Reconciling seasonal hydraulic risk and plant water use through probabilistic soil-plant dynamics.
- Author(s): Feng, Xue
- Dawson, Todd E
- Ackerly, David D
- Santiago, Louis S
- Thompson, Sally E
- et al.
Published Web Locationhttps://doi.org/10.1111/gcb.13640
Current models used for predicting vegetation responses to climate change are often guided by the dichotomous needs to resolve either (i) internal plant water status as a proxy for physiological vulnerability or (ii) external water and carbon fluxes and atmospheric feedbacks. Yet, accurate representation of fluxes does not always equate to accurate predictions of vulnerability. We resolve this discrepancy using a hydrodynamic framework that simultaneously tracks plant water status and water uptake. We couple a minimalist plant hydraulics model with a soil moisture model and, for the first time, translate rainfall variability at multiple timescales - with explicit descriptions at daily, seasonal, and interannual timescales - into a physiologically meaningful metric for the risk of hydraulic failure. The model, parameterized with measured traits from chaparral species native to Southern California, shows that apparently similar transpiration patterns throughout the dry season can emerge from disparate plant water potential trajectories, and vice versa. The parsimonious set of parameters that captures the role of many traits across the soil-plant-atmosphere continuum is then used to establish differences in species sensitivities to shifts in seasonal rainfall statistics, showing that co-occurring species may diverge in their risk of hydraulic failure despite minimal changes to their seasonal water use. The results suggest potential shifts in species composition in this region due to species-specific changes in hydraulic risk. Our process-based approach offers a quantitative framework for understanding species sensitivity across multiple timescales of rainfall variability and provides a promising avenue toward incorporating interactions of temporal variability and physiological mechanisms into drought response models.