In the last decades, the Eddy Covariance (EC) technique has become a standard method to measure net ecosystem CO2 exchange (NEE), but it doesn’t let to distinguish between Gross Primary Production (GPP) and Ecosystem Respiration (Reco). Olive (Olea europea L.) is one of the most important agrosystems on the Mediterranean basin, covering 9.5Mha and accounting for 98% of olive groves global surface. In this study we analyze the EC fluxes from an olive orchard of SE Spain with two soil treatments: 1) leaving spontaneous weed cover (WC) growing on the soil, and 2) inhibiting this growth with a glyphosate-based herbicide (WF). These two different treatments provide high differences in NEE, but the contribution of each component (trees, weed and soil) in the NEE require a better understanding. In this study, we analyze Eddy Covariance fluxes from an olive orchard in SE Spain at different altitudes (above and below the olive trees). To study carbon fluxes contribution of weed in the olive orchard 4 EC towers were installed, placing them on two different areas: one area in WC treatment and the other in WF treatment. On each area, a canopy tower and a subcanopy tower were installed. After a data-filtering during the growth season in which only wind directions coming from olive orchard alleys were accepted, preliminary results from the subcanopy towers show that there are prevailing CO2 emission values from the soil in the WF area and CO2 fixation from the weed in the WC area. On the other hand, during senescence period, CO2 emission fluxes were obtained from both subcanopy towers. These results layout the relevant place of subcanopy towers to understand the role in carbon cycle of the different components in an ecosystem.