Abstract: The fixation of carbon in tropical forests mainly occurs through the production of wood and leaves, both being the principal components of net primary production. Currently field and satellite observations are independently used to describe the forest carbon cycle, but the link between satellite-derived forest phenology and field-derived forest productivity remains opaque. We used a unique combination of a MODIS enhanced vegetation index (EVI) dataset, a wood production model based on climate data and direct litterfall observations at an intra-annual timescale in order to question the synchronism of leaf and wood production in tropical forests. Even though leaf and wood biomass fluxes had the same range (respectively 2.4 ± 1.4 and 2.2 ± 0.4 Mg C ha-1 yr-1), they occurred separately in time. EVI increased with leaf renewal at the beginning of the dry season, when solar irradiance was at its maximum. At this time, wood production stopped. At the onset of the rainy season, when new leaves were fully mature and water available again, wood production quickly increased to reach its maximum in less than a month, reflecting a change in carbon allocation from short-lived pools (leaves) to long-lived pools (wood). The time lag between peaks of EVI and wood production (109 days) revealed a substantial decoupling between the leaf renewal assumed to be driven by irradiance and the water-driven wood production. Our work is a first attempt to link EVI data, wood production and leaf phenology at a seasonal timescale in a tropical evergreen rainforest and pave the way to develop more sophisticated global carbon cycle models in tropical forests. © 2013 Author(s).

Abstract: The FLUXNET2015 dataset provides ecosystem-scale data on CO2, water, and energy exchange between the biosphere and the atmosphere, and other meteorological and biological measurements, from 212 sites around the globe (over 1500 site-years, up to and including year 2014). These sites, independently managed and operated, voluntarily contributed their data to create global datasets. Data were quality controlled and processed using uniform methods, to improve consistency and intercomparability across sites. The dataset is already being used in a number of applications, including ecophysiology studies, remote sensing studies, and development of ecosystem and Earth system models. FLUXNET2015 includes derived-data products, such as gap-filled time series, ecosystem respiration and photosynthetic uptake estimates, estimation of uncertainties, and metadata about the measurements, presented for the first time in this paper. In addition, 206 of these sites are for the first time distributed under a Creative Commons (CC-BY 4.0) license. This paper details this enhanced dataset and the processing methods, now made available as open-source codes, making the dataset more accessible, transparent, and reproducible.

Abstract: The gelatinous layer (G-layer) of tension wood fibres in hardwood contributes to the mechanical function of the living tree and has significant consequences on properties of solid wood. Its size, shape and structure observed by optical or electron microscopy exhibits characteristic anatomical features. However, we found that sectioning of non-embedded wood samples results in an uncontrolled swelling of the G-layer. In order to assess this artefact, the shape and thickness of the G-layer was monitored by serial sections from an embedded wood sample, from its trimmed transverse face to that located several hundreds of micrometres deep. The results revealed that the initial cutting before embedding produced a border effect responsible for the swollen nature, which is similar to sections from non-embedded material. After a conventional embedding technique was applied, a section of at least 30 micrometres below the trimming surface is required to observe an un-swollen G-layer.

Abstract: The genus Phytophthora represents a group of plant pathogens with broad global distribution. The majority of them cause the collar and root-rot of diverse plant species. Little is known about Phytophthora communities in forest ecosystems, especially in the Neotropical forests where natural enemies could maintain the huge plant diversity via negative density dependence. We characterized the diversity of soil-borne Phytophthora communities in the North French Guiana rainforest and investigated how they are structured by host identity and environmental factors. In this little-explored habitat, 250 soil cores were sampled from 10 plots hosting 10 different plant families across three forest environments (Terra Firme, Seasonally Flooded and White Sand). Phytophthora diversity was studied using a baiting approach and metabarcoding (High-Throughput Sequencing) on environmental DNA extracted from both soil samples and baiting-leaves. These three approaches revealed very similar communities, characterized by an unexpected low diversity of Phytophthora species, with the dominance of two cryptic species close to Phytophthora heveae. As expected, the Phytophthora community composition of the French Guiana rainforest was significantly impacted by the host plant family and environment. However, these plant pathogen communities are very small and are dominated by generalist species, questioning their potential roles as drivers of plant diversity in these Amazonian forests. © 2020 Society for Applied Microbiology and John Wiley & Sons Ltd.

Abstract: The high species diversity of mixed tropical forests hinders the development of forest dynamic models. A solution commonly adopted is to cluster species in groups. There are various methods for grouping species that can be linked to three strategies (i) the ecological subjective strategy, (ii) the ecological data-driven strategy, and (iii) the dynamic process strategy. In the first two strategies a species will be assigned to a single group while in the latter strategy, a specific grouping is defined for each process of population dynamics ( typically based on recruitment, growth, mortality). Little congruency or convergence is observed in the literature between any two classifications of species. This may be explained by the independence between the sets of tree characters used to build species groups, or by the intra-specific variability of these characters. We therefore recommend the dynamic process strategy as the most convenient strategy for building groups of species.