How similar is the zoning of different vegetation indices: Defining the optimal framework for monitoring grapevines’ growth within vigorous vineyards
The spatial patterns of grapevines’ vigor across vigorous vineyards were assessed through distinct vegetation indices (VI) for comparison purposes. This survey also aids the selection of the most sensitive VI to monitor the vineyards during a given optimal stage of the last period of the growing sea...
| Autores: | , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2023 |
| País: | Brasil |
| Institución: | Universidade Estadual Paulista (UNESP) |
| Repositorio: | Repositório Institucional da UNESP |
| Idioma: | inglés |
| OAI Identifier: | oai:repositorio.unesp.br:11449/299023 |
| Acceso en línea: | http://dx.doi.org/10.1016/j.scienta.2023.112404 https://hdl.handle.net/11449/299023 |
| Access Level: | acceso abierto |
| Palabra clave: | Grapevine vigor Map comparison Precision viticulture Vitis vinifera L. |
| Sumario: | The spatial patterns of grapevines’ vigor across vigorous vineyards were assessed through distinct vegetation indices (VI) for comparison purposes. This survey also aids the selection of the most sensitive VI to monitor the vineyards during a given optimal stage of the last period of the growing season. The canopy reflectance (ρ) of grapevines cv. Cabernet Franc and Cabernet Sauvignon (Vitis vinifera L.) was measured throughout their berry development and ripening stages. Georeferenced data of ρ at the red (680 nm), red edge (730 nm), and near-infrared (780 nm) wavebands were taken using a hand-held active sensor to compute the normalized difference red edge index (NDRE), and the normalized difference vegetation index (NDVI). Spatial predictions of both VI were estimated by geostatistical interpolation technique (kriging) and then categorized into homogeneous zones (HZ) through the Jenks natural breaks optimization method to delimit low and high vigor regions. Statistics of the inter-rater reliability between each VI zoning were calculated afterward. The relative dispersion of data around their mean was assumed as a criterion to assess the sensibility of each VI in detecting vigor variability under a given condition. The most appropriate stage to monitor the vineyards was determined based on the peak of vegetative growth inferred from the values of the VI we presumed as most sensitive. Differences between the HZ defined from the best combination of VI and stage in terms of yield, cane weight, and crop load were also examined. We demonstrated that NDRE was more sensitive than NDVI in detecting the vigor variability of high-density vegetation due to the relative dispersion of its datasets and the saturation effect of NDVI at the late stages of the grapevine growing season. The lowest and higher agreement between the VI zoning was usually detected at the berry touch and berry softening growth stages, respectively. Although there were distinct levels of concordance between NDRE and NDVI maps, their overall agreement was higher than 70% in most surveys. The disagreement in quantity or spatial allocation of vigor categories provided by each VI relied on the vineyard, growth stage, and growing season. Nevertheless, the early definition of HZ from NDRE monitoring carried out at veraison could be suitable to identify differences in cane weight at pruning, with a moderate effect size. |
|---|