Disease warning systems for rational management of Asian soybean rust in Brazil

• Ano de defesa: 2018
• Autor(a) principal: Beruski, Gustavo Castilho
• 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: Glycine max; Phakopsora pachyrhizi; Chuva; Integrated disease management; Manejo integrado de pragas; Rainfall
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/11/11152/tde-25072018-163838/

The Asian soybean rust (ASR), caused by the fungus Phakopsora pachyrhizi, may promote significant damages in soybean crop. The disease is mainly controlled by sequential applications of fungicides following a calendarbased system. However, this practice disregards the weather favorability to recommend spraying to ASR control. The proposition of fungicide schemes to make the ASR control more efficient can be reached by the use of diseasewarning systems. Thus, the current study aimed to assess the performance of different disease-warning systems to determine better fungicide spraying schemes for the ASR control. The experiment was conducted in Piracicaba, SP, Ponta Grossa, PR, Campo Verde and Pedra Preta, MT, Brazil, over the 2014/2015 and 2015/2016 soybean growing seasons. The treatments were: Unsprayed check treatment; Calendar-based sprays in a 14-day interval from R1 stage (CALEND); Disease warning system based on rainfall data with less conservative threshold (PREC_1 - 80% severity cut-off); and more conservative threshold (PREC_2 - 50% severity cut-off); Disease warning system based on air temperature and leaf wetness duration with less conservative threshold (TLWD_1 - 6 lesions cm-2) and more conservative threshold (TLWD_2 - 9 lesions cm-2). The results confirmed that weather conditions in the field trials were favorable to ASR progress. Among the weather elements correlated to severity leaf wetness duration, cumulative rainfall and air temperature during leaf wetness duration influenced positively the ASR. By testing warning systems to control ASR it ones was evidenced that those based on rainfall data presented highest performances. PREC_2 showed a high performance considering all sowing dates; whereas, PREC_1 was better treatment during sowing dates between October and November. The TLWD diseasewarning systems, with both thresholds, overestimated the ASR, recommending more sprays compared to other treatments. Empirical models were efficient for estimation of LWD in Ponta Grossa, Campo Verde and Pedra Preta. High performances in estimating LWD were identified by using number of hours with relative humidity above 90% (NHRH>=90%), being these able to be use as input in the disease-warning systems (RMSE less than 2.0 h). The profitability of use rainfall based warning systems was conditioned by variations in the rainfalls regimes at the studied sites. PREC_1 and PREC_2 presented the highest relative yield gains in relation to CALEND during the period with the highest rainfalls in Piracicaba, Campo Verde and Pedra Preta. However, in Ponta Grossa, the rainfall based warning systems were not effective to control ASR.