2. Global budgets based on models of oceanic carbonuptakeModels of the oceans carbon cycle (including carbon chem-istry and oceanic circulation) can be used with temporal pat-terns of atmospheric CO2and fossil fuel emissions to infer thenet Terrestrial sources and sinks of carbon . This approachyields a global value (no regional distinctions), but can beapplied over the last two centuries.

8 CO2concentrationsbefore 1957, when direct measurements began, areobtained from analysis of the air trapped in bubbles inGreenland and Antarctic Regional carbon budgets constructed from forest invento-riesMost of the developed nations have national forest invento-ries that monitor the volume of wood in forests. Changes involumes over time indicate sources or sinks of information can be used to infer associatedchanges in woody debris, wood products, and soil carbon those components that are not directly measured in the Stand-level direct measurement of CO2flux (from towers)Direct measurement of CO2flux over an ecosystem isobtained by linking CO2concentrations in air with the upwardor downward movement of that air (the eddy covariancetechnique). The approach yields hourly estimates of flux forareas of a few km2, but is difficult to scale up to largerregions. Because net fluxes of carbon are generally smallrelative to the gross daytime and night time fluxes, theapproach is better suited for determining short-termresponses of photosynthesis and respiration to variations intemperature and moisture than to determining annualcarbon Physiologically-based process models of ecosystemsEcosystem- and global-scale models simulate changes incarbon storageas a result of environmentally inducedchanges in photo-synthesis, respiration, growth, litter fall and advantage of such process-based models is that they canbe used to distinguish the effects of different factors on carbonstorage.

9 The disadvantage is that it is difficult to know whetherthey include all of the important carbon models based on changes in land useThese models are based on the changes in vegetation andsoil that accompany a change in land use. When forests areconverted to agricultural land, for example, the amount ofcarbon in the vegetation and soil is reduced. When openlands are afforested, carbon stocks increase. The rates ofloss and increase vary with land use and type of ecosystem,but they are documented for many systems. Most of theerror in the calculated flux of carbon results from uncertain -ties in rate of deforestation and wood of carbon flux based on changes in land use cantheoretically identify sources and sinks resulting from directhuman activity, but to do so requires baseline information onthe rates of forest growth that would have existed in theabsence of changes in CO2, N, climate, or otherenvironmental variables.

10 Such data are generally unavail-able. Nevertheless, the method, in combination with meth-ods that measure a net source or sink (methods 1 4 above),has the potential to distinguish sinks resulting from directBox 1. Methods that are used to measure or infer carbon sinksBiologist (2002) 49 (4)157either plants or soils. Annual plants and most crops do notaccumulate carbon for more than a year. Furthermore, thestimulatory effects of CO2have generally been observed inshort-term experiments, while over longer intervals the effectsare often reduced or absent. For example, plants often accli-mate to higher concentrations of CO2so that their rates ofphotosynthesis and growth return to the rates observedbefore the concentration was raised (Oren et al., 2001). Thefew studies conducted at higher levels of integration orcomplexity, such as mature trees or whole ecosystems,including soils as well as vegetation, suggest much reducedresponses (Table 3).