Matches in SemOpenAlex for { <https://semopenalex.org/work/W2013815193> ?p ?o ?g. }
- W2013815193 endingPage "2041" @default.
- W2013815193 startingPage "2039" @default.
- W2013815193 abstract "Less than half of anthropogenic carbon emissions are accumulating in the atmosphere, due to large net fluxes into both the oceans and the land (Le Quéré et al., 2012). The land sink in particular has increased markedly, doubling in strength since the 1960s, to reach 26 petagrams of carbon in the latest decade. However, the location and drivers of this large terrestrial sink are still relatively poorly constrained by atmospheric measurements (Ciais et al., ). Pan et al. (2011) recently utilized >1 million forest inventory plots to provide summaries of forest carbon stocks, and the first global bottom-up estimates of carbon fluxes for the world's forest biomes for the period 1990–2007. One key result was that almost all the residual global terrestrial carbon sink (i.e. carbon uptake after accounting for land-use change), some 2.4 ± 0.4 Pg of carbon per year, is located in the world's established forests (Pan et al., 2011). The sink is distributed worldwide, with globally significant net fluxes into boreal and temperate forests, and a large sink in intact tropical forest, albeit with large uncertainty. Furthermore, Pan et al. (2011) showed that this tropical intact forest sink may have faded from an estimated annual 1.3 ± 0.4 Pg C in the 1990s to 1.0 ± 0.5 Pg C for 2000–2007. The tropical intact forest sink is offset by a net land-use emission [1.5 Pg C yr−1 (1990–1999)] declining to [1.1 Pg C yr−1 (2000–2007)], and as a consequence aircraft measurements and inverse modelling studies indicate the tropics to be close to neutral in terms of net carbon fluxes (reviewed by Ciais et al., ). While the intact tropical forest sink values represent updates from similar values published previously (e.g. Lewis et al., 2009), the fact that almost the entire residual terrestrial carbon sink is accounted for by the forests of the world was a notable discovery. Evidence from the ground now points to established forests being a net sink across almost every major forest region, including all extra-tropical forest regions analysed. In a recent letter to this journal, Wright (2013) contested only the report of a sink in intact tropical forests. In so doing, he focused on only one aspect of the paper, ignoring the global result from Pan et al. (2011). Wright made three specific claims. Firstly, that the reported tropical forest sink is too large because of our definition of ‘tropical forest’. Secondly, that we ignored vital data in our analyses. Finally, he devotes most space to arguing that carbon dioxide fertilization cannot be substantially responsible for the global forest sink. We briefly address each in turn. Wright argues that by using a globally consistent forest definition that includes more open forests (FAO, ), we overestimate the carbon sink. Pan et al. (2011) is a global-scale report, so we use the widely used global definition. Furthermore, there is no generally accepted single definition of tropical forest (see Torello-Raventos et al., 2013). We agree the network of forest monitoring plots used in the analysis is certainly biased away from more open forests. The key question is therefore: will applying mean rates of increase in carbon storage from more closed forest to more open forest overestimate the total sink? Wright considers this to be the case, explaining that the ‘changes in carbon stocks are potentially much larger in tall, closed canopy forests than those in stunted, open forests.’ This is not necessarily so. There are numerous reports of significant closed forest advance and large biomass gains at the dry margins of forests in regions not captured by our networks [e.g., Bowman et al., 2010 (Australia); Mitchard et al., 2009 (central Africa); Buitenwerf et al., 2012 (southern Africa)]. More fundamentally, if carbon dioxide is a driver of the increases in biomass then drier forests may well show greater sensitivity as they are more water limited and carbon dioxide increases water use efficiency (e.g., Keenan et al., 2013). How much is the undersampling of dry forests likely to matter for the total value of the pan-tropical intact forest sink? An alternative method that adopts a much more conservative assumption about dry-margin forest biomass increases (Malhi, 2010) produced a similar intact tropical forest sink estimate to ours, within our stated uncertainty in Pan et al. (2011). And like ours, Malhi's estimate