Does increased spatial replication above heterogeneous agroforestry improve the representativeness of eddy covariance measurements?

Abstract. Spatial heterogeneity in terrestrial ecosystems compromises the accuracy of eddy covariance measurements. An example of heterogeneous ecosystems are temperate agroforestry systems, that have been poorly studied by eddy covariance. Agroforestry systems get an increasing attention due to their potential environmental benefits, e.g. a higher carbon sequestration, enhanced microclimate and erosion reduction compared to monocropping agricultural systems. Lower-cost eddy covariance setups might offer an opportunity to reduce this bias by allowing for more spatial replicates of flux towers. The aim of this study was to quantify the spatial variability of carbon dioxide (FC), latent heat (LE) and sensible heat (H) fluxes above a heterogeneous agroforestry system in northern Germany using a distributed network of three lower-cost eddy covariance setups across the agroforestry systems. Fluxes from the three towers in the agroforestry were further compared to fluxes from an adjacent monocropping site. The campaign took place from March 2023 until September 2024. The results indicated that the spatial variability of fluxes was largest for FC, attributed to the effect of different crops (rapeseed, corn and barley) within the flux footprints contributed to the measured fluxes. Differences between fluxes across towers were enhanced after harvest events. However, the temporal variability due to the seasonality and diurnal cycles during the campaign was larger than the spatial variability across the three towers. When comparing fluxes between the agroforestry and the monocropping systems, weekly sums of carbon and evapotranspiration fluxes followed similar seasonality, with peak values during the growing season of-50 g C m⁻² week⁻¹ and 40 mm week⁻¹, respectively. The variation of the magnitude depended on the phenology of the different crops. The effect size, which is an indicator of the representativeness of the fluxes across the distributed network of three eddy covariance towers against only one, showed in conjunction with the other results that the spatial heterogeneity across the agroforestry was better captured by the network of three stations. This supports previous findings that spatial heterogeneity should be taken into account in eddy covariance studies, and that lower-cost setups may offer the opportunity to bridge this gap and improve the accuracy of eddy covariance measurements above heterogeneous ecosystems.