European Journal of Sustainable Agroecosystems

Background: The increasing scarcity of water is limiting agricultural productivity in irrigated and dryland agroecosystems, which has made deficit irrigation (DI) one of the most widely used techniques to optimize limited water supplies and stable crop production. Since the rate of stomatal conductance (gs) determines the simultaneous fluxes of water loss and carbon assimilation, sensor-based methods that identify soil, plant, and canopy signals have been suggested for DI threshold calibration. 
Purpose: The purpose of this review is to summarize peer-reviewed studies published primarily between the years of 2015 and 2025 in relation to the impact of sensor-based DI techniques on gs and productivity, and to conduct a comparative analysis of soil-, plant-, and remote/canopy-based approaches to sensing. 
Research procedure: A structured search was conducted in Scopus, Web of Science, and PubMed, as well as grey literature sources (FAO/AQUASTAT and other reports) utilizing combinations of the following terms: 'deficit irrigation', 'stomatal conductance', 'sensor', 'crop water stress index' и 'precision irrigation'. Out of the completed search, sixty-three articles matched the inclusion criteria for the thematic analysis, and forty-two of them played a key role in comparative analysis.
Summary of conclusions: Soil moisture sensors can help with inexpensive scheduling of irrigation, but they do not indicate plant stress levels; thermal sensors of canopy temperature correlate well with gs but are affected by temperature shifts of other environmental factors; methods that are directly aimed at measuring plants’ water consumption, including porometers and dendrometers, are associated with the highest degree of reliability but are difficult to use due to the high costs; moreover, PRD, or partial root-zone drying, uses roots for delivering stress signals for the purpose of limiting gs and at the same time supporting high productivity of plants including greening potential. In the studied cases, it was indicated that the levels of stress that are equal to 60-80% of evapotranspiration were sufficient for achieving good yield of the crops and at the same time allowing water consumption decrease by 15-35%. 
Research gaps: Some of the existing gaps are the lack of calibration standards that cross-genotypes and crops, the low level of validation of multisensory data fusion algorithms throughout the seasons, the low integration of AI decision-making tools with hardware easily accessible to smallholder farmers, and the limited socio-economic assessment of barriers to adoption.
Conclusion: While deficit irrigation based on sensors provides a scientifically based approach to saving water in agriculture, its successful implementation requires the establishment of national standards, the development of accessible multimodal sensing systems, and long-term testing of results in several agricultural seasons in a variety of contexts.