Navigating complexity to breed disease-resistant crops.

Plant diseases are responsible for substantial crop losses each year and pose a threat to global food security and agricultural sustainability. Improving crop resistance to pathogens through breeding is an environmentally sound method for managing disease and minimizing these losses. However, it is challenging to breed varieties with resistance that is effective, stable and broad-spectrum. Recent advances in genetic and genomic technologies have contributed to a better understanding of the complexity of host-pathogen interactions and have identified some of the genes and mechanisms that underlie resistance. This new knowledge is benefiting crop improvement through better-informed breeding strategies that utilize diverse forms of resistance at different scales, from the genome of a single plant to the plant varieties deployed across a region.

Clues to an Evolutionary Mystery: The Genes for T-Toxin, Enabler of the Devastating 1970 Southern Corn Leaf Blight Epidemic, Are Present in Ancestral Species, Suggesting an Ancient Origin.

The Southern corn leaf blight (SCLB) epidemic of 1970 devastated fields of T-cytoplasm corn planted in monoculture throughout the eastern United States. The epidemic was driven by race T, a previously unseen race of Cochliobolus heterostrophus. A second fungus, Phyllosticta zeae-maydis, with the same biological specificity, appeared coincidentally. Race T produces T-toxin, while Phyllosticta zeae-maydis produces PM-toxin, both host-selective polyketide toxins necessary for supervirulence. The present abundance of genome sequences offers an opportunity to tackle the evolutionary origins of T- and PM- toxin biosynthetic genes, previously thought unique to these species. Using the C. heterostrophus genes as probes, we identified orthologs in six additional Dothideomycete and three Eurotiomycete species. In stark contrast to the genetically fragmented race T Tox1 locus that encodes these genes, all newly found Tox1-like genes in other species reside at a single collinear locus. This compact arrangement, phylogenetic analyses, comparisons of Tox1 protein tree topology to a species tree, and Tox1 gene characteristics suggest that the locus is ancient and that some species, including C. heterostrophus, gained Tox1 by horizontal gene transfer. C. heterostrophus and Phyllosticta zeae-maydis did not exchange Tox1 DNA at the time of the SCLB epidemic, but how they acquired Tox1 remains uncertain. The presence of additional genes in Tox1-like clusters of other species, although not in C. heterostrophus and Phyllosticta zeae-maydis, suggests that the metabolites produced differ from T- and PM-toxin.

Determinants of Virulence and In Vitro Development Colocalize on a Genetic Map of Setosphaeria turcica.

Generating effective and stable strategies for resistance breeding requires an understanding of the genetics of host-pathogen interactions and the implications for pathogen dynamics and evolution. Setosphaeria turcica causes northern leaf blight (NLB), an important disease of maize for which major resistance genes have been deployed. Little is known about the evolutionary dynamics of avirulence (AVR) genes in S. turcica. To test the hypothesis that there is a genetic association between avirulence and in vitro development traits, we (i) created a genetic map of S. turcica, (ii) located candidate AVRHt1 and AVRHt2 regions, and (iii) identified genetic regions associated with several in vitro development traits. A cross was generated between a race 1 and a race 23N strain, and 221 progeny were isolated. Genotyping by sequencing was used to score 2,078 single-nucleotide polymorphism markers. A genetic map spanning 1,981 centimorgans was constructed, consisting of 21 linkage groups. Genetic mapping extended prior evidence for the location and identity of the AVRHt1 gene and identified a region of interest for AVRHt2. The genetic location of AVRHt2 colocalized with loci influencing radial growth and mycelial abundance. Our data suggest a trade-off between virulence on Ht1 and Ht2 and the pathogen's vegetative growth rate. In addition, in-depth analysis of the genotypic data suggests the presence of significant duplication in the genome of S. turcica.

Computer Vision for High-Throughput Quantitative Phenotyping: A Case Study of Grapevine Downy Mildew Sporulation and Leaf Trichomes.

Quantitative phenotyping of downy mildew sporulation is frequently used in plant breeding and genetic studies, as well as in studies focused on pathogen biology such as chemical efficacy trials. In these scenarios, phenotyping a large number of genotypes or treatments can be advantageous but is often limited by time and cost. We present a novel computational pipeline dedicated to estimating the percent area of downy mildew sporulation from images of inoculated grapevine leaf discs in a manner that is time and cost efficient. The pipeline was tested on images from leaf disc assay experiments involving two F1 grapevine families, one that had glabrous leaves (Vitis rupestris B38 × 'Horizon' [RH]) and another that had leaf trichomes (Horizon × V. cinerea B9 [HC]). Correlations between computer vision and manual visual ratings reached 0.89 in the RH family and 0.43 in the HC family. Additionally, we were able to use the computer vision system prior to sporulation to measure the percent leaf trichome area. We estimate that an experienced rater scoring sporulation would spend at least 90% less time using the computer vision system compared with the manual visual method. This will allow more treatments to be phenotyped in order to better understand the genetic architecture of downy mildew resistance and of leaf trichome density. We anticipate that this computer vision system will find applications in other pathosystems or traits where responses can be imaged with sufficient contrast from the background.

