Long-term nitrogen and phosphorus additions alter soil and tree stem methane and nitrous oxide fluxes under contrasting soil water conditions in a tropical forest

Globally, tropical forests are thought to be an important source of atmospheric nitrous oxide (N₂O) and a sink for methane (CH₄), with small biome-wide changes in the structure, dynamics and environment of these forests either mitigating or exacerbating increases in atmospheric concentrations of these major greenhouse gases (GHGs). Anthropogenic activities have dramatically increased nitrogen (N) and phosphorus (P) inputs to the biosphere, potentially altering soil biogeochemical cycles. However, the effects of N and P addition on soil CH₄ and N₂O fluxes in tropical forest ecosystems are not yet understood. Besides soils, tree-mediated transport can also contribute significantly to GHG exchange in forests. In the soil, CH₄ and N₂O produced can be absorbed by roots and transported into aboveground tree tissues. In addition, these gases can be produced in trees by microorganisms living in the tissues or by physiological and photochemical processes. Yet observations of CH₄ and N₂O fluxes in tropical forests, particularly in tree stems, are still limited and have not been described in the context of long-term nutrient addition experiments.

Here we report data derived from measurements of soil and stem fluxes and environmental variables in N (+N) and P (+P) addition plots over seven years in a tropical forest of the north-eastern Amazon, French Guiana. In each plot (+N, +P, +NP and controls), CH₄ and N₂O fluxes, soil water content (SWC), soil and air temperature, total N and carbon content and available P were measured at five different locations combining a tree stem (> 30 cm diameter) and its surrounding soil. These measurements were made in plots located in three contrasting habitats, a well-drained, nutrient-poor soil at the top of the hill (upland area) and two waterlogged, nutrient-rich soils at the bottom of the hill (seasonally and permanently flooded areas).

We found that soil and stem CH₄ and N₂O fluxes were highly spatially variable in situ. In the control plots, soil CH₄ uptake and N₂O emissions decreased with increasing SWC (i.e. from the hill-top to the wettest hill-bottom). Regardless of the forest habitat, N additions (+N and +NP) resulted in substantially higher soil N₂O fluxes, whereas P additions (+P) resulted almost exclusively in soil CH₄ uptake. This suggests that N addition increases soil N beyond microbial immobilisation and plant nutritional requirements, with the excess being nitrified or denitrified, while P addition stimulates soil methanotrophic activity. In the control plots, stems growing in the waterlogged soils of the permanently flooded area were moderate and strong emitters of N₂O and CH₄, respectively. For both gases, CH₄ and N₂O, higher stem fluxes resulted from P addition (+P and +NP) in hill-bottom plots.

Our results highlight (i) the key role of N and P in CH₄ and N₂O cycling in tropical forest soils and (ii) the substantial CH₄ and N₂O source potential of tree stems in highly waterlogged areas. This underlines the importance of including processes related to water, N and P availability in GHG flux modelling in tropical forests.