ignores the well-documented thickening and woody encroachment into savannas, and assumes zero net sink in tropical soils – in spite of the forest productivity increases. In sum, while the plot networks are of course not a stratified-random sample of all of today's undisturbed tropical forest, and therefore there is uncertainty as to the magnitude of the increase in carbon storage in more open intact tropical forests, there are clearly modest potential biases in both directions. We used 215 plots from across Amazonia and Africa to estimate the intact tropical forest carbon sink. Wright claims that our ‘omission’ of 12 forest plots from another inventory network reported by Chave et al. (2008) caused us to overestimate the carbon sink. Certainly data from other networks are valuable and alternative analyses possible. However, Wright may have misunderstood both the rationale of the analysis and the exact plots utilized. Thus, ours was a continental-scale analysis (rather than aimed at a single pan-tropical mean value as in Lewis et al., 2009; which did include all the Chave et al. plots). Furthermore, in the Pan et al. paper the four African plots from Chave et al. (2008) were in fact all included. For the tropical Americas we already had 136 standardized plots from an Amazon-wide analysis based on tree-by-tree data (Phillips et al., 2009), but lacked access to the additional tree-by-tree data analysed in Chave et al. (2008). Including those plots would have reduced the tropical American sink estimate by ~3%. Finally, for Southeast Asian forests we concluded that the six published plots in Chave et al. (2008) were insufficient to draw robust regional-scale conclusions. We therefore applied the mean rate of carbon storage increase measured from across Africa and Amazonia. This limitation was clearly stated (‘we lack sufficient data in Southeast Asia to make robust estimates’). If tropical Asian forests turn out to have been behaving differently from other tropical forests it would be very interesting for ecologists to know why. However, it matters little for the global carbon balance: heavily deforested and fragmented, Asia only represents 9% of total intact tropical forest area. In Pan et al. they contributed 0.1 Pg C yr−1 to the global forest sink, well within our stated uncertainty. Finally, Wright devotes most effort to denying the plausibility of atmospheric carbon dioxide as a significant driver of the forest sink. Yet, the possibility of carbon dioxide fertilization was not broached in Pan et al. (2011). The important question of mechanism(s) was largely beyond the scope of that paper and is too complex to be treated in full here too. For brevity, we here focus on two points. First, it is the rate of increase in a forest's woody production, together with the typical woody carbon residence time, that will determine the magnitude of the sink at any given time in an old-growth forest. Thus, other things being equal, and assuming CO2 is the driver, the magnitude of the C sink should scale with the rate of increase in atmospheric CO2 (averaged over the relevant woody biomass residence time), and not its absolute value as assumed by Wright in his calculations. Using first principles (Taylor & Lloyd, 1992; Lloyd & Farquhar, 1996), when the appropriate simulations are performed using Dynamic Global Vegetation Models, then the magnitude of the modelled residual terrestrial C sink is surprisingly close to that measured on the ground. The mean tropical forest sink (net biome productivity) across nine models was 1.1 ± 0.4 Pg C yr−1 from 1980 to 2009 (Piao et al., 2013), indistinguishable from the 1.2 ± 0.4 Pg C yr−1 from 1990 to 2007 reported in Pan et al. (2011). Furthermore, the response of land vegetation to carbon dioxide fertilization is considered by vegetation modellers to be greatest in tropical forests (Ciais et al., ; fig. 6.22). Of course, there are significant uncertainties and simplifications in these models, such as the lack of accounting for phosphorus. However, when simulations are run against experiments elevating CO2 in (temperate) forest stands, the results are similar: Piao et al. (2013) report a 16% increase in net primary productivity per 100 ppm CO2 in simulations, vs. 13% in experiments. We conclude that it is highly likely that a substantial fraction of the long-term large-scale increases in established forest carbon density is driven by rising atmospheric carbon dioxide concentrations. Wright's second line of argument is that the rise in atmospheric carbon dioxide has been so large over