Effect of the water stress on gross primary production modeling of a Mediterranean oak savanna ecosystem

Abstract. Dehesa ecosystem consists of widely-spaced oak trees combined with crops, pasture and Mediterranean shrubs. It is located in the southwest of the Iberian Peninsula, where water scarcity is recurrent, severely affecting the multiple productions and services of the ecosystem. Upscaling in situ Gross Primary Production (GPP) estimates in these areas is challenging for regional and global studies, given the significant spatial variability of plant functional types and the vegetation stresses usually present. The estimation of GPP is often addressed using light use efficiency models (LUE-models). Under soil water deficit conditions, biomass production is reduced below its potential rate. This work investigates the effect of different parameterizations to account for water stress on GPP estimates and their agreement with observations. Ground measurements of GPP are obtained using an Eddy Covariance (EC) system installed over an experimental site located in Cordoba, Spain. GPP is estimated with a LUE-model in the footprint of the EC tower using several approaches: a fixed value taken from previous literature; a fixed value modified by daily weather conditions; and both formulations modified by an additional coefficient to explicitly consider the vegetation water stress. The preliminary results obtained during two hydrological years (2015/2016 and 2016/2017) are compared, focusing on specific wet and dry periods.