Physiological trait networks enhance understanding of crop growth and water use in contrasting environments.

Plant function arises from a complex network of structural and physiological traits. Explicit representation of these traits, as well as their connections with other biophysical processes, is required to advance our understanding of plant-soil-climate interactions. We used the Terrestrial Regional Ecosystem Exchange Simulator (TREES) to evaluate physiological trait networks in maize. Net primary productivity (NPP) and grain yield were simulated across five contrasting climate scenarios. Simulations achieving high NPP and grain yield in high precipitation environments featured trait networks conferring high water use strategies: deep roots, high stomatal conductance at low water potential ("risky" stomatal regulation), high xylem hydraulic conductivity and high maximal leaf area index. In contrast, high NPP and grain yield was achieved in dry environments with low late-season precipitation via water conserving trait networks: deep roots, high embolism resistance and low stomatal conductance at low leaf water potential ("conservative" stomatal regulation). We suggest that our approach, which allows for the simultaneous evaluation of physiological traits, soil characteristics and their interactions (i.e., networks), has potential to improve our understanding of crop performance in different environments. In contrast, evaluating single traits in isolation of other coordinated traits does not appear to be an effective strategy for predicting plant performance.

Stream depletion in alluvial valleys using the SDF semianalytical model.

A semianalytical method commonly used for quantifying stream depletion caused by ground water pumping was reviewed for applicability in narrow alluvial aquifers. This stream depletion factor (SDF) method is based on the analytic Glover model, but uses a numerical model-derived input parameter, called the SDF, to partly account for mathematically nonideal conditions such as variable transmissivity and nearby aquifer boundaries. Using the SDF can improve and simplify depletion estimates. However, the method's approximations introduce error that increases with proximity to the impermeable aquifer boundary. This article reviews the history of the method and its assumptions. New stream depletion response curves are presented as functions of well position within bounded aquifers. A simple modification to modeled SDF values is proposed that allows the impermeable boundary to be accounted for with image wells, but without overaccounting for boundary effects that are already reflected in modeled SDFs. It is shown that SDFs for locations closer to the river than to the aquifer boundary do not reflect impermeable-boundary effects, and thus need no modification, and boundary effects in the other portion of the aquifer follow a predictable removable pattern. This method is verified by comparing response curves using modified SDFs with response curves from an extensively calibrated numerical model of a managed ground water recharge site. The modification improves SDF-based stream depletion estimates in bounded aquifers while still benefiting from the additional information contained in SDF maps and retaining their value as standardized references for water rights administration.