To better describe the magnitudes and patterns of GHG in tropical forests, an automated system continuously measuring in situ soil fluxes of CO2, CH4 and N2O simultaneously, was elaborated and tested in the experimental study site of Paracou, French Guiana. The system was a combination of soil GHG chambers connected to two gas analysers running in parallel to enable the continuous long-term measurements of the three gases. Subsequently, optimal chamber closure time for each GHG were examined. Our automated soil GHG flux system run successfully during two years and this experimental setup was, for the first time, fully described and tested under tropical field conditions (Courtois et al. 2019). In this study we demonstrated that short closure time of 2 min was sufficient for reliable estimations of soil CO2 and CH4 fluxes whereas a closure time of 25 min was more appropriate to estimate soil N2O fluxes. Then, an original extension was designed to better quantify and understand the role of tropical tree stems in GHG exchanges in tropical forests, consisting of a flexible custom-made stem chamber system connected and controlled by the existing soil GHG flux system. Based on this system and an eleven-months period, we hypothesized that, as for the soil, the accuracy of CO2, CH4 and N2O flux estimates in tree stems is gas-specific, where the rate of gas diffusion and build-up within the chamber determines the minimum duration of the measurement cycle needed to compute fluxes most accurately. In our study, we presented the first automated flexible chamber system for continuously measuring in situ stem GHG fluxes at high temporal resolution (every second) under the warm and extremely humid environment of a tropical forest. We also demonstrated that, contrary to CO2, lengthening the stem chamber closure time not only improved the flux measurements but also largely affected final tree stem flux estimates of CH4 and N2O.
In addition, we proved that our new automated system for continuous stem and soil GHG flux measurements was not only able to capture the seasonal variations, but also the diel variations of stem and soil CO2, and to a lesser extent, CH4 and N2O fluxes. Our results showed clearly circadian rhythm in the stem CO2 efflux, as expected from temperature-driven Arrhenius kinetics. Opposite diel patterns between CO2 efflux from the stem and those emitted at an adjacent soil location were also noticeable, suggesting that in tropical forests stem and soil respiratory physiology is not necessarily governed by the same environmental drivers. No circadian rhythms were observed for CH4 and N2O on either the soil and the tree stem, likewise denoting decoupled regulation from that of diel CO2 fluxes.
In parallel to the abovementioned measurements and to disentangle environmental drivers, e.g. precipitations and nutrient availability, that may explain the spatial heterogeneity in soil GHG fluxes in tropical forests, an experiment combining drought and fertilisation treatments was set up in the Paracou study site, French Guiana (Bréchet et al. 2019). Our study revealed contrasted responses in soil fluxes of GHG, CO2 and CH4 in particular, to the treatments, where (i) nitrogen and phosphorus additions, mitigated by soil water content via imposed drought conditions, had a positive effect on CO2 efflux, and (ii) soil water content only strongly and positively affected CH4 fluxes. Surprisingly, fertilization only affected soil CO2 efflux, and drought caused soil to become sources of CH4 instead of sinks. These results suggested that changes in nutrients and water contents in soils most likely influence the complex processes of CO2 and CH4 exchanges, which are controlled by multiple biophysical and biogeochemical conditions, e.g. methanotroph activities.
