Evaluation of water-use efficiency in foxtail millet (Setaria italica) using visible-near infrared and thermal spectral sensing techniques

摘要

Water limitations decrease stomatal conductance (g(s)) and, in turn, photosynthetic rate (A(net)), resulting in decreased crop productivity. The current techniques for evaluating these physiological responses are limited to leaf-level measures acquired by measuring leaf-level gas exchange. In this regard, proximal sensing techniques can be a useful tool in studying plant biology as they can be used to acquire plant level measures in a high-throughput manner. However, to confidently utilize the proximal sensing technique for high-throughput physiological monitoring, it is important to assess the relationship between plant physiological parameters and the sensor data. Therefore, in this study, the application of rapid sensing techniques based on thermal imaging and visual-near infrared spectroscopy for assessing water-use efficiency (WUE) in foxtail millet (Setaria italica (L.) P. Beauv) was evaluated. The visible-near infrared spectral reflectance (350-2500 nm) and thermal (7.5-14 mu m) data were collected at regular intervals from well-watered and drought-stressed plants in combination with other leaf physiological parameters (transpiration rate-E, A(net), g(s), leaf carbon isotopic signature-delta C-13(leaf) WUE). Partial least squares regression (PLSR) analysis was used to predict leaf physiological measures based on the spectral data. The PLSR modeling on the hyperspectral data yielded accurate and precise estimates of leaf E, g(s), delta C-13(leaf), and WUE with coefficient of determination in a range of 0.85-0.91. Additionally, significant differences in average leaf temperatures (similar to 1 degrees C) measured with a thermal camera were observed between well-watered plants and drought-stressed plants. In summary, the visible-near infrared reflectance data, and thermal images can be used as a potential rapid technique for evaluating plant physiological responses such as WUE. (C) 2016 Elsevier B.V. All rights reserved.