Abstract: Understanding how plants are constructed; i.e., how key size dimensions and the amount of mass invested in different tissues varies among individuals; is essential for modeling plant growth, estimating carbon stocks, and mapping energy fluxes in the terrestrial biosphere. Allocation patterns can differ through ontogeny, but also among coexisting species and among species adapted to different environments. While a variety of models dealing with biomass allocation exist, we lack a synthetic understanding of the underlying processes. This is partly due to the lack of suitable data sets for validating and parameterizing models. To that end, we present the Biomass and allometry database (BAAD) for woody plants. The BAAD contains 259 634 measurements collected in 176 different studies, from 21 084 individuals across 678 species. Most of these data come from existing publications. However, raw data were rarely made public at time of publication. Thus the BAAD contains individual level data from different studies, transformed into standard units and variable names. The transformations were achieved using a common workflow for all raw data files. Other features that distinguish the BAAD are: (i) measurements were for individual plants rather than stand averages; (ii) individuals spanning a range of sizes were measured; (iii) inclusion of plants from 0.01-100 m in height; and (iii) biomass was estimated directly, i.e., not indirectly via allometric equations (except in very large trees where biomass was estimated from detailed subsampling). We included both wild and artificially grown plants. The data set contains the following size metrics: total leaf area; area of stem crosssection including sapwood, heartwood, and bark; height of plant and crown base, crown area, and surface area; and the dry mass of leaf, stem, branches, sapwood, heartwood, bark, coarse roots, and fine root tissues. We also report other properties of individuals (age, leaf size, leaf mass per area, wood density, nitrogen content of leaves and wood), as well as information about the growing environment (location, light, experimental treatment, vegetation type) where available. It is our hope that making these data available will improve our ability to understand plant growth, ecosystem dynamics, and carbon cycling in the world's vegetation.

Abstract: Background: Managed forests are a major component of tropical landscapes. Production forests as designated by national forest services cover up to 400 million ha, i.e. half of the forested area in the humid tropics. Forest management thus plays a major role in the global carbon budget, but with a lack of unified method to estimate carbon fluxes from tropical managed forests. In this study we propose a new time- and spatially-explicit methodology to estimate the above-ground carbon budget of selective logging at regional scale. Results: The yearly balance of a logging unit, i.e. the elementary management unit of a forest estate, is modelled by aggregating three sub-models encompassing (i) emissions from extracted wood, (ii) emissions from logging damage and deforested areas and (iii) carbon storage from post-logging recovery. Models are parametrised and uncertainties are propagated through a MCMC algorithm. As a case study, we used 38 years of National Forest Inventories in French Guiana, northeastern Amazonia, to estimate the above-ground carbon balance (i.e. the net carbon exchange with the atmosphere) of selectively logged forests. Over this period, the net carbon balance of selective logging in the French Guianan Permanent Forest Estate is estimated to be comprised between 0.12 and 1.33 Tg C, with a median value of 0.64 Tg C. Uncertainties over the model could be diminished by improving the accuracy of both logging damage and large woody necromass decay submodels. Conclusions: We propose an innovating carbon accounting framework relying upon basic logging statistics. This flexible tool allows carbon budget of tropical managed forests to be estimated in a wide range of tropical regions. © 2016 The Author(s).

