This study summarizes the spatiotemporal spread of citrus Huanglongbing (HLB, or greening) in Minas Gerais state since its first detection in 2005 using data gathered from the state agency database of official reports as part of the eradication program. In total, 118 municipalities and 1487 orchards have been inspected up to 2018. The overall prevalence of HLB-affected orchards was 57.2%, detected in 64.4% of inspected municipalities. A total of 459,254 plants, mainly mandarins (287,978) and sweet oranges (163,823), have been eradicated due to the HLB. Prevalence varied varied between sweet orange in the West (20.0%) and in the south (64.8%) and mandarins in the South (80.0%), and in the central MG (40.4%). The cumulative incidence (% eradicated plants) were generally higher in mandarins than sweet oranges, which accounted for 96.5% of the orchards, besides lemon (0.47%) and acid lime (3.03%). From the intial focus in the south, HLB spread to the west in 2010 and to central MG in 2015. Thereafter, HLB has spread rapidly due to a lack of HLB-oriented management and proximity to small orchards. HLB incidence across orchards is still lower on average (0.47%) in the central than south (5.6%). Spatial analysis confirmed stronger aggregation the spatial dependence of HLB-affected mandarins orchards within 25 km of distance, while randomness and spatial dependence at much shorter distances was predominant for sweet oranges. Our study provides new data and insights into the large scale HLB epidemiology and information useful for decision-makers on how to best deploy resources for further monitoring and assessing the risk in regions where HLB is still absent.

Soybean rust (SBR) in Brazil is controlled with fungicides, which have shown variable, eventually declining, efficacy. A Monte Carlo simulation framework was proposed to approximate profitability depending on the fungicide program's efficacy and total cost. Probability distributions were fitted to slopes and intercepts of the disease-yield relationship and severity in the untreated plots reported in the literature, as well as historical records of soybean price. Simulations of disease reduction conditioned to predefined control efficacy and total application costs were split into scenarios that combined two categories of severity (high and low) and two attainable yield classes (high and low). These categories were defined based on the median of severity (57.8%) and median of the intercept (yield when severity is zero, 2995.1 kg/ha). Probability matrices were constructed relating fungicide efficacy and costs. A higher frequency of break-even events occurred in scenarios of high disease pressure and higher yield. Yearly simulations, starting with 79.4% efficacy, assuming two rates of decline determined for tebuconazole (high decline), showed that the program may remain profitable during the first 5 to 7 years of use. Contrasting to cyproconazole, a fungicide that would be profitable during the entire decade. These simulations can be useful to aid in decision-making when planning fungicide programs. This approach can be adapted to other diseases of soybean and other crops as long as damage functions are available. An interactive web app was developed to perform the simulations accessible at alvesks.shinyapps.io/rusty-profits/.

Soybean rust (SBR; Phakopsora pachyrhizi) is one of the most damaging fungal diseases of soybean, particularly in Brazil where farmers rely on fungicide sprays. Disease risk is variable among regions, as well as between years; the main drivers include the availability of inoculum early in the season and both pre- and within-season weather conditions. El Niño-southern oscillation (ENSO) is known to affect global climate. In southern Brazil, precipitation levels are above normal during spring and summer in the El Niño phase. Here, we describe the temporal and spatial features of SBR spread across municipalities in Southern Brazil and evaluate the effects of ENSO events and anomalies on the sea surface temperature (SST) in the El Niño 3.4 region in the central Pacific on those features. We used data provided by the 'Consórcio Antiferrugem' on disease reports during 17 crop seasons (2004/05 to 2020/21). The temporal and spatial aspects of SBR epidemics were variable among the years and significantly affected by El Niño events. A spatially structured Cox proportional Bayesian model suggested that the anomalies in the normal SST during May-June-July and October-November-December trimesters increase the risk of early SBR onset in municipalities. In general, epidemics are expected to establish and spread early and faster across the entire region of southern Brazil during the El Niño seasons. Results of this study provide insights into the large-scale spread of soybean rust across southern Brazil, which may be useful for seasonal outlook and decision-making when planning disease management.

Huanglongbing (HLB) is one of the most important diseases for the citriculture in the world. Knowledge of climatic factors linked to HLB risk at the large spatial scale is limited. We gathered HLB presence/absence data from official surveys conducted in the state of Minas Gerais, Brazil, for 13 years. The total count of orange and mandarin orchards, and mean orchard area, normalized to a spatial grid of 60 cells (55 x 55 km), were derived from the same database. The monthly climate normal (1984 to 2013) on rainfall, mean temperature, and wind speed were split into rainy (September to April) and dry (May to August) seasons (annual summary was retained) were also obtained for each grid cell. Two hierarchical Bayesian modeling approaches were evaluated both based on the integrated nested Laplace approximation methodology. The first, the climate covariates model (CC model), used orchard, climate, and the spatial effect as covariates. The second, principal components (PC model), used the first three components from a PCA of all variables and the spatial effect as covariates. Both models showed an inverse relationship between posterior prevalence and mean temperature during dry season across the grid cells. Annual wind speed, as well as annual and rainy season rainfall, contributed significantly towards HLB risk, in the CC and PC models, respectively. A partial influence of neighboring regions on HLB risk was observed. These results should assist policymakers in defining regions at HLB risk and monitoring strategies to avoid further spread in the target region.

Wheat blast, caused by Pyricularia oryzae Triticum (PoT) lineage, is a major constraint to wheat production, mainly in the tropics of Brazil where severe epidemics are more frequent. We analyzed disease and wheat yield data from 42 uniform field trials conducted during nine years (2012 to 2020) in order to assess whether the percent control and yield response were influenced by fungicide type, region (tropical or subtropical), and year. Six treatments were selected, all evaluated in at least 19 trials. Two fungicides were applied as solo active ingredients: MANCozeb, and TEBUconazole, and four were premixes: AZOXistrobin + TEBU, TriFLoXistrobin + PROThioconazole, TFLX + TEBU, and PYRAclostrobin + EPOXiconazole. Percent control, calculated from back-transforming estimates by a meta-analysis network model fitted to the log of the means, ranged from 43% to 58%, with all but PYRA + EPOX showing efficacy greater than 52% on average, not differing among them. The variation in both efficacy and yield response were explained by region and all but TEBU performed better in the subtropics than in the tropics. Yield response from using three sequential sprays was around two times greater in the subtropics (319 to 532 kg/ha) than in the tropics (149 to 241.3 kg/ha). No significant decline in fungicide efficacy or yield response were observed in nine years of study for any of the fungicides. Our results reinforce the need to improve control by adopting an integrated management approach in the tropics given the poorer performance and lower profitability, especially for the premixes, than in the subtropics.