Abstract: Atmospheric carbon dioxide records indicate that the land surface has acted as a strong global carbon sink over recent decades, with a substantial fraction of this sink probably located in the tropics, particularly in the Amazon. Nevertheless, it is unclear how the terrestrial carbon sink will evolve as climate and atmospheric composition continue to change. Here we analyse the historical evolution of the biomass dynamics of the Amazon rainforest over three decades using a distributed network of 321 plots. While this analysis confirms that Amazon forests have acted as a long-term net biomass sink, we find a long-term decreasing trend of carbon accumulation. Rates of net increase in above-ground biomass declined by one-third during the past decade compared to the 1990s. This is a consequence of growth rate increases levelling off recently, while biomass mortality persistently increased throughout, leading to a shortening of carbon residence times. Potential drivers for the mortality increase include greater climate variability, and feedbacks of faster growth on mortality, resulting in shortened tree longevity. The observed decline of the Amazon sink diverges markedly from the recent increase in terrestrial carbon uptake at the global scale, and is contrary to expectations based on models. © 2015 2015 Macmillan Publishers Limited.

Abstract: Tension wood of Laetia procera (Poepp.) Eichl. (Flacourtiaceae), a neo-tropical forest species, shows a peculiar secondary wall structure, with an alternance of thick and thin layers, while opposite wood of this species has a typical secondary wall structure (S1 + S2 + S3). Samples for the study of microstructural properties were collected upon the estimation of growth stresses in the living tree, in order to analyze the correlation of the former with the latter. Investigation using optical microscopy, scanning electron microscopy and UV microspectrophotometry allowed the description of the anatomy, ultra-structure and chemistry of this peculiar polylaminate secondary wall. In the thick layers, cellulose microfibril angle is very low (i.e., microfibril orientation is close to fibre axis) and cellulose microfibrils are well organized and parallel to each other. In the thin layers, microfibrils (only observable in the inner layer) are less organized and are oriented with a large angle relative to the axis of the cell. Thick layers are lightly lignified although thin layers show a higher content of lignin, close to that of opposite wood secondary wall. The more the wood was under tensile stress, the less the secondary wall was lignified, and lower the syringyl on guaiacyl lignin units' ratio was. The innermost layer of the secondary wall looks like a typical S3 layer with large microfibril angle and lignin occurrence. The interest of this kind of structure for the understanding of stress generation is discussed.

Abstract: A generalisation of existing mechanical models is proposed to account for the relation between wood macroscopic properties and fibre microstructure and chemical composition. It is applied to understanding of the origin of anisotropic maturation strains measured at the outermost surface of the xylem. Various assumptions are considered for boundary conditions of the fibre during the progressive maturation process and are applied to experimental data from the literature. Assumptions that the fibre is fully restrained in displacement, or fully unrestrained or unrestrained in the transverse direction only are all incompatible with observations. Indeed, within the tree, the fibre is restrained in the longitudinal and tangential directions, but unrestrained in the radial direction towards the bark. Mixed boundary conditions must be introduced to correctly simulate both longitudinal and tangential maturation strains. In the context of an analytical axisymmetric model, this is estimated by considering a parameter of partial release of tangential stress during maturation. Consistence with data and with finite element computation in the case of a square fibre confirmed that, because of the unrestrained radial condition, a large part of the tangential maturation stress is released in situ.

Abstract: Tank-bromeliads are discrete habitats which contain distinct aquatic communities, and which commonly occur in the neotropics. Because they span a broad range of ecological gradients in terms of habitat structure and amount of resources, researchers study the associations between the biodiversity of communities and these gradients in rainforests, where the very high species richness and densities within other continuous habitats makes it difficult to quantify animal communities. We analysed the diversity of aquatic insect communities in relation to different tank-bromeliad species in a primary rainforest (French Guiana) using artificial intelligence and complex optimization techniques to classify communities and model their determinants. First, the self-organizing map (neural network) was used to classify 158 bromeliads according to the quantitative structure of the insect communities. Catopsis berteroniana and Guzmania lingulata formed separate clusters of plants on the virtual map, while Vriesea splendens, Vriesea pleiosticha and Aechmea melinonii were grouped together in the remaining clusters. Some insect taxa occurred in all bromeliads, while other taxa were specific to a given species. Second, general linear modelling allowed us to specify the influence of the bromeliad species; water volume and volume of fine particulate organic matter inside of the tank; elevation above the ground; and sampling site on taxonomic richness and insect abundance. The number of taxa and individuals per plant increased with greater water volume, but the slope of the relationships depended on the bromeliad species. The significant influence of bromeliad species suggested that at similar water volumes different plant species had different taxon richness and insect abundance. Greater amounts of fine particulate organic matter were detrimental to community diversity in the tanks, probably because they decreased available space by clogging it and/or affected oxygen concentrations. The influence of tank-bromeliad species on the aquatic insect community was primarily related to their physical (and probably chemical) features rather than to species-specific associations sensu stricto. The classification of bromeliads with respect to animal species is likely to provide referential schemes for those biodiversity patterns to be expected under certain conditions, and may help to target model communities for subsequent experimental research. (C) 2008 Elsevier B,V. All rights reserved.

Abstract: One third of contemporary tropical forests is designated by national forest services for timber production. Tropical forests are also increasingly affected by anthropogenic disturbances. However, there is still much uncertainty around the capacity of tropical forests to recover their timber volume after logging as well as other disturbances such as fires, large blow-downs and extreme droughts, and thus on the long-term sustainability of logging. We developed an original Bayesian hierarchical model of Volume Dynamics with Differential Equations (VDDE) to infer the dynamic of timber volumes as the result of two ecosystem processes: volume gains from tree growth and volume losses from tree mortality. Both processes are expressed as explicit functions of the forest maturity, i.e. the overall successional stage of the forest that primarily depends on the frequency and severity of the disturbances that the forest has undergone. As a case study, the VDDE model was calibrated with data from Paracou, a long-term disturbance experiment in a neotropical forest where over 56 ha of permanent forest plots were logged with different intensities and censused for 31 years. With this model, we could predict timber recovery at Paracou at the end of a cutting cycle depending on the logging intensity, the rotation cycle length, and the proportion of commercial volume. The VDDE modelling framework developed presents three main advantages: (i) it can be calibrated with large tree inventories which are widely available from national forest inventories or logging concession management plans and are easy to measure, both on the field and with remote sensing; (ii) it depends on only a few input parameters, which can be an advantage in tropical regions where data availability is scarce; (iii) the modelling framework is flexible enough to explicitly include the effect of other types of disturbances (both natural and anthropogenic: e.g. blow-downs, fires and climate change) on the forest maturity, and thus to predict future timber provision in the tropics in a context of global changes. © 2018 Elsevier B.V.