Discovering weather periods and crop properties favorable for coffee rust incidence from feature selection approaches

Abstract

Coffee Leaf Rust (CLR) is a disease that leads to considerable losses in the worldwide coffee industry; as those that have been reported recently in Colombia and Central America. The early detection of favorable conditions for epidemics could be used to improve decision making for the coffee grower and thus reduce the losses due to the disease. Researchers tried to predict the occurrence of the disease earlier through statistical and machine learning models from crop properties, disease indicators and weather conditions. These studies considered the impact of weather variables in a common period for all. Assuming that the dynamics of weather that most impact the development of the disease occur in the same time periods is simplistic. We propose an approach to discover the time period (window) for each weather variables and crop related features that most explain a future observed CLR incidence, in order to obtain a prediction model through machine learning. The selection of the variables more related with coffee rust incidence and rejection of the features with no significant contribution of information in machine learning tasks were approached from Feature Selection methods (Filter, Wrapper, Embedded). In this way, a CLR incidence prediction model based on the features with the greatest impact on the development of the disease was obtained. Moreover, the use of SHapley Additive exPlanations allowed us to identify the impact of features in the model prediction. The monitoring of coffee rust incidence is the most important predictor, since it provides information about current inoculum and this determines how much can the incidence grow or decrease. Temperature is a determining driver for germination and penetration phases in days 9 to 6 and 4 to 1 before the date of prediction. Additionally, the amount of rain determines whether uredospore dispersal or washing conditions occurred. The mean absolute error expected in the model is 6.94% of incidence, trained with XGBoost algorithm and the dataset reduced by Embedded method. The estimation of the disease incidence 28 days later can be used to improve decision making in control and nutrition practices.

Characterization of Coffea arabica L. parent plants and physicochemical properties of associated soils, Peru

“…The difference found at the provincial level could be associated with the shade used in each location since it has been shown that this could have antagonistic effects on coffee rust ( López-Bravo et al., 2012 ). In addition, climatic variables such as temperature, which is a determining factor for the germination and penetration phases of the fungus, precipitation, which contributes to dispersal conditions and the washing of uredospores, and relative humidity also play a role ( Lasso et al., 2020 ).…”

Agricultural Shocks, Coping Policies and Deforestation: Evidence from the Coffee Leaf Rust Epidemic in Mexico

“…1 Unlike temperatures, precipitations have an ambiguous impact on CLR. The disease seems to be favored by wetness but is washed-off by intense rainfall (Merle et al, 2020;Lasso et al, 2020). The timing of precipitations also matters, as suggested by Avelino et al (2015) who observe positive early rainfall anomalies before the CLR epidemic in Guatemala, Honduras, and Nicaragua.…”

FramePests: A Comprehensive Framework for Crop Pests Modeling and Forecasting

“…The detailed process of this modeling, as well as the analysis of the results was published in [57]. Next we will present a summary of the results, framing them in the DM phases of FramePests.…”

“…We relied on the weather 14 days before DP, similar to KM. We used the concept of time windows [29], [57] to generate consecutive subperiods of each climatic variable within the main period of 14 days before DP.…”

“…Each subset had 439 instances, and the dimension depended on window size: 14D had 13 variables (8 related to climate), 7D had 69 features (64 related to climate), 4D had 93 features (88 related to climate), and 3D had 101 features (96 related to climate). The datasets can be downloaded at [58].…”