Resolving the links between wood production, leaf phenology and whole-ecosystem carbon fluxes in temperate forests

Wood production is an essential component of terrestrial carbon dynamics, but we only have a limited understanding of the environmental cues that trigger wood production to start and stop during the growing season and how these vary among temperate tree species. Moreover, we lack a clear picture of how the seasonal timings of wood production relate to leaf phenology and whole-ecosystem carbon fluxes - severely limiting our ability to estimate woody productivity from remote sensing or eddy covariance flux tower data. To address this knowledge gap, between 2023 and 2024 we used automated dendrometers to take hourly measurements of stem diameter variations across 160 trees representing seven locally-dominant broadleaf and coniferous species in the Wytham Woods 18-ha ForestGEO plot in the UK. We combined these with overlapping flux tower measurements of gross primary production (GPP) and net ecosystem exchange (NEE), NDVI time-series generated from Sentinel-2 to capture canopy phenology and local microclimate data. Using these complementary datasets we found that wood growth started later and ended much earlier than one might estimate from NDVI alone. Moreover, temporal trends in wood production (including the onset, maximum rate and cessation of growth) varied significantly between species - with beech and oak trees growing almost 60 days longer per year than sycamore and ash. This variation in wood phenology across species significantly complicates any attempts to infer wood production from flux tower measurements of GPP and NEE. Our study advances our understanding of the synchronization and mismatches between ecosystem carbon uptake and investment in wood production in temperate forests. We highlight the potential of combining remote sensing, flux tower and high-resolution dendrometer data to improve our ability to track terrestrial carbon cycling at scale and predict its responses to climate change.