the past two centuries that the implied increases in aboveground biomass and/or numbers of trees per hectare over such a timescale are simply implausible. Unfortunately, of course, we lack direct measurements of the preindustrial ecology of tropical forests, but we do have a rapidly improving understanding of present-day variation with the biome. Even humid forests today vary greatly, beyond the differences Wright suggests are implausible. For example, both Asian (Slik et al., 2010) and African (Lewis et al., 2013) forests, on average, store >30% more biomass per hectare than ‘typical’ Amazon forests, whereas Amazon forests themselves vary enormously across resource gradients (e.g. Quesada et al., 2012). Asian and Amazon forests also have a ca. 40% greater density of trees >10 cm diameter than African forests (Lewis et al., 2013). Thus, differing environmental conditions across space coincide with large differences in forest structure, so the possibility that tropical forest biomass has increased markedly in response to global environmental changes is reasonable. These drivers of biomass increases could of course extend beyond the biological effects of carbon dioxide. Additional global mechanisms may include the effect of higher air temperatures that decrease water viscosity (Roderick & Berry, 2001), so benefiting tropical species with higher wood density (Lloyd, personal communication). Notably, wood density increases with temperature across contemporary thermal gradients in Amazonia (Quesada et al., 2012) and Africa (Lewis et al., 2013). Elsewhere, Wright and colleagues recently also concluded that a plot-based annual increase in tropical biomass of 0.6 Mg C ha−1 is indeed consistent with the recent carbon balance of the Earth system (Cernusak et al., 2013). Similarly, a substantial intact forest sink is required to obtain a net balance of near zero for the tropics given known land-use emissions (Ciais et al., ). So, while the precise contributions of the drivers of this increase in intact forest carbon stocks remain to be quantified, perhaps the long-running debate about the existence of a significant sink is now over (c.f. Wright, 2006). We do agree with Wright (2013) that there remains significant uncertainty as to how large the sink is, particularly (i) as no ground-based monitoring programme is fully randomized at continental and global scales; and (ii) changes belowground remain unassessed. We suggest that these uncertainties may result in either overestimating or underestimating the tropical forest carbon sink. Nevertheless the key points from Pan et al. (2011) bear repeating: atmospheric and other records show a very large amount of carbon being sequestered on the Earth's land surface, and ground measurements independently show that almost all forest regions – tropical forests included – have contributed to this remarkable multidecadal terrestrial carbon sink. This work was supported by the EU (an ERC Advanced Grant, and the GEO-CARBON consortium). OLP is supported by a Royal Society Wolfson Research Merit Award, SL by a Royal Society Research Fellowship. We thank Jon Lloyd and two anonymous reviewers for their helpful comments." @default.
- W2013815193 created "2016-06-24" @default.
- W2013815193 creator A5032473063 @default.
- W2013815193 creator A5044312224 @default.
- W2013815193 date "2014-04-19" @default.
- W2013815193 modified "2023-10-17" @default.
- W2013815193 title "Evaluating the tropical forest carbon sink" @default.
- W2013815193 cites W2014770500 @default.
- W2013815193 cites W2020507833 @default.
- W2013815193 cites W2047189523 @default.
- W2013815193 cites W2048518869 @default.
- W2013815193 cites W2062526709 @default.
- W2013815193 cites W2062828898 @default.
- W2013815193 cites W2101776010 @default.
- W2013815193 cites W2102487370 @default.
- W2013815193 cites W2109631166 @default.
- W2013815193 cites W2111234726 @default.
- W2013815193 cites W2116673487 @default.
- W2013815193 cites W2121361055 @default.
- W2013815193 cites W2121679532 @default.
- W2013815193 cites W2128183889 @default.
- W2013815193 cites W2136088536 @default.
- W2013815193 cites W2146377274 @default.
- W2013815193 cites W2158971682 @default.
- W2013815193 cites W2160051745 @default.
- W2013815193 cites W2168928730 @default.
- W2013815193 cites W2326954126 @default.
- W2013815193 cites W4211228737 @default.
- W2013815193 doi "https://doi.org/10.1111/gcb.12423" @default.
- W2013815193 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/24123580" @default.
- W2013815193 hasPublicationYear "2014" @default.
