Over the last 20 years, begomoviruses have emerged as devastating pathogens, limiting the production of different crops worldwide. Weather conditions increase vector populations, with negative effects on crop production. In this work we evaluate the relationship between the incidence of begomovirus and weather before and during the crop cycle. Soybean and bean fields from north-western (NW) Argentina were monitored between 2001 and 2018 and classified as moderate (≤50%) or severe (>50%) according to the begomovirus incidence. Bean golden mosaic virus (BGMV) and soybean blistering mosaic virus (SbBMV) were the predominant begomovirus in bean and soybean crops, respectively. Nearly 200 bio-meteorological variables were constructed by summarizing climatic variables in 10-day periods from July to November of each crop year. The studied variables included temperature, precipitation, relative humidity, wind (speed and direction), pressure, cloudiness, and visibility. For bean, high maximum winter temperatures, low spring humidity, and precipitation 10 days before planting correlated with severe incidence. In soybeans, high temperatures in late winter and in the pre-sowing period, and low spring precipitations were found to be good predictors of high incidence of begomovirus. The results suggest that temperature and pre-sowing precipitations can be used to predict the incidence status [predictive accuracy: 80% (bean) and 75% (soybean)]. Thus, these variables can be incorporated in early warning systems for crop management decision-making to reduce the virus impact on bean and soybean crops.

Nowadays it is possible to easily access large volumes of georeferenced climatic data. These data can be used to model the relationship between climatic conditions and disease from multiple meteorological variables, usually correlated and redundant. The selection of variables allows the identification of a subset of relevant regressors to build predictive models. Stepwise, Boruta, and LASSO are variable selection procedures of different nature, so their relative performance has been scarcely explored. The objective of this work was the comparison of these methods simultaneously applied in the construction of regression models to predict disease risk from climatic data. Three georeferenced databases were used with presence/absence values of different pathogens in maize crops in Argentina. For each scenario, climatic variables from the period prior to sowing until harvest were obtained. The three variable selection methods obtained models with accuracy close to 70 %. However, LASSO produced the best predictive model, selecting an intermediate number of variables with respect to Stepwise (lower number) and Boruta (higher number). The results could be extended to other pathosystems and inspire the construction of alarm systems based on climatic variables. Hoy es posible acceder fácilmente a cuantiosos volúmenes de datos climáticos georreferenciados. Estos pueden ser usados para modelar la relación entre condiciones climáticas y enfermedad, para lo cual es necesario usar múltiples variables meteorológicas, usualmente correlacionadas y redundantes. La selección de variables permite identificar un subconjunto de regresoras relevantes para construir modelos predictivos. Stepwise, Boruta y LASSO son procedimientos de selección de variables de distinta naturaleza por lo que su desempeño relativo ha sido poco explorado. El objetivo de este trabajo fue la comparación de estos métodos aplicados simultáneamente en la construcción de modelos de regresión para predecir riesgo de enfermedad desde datos climáticos. Se utilizaron tres bases de datos georreferenciados con valores de presencia/ausencia de distintos patógenos en cultivos de maíz en Argentina. Para cada escenario se obtuvieron variables climáticas del periodo previo a la siembra hasta la cosecha. Con los tres métodos se generaron modelos predictivos con precisión de clasificación cercana al 70 %. LASSO produjo mejor predicción, seleccionando una cantidad intermedia de variables respecto a Stepwise (menor cantidad) y a Boruta (mayor). Los resultados podrían extenderse a otros patosistemas y contribuir a la construcción de sistemas de alarma basados en variables climáticas.

Over the last 20 years, begomoviruses have emerged as devastating pathogens, limiting the production of different crops worldwide. Weather conditions increase vector populations, with negative effects on crop production. In this work we evaluated the relationship between the in-cidence of begomovirus and climatic conditions before and during the crop cycle. Soybean and bean fields from the northwest (NW) of Argentina were monitored for 14 years and classified as moderate (≤50%) and severe (> 50%) according to the relative incidence. Two hundred bio-meteorological variables were constructed, summarizing meteorological data in 10-day peri-ods from June to March of each crop year. The studied variables included temperature, precipi-tation, relative humidity, wind (speed and direction), pressure, cloudiness and visibility. For bean, high maximum winter temperatures, low spring humidity and precipitation 10 days before planting correlated with severe incidence. In soybeans, high late winter and pre-planting tem-peratures, and low spring precipitations were found to be good predictors of high incidence of begomovirus presence. The results suggest that temperature and pre-sowing precipitations can be used to predict incidence status [predictive accuracy: 82% (bean) and 75% (soybean)]. Thus, these variables can be incorporated in early warning systems for crop management deci-sion-making to reduce the virus impact on bean and soybean crops.