Identification of the best spectral indices to remotely trace the diurnal course of water use efficiency of Tamarix ramosissima in the Gurbantunggut Desert, China

Abstract

Water availability is one of the most important factors limiting photosynthetic assimilation of carbon dioxide and growth of individual plants in terrestrial ecosystems. Water use efficiency (WUE) of plants has been widely assessed using ecological methods in field measurements; however, approaches for remotely sensing WUE are still lacking, particularly in arid ecosystems. In this study, a comprehensive analysis of diurnal WUE via spectral indices in arid ecosystems was assessed. Analyses were conducted on a native dominant desert shrub, Tamarix ramosissima, in its original habitat on the southern periphery of the Gurbantunggut Desert, China. Based on diurnal measurements of spectral reflectance, photosynthesis, and micrometeorological variables, simple and useful spectral indices for estimating diurnal WUE at the assimilative organ scale were explored. From six types of spectral indices, ranging from simple to sophisticated, the best wavelength domains for a given type of index were determined by screening all combinations using correlation analysis. The coefficient of determination (R 2 ), ranging from 0.19 to 0.60, for WUE was calculated for all indices derived from spectra taken from the assimilative organs.With only two wavelengths and a significant correlation coefficient (R 2 = 0.60, P \ 0.001), the simple ratio (SR) type index was the most sensitive to WUE among all of the indices. Furthermore, SR is a useful indicator to determine the dynamic and diurnal processes of photosynthesis and transpiration of T. ramosissima. Although it has a few weaknesses, SR serves as a simple and robust indicator of WUE in arid ecosystems.

Machine Learning Techniques for Predicting Crop Photosynthetic Capacity from Leaf Reflectance Spectra

“…Reflectance spectra from crop canopies facilitate the assessment of the plants' composition, nutritional status, as well as certain stress parameters. This includes estimates of nutrient composition (Gillon et al, 1999;Zhai et al, 2013), water content and water use efficiency (Wang et al, 2012a), chlorophyll content (Atzberger et al, 2010), cell wall composition (Penning et al, 2009), content of secondary metabolites (Couture et al, 2013;Jia et al, 2013), heavy metal content (Liu et al, 2010a), disease expansion (Xu et al, 2006;Mahlein et al, 2012), and species composition (Borregaard et al, 2000;Manevski et al, 2011). Reflectance-based evaluation of the nitrogen status of crop plants has been particularly successful; data for specific wavelengths (Alchanatis et al, 2005) or indices can be correlated to the nitrogen status of the plants (Zhao et al, 2005;Wang et al, 2012b;Bao et al, 2013;Ecarnot et al, 2013) over a wide range of species.…”

Remote sensing of assimilating branch light use efficiency using the photochemical reflectance index inHaloxylon ammodendronforest

“…Remote sensing has been useful for predicting changes in plant structure and function within various ecosystems through application of spectrally derived efficiencies to monitor water use efficiency, 8,10,11 photosynthetic nitrogen use efficiency, 12 light use efficiency (LUE), 13 and photosynthetic efficiency 14 at leaf, canopy, and ecosystem scales. Although the photochemical reflectance index (PRI) has been widely applied in the estimation of LUE in most species, 2,14,15 none of the studies had considered its applicability in Haloxylon ammodendron, especially under varying light within canopies.…”

A Review on the Relevant Applications of Machine Learning in Agriculture

“…Reflectance spectra from crop canopies helped in the assessment of the plants' composition, nutritional status, as well as certain stress parameters. This includes estimates of nutrient components and macronutrient concentration (Gillon et al, 1999 [59]; Zhai et al, 2013[60]), water content and water use efficiency(Wang et al, 2012a(Wang et al, , 2012b [61,62], chlorophyll content(Atzberger et al, 2010)[63], cell wall composition(Penning et al, 2009) [64], content of secondary metabolites(Couture et al, 2013 [65];Jia et al, 2013 [66]), heavy metal content(Liu et al, 2010) [67], disease expansion(Xu et al, 2006 [68];Mahlein et al, 2012 [69), and species structure(Borregaard et al, 2000 [70];Manevski et al, 2011[71]) Heckmann et al (2017).…”