Tropical forests play a disproportionate role in the global carbon cycles. However, the seasonal variability in carbon fluxes between tropical rainforests and atmosphere is still poorly understood.
Leaf phenological patterns have been suggested to play an important role in the variations of productivity, yet, the drivers of leaf shedding and flushing remains unclear. While top-canopy trees likely experience similar aboveground climatic conditions at a given site, their access to water and nutrients may differ due to small-scale variations in edaphic properties. However, conventional methods for characterizing soil physical properties are often labor-intensive, destructive, and provide limited spatial coverage, and this prevents from exploring the role of soil spatial heterogeneity in driving the observed variability in tree functioning.
The SOHO project is an inter-disciplinary project, at the intersection of forest ecology, geophysics, and soil sciences, and aims at characterizing the fine-scale spatial soil heterogeneity to understand observed variability in co-occurring tree phenology. The objectives are 1) to provide a first test case to implement geophysical approaches in a tropical forest in French Guiana to describe variability in soil properties, and 2) to complement the pioneering aboveground monitoring of canopy dynamics provided by the CEBA-funded PhenObs project at individual crown level with a belowground continuous information. More specifically, the project consists in implementing three complementary approaches in four 0.5-ha plots distributed along a topographic gradient across the Guyaflux tower footprint and comparing them to assess spatio-temporal variations in soil properties at fine scales. These methods are i) soil coring and texture analysis; ii) soil water content monitoring using automatic probes; iii) two geophysical approaches, electromagnetic induction (EMI) and electrical resistivity tomography (ERT) to spatially extrapolate the punctual information provided by (i) and (ii). The role of soil variations in driving the observed variability in the timing of leaf shedding and flushing across individuals within the PhenObs dataset will then be tested. In doing so, it will provide an unprecedented opportunity to unravel the diversity of trees functioning within a tropical forest through an integrative way.

A small video made by a French journalist during our field trip can be found here:

https://leblob.fr/videos/la-foret-guyanaise-source-ou-puits-de-carbone