Understanding spatial variations in tree water stress across species and hillslope gradients

Understanding the spatial variability of tree water stress in complex terrain remains challenging due to strong heterogeneity in subsurface hydrology, plant water availability, and species-specific physiological responses. Along hillslopes, topographic gradients influence soil moisture redistribution, while tree species differ in their hydraulic regulation strategies and sensitivity to drought. Although field-based measurements provide detailed insights into individual-tree water relations, scaling these observations to heterogeneous landscapes remains limited. 

In this study, we investigate how hillslope position, associated with spatial heterogeneity in water stress exposure driven by slope and soil depth,  and species identity, modulate tree water stress across two forested sites in Valais, Switzerland: Saillon and Lens. Saillon is a mixed Quercus robur-Fagus sylvatica forested hillslope ranging from 645 to 1110 m a.s.l., with study trees located between 830 and 960 m a.s.l. Along this hillslope, soil depth varies strongly, with deep loess deposits upslope and shallower soils downslope, providing pronounced spatial heterogeneity in water stress exposure. The Lens site is a Pinus sylvestris-dominated forest with a south-facing hillslope ranging from 1057-1197 m.a.s.l. Tree water deficit was monitored using stem dendrometers installed on individual trees (7 oak, 8 beech, and 2 representative pine trees), complemented by in situ measurements of soil water potential and meteorological variables. To capture spatial and temporal variability in canopy conditions and link point-scale physiological measurements to landscape-scale patterns, unmanned aerial vehicles (UAV)-based multispectral and thermal imagery were acquired repeatedly over two consecutive growing seasons (2024 and 2025).

Preliminary analyses from the 2024 season indicate clear spatial and temporal patterns in tree water status across species and hillslope positions at the Saillon site, with downslope trees generally exhibiting higher water deficit and reduced canopy greenness compared to upslope trees, particularly during mid to late summer, reflecting the site-specific soil depth gradient, with shallower soils downslope and deeper loess deposits upslope. Relationships among UAV-derived spectral and thermal metrics, tree water deficit, and soil water potential were examined across species and hillslope positions. This integrated, multi-scale observational framework aims to improve the interpretation and spatial scaling of plant water stress across heterogeneous landscapes and complex terrain.