The expanding stream of remote sensing data offers new opportunities to understand the effects of plant hydraulics on the global carbon and water cycles. Response-based metrics, such as isohydricity, seek to characterize plant water use strategy from transient observations of broadly observable ecosystem variables. Such metrics have particular appeal at the landscape or greater scale, due to the uncertainty involved in aggregating field-derived trait data. The utility of isohydricity has been discussed in the recent literature[1], asking whether the emergent relationship between transient soil and leaf water potentials can effectively diagnose a meaningful facet of vegetation water use strategy. Adopting a model-based approach, we ask: what trait or combination of traits do two common isohydricity metrics (σ and hydroscape area) effectively measure, and how might this be confounded by prevailing meteorological conditions? We propose an alternative metric that offers a narrower trait definition, comparing it with isohydricity in a set of synthetic test cases.
Results/Conclusions
Using a single hypothetical plant species, we found that both σ and hydroscape area returned four distinct values, when tested across four levels of vapor pressure deficit, despite identical hydraulic traits. Our alternative metric, relative isohydricity, effectively controls for vapor pressure deficit, and returned a single value for the four simulations. Because ecological response (e.g. a reduction in transpiration) is a product of forcing (e.g. reduced soil moisture) and water use strategy, we must take care to control for meteorological conditions when inferring water use strategy from transient observations of response variables. We discuss some of the challenges with response-based metrics, including data requirements, linearity assumptions, and covarying meteorological variables, suggesting a set of practical recommendations for diagnostic schemes.
[1] Hochberg et al. 2017, Feng et al. 2019, Ratzmann et al. 2019, Novick et al. 2019