Agrometeorological modeling of citrus huanglongbing and climatic risk of its occurrence in Brazil and in the United States of America

• Ano de defesa: 2021
• Autor(a) principal: Brand, Silvane Isabel
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Disease models; HLB; Meteorological variables; Modelos de doenças; Psilídeo; Psyllid; Variáveis meteorológicas
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/11/11152/tde-18062021-110454/

Huanglongbing (HLB) is the main current citrus disease and has been promoting a significant reduction in orchards productivities. HLB is a pathosystem that has the psyllid as a transmitting agent for the bacterium, Candidatus Liberibacter asiaticus. This vector is attracted to citrus plants when new sprouts are available and under suitable climatic conditions for their development. Therefore, the determination of the best conditions for the attack of this insect and transmission of HLB is fundamental for the rational control of this disease. Therefore, the objectives of this study were: i) to develop, calibrate and validate an agrometeorological model to predict the occurrence of the psyllid based on data from captures of the insect in citrus orchards in the states of São Paulo, Brazil, and Florida, United States; ii) to develop risk maps for the occurrence of psyllids for the states of São Paulo, as well as for the American states of California, Florida and Texas; iii) to develop, calibrate and validate an agrometeorological model to estimate the proportion of plants with symptoms of HLB based on the occurrence of the vector and meteorological conditions that favor the development of the bacterium and make risk maps for that in the state of São Paulo; and iv) to evaluate the effect of the El Niño Southern Oscillation (ENSO) phases on the risk of occurrence of psyllid and HLB for the states of São Paulo and Florida. To model the occurrence of the psyllid, data from eight farms in the State of São Paulo were used, being five for model calibration and three for model validation. This model was also validated with data from psyllid captures from 53 farms located in the state of Florida. The meteorological data (maximum, minimum and average temperature, relative humidity, wind speed and rain) used in the models were obtained from the following databases: Xavier grid database and stations installed close to the Brazilian farms; grid database of PRISM and NASA/POWER systems for the American states. Once the psyllid capture estimation model was calibrated and validated, it was used to spatialize the risk of infestation of this vector in the orchards of the main producing states in Brazil and the United States. The estimation model of the proportion of plants infected by HLB was based on the number of plants eradicated in the eight farms in the state of São Paulo, with data from five of them used in calibration and three of them in validation. This model had as independent variables the number of captured psyllids, resulting from the model proposed here, and the air temperature that favors the development of the disease-causing bacterium. After calibration and validation, this model was used to spatialize the risk of occurrence of HLB for the state of São Paulo. Finally, the effect of the different phases of ENSO, with low, moderate, and high intensities, on the risk of occurrence of psyllid and HLB was evaluated. The occurrence of the psyllid was based on the estimated of citrus sprouting, which always occurs when there is a water deficit of 5 mm or more, followed by rainfall of 5 mm and temperature between 16 and 28 °C. This model presented 90% correctness for the states of São Paulo and Florida. The risk of occurrence of the psyllid was higher in the center, east and part of the north of the state of São Paulo, while in the American states there is a greater risk in the south. The risk is greatest in the state of Florida, followed by the states of São Paulo, Texas and California. Over the months, the risk is greater in the spring in the State of São Paulo, starting to increase in September and lasting until December. In the United States, the risk is greatest in the summer. The model that estimates the occurrence of HLB was based on the population of psyllid captured 18 months before the estimation and on the temperature of the month of eradication, presenting a good performance (average around 70% calibration and 66% validation). The risk of HLB occurrence was higher in the south and center of the state of São Paulo, but it varied throughout the year being bigger in the central, south and east regions of the state, being higher in autumn and part of winter. The risk of occurrence of the psyllid associated with the different phases of ENOS is greater in the years of moderate La Niña in the state of São Paulo and strong La Niña in the state of Florida, with the former having the greatest risk in spring and summer and for the latter in the summer. When the intensity of the ENOS phase is not considered, the greatest risk in São Paulo occurs in the Neutral years, while in Florida in the La Niña years. When assessing the risk associated with the occurrence of HLB, it is higher in ENSO Neutral years for the state of São Paulo, with the autumn and winter months having the greatest risk.