(A)Across 123 Amazonian based on initial and final sta biomass plots. estimates calculated using an allometric equation relating individual tree di and wood density to biomass. (B)Across 79 plots from Africa, including estimated tree eight for each stem, addition to diameter and wood density, to estimate biomass, with uncertainty in the height and diameter measurement both propagated final biomass change estimates. As would be expected in a random sample of small plots measured for a finite period, some sites show a decline in biomass during that period indicating that at that particular point in space and time tree mortality has exceeded tree growth. However, the mean and median are shifted significantly to the right for both datasets(P< 0.01)insensitive to different weightings based on measurement interval and plot area(supplementary information in Phillips et al. 2009). Using the same approach, we recently discovered a comparable phenomenon in African for- ests. Here, we measured a similar net sink in trees 210 cm diameter with a mean of 0.63(bootstrapped 95% confidence interval, CI, o.22-0.94) tonnes of carbon per hectare per year(n 79 plots, mean start date 1987 and mean end date 1996; Lewis et al. 2009a). The distribution is left-skewed and shifted to the right of zero(Figure 4.3B). Resampling shows that obtaining such a sample of increasing biomass from a domain that was not increasing in biomass is highly unlikely 1; Lewis et al. 2009a). African forests have greater biomass per unit area than amazon forests ; once this difference is accounted for, both forest blocks have been gaining net biomass at the same relative rate(0.30% per year for Amazonia, o.29% per year for Africa). There are various possible ways by which these plot-based measures can be scaled to tropical forests across Amazonia, South America and Africa. Here adopt a relatively simple approach, given the various and not always quantifiable uncertainties, for example in terms of stems smaller than those we measure belowground(root) biomass carbon, carbon in dead trees and litter area of each forest type and degree ofhuman disturbance. Thus, we assume that our measurements are representative of the wider forest landscape, and that other biomass and necromass components also increase proportionally, but that soil carbon stocks are static. We estimate the magnitude of the sink in each continent by multiplying the plot-based net carbon gain rate by a series of correction factors to account for biomass of lianas, trees <10 cm diameter, necromass and belowground carbon, and a mid range estimate of the surviving forest area for year 2000(Table 4.1). For the 1990s this yields a total estimated South American forest sink of 0.65 t 0.17 PgC yr and a corresponding sink in African forests of 0.53 +0.30 PgC yr 1. Meanwhile, 0.14+0.04 PgC yr 1 in mature undisturbed Asian forests may be assumed if these responded as African and South American forests did(Pan et al. 2011) Thus the combined old-growth tropical forest sink in the 1990s is estimated to have been 1.3 0.35 PgC yr before allowing for any possible net change in soil carbon stock.
This is v similar to the figure given by Lewis et al.(2009), of 1.3 PgCyr i(bootstrapped o 8-1.6) using plots with a mean time interval of 1987-97 methodology(Tropical America, 0.62; Tropical Africa, 0.44; and 0.25 Pgo yr 1).
In the subsequent decade the American tropical a result of the 2005 Amazon drought, we discuss this below.
Clearly these estimates depend on: (i) measurement techniques; (ii) how representative the plots are of forests in South America, and the rest of the tropics: (iii) assumptions about the extent of mature forest remaining, and(iv) the extent to which we have sampled the regional-scale matrix of natural disturbance and recovery.
Moreover, they represent average annual estimates for the late twentieth century forest plots are not measured sufficiently frequently in enough places to estimate biome carbon balance on a year-by year basis.
However, they are consistent with independent evidence from recent inversion-based studies, showing that the tropics are either carbon neutral or carbon sink regions, despite widespread deforestation(Denman et al. 2007, p. 522: Stephens et al. 2007), and the fact that the terrestrial biosphere as a whole has been acting as a very large net sink for decades now(e.g. Quéré 2009).
การแปล กรุณารอสักครู่..
