Modeling root water uptake depth driven by climate and soil texture using a simple bucket model approach

Root water uptake (RWU) in grasslands is determined by species composition, climate and soil hydraulic properties. Generally, plant communities are adapted to their environment, showing different rooting patterns along climate gradients. Due to climate change, ecosystems are exposed to shifts in precipitation patterns and rising temperatures, causing the need to adapt rooting strategies. RWU is mainly driven by plant transpiration and soil hydraulic status in the rooting zone. Soil hydraulic properties depend strongly on soil texture, which has been observed to influence rooting depth, increasing the root length from fine to coarse soils. Secondly, precipitation patterns affect the typical soil moisture status, and subsequently the rooting depth. Global models suggest that in dry environments RWU should move deeper, to enhance the plant available soil water. However, few studies have at the same time considered the effect of climate and soil properties on RWU depth, although soil properties vary substantially and probably more than precipitation patterns due to climate change.

Biogeochemical models suffer from uncertainty in subsurface hydrological processes, RWU being an important part of it. Thus, ecohydrological models are needed for an integration in larger context biogeochemical models. The trend of ecological models is towards high parameterized models, implying high uncertainty and challenging calibration for those parameters. Especially in the subsurface, parameters are often unknown and are usually impossible to derive from direct measurements. In this project, a simple, parsimonious bucket model was implemented, solving the water balance equation for a multi-layer soil profile. The objective of this work is to predict maximum required RWU depth required to satisfy potential evapotranspiration across established experimental grassland sites with different climate and soil water retention properties. For this we use soil moisture measurements, textures and hydraulic properties determined in three grassland sites of the Nutrient-Network (NutNet) across a climate gradient. We test the sensitivity of the model towards climate and soil hydraulic parameters. First model results show a high sensitivity of RWU depth besides to dynamics to climate, also to soil water retention determined by texture and organic matter content in the soils.