Flexible regression model for predicting the dissemination of Candidatus Liberibacter asiaticus under variable climatic conditions
Julio Cezar Souza Vasconcelos,Silvio Aparecido Lopes,Juan Camilo Cifuentes Arenas,Maria Fátima das Graças Fernandes da Silva
DOI: https://doi.org/10.1016/j.idm.2024.09.005
2024-09-13
Abstract:Greening, or Huanglongbing (HLB), poses a severe threat to global citrus cultivation, affecting various citrus species and compromising fruit production. Primarily transmitted by psyllids during phloem feeding, the bacterium Candidatus Liberibacter induces detrimental symptoms, including leaf yellowing and reduced fruit quality. Given the limitations of conventional control strategies, the search for innovative approaches, such as resistant genotypes and early diagnostic methods, becomes essential for the sustainability of citrus cultivation. The development of predictive models, such as the one proposed in this study, is essential as it enables the estimation of the bacterium's concentration and the vulnerability of healthy plants to infection, which will be instrumental in determining the risk of HLB. This study proposes a prediction model utilizing environmental factors, including temperature, humidity, and precipitation, which play a decisive role in greening epidemiology, influencing the complex interaction among the pathogen, vector, and host plant. In the proposed modeling, it addresses non-linear relationships through cubic smoothing splines applications and tackles imbalanced categorical predictor variables, requiring the use of a random-effects regression model, incorporating a random intercept to account for variability across different groups and mitigate the risk of biased predictions. The model's ability to predict HLB incidence under varying climatic conditions provides a significant contribution to disease management, offering a strategic tool for early intervention and potentially reducing the spread of HLB. Using climatological and environmental data, the research aims to develop a predictive model, assessing the influence of these variables on the spread of Candidatus Liberibacter asiaticus, essential for effective disease management. The proposed flexible model demonstrates robust predictions for both training and test data, identifying climatological and environmental predictors influencing the dissemination of Candidatus Liberibacter asiaticus, the vascular bacterium associated with Huanglongbing (HLB) or greening.