Prediction of plant diseases through modelling and monitoring airborne pathogen dispersal.

Abstract

Many plant diseases that spread by airborne inocula have had major economic and social impacts worldwide. Plant diseases account for 16% of the yield losses in eight of the most important food and cash crops. Numerical models and monitoring networks have been developed to forecast the spread of these diseases both locally and over long distances. The epidemics of these airborne diseases depend on production of infectious propagules, their aerial transport, specific infectiousness and finally their reproduction. This article first reviews current understanding of these processes with an emphasis on their treatments in disease forecast models as well as the uncertainties the treatments introduce. Then we discuss the concepts and frameworks of forecast models that are broadly classified into epidemic models and aerobiological models and, finally, we present an example of the application of these modelling approaches to soybean rust forecasting.

Modification of the TOMCAST Model with Aerobiological Data for Management of Potato Early Blight

“…Tools to forecast the risk of the disease allow to implement treatments and to control it effectively. Many numerical models and monitoring networks have been developed to forecast the spread of some diseases locally and over long distances [24,43,44]. Frequently, these models use weather variables and plant phenology but do not consider the presence of pathogen in field, being this the third support for the disease.…”

“…Prediction models based on meteorological variables are an opportunity for the environmentally friendly application of chemical products. However, the precise modelling of plant disease is particularly difficult because it requires specialist staff to identify critical biophysical processes driving disease spread based on time and location [24].…”

“…Epidemiological models using mathematical descriptions of the interaction between the environment, host and pathogen to forecast when the disease will occur and to improve the use of control measures were developed [24]. Many models are based exclusively on meteorological factors to calculate accumulated risk units when the conditions are favorable for the infection.…”

A system approach to the modeling of fungal infections of the wheat leaf

“…В опубликованных моделях развития грибной инфекции в посевах число рассматриваемых стадий развития патогена и/или инфекции варьирует от 1 до 8 с медианным значением 3.5 (De Wolf, Isard, 2007). Среди них моделируются распространение спор патогена (McCartney et al, 2006;Pan et al, 2010), попадание их на поверхность листа и прикрепление, прорастание споры и проникновение через устьица в субэпидермальное про странство листа, развитие гриба до формирования но вого поколения спор (Lew, 2011;Balmant et al, 2015;SugaiGuérios et al, 2016). В других моделях некоторые биологические стадии «сворачиваются» и рассматриваются процессы инфицирования, латентного периода и по явление пустул (раскрытых мешочков с новыми спора ми) (Audsley et al, 2005).…”

“…Масштабы моделируемых процессов могут варьировать в широких пределах. Так, в некоторых моделях процессы формирования микроклимата и воздушных потоков в посевах и над ними, важные для описания роста, развития и распространения патогенных организмов, имеют пространственную детализацию в один метр при размере моделируемого посева до нескольких километров (McCartney et al, 2006;Pan et al, 2010). В случае же описания развития патогенеза на растении пространственное поведение прорастающей споровой трубки на поверхности листа моделируется с точностью до нескольких микрон (Setten et al, 2015).…”