Characterizing the Response of Puccinia striiformis f. sp. tritici to Periods of Heat Stress that Are Common in Kansas and the Great Plains Region of North America.

Stripe rust of wheat, caused by Puccinia striiformis f. sp. tritici, is considered a disease of cool environments, and it has been observed that high temperatures can suppress disease development. However, recent field observations in Kansas suggest that the pathogen may be recovering from heat stress more quickly than expected. Previous research indicates that some strains of this pathogen were adapted to warm temperature regimes but did not consider how the pathogen responds to periods of heat stress that are common in the Great Plains region of North America. Therefore, the objectives of this study were to characterize the response of contemporary isolates of P. striiformis f. sp. tritici to periods of heat stress and to look for evidence of temperature adaptations within the pathogen population. These experiments evaluated nine isolates of the pathogen: eight isolates collected in Kansas between 2010 and 2021 and a historical reference isolate. Treatments compared the latent period and colonization rate of isolates given a cool temperature regime (12 to 20°C) and as they recovered from 7 days of heat stress (22 to 35°C). Results documented that contemporary isolates of the pathogen had similar latent periods and colonization rates as the historical reference under the cool temperature regime. Following exposure to 7 days of heat stress, the contemporary isolates had shorter latent periods and higher colonization rates than the historical isolate. There was also variability in how the contemporary isolates recovered from heat stress, with some isolates collected during 2019 to 2021 recovering sooner than those collected just 5 to 10 years ago.

Into the Trees: Random Forests for Predicting Fusarium Head Blight Epidemics of Wheat in the United States.

Constructing models that accurately predict Fusarium head blight (FHB) epidemics and are also amenable to large-scale deployment is a challenging task. In the United States, the emphasis has been on simple logistic regression (LR) models, which are easy to implement but may suffer from lower accuracies when compared with more complicated, harder-to-deploy (over large geographies) model frameworks such as functional or boosted regressions. This article examined the plausibility of random forests (RFs) for the binary prediction of FHB epidemics as a possible mediation between model simplicity and complexity without sacrificing accuracy. A minimalist set of predictors was also desirable rather than having the RF model use all 90 candidate variables as predictors. The input predictor set was filtered with the aid of three RF variable selection algorithms (Boruta, varSelRF, and VSURF), using resampling techniques to quantify the variability and stability of selected variable sets. Post-selection filtering produced 58 competitive RF models with no more than 14 predictors each. One variable representing temperature stability in the 20 days before anthesis was the most frequently selected predictor. This was a departure from the prominence of relative humidity-based variables previously reported in LR models for FHB. The RF models had overall superior predictive performance over the LR models and may be suitable candidates for use by the Fusarium Head Blight Prediction Center.

Early Season Environmental Indicators of Wheat Stripe Rust Epidemics in Kansas and the Central Great Plains Region of the United States.

During the past two decades, the wheat-producing areas of the Great Plains region in North America experienced frequent, severe yield losses to stripe rust (Puccinia striiformis f. sp. tritici). In general, outbreaks of rust diseases in the Southern Great Plains region often precede disease problems in the Central and Northern Great Plains. However, these generalizations provide little information, and our objective for this study was to identify weather variables, geographical areas, and time periods that influence the early stages of stripe rust epidemics in the Great Plains. Data used in this analysis consisted of monthly summaries of temperature, precipitation, and soil moisture from 10 climate districts in Texas of the United States. These environmental variables were paired with estimates of wheat yield losses to stripe rust in Kansas from 2000 to 2019, with yield loss coded as a binary variable (1 = >4% statewide yield loss). An ensemble of simple models representing weather variables, time periods, and geographical locations were hypothesized to be influential in the development of stripe rust epidemics. Model performance was verified with observations not used in model development. Results of this study indicated that soil moisture within two to three climate districts in Texas were particularly influential in regional disease development. These areas of Texas were 700 to 1,000 km away from locations in Kansas where the disease-related yield losses were observed, and they often preceded disease losses by 3 to 6 months. In the future, these models could help establish priority locations and time periods for disease scouting and inform regional estimates of disease risk.

Accuracy in the prediction of disease epidemics when ensembling simple but highly correlated models.

Ensembling combines the predictions made by individual component base models with the goal of achieving a predictive accuracy that is better than that of any one of the constituent member models. Diversity among the base models in terms of predictions is a crucial criterion in ensembling. However, there are practical instances when the available base models produce highly correlated predictions, because they may have been developed within the same research group or may have been built from the same underlying algorithm. We investigated, via a case study on Fusarium head blight (FHB) on wheat in the U.S., whether ensembles of simple yet highly correlated models for predicting the risk of FHB epidemics, all generated from logistic regression, provided any benefit to predictive performance, despite relatively low levels of base model diversity. Three ensembling methods were explored: soft voting, weighted averaging of smaller subsets of the base models, and penalized regression as a stacking algorithm. Soft voting and weighted model averages were generally better at classification than the base models, though not universally so. The performances of stacked regressions were superior to those of the other two ensembling methods we analyzed in this study. Ensembling simple yet correlated models is computationally feasible and is therefore worth pursuing for models of epidemic risk.

