Inter-specific variations in tree stem methane and nitrous oxide exchanges in a tropical rainforest

<p>Tropical forests are the most productive terrestrial ecosystems, global centres of biodiversity and important participants in the global carbon and water cycles. The Amazon, which is the most extensive tropical forest, can contain more than 600 trees (diameter at breast height above 10 cm) and up to 200 tree species in only one hectare of forest. In upland forest, tropical soils are known to be a methane (CH<sub>4</sub>) sink and a weak source of nitrous oxide (N<sub>2</sub>O), which are both major greenhouse gases (GHG). Most of researches on GHG fluxes have been conducted on the soil compartment but recent works reported that tree stems of some tropical forests can be a substantial source of CH<sub>4</sub> and, a to lesser extend of N<sub>2</sub>O. Tropical tree stems can act as conduits of soil-produced GHG but biophysical mechanisms controlling GHG fluxes and differences among tree species are not yet fully understood.</p><p>In order to quantify CH<sub>4</sub> and N<sub>2</sub>O fluxes of different tropical tree species, we took gas samples in 101 mature tree stems of twelve species with the manual chamber technique during the wet season 2020, in a French Guiana forest. Tree species were selected because of their abundance and their habitat preference. We chose trees belonging to two contrasted forest habitats, the hill-top and hill-bottom, which are respectively characterized by aerobic conditions and seasonal anaerobic conditions. Simultaneously with sampling GHG, we measured bark moisture and tree diameter. Four tree species were found in both habitats whereas the eight others were only present in one of these two habitats.</p><p>Among the 101 tree stems, 78.6% were net sources of CH<sub>4</sub> with a greater proportion in hill-bottom than hill-top. Overall, stem CH<sub>4</sub> fluxes were significantly and positively correlated with the wood density (χ<sup>2</sup> = 28.0; p < 0.01; N = 75) but neither with the habitat, bark moisture or tree size. We found a significant effect of the tree species on stem CH<sub>4</sub> fluxes (F = 3.7, p < 0.001) but no interactions between the tree species and habitats.</p><p>Among 43.0% of the stem N<sub>2</sub>O fluxes that were different from zero, half were from trees that were net sources of N<sub>2</sub>O mainly located in hill-top. Stem N<sub>2</sub>O fluxes are not significantly correlated with habitat, as also with the tree size, wood density or bark moisture. Unlike stem CH<sub>4</sub> fluxes, tree species did not significantly influence stem N<sub>2</sub>O fluxes.</p><p>Our study revealed that, in tropical forest, spatial variations in GHG fluxes would not only depend on soil water conditions, but also on tree species. Specific tree traits such as the wood density can favour stem CH<sub>4</sub> emissions by providing more or less effective pore space for CH<sub>4</sub> diffusion but seems to have a limited influence on stem N<sub>2</sub>O fluxes maybe because of the lower diffusive and ebullitive transport of N<sub>2</sub>O compared to CH<sub>4</sub>. Further investigation linking tree species traits and tree GHG fluxes are, however, necessary to elucidate the processes and mechanisms behind tree CH<sub>4</sub> and N<sub>2</sub>O exchanges.</p>