Assessment of tree-induced water regime - D5.3

This delivery reports on the analysis of soil water regime and the influence of trees and shrubs om this process. The presence of trees and shrubs in crop and pasture fields (agroforestry) is highly promoted to buffer climatic variability and water constraint in dry areas (Van Noordwijk et al. 2014). It is assumed that they will positively influence the soil water regime through; i. long term effects on soil properties such as as density/porosity, infiltration and water retention and ii. short term effects by changing the water balance components as preferential infiltration along coarse roots, soil evaporation, water uptake, water table recharge, root hydraulic redistribution, surface soil water availability for crops. The objectives of the task were to analyse these potential effects, and to provide inputs for modeling tree influence at different scales (WP4 & WP7). The approach combines large scale collection of indicators in agroforestry systems of farmer fields in Burkina, Mali and Sénégal and in depth studies using innovative devices in the Faidherbia Flux site of Sénégal. The data analysis is still running but several key results have been obtained: Improvement of soil bulk density (decrease) and water infiltration (increase) are generally observed close to trees and shrubs. However the difference may be slight and variable between sites and systems. The decrease or buffering of crop water stress at tree proximity is evidenced but also variable. For short term effects on the water balance, several insights were observed in Faidherbia-Flux site: Preferential infiltration along coarse roots after stemflow Tree water use reach up to 300 l/day per individual, with seasonal course highly related to the inverted phenology. Despite annual water use estimated about 40 000 L/tree, it represents only 5 to 10% of the annual rainfall when the relatively low tree density is taken into account. Knowing the benefits on crop productivity, this suggest the interest of increasing tree density without risk for the water resource. However the estimated recharge of the water table as assessed through isotopes signature is low and of similar order than tree water use. Hence an increase of tree density could have a significant impact on recharge and related constraints such as salinity. For the first time, root hydraulic redistribution has been evidenced under Faidherbia albida tree. The daily water increment at night can reach up to 1% at 100 cm soil depth. However the phenomenon appears heterogeneous both in space and time. GPR scanning revealed that all the area of the field is colonized by tree superficial roots even 20 m from trees and mainly located between 40 and 140 cm with an average around 90 cm. Knowing that root hydraulic redistribution has also been evidenced in the past on several sahelian species of the project as A. tortilis, P. reticulatum and G. senegalensis, our results with Faidherbia trees confirm the generality of the process and points out the interest in taking into account this mechanism in future modeling works. The results and experiences also provide some methodological insights: The impact of innovative tools or applications to understand processes (sap flow measurements with TIMFLOW prototype, automatic infiltrometer, georadar) but with practical limitation to extend in many fields The observation of the large tree root colonization in open spaces of agroforestry systems, which somehow challenges the approach comparing plots close and far from trees reveal the tree effects. Moreover, processes that occurred right under tree or shrubs collar as preferential infiltration seen by GPR, cannot be catched by classical measurements. Importance (and difficulty) of modelling hydrodynamics in sandy soils Validation of the multi-scale and parallel approach adopted in this project which combines simple indicators collection in the different sites, and comprehensive in-depth study of underlying processes in selected instrumented test sites.