Occurrence of Viruses and Associated Grain Yields of Paired Symptomatic and Nonsymptomatic Tillers in Kansas Winter Wheat Fields.

Vector-borne virus diseases of wheat are recurrent in nature and pose significant threats to crop production worldwide. In the spring of 2011 and 2012, a state-wide sampling survey of multiple commercial field sites and university-managed Kansas Agricultural Experiment Station variety performance trial locations spanning all nine crop-reporting regions of the state was conducted to determine the occurrence of Barley yellow dwarf virus-PAV (BYDV-PAV), Cereal yellow dwarf virus-RPV, Wheat streak mosaic virus (WSMV), High plains virus, Soilborne wheat mosaic virus, and Wheat spindle streak mosaic virus using enzyme-linked immunosorbent assays (ELISA). As a means of directly coupling tiller infection status with tiller grain yield, multiple pairs of symptomatic and nonsymptomatic plants were selected and individual tillers were tagged for virus species and grain yield determination at the variety performance trial locations. BYDV-PAV and WSMV were the two most prevalent species across the state, often co-occurring within location. Of those BYDV-PAV- or WSMV-positive tillers, 22% and 19%, respectively, were nonsymptomatic, a finding that underscores the importance of sampling criteria to more accurately assess virus occurrence in winter wheat fields. Symptomatic tillers that tested positive for BYDV-PAV produced significantly lower grain yields compared with ELISA-negative tillers in both seasons, as did WSMV-positive tillers in 2012. Nonsymptomatic tillers that tested positive for either of the two viruses in 2011 produced significantly lower grain yields than tillers from nonsymptomatic, ELISA-negative plants, an indication that these tillers were physiologically compromised in the absence of virus-associated symptoms. Overall, the virus survey and tagged paired-tiller sampling strategy revealed effects of virus infection on grain yield of individual tillers of plants grown under field conditions and may provide a complementary approach toward future estimates of the impact of virus incidence on crop health in Kansas.

Effects of Integrating Cultivar Resistance and Fungicide Application on Fusarium Head Blight and Deoxynivalenol in Winter Wheat.

Fusarium head blight (FHB) or scab, incited by Fusarium graminearum, can cause significant economic losses in small grain production. Five field experiments were conducted from 2007 to 2009 to determine the effects on FHB and the associated mycotoxin deoxynivalenol (DON) of integrating winter wheat cultivar resistance and fungicide application. Other variables measured were yield and the percentage of Fusarium-damaged kernels (FDK). The fungicides prothioconazole + tebuconazole (formulated as Prosaro 421 SC) were applied at the rate of 0.475 liters/ha, or not applied, to three cultivars (experiments 1 to 3) or six cultivars (experiments 4 and 5) differing in their levels of resistance to FHB and DON accumulation. The effect of cultivar on FHB index was highly significant (P < 0.0001) in all five experiments. Under the highest FHB intensity and no fungicide application, the moderately resistant cultivars Harry, Heyne, Roane, and Truman had less severe FHB than the susceptible cultivars 2137, Jagalene, Overley, and Tomahawk (indices of 30 to 46% and 78 to 99%, respectively). Percent fungicide efficacy in reducing index and DON was greater in moderately resistant than in susceptible cultivars. Yield was negatively correlated with index, with FDK, and with DON, whereas index was positively correlated with FDK and with DON, and FDK and DON were positively correlated. Correlation between index and DON, index and FDK, and FDK and DON was stronger in susceptible than in moderately resistant cultivars, whereas the negative correlation between yield and FDK and yield and DON was stronger in moderately resistant than in susceptible cultivars. Overall, the strongest correlation was between index and DON (0.74 ≤ R ≤ 0.88, P ≤ 0.05). The results from this study indicate that fungicide efficacy in reducing FHB and DON was greater in moderately resistant cultivars than in susceptible ones. This shows that integrating cultivar resistance with fungicide application can be an effective strategy for management of FHB and DON in winter wheat.

Disease cycle approach to plant disease prediction.

Plant disease cycles represent pathogen biology as a series of interconnected stages of development including dormancy, reproduction, dispersal, and pathogenesis. The progression through these stages is determined by a continuous sequence of interactions among host, pathogen, and environment. The stages of the disease cycle form the basis of many plant disease prediction models. The relationship of temperature and moisture to disease development and pathogen reproduction serve as the basis for most contemporary plant disease prediction systems. Pathogen dormancy and inoculum dispersal are considered less frequently. We found extensive research efforts evaluating the performance of prediction models as part of operation disease management systems. These efforts appear to be greater than just a few decades ago, and include novel applications of Bayesian decision theory. Advances in information technology have stimulated innovations in model application. This trend must accelerate to provide the disease management strategies needed to maintain global food supplies.