Use of proximal sensing and vegetation indexes to detect the inefficient spatial allocation of drip irrigation in a spot area of tomato field crop

Abstract

Hyperspectral vegetation indexes (VIs) were used to detect stressed crop areas in drip irrigated tomato subjected to waterlogging. The crop was quite uniform throughout the field until the beginning of flowering, as confirmed by spectroradiometric readings and agronomic traits. From 78 days after transplanting (DAT) (42 days before harvest), a spot area of 500 m2 showed increasing excess soil moisture due to topsoil depression, which induced evident waterlogging. Leaves first yellowed (90 DAT) and eventually plants died (100 DAT). The plants surrounding this spot area were affected in their physiological, spectroradiometric and productive responses. Regressions among spectral VIs and crop yield, and photosynthesis (A) and stomatal conductance (gs) were highly significant. The best relationships were found with Soil-Adjusted Vegetation Index, Optimized Soil-Adjusted Vegetation Index, Transformed Soil-Adjusted Vegetation Index, Structure Intensive Pigment Index and Normalized Difference Vegetation Index. Maps of photosynthesis and VIs were roughly similar to the spatial distribution of crop yield. Spectroradiometry was proved efficient as early warning tool for detecting over-irrigation at the field scale. Proximal sensing techniques may contribute to improve (i) irrigation efficiency, with positive effects on tomato crop productivity and water saving, and (ii) the accuracy of remote sensing surveys aimed at estimating tomato crop yield.