Application of a spatial risk model of the crystalline spider mite (Oligonychus sp.) to avocado crop damage using remote sensing

The avocado is one of the most consumed foods in the world and it is affected by the mite Oligonychus sp., which affects the generation of chlorophyll by the plant, resulting in a decrease in productivity. Given the economic importance of the avocado, a spatial statistical methodology was used to an...

Full description

Saved in:
Bibliographic Details
Published inRevista Facultad Nacional de Agronomía, Medellín Vol. 76; no. 2; pp. 10309 - 10321
Main Authors Báñez Aldave, Harry Wilson, Cuesta Herrera, Ledyz, López Hernández, Juan Ygnacio, Andrades Grassi, Jesús Enrique, Torres Mantilla, Hugo Alexander
Format Journal Article
LanguageEnglish
Published Bogota Universidad Nacional de Colombia 01.05.2023
Subjects
Chlorophyll
Crop damage
Crop diseases
Data analysis
Economic importance
Food plants
Geomorphology
Geostatistics
Indicators
Information systems
Leaves
Mites
Pests
Predictions
Remote sensing
Simulation
Spatial discrimination
Spatial resolution
Stationary processes
Statistical methods
Peru
Online AccessGet full text

Cover

More Information
Summary:The avocado is one of the most consumed foods in the world and it is affected by the mite Oligonychus sp., which affects the generation of chlorophyll by the plant, resulting in a decrease in productivity. Given the economic importance of the avocado, a spatial statistical methodology was used to analyze the risk of a pest in its crops. A total of 202 observations of a 1.1 ha avocado farm were used to measure the number of mites per leaf in the area of Barranca, Perú. Predictive geostatistical methods and indicators were applied. A Spherical semivariogram was adjusted to estimate a Univariate Ordinary Kriging, covariates such as vegetation indicators and geomorphometric variables were used to improve the spatial resolution of the covariates and geostatistical simulation was used and linear co-regionalization models were adjusted with which pest predictions were made with co-Kriging. Finally, the predictions were transformed into a risk model using Kriging Indicator. The results obtained show that the mite presents a stationary process in second order with spatial dependence of less than 10 m, in which univariante Ordinary Kriging was the most efficient. Despite the results, the linear co-regionalization models are consistent, but the geostatistical simulation was not enough to improve the predictions. Covariate data should be incorporated at a higher level of detail and small-scale variations should be analyzed. It is suggested to incorporate covariate data with a higher level of detail and analyze small-scale variations.
ISSN:0304-2847
2248-7026
DOI:10.15446/rfnam.v76n2.102479