Spectral vegetation indexes as indicators of leaf pigment content in apple (Malus domestica Borkh.)

Methods of operational remote (satellite and unmanned) agricultural monitoring are currently based on the use of spectral vegetation indices as some integral indicators of plant condition. Since the first of them (Normalized Difference Vegetation Index - NDVI) appeared in the early 1970’s, rich experience has been accumulated in their use to detect various properties of agricultural plants and agrophytocenoses as a whole. About a hundred different indices have been proposed to detect different properties, e.g., moisture, leaf structure, architecture of plants in crops, the content of various substances, including pigments regulating photosynthesis and plant productivity. In many cases, the proposed indices function reliably for specific plants or for the vegetation as a whole. For fruit crops and, in particular, for apple-tree, there are practically no such indices. In this paper, it is shown for the first time that the spectral vegetation indices proposed for the detection of pigments in agricultural plants need to be refined when they are used for similar detection of pigments in the leaves of an apple tree of a particular variety. Our goal was to analyze the relationship between the spectral vegetation indices calculated for the leaves of the Imrus apple tree ( Malus domestica Borkh.) with the leaf content of chlorophyll and carotenoids. We evaluated the applicability of several dozen vegetation indices proposed for determining the content of chlorophylls and carotenoids in the leaves of various plants to the non-contact determination of these pigments in the leaves of the Imrus apple tree. On October 19, 2021, leaves were collected at noon randomly from 2-5-year old branches of the middle part of the crown of model Imrus trees grown from 2011 at the test plot (Stupino District, Moscow Province, Russia). In total, 26 mixed leaf samples were collected for pigment content analysis. The content of chlorophylls a + b was determined in the laboratory by the Wintermans-De Mots method, carotenoids by the von Wetshtein method. For the same leaves, spectral reflectance was measured under field conditions using a SR-6500 field spectroradiometer (Spectral Evolution, USA), which operates in the 350-2500 nm range with a resolution of 1 nm. Spectral reflectivity curves were plotted in 5 replicates for the upper surface of the leaves, averaged for each leaf, and then for each of the 26 mixed groups of leaves. Based on the averaged spectral reflectance curves, the most common spectral vegetation indices were calculated, followed by an analysis of the relationship between the values of the spectral vegetation indices and the content of pigments in the leaves. It has been established that the previously proposed numerous vegetation indices cannot be used for non-contact detection of the content of chlorophyll and carotenoids in the leaves of the Imrus apple tree. There is practically no connection between the index value and pigment content. It is also not possible to group the analyzed leaves according to the content of pigments based on the construction of a dendrogram of the similarity between the spectral reflectance curves of leaves in the range of 350-2500 nm. Based on the correction of the indices that showed the most accurate dependence, new vegetation indices were proposed for non-contact detection of the content of carotenoids and chlorophyll in apple leaves, which make it possible to obtain regression models with R2 above 0.65. Before widespread use, they must be tested for leaves of apple trees of other varieties, as well as for leaves at different stages of development.