Assessing the impacts of recent-past climatic constraints on potential wheat yield and adaptation options under Mediterranean climate in southern Portugal

Wheat yield potentials under rainfed Mediterranean conditions have been long limited by late-in-season occurrence of enhanced water deficits and high temperatures, coinciding with sensitive reproductive stages. Present study aims to quantify and separate the impacts of two main abiotic stresses (drought & heat) on potentially attainable wheat yields, in a typical Mediterranean environment of southern Portugal (Alentejo) over 1986—2015. We also evaluate how possible adaptation options could mitigate potential yield losses (reduce the gap between actual and potential yield). Previously calibrated STICS soil-crop model is used for these purposes, which has been satisfactorily evaluated herein for yield simulations using additional field data before running at regional level. By coupling with high-resolution gridded soil and climate datasets, STICS simulations reliably reproduce the inter-annual variability of 30-year regional yield statistics, together with reasonable estimations of experimental potential yields. Therefore, the model is useful to explore the source of yield gap in the region. The quantified impacts, though with some uncertainties, identify the prolonged terminal drought stress as the major cause of yield gap, causing 40–70% mean potential yield losses. In contrast, a short-duration of crop heat stress (≥38 °C) during late grain-filling phase only results in small-to-moderate reductions (up to 20%). Supplemental Irrigation (SI) during reproductive stages provides good adaptive gains to recover potential yield losses by 15–30%, while the proposed early-flowering cultivar is more useful in escaping the terminal heat stress (5–15% adaptive gains) than avoiding prolonged drought stress. In addition, advancing sowing date generally favours wheat production with a robust spatial-temporal pattern. Therefore, combined options based on application of SI, using balanced early-flowering cultivar and early sowing date, may contribute to considerably reduce local yield gap, where current yields can account for 60% of potential yields (26–32% without adaptation). Regional impact assessment and adaptation modelling studies are essential to support agricultural policy development under climate change and variability. The recommended combined adaptation may also represent a promising adaptation strategy for rainfed wheat cropping system in other regions with similar Mediterranean conditions. However, the existing spatial-temporal variability of adaptation response highlights the need to address adaptation strategies at a more detailed local scale with better flexible design.