Abstract: The Amazon Basin has experienced more variable climate over the last decade, with a severe and widespread drought in 2005 causing large basin-wide losses of biomass. A drought of similar climatological magnitude occurred again in 2010; however, there has been no basin-wide ground-based evaluation of effects on vegetation. We examine to what extent the 2010 drought affected forest dynamics using ground-based observations of mortality and growth from an extensive forest plot network. We find that during the 2010 drought interval, forests did not gain biomass (net change: −0.43 Mg ha−1, confidence interval (CI): −1.11, 0.19, n = 97), regardless of whether forests experienced precipitation deficit anomalies. This contrasted with a long-term biomass sink during the baseline pre-2010 drought period (1998 to pre-2010) of 1.33 Mg ha−1 yr−1 (CI: 0.90, 1.74, p < 0.01). The resulting net impact of the 2010 drought (i.e., reversal of the baseline net sink) was −1.95 Mg ha−1 yr−1 (CI:−2.77, −1.18; p < 0.001). This net biomass impact was driven by an increase in biomass mortality (1.45 Mg ha−1 yr−1 CI: 0.66, 2.25, p < 0.001) and a decline in biomass productivity (−0.50 Mg ha−1 yr−1, CI:−0.78, −0.31; p < 0.001). Surprisingly, the magnitude of the losses through tree mortality was unrelated to estimated local precipitation anomalies and was independent of estimated local pre-2010 drought history. Thus, there was no evidence that pre-2010 droughts compounded the effects of the 2010 drought. We detected a systematic basin-wide impact of the 2010 drought on tree growth rates across Amazonia, which was related to the strength of the moisture deficit. This impact differed from the drought event in 2005 which did not affect productivity. Based on these ground data, live biomass in trees and corresponding estimates of live biomass in lianas and roots, we estimate that intact forests in Amazonia were carbon neutral in 2010 (−0.07 Pg C yr−1 CI:−0.42, 0.23), consistent with results from an independent analysis of airborne estimates of land-atmospheric fluxes during 2010. Relative to the long-term mean, the 2010 drought resulted in a reduction in biomass carbon uptake of 1.1 Pg C, compared to 1.6 Pg C for the 2005 event. ©2016. American Geophysical Union. All Rights Reserved.

Abstract: Fish smoking with local wood species is a traditional practice in French Guiana. We evaluated the carcinogenic risk and the smoky taste in acoupa weakfish (Cynoscion acoupa) smoked by a small Guianese company specializing in fish smoking using local wood species. The goal of this study is to promote regional economic development by encouraging the establishment of small companies offering fish smoked with local wood species in agreement with the European health norms in terms of polycyclic aromatic hydrocarbon (PAH) content in smoked fish. The fish smoked with three species of wood traditionally used in French Guiana, Parinari campestris, Caesaria grandiflora, and Laetia procera, conformed to European standards for PAH content. Their phenol contents (correlated with the smoky taste) were close to smoked salmon, the reference in Europe. Given the low rate of extractable compounds in these woods, other flavors had little chance of predominating on the smoky taste. These three tropical wood species might be used for the production of cold smoked fish in compliance with European standards for PAH residues. The flavor and consumer's acceptance of the smoked fish should now be investigated to characterize the added typicity of local woods in comparison to the commonly used European woods. © 2016 Society of Ethnobiology.

Abstract: Mapping forest aboveground biomass (AGB) has become an important task, particularly for the reporting of carbon stocks and changes. AGB can be mapped using synthetic aperture radar data (SAR) or passive optical data. However, these data are insensitive to high AGB levels (>150 Mg/ha, and >300 Mg/ha for P-band), which are commonly found in tropical forests. Studies have mapped the rough variations in AGB by combining optical and environmental data at regional and global scales. Nevertheless, these maps cannot represent local variations in AGB in tropical forests. In this paper, we hypothesize that the problem of misrepresenting local variations in AGB and AGB estimation with good precision occurs because of both methodological limits (signal saturation or dilution bias) and a lack of adequate calibration data in this range of AGB values. We test this hypothesis by developing a calibrated regression model to predict variations in high AGB values (mean >300 Mg/ha) in French Guiana by a methodological approach for spatial extrapolation with data from the optical geoscience laser altimeter system (GLAS), forest inventories, radar, optics, and environmental variables for spatial inter- and extrapolation. Given their higher point count, GLAS data allow a wider coverage of AGB values. We find that the metrics from GLAS footprints are correlated with field AGB estimations (R2 = 0.54, RMSE = 48.3 Mg/ha) with no bias for high values. First, predictive models, including remote-sensing, environmental variables and spatial correlation functions, allow us to obtain “wall-to-wall” AGB maps over French Guiana with an RMSE for the in situ AGB estimates of ∼50 Mg/ha and R2 = 0.66 at a 1-km grid size. We conclude that a calibrated regression model based on GLAS with dependent environmental data can produce good AGB predictions even for high AGB values if the calibration data fit the AGB range. We also demonstrate that small temporal and spatial mismatches between field data and GLAS footprints are not a problem for regional and global calibrated regression models because field data aim to predict large and deep tendencies in AGB variations from environmental gradients and do not aim to represent high but stochastic and temporally limited variations from forest dynamics. Thus, we advocate including a greater variety of data, even if less precise and shifted, to better represent high AGB values in global models and to improve the fitting of these models for high values. © 2016 Elsevier B.V.