Effect of root and water distribution in lysimeters and in the field on the onset of crop water stress

Summary Lysimeters have been frequently used to study crop response to the onset of water stress. To test the representativeness of lysimeter derived criteria for the onset of crop water stress, spring wheat (Triticum aestivum L.) was grown in two field plots with 1.0 m deep lysimeters in the center of each plot. One plot was well-watered while the second was subjected to a drying period with no irrigation. Crop water stress was assessed by monitoring leaf water potential ( _l ), stomatal diffusive resistance (r _s ), canopy temperature (CT), evapotranspiration (ET), and soil water content in both plots and lysimeters. The rate of change of all these measured parameters, when compared to the well-watered field control-plot revealed that the field-grown plants showed signs of water stress long before the lysimeter-grown plants. Water stress developed gradually for the field crop, but the transition from the well-watered to the stressed condition happened abruptly for the lysimeter-grown plants. Once this transition occurred, the lysimeter-grown plants were more drought stressed than the field-grown plant. Water profiles measured inside the lysimeter were different from those measured in the adjacent plots. An increase in root length density with depths below 0.6 m was observed in the lysimeters as opposed to a quasimonotonic decrease with depth in the field. The response of the lysimeter-grown plants was a result of the anomalous water content and root distribution. We conclude that threshold values of ET, _l , r _s , and CT for the onset of water stress obtained when deep-rooted crops grown in a shallow lysimeter are subjected to drought periods may not be directly applicable to field situations.