A long-standing ambition in ecosystem science has been to understand the relationship between ecosystem community composition, structure and function. Differential water use and hydraulic redistribution have been proposed as one mechanism that might allow for the coexistence of overstory woody plants and understory grasses. Here, we investigated how patterns of hydraulic redistribution influence overstory and understory ecophysiological function and how patterns vary across timescales of an individual precipitation event to an entire growing season. To this end, we linked measures of sap flux within lateral and tap roots, leaf-level photosynthesis, ecosystem-level carbon exchange and soil carbon dioxide efflux with local meteorology data. The hydraulic redistribution regime was characterized predominantly by hydraulic descent relative to hydraulic lift. We found only a competitive interaction between the overstory and understory, regardless of temporal time scale. Overstory trees used nearly all water lifted by the taproot to meet their own transpirational needs. Our work suggests that alleviating water stress is not the reason we find grasses growing in the understory of woody plants; rather, other stresses, such as excessive light and temperature, are being ameliorated. As such, both the two-layer model and stress gradient hypothesis need to be refined to account for this coexistence in drylands.

A major challenge in critical zone science is to understand and predict the interaction between above‐ground and below‐ground ecohydrologic processes. One process that facilitates this connection is hydraulic redistribution, a phenomenon by which roots serve as preferential pathways for water movement from wet to dry soil layers. We use a multilayer canopy model in conjunction with experimental data to quantify the influence of hydraulic redistribution on ecohydrologic processes in order to characterize the competitive and facilitative interaction between mesquite trees and bunchgrasses in a semiarid savanna. Both measured and simulated results show that hydraulic descent dominates during the wet monsoon season, whereas hydraulic lift occurs between precipitation events. For 2015 year‐long simulation, we find about 17% of precipitation is absorbed as soil moisture, with the rest of the precipitation returning to the atmosphere as evapotranspiration. In the wet season, 13% of precipitation is transferred to deep soil (>1.5 m) through roots, and in the dry season, 9% of this redistributed water is then transported back to shallow soil depths (<0.5 m). Assuming water supplied through hydraulic redistribution is well‐mixed with moisture transported directly through the soil matrix and supports vegetation evapotranspiration, hydraulic redistribution supports 47% of mesquite transpiration and 9% of understory transpiration. Through modeling and experimental synthesis, this study demonstrates that in semiarid savanna ecosystems, mesquite exhibits a competitive advantage over understory bunchgrass through hydraulic redistribution. This analysis evaluates the relationship between two coexisting vegetation types that could be expanded to multiple vegetation species sharing resources in an ecosystem.

No abstract

Anticipating the ability of ecosystems to maintain functional integrity across predicted altered precipitation regimes remains a grand ecohydrological challenge. Overstory trees and understory grasses within semiarid savannas vary in their structure and sensitivity to environmental pressures, underscoring the need to examine the ecohydrological implications of this climatic variability. Whereas precipitation has long been recognized as a key driver of landscape ecohydrology, understanding a site's hydraulic redistribution regime (the balance in downward and upward movement of water and the seasonality of these bidirectional flows) may be equally important to understanding moisture availability to vegetation in these dryland ecosystems. As a result, we linked measures of ecosystem‐scale carbon exchange, overstory tree sap flux and leaf‐level gas exchange to understory whole‐plot and leaf‐level carbon and water exchange within intact and trenched plots (isolating trees from grasses) in a riparian savanna ecosystem. We maintained measurements across 2 years with distinct precipitation regimes. We found that interannual precipitation variability yielded a categorical shift in the directionality and magnitude of the hydraulic redistribution regime—even within this single site. Additionally, we found that connectivity between overstory trees and understory grasses through hydraulic redistribution created a short period of competition within an average rain year but that facilitation of understory function by overstory trees was much greater and lasted longer during drier years. Together, these findings suggest that hydraulic redistribution can serve as a hydrologic buffer against interannual precipitation variability. Given current climate projections of more variable precipitation within and across years, understanding how hydraulic redistribution regimes vary through time will greatly enhance our capacity to anticipate future ecohydrological function.

In plant communities, there have been efforts to improve understanding of the species coexistence in terms of ecological paradigms (Rodriquez-Robles et al., 2020). The major interests in ecohydrology have been linked to how different vegetation types coexist, how woody plant encroachment alters dynamics of water and carbon dioxide exchanges, and how coexisting vegetation species share or compete for the same resources (Barron-Gafford et al.