Previously, tree water storage was inferred through high frequency micrometer scale dendrometer or psychrometer measurements. However, novel use of frequency domain reflectometry (FDR) sensors allows for the continuous, direct measurements of wood water content in situ in large trees of differing vascular structures. We present results from a two year study of stem water storage dynamics and sap flow in five species in a mixed deciduous forest in northern lower Michigan. The forest receives ~800mm of rainfall annually, but water potential in the site’s well-drained sandy soil nears the wilting point several times a year during long interstorm periods. We demonstrate radical differences in stem water storage use, or storage withdrawal, between drought tolerant and intolerant species. Red maple, a drought intolerant species, showed a strong dependence on stored water for transpiration during both wet and dry periods. Red oak, a more drought hearty species, was much less reliant on storage during all conditions. During well-watered conditions, withdrawal from storage by red maple was ~10 kg day-1, while storage withdrawal from similarly sized red oaks was ~1 kg day-1. Red oaks only drew strongly upon stem water storage during periods of extremely dry soils (< 4% volumetric water content).
The dynamic change in stem water storage throughout the course of the day is largely responsible for the biotic control of the hysteretic behavior of transpiration. Hysteresis in transpiration is governed abiotically through the time lag between photosynthetically active radiation (PAR) and vapor pressure deficit (VPD), and biotically through dynamic regulation of transpiration by vegetation. For a given set of atmospheric conditions, namely PAR and VPD, a tree will transpire more during the morning hours when nocturnal recharge has replenished stem water storage and hydraulic capacitance, than for the same set of atmospheric conditions in the afternoon, when storage reserves have been depleted by transpiration. We find that trees such as red maple, which demonstrate strong reliance on stem-stored water as a buffer for transpiration, have larger diurnal hysteresis than red oak, which do not rely heavily on storage flux.
The newest generation of advanced plant hydrodynamics models, such as FETCH2, mechanistically simulate water flow through the xylem system of trees for the purpose of more accurate predictions of transpiration. These hydrodynamic models have the capability to resolve the dynamics of stem water storage and resolve the hysteresis of transpiration at a level currently unattainable through the semi-empirical functions within land-surface and dynamic global vegetation models. The availability of these new model and measurement technologies will enable the improvement of our understanding of species’ water status regulation in response to drying soil and drought.