Inverse modelling to estimate plant hydraulic traits and water use strategies in a Costa Rican tropical dry forest

This study aims to understand water uptake depths and dynamics of tropical dry forest trees, namely RonRon (Astronium graveolens), Guacimo (Guazuma ulmifolia), Guapinol (Hymenaea courbaril) and Caoba (Swietenia macrophylla), at Estación Experimental Forestal Horizontes, an intensively monitored research site in northwestern Costa Rica. The climate of this region is driven by the El Niño–Southern Oscillation, which results in distinct wet (June to December) and dry seasons (December to May). We combine field measurements of sap flow and soil moisture, collected between December 2020 to December 2021, to estimate plant hydraulic traits through inverse modelling with a differential evolution approach using the mechanistic plant hydraulic model SurEau-Ecos. This allowed the estimate of 12 plant hydraulic parameters for each of these four species. We compare these inversely estimated traits across species and link them to observed soil moisture and sap flow dynamics. Our results show distinct hydraulic strategies for each plant, which feedback into the spatiotemporal dynamics of soil moisture. Caoba and Guacimo conserve water through early stomatal closure to inhibit transpiration during the dry season, while RonRon and Guapinol keep their stomata open and sustain a relatively higher transpiration rate even with limited soil moisture. Overall, clear differences in drought-response strategies among species, including general isohydric and anisohydric behavior, are shown both in estimated plant hydraulic traits and in ecohydrological signatures. A limitation of this study is that interspecies interactions have been neglected. Nevertheless, fair agreement between model results and field observations has been achieved. Our findings contribute to the mechanistic understanding of hydraulic strategies of tropical dry forests, which are currently understudied ecosystems.