Differential use of shallow and deep soil moisture in a semiarid shrubland: Linking sap flow and stable isotope techniques to quantify temporal variability

Abstract:

Semiarid shrublands and other dryland ecosystems are highly responsive to precipitation pulses that, depending on their size, differentially influence moisture distribution in the soil profile. Recent dryland ecosystem field studies have shown that transpiration dynamics are largely a function of deep soil moisture available after large precipitation events, regardless of where the majority of plant roots occur. This is contrary to some ecological frameworks, such as Walter’s two-layer hypothesis, which suggest the physical location of plant roots result in plant-specific niches of water availability. We suggest that adopting a hydrologically defined two-layer conceptual framework of the soil profile is more appropriate for understanding dryland plant water use than a rooting-depth-based framework. We make the assumption that shallow and deep soil layers have different isotopic signatures and use this framework to show how dryland transpiration dynamics vary with deep soil moisture availability and how the source of that moisture varies over time. We present continuous eddy covariance, sap flow transpiration and soil moisture data with discrete isotopic samples of precipitation, soil, and stems taken over 18 months at a creosotebush-dominated shrubland ecosystem in southern Arizona. We found that the shallow and deep soils were isotopically distinct; in particular, on each sampling day, shallow soil δ²H values tended to be more positive than deep soil δ²H values, except on days where a large, isotopically depleted storm wetted the whole soil profile. Our data indicate that transpiration was generally more strongly correlated with deep moisture, and that stem samples were isotopically similar to deep soil samples. Using a combined approach to understanding plant water use in semiarid areas offers more insights than using sap flow or isotopic techniques alone; this is in part because of the complexity in rainfall isotopic patterns and the conditions for fractionation. Using our combined approach, we contend that semiarid shrubs depend on deep moisture for growth and functioning and are therefore vulnerable to shifts in precipitation, such as a decrease in the number of large storms, which would limit available soil moisture to the shallow surface layer.