No abstract

AimThe accurate mapping of forest carbon stocks is essential for understanding the global carbon cycle, for assessing emissions from deforestation, and for rational land-use planning. Remote sensing (RS) is currently the key tool for this purpose, but RS does not estimate vegetation biomass directly, and thus may miss significant spatial variations in forest structure. We test the stated accuracy of pantropical carbon maps using a large independent field dataset.LocationTropical forests of the Amazon basin. The permanent archive of the field plot data can be accessed at: http://dx.doi.org/10.5521/FORESTPLOTS.NET/2014_1MethodsTwo recent pantropical RS maps of vegetation carbon are compared to a unique ground-plot dataset, involving tree measurements in 413 large inventory plots located in nine countries. The RS maps were compared directly to field plots, and kriging of the field data was used to allow area-based comparisons.ResultsThe two RS carbon maps fail to capture the main gradient in Amazon forest carbon detected using 413 ground plots, from the densely wooded tall forests of the north-east, to the light-wooded, shorter forests of the south-west. The differences between plots and RS maps far exceed the uncertainties given in these studies, with whole regions over- or under-estimated by > 25%, whereas regional uncertainties for the maps were reported to be < 5%.Main conclusionsPantropical biomass maps are widely used by governments and by projects aiming to reduce deforestation using carbon offsets, but may have significant regional biases. Carbon-mapping techniques must be revised to account for the known ecological variation in tree wood density and allometry to create maps suitable for carbon accounting. The use of single relationships between tree canopy height and above-ground biomass inevitably yields large, spatially correlated errors. This presents a significant challenge to both the forest conservation and remote sensing communities, because neither wood density nor species assemblages can be reliably mapped from space.

Question: In Amazonian moist forest, four questions arose: 1. Do tree species differ in their susceptibility to lianas? 2. What host tree traits (branch‐free bole height, growth rate, bark type, leaf length and adult stature) are correlated with the susceptibility of tree species to lianas infesting the trunk and the crown? 3. To what extent do spatial variables (proximity to liana‐infested trees and the light environment of the tree crown) affect the likelihood of liana infestation? 4. Are spatial variables or tree traits relatively more important in influencing the susceptibility of trees to lianas? We address all questions separately for trunk and crown infestation. Location: Tambopata Nature Reserve, Peru. Methods: We collected information on liana infestation, tree morphological traits, growth, light‐environment and position for 3675 trees in seven 1‐ha permanent sample plots. We separated trunk from crown infestation and used correlation and logistic regression analyses for tree species and individual tree‐level analyses, respectively. Results: Half of all trees were colonised by at least one liana. Of 41 relatively common dicot tree species, at least five have significantly greater and three significantly lower crown infestation rates than expected by chance. Trunk and crown infestation are influenced by different host traits – trunk infestation was only affected by bark type, while crown infestation is reduced when trees are fast‐growing, tall, have low‐density wood, long branch‐free boles and long leaves. The likelihood of both trunk and crown infestation increases for trees growing in close proximity to another liana‐infested tree, but is invariant with the light environment of tree crowns. Conclusion: Crown and trunk infestation have not been properly distinguished before; it is important to do so as the factors determining the different modes of infestation differ fundamentally. The association between crown infestation and tree traits suggests that increases in liana dominance in Amazonian forests could cause changes in forest composition, including favouring faster growing tree species with low density wood, potentially reducing the carbon stored by mature forests.

Above-ground tropical tree biomass and carbon storage estimates commonly ignore tree height. We estimate the effect of incorporating height (<i>H</i>) on forest biomass estimates using 37 625 concomitant <i>H</i> and diameter measurements (<i>n</i> = 327 plots) and 1816 harvested trees (<i>n</i> = 21 plots) tropics-wide to answer the following questions: <br><br> 1. For trees of known biomass (from destructive harvests) which <i>H</i>-model form and geographic scale (plot, region, and continent) most reduces biomass estimate uncertainty? <br><br> 2. How much does including <i>H</i> relationship estimates derived in (1) reduce uncertainty in biomass estimates across 327 plots spanning four continents? <br><br> 3. What effect does the inclusion of <i>H</i> in biomass estimates have on plot- and continental-scale forest biomass estimates? <br><br> The mean relative error in biomass estimates of the destructively harvested trees was half (mean 0.06) when including <i>H</i>, compared to excluding <i>H</i> (mean 0.13). The power- and Weibull-<i>H</i> asymptotic model provided the greatest reduction in uncertainty, with the regional Weibull-<i>H</i> model preferred because it reduces uncertainty in smaller-diameter classes that contain the bulk of biomass per hectare in most forests. Propagating the relationships from destructively harvested tree biomass to each of the 327 plots from across the tropics shows errors are reduced from 41.8 Mg ha⁻¹ (range 6.6 to 112.4) to 8.0 Mg ha⁻¹ (−2.5 to 23.0) when including $H$. For all plots, above-ground live biomass was 52.2±17.3 Mg ha⁻¹ lower when including <i>H</i> estimates (13%), with the greatest reductions in estimated biomass in Brazilian Shield forests and relatively no change in the Guyana Shield, central Africa and southeast Asia. We show fundamentally different stand structure across the four forested tropical continents, which affects biomass reductions due to $H$. African forests store a greater portion of total biomass in large-diameter trees and trees are on average larger in diameter. This contrasts to forests on all other continents where smaller-diameter trees contain the greatest fractions of total biomass. After accounting for variation in $H$, total biomass per hectare is greatest in Australia, the Guyana Shield, and Asia and lowest in W. Africa, W. Amazonia, and the Brazilian Shield (descending order). Thus, if closed canopy tropical forests span 1668 million km² and store 285 Pg C, then the overestimate is 35 Pg C if <i>H</i> is ignored, and the sampled plots are an unbiased statistical representation of all tropical forest in terms of biomass and height factors. Our results show that tree $H$ is an important allometric factor that needs to be included in future forest...