The integrated nature of global and regional studies means that they are of limited use in advancing specific understanding of controls and drivers of ecosystem productivity. For this smaller-scale studies are required. The application of eddy covariance systems, capable of continuous measurements of NEE and integrated measurements of NEP, to terrestrial ecosystems, has enabled insights into carbon cycling through terrestrial ecosystems at temporal and spatial scales, which was previously impossible.
The number of sites and years of data collected is enabling syntheses that are assessing the extent to which carbon cycling through different ecosystems responds similarly to biotic and abiotic drivers. Among these studies several stand out as providing genuinely new insights into the factors controlling ecosystem productivity.
As already detailed in the preceding section, Valentini et al. (2000) highlighted how the relatively young, largely managed forests of Europe were significant sinks for carbon. Importantly, Valentini et al. (2000) concluded that it was variation in respiration rather than GEP that was the key control on the net size of the carbon sink of European forest systems, with respiration rates of European forests well correlated with latitude, and, surprisingly, the control exerted by respiration on NEP being greater in colder northern latitudes.
It has long been thought that productivity of trees and forest stands declines with age. The theory proposed by Odum (1969) is one of several commonly cited as a reason for this. It proposes that as trees or forests age, leaf area decreases relative to woody biomass; the consequence is that NPP slowly declines, tending towards zero as ultimately photosynthesis is balanced by autotrophic respiration (Odum, 1969). As a result, mature forest stands have not been thought to be significant sinks for carbon. Recent studies have begun to challenge this idea. Although Pregitzer and Euskirchen (2004) showed clear declines in NPP of tropical, temperate and boreal forests with age, forests older than 120 years still exhibited significant NPP. A study by Carey et al. (2001) discovered that productivity of subalpine forest stands only began to plateau after 500 years and that these 500-year-old stands had an NPP of close to 6 t C/ha. A study by Malhi et al. (2004) showed that the lowland new world humid tropical forests are accumulating carbon in woody biomass at an average rate of 3.1 t C/ha/year. One characteristic of mature unmanaged forest stands is a large dead wood pool. Carbon losses from this pool can be significant (Saleska et al., 2003) and will reduce NEP in relation to NPP. However, studies that have measured NEP in the Amazon have shown significant NEP (Grace et al., 1995; Malhi et al., 1998). Similarly Knohl et al. (2003) measured NEP of ~500 g C/year in a 250-year-old beech forest in Germany.
Major disturbances, either managed or natural, as a consequence of fire, wind, pest or disease are a central feature of forest ecosystems. New quantitative insights into the effect of these disturbances on productivity immediately after and during recovery are being achieved by the use of chronosequences of forest stands at different ages, thereby replacing space for time. As expected, disturbance decreases photo-synthetic capacity, completely in the case of logging, and typically converts a forest carbon sink into a carbon source (Knohl et al., 2002; Kowalski et al., 2004). Recovery is characterized by increases in NPP over time. However, of key importance from the perspective of carbon sequestration and its management is the time required for a stand to become carbon-neutral and ultimately sequester carbon after the initial disturbance. Studies are beginning to provide insight into this. Thornton et al. (2002) combined measurements with an ecosystem process model and found this time to be highly variable, depending on type of disturbance, intensity and postdisturbance management for evergreen forests. Howard et al. (2004) found it to be 10 years for a jack pine stand in Canada, Litvak et al. (2003) recorded 11 years for a black spruce stand in Manitoba, while NEP of a Siberian Scots pine forest recovered to zero 12 and 24 years after a stand replacing fire in forests with a Vaccinium sp. and lichen understo-rey, respectively (Knohl et al., 2002).
Important new environmental drivers of productivity have also been uncovered. Instantaneous measurements of NEE of forest ecosystems at a given PPFD have been shown to be dependent on the ratio of diffuse to direct PPFD, with NEE increasing as the diffuse component increases and the effect becoming greater as PPFD increases (Gu et al., 2002; Law et al., 2002). While diffuse light will penetrate a forest canopy to a greater extent than direct light, increasing photosynthesis in shade leaves, it will also change the canopy microclimate, with likely decreases in both air and leaf temperature, and changes in leaf air vapour pressure deficit. All are key drivers of either photosynthesis or respiration, or both. The net effect of all these changes is increased net photosynthesis (Law et al., 2002).
These important advances aside, there is still much uncertainty in our understanding of the biosphere-atmosphere feedbacks that drive spatial and temporal variability in carbon cycling through specific ecosystems and biomes. We highlight the need to better understand: (i) how representative current study sites are of larger biomes, which has important consequences for scaling up using biogeochemical models and 'ground truth-ing' the algorithms used for satellite-based global productivity estimates; (ii) the role that large-scale disturbance and extreme climatic events play in driving ecosystem productivity in the short and long term; and (iii) the extent to which an understanding of the drivers of current carbon cycling through terrestrial ecosystems can inform models predicting carbon cycling into the next century.
Was this article helpful?