Temporal clustering detection of disease in agricultural crops

Abstract

Information about temporal dynamics of plant diseases is of paramount importance for appropriate technologies development for diseases management in production systems. The major interest when studying a temporal point pattern is to detect temporal clustering of events. There are some methods available for events cluster detection over time. The majority of these methods has been developed to detect temporal clustering inhuman diseases. The temporal patterns analysisfor plant diseases are not very well described in the literature. In this study, we aimed to propose new methods, based on both empirical distribution function and Monte Carlo simulation, for testing the null hypothesis that a temporal point pattern is purely random. These methods are compared to the time K-function for detecting temporal clustering for incidence of citrus sudden death disease in orange trees. All methodologies were found to show good performance for analyzing temporal point patterns and they led to the detection of temporal clustering of the citrus sudden death disease in an orange trees planting.