- W2013815193 type Work @default.
- W2013815193 sameAs 2013815193 @default.
- W2013815193 citedByCount "36" @default.
- W2013815193 countsByYear W20138151932015 @default.
- W2013815193 countsByYear W20138151932016 @default.
- W2013815193 countsByYear W20138151932017 @default.
- W2013815193 countsByYear W20138151932018 @default.
- W2013815193 countsByYear W20138151932019 @default.
- W2013815193 countsByYear W20138151932020 @default.
- W2013815193 countsByYear W20138151932021 @default.
- W2013815193 countsByYear W20138151932022 @default.
- W2013815193 countsByYear W20138151932023 @default.
- W2013815193 crossrefType "journal-article" @default.
- W2013815193 hasAuthorship W2013815193A5032473063 @default.
- W2013815193 hasAuthorship W2013815193A5044312224 @default.
- W2013815193 hasBestOaLocation W20138151931 @default.
- W2013815193 hasConcept C108216600 @default.
- W2013815193 hasConcept C110872660 @default.
- W2013815193 hasConcept C132651083 @default.
- W2013815193 hasConcept C14168384 @default.
- W2013815193 hasConcept C143050476 @default.
- W2013815193 hasConcept C149095469 @default.
- W2013815193 hasConcept C149677717 @default.
- W2013815193 hasConcept C18903297 @default.
- W2013815193 hasConcept C205649164 @default.
- W2013815193 hasConcept C22884784 @default.
- W2013815193 hasConcept C2776285232 @default.
- W2013815193 hasConcept C39432304 @default.
- W2013815193 hasConcept C530467964 @default.
- W2013815193 hasConcept C54286561 @default.
- W2013815193 hasConcept C58640448 @default.
- W2013815193 hasConcept C6939412 @default.
- W2013815193 hasConcept C86803240 @default.
- W2013815193 hasConceptScore W2013815193C108216600 @default.
- W2013815193 hasConceptScore W2013815193C110872660 @default.
- W2013815193 hasConceptScore W2013815193C132651083 @default.
- W2013815193 hasConceptScore W2013815193C14168384 @default.
- W2013815193 hasConceptScore W2013815193C143050476 @default.
- W2013815193 hasConceptScore W2013815193C149095469 @default.
- W2013815193 hasConceptScore W2013815193C149677717 @default.
- W2013815193 hasConceptScore W2013815193C18903297 @default.
- W2013815193 hasConceptScore W2013815193C205649164 @default.
- W2013815193 hasConceptScore W2013815193C22884784 @default.
- W2013815193 hasConceptScore W2013815193C2776285232 @default.
- W2013815193 hasConceptScore W2013815193C39432304 @default.
- W2013815193 hasConceptScore W2013815193C530467964 @default.
- W2013815193 hasConceptScore W2013815193C54286561 @default.
- W2013815193 hasConceptScore W2013815193C58640448 @default.
- W2013815193 hasConceptScore W2013815193C6939412 @default.
- W2013815193 hasConceptScore W2013815193C86803240 @default.
- W2013815193 hasIssue "7" @default.
- W2013815193 hasLocation W20138151931 @default.
- W2013815193 hasLocation W20138151932 @default.
- W2013815193 hasOpenAccess W2013815193 @default.
- W2013815193 hasPrimaryLocation W20138151931 @default.
- W2013815193 hasRelatedWork W1999753309 @default.
- W2013815193 hasRelatedWork W2060657782 @default.
- W2013815193 hasRelatedWork W2098303304 @default.
- W2013815193 hasRelatedWork W2102487370 @default.
- W2013815193 hasRelatedWork W2136088536 @default.
- W2013815193 hasRelatedWork W2358714844 @default.
- W2013815193 hasRelatedWork W2997476521 @default.
- W2013815193 hasRelatedWork W3175738537 @default.
- W2013815193 hasRelatedWork W3212546445 @default.
- W2013815193 hasRelatedWork W4238580864 @default.
- W2013815193 hasVolume "20" @default.
- W2013815193 isParatext "false" @default.