Automated Identification of Northern Leaf Blight-Infected Maize Plants from Field Imagery Using Deep Learning.

Northern leaf blight (NLB) can cause severe yield loss in maize; however, scouting large areas to accurately diagnose the disease is time consuming and difficult. We demonstrate a system capable of automatically identifying NLB lesions in field-acquired images of maize plants with high reliability. This approach uses a computational pipeline of convolutional neural networks (CNNs) that addresses the challenges of limited data and the myriad irregularities that appear in images of field-grown plants. Several CNNs were trained to classify small regions of images as containing NLB lesions or not; their predictions were combined into separate heat maps, then fed into a final CNN trained to classify the entire image as containing diseased plants or not. The system achieved 96.7% accuracy on test set images not used in training. We suggest that such systems mounted on aerial- or ground-based vehicles can help in automated high-throughput plant phenotyping, precision breeding for disease resistance, and reduced pesticide use through targeted application across a variety of plant and disease categories.

Maize Introgression Library Provides Evidence for the Involvement of liguleless1 in Resistance to Northern Leaf Blight.

Plant disease resistance is largely governed by complex genetic architecture. In maize, few disease resistance loci have been characterized. Near-isogenic lines are a powerful genetic tool to dissect quantitative trait loci. We analyzed an introgression library of maize (Zea mays) near-isogenic lines, termed a nested near-isogenic line library for resistance to northern leaf blight caused by the fungal pathogen Setosphaeria turcica The population was comprised of 412 BC5F4 near-isogenic lines that originated from 18 diverse donor parents and a common recurrent parent, B73. Single nucleotide polymorphisms identified through genotyping by sequencing were used to define introgressions and for association analysis. Near-isogenic lines that conferred resistance and susceptibility to northern leaf blight were comprised of introgressions that overlapped known northern leaf blight quantitative trait loci. Genome-wide association analysis and stepwise regression further resolved five quantitative trait loci regions, and implicated several candidate genes, including Liguleless1, a key determinant of leaf architecture in cereals. Two independently-derived mutant alleles of liguleless1 inoculated with S. turcica showed enhanced susceptibility to northern leaf blight. In the maize nested association mapping population, leaf angle was positively correlated with resistance to northern leaf blight in five recombinant inbred line populations, and negatively correlated with northern leaf blight in four recombinant inbred line populations. This study demonstrates the power of an introgression library combined with high density marker coverage to resolve quantitative trait loci. Furthermore, the role of liguleless1 in leaf architecture and in resistance to northern leaf blight has important applications in crop improvement.

Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data.

Computer vision models that can recognize plant diseases in the field would be valuable tools for disease management and resistance breeding. Generating enough data to train these models is difficult, however, since only trained experts can accurately identify symptoms. In this study, we describe and implement a two-step method for generating a large amount of high-quality training data with minimal expert input. First, experts located symptoms of northern leaf blight (NLB) in field images taken by unmanned aerial vehicles (UAVs), annotating them quickly at low resolution. Second, non-experts were asked to draw polygons around the identified diseased areas, producing high-resolution ground truths that were automatically screened based on agreement between multiple workers. We then used these crowdsourced data to train a convolutional neural network (CNN), feeding the output into a conditional random field (CRF) to segment images into lesion and non-lesion regions with accuracy of 0.9979 and F1 score of 0.7153. The CNN trained on crowdsourced data showed greatly improved spatial resolution compared to one trained on expert-generated data, despite using only one fifth as many expert annotations. The final model was able to accurately delineate lesions down to the millimeter level from UAV-collected images, the finest scale of aerial plant disease detection achieved to date. The two-step approach to generating training data is a promising method to streamline deep learning approaches for plant disease detection, and for complex plant phenotyping tasks in general.

Image set for deep learning: field images of maize annotated with disease symptoms.

OBJECTIVES: Automated detection and quantification of plant diseases would enable more rapid gains in plant breeding and faster scouting of farmers' fields. However, it is difficult for a simple algorithm to distinguish between the target disease and other sources of dead plant tissue in a typical field, especially given the many variations in lighting and orientation. Training a machine learning algorithm to accurately detect a given disease from images taken in the field requires a massive amount of human-generated training data. DATA DESCRIPTION: This data set contains images of maize (Zea mays L.) leaves taken in three ways: by a hand-held camera, with a camera mounted on a boom, and with a camera mounted on a small unmanned aircraft system (sUAS, commonly known as a drone). Lesions of northern leaf blight (NLB), a common foliar disease of maize, were annotated in each image by one of two human experts. The three data sets together contain 18,222 images annotated with 105,705 NLB lesions, making this the largest publicly available image set annotated for a single plant disease.

Multiple Disease Resistance in Plants.

Many plants, both in nature and in agriculture, are resistant to multiple diseases. Although much of the plant innate immunity system provides highly specific resistance, there is emerging evidence to support the hypothesis that some components of plant defense are relatively nonspecific, providing multiple disease resistance (MDR). Understanding MDR is of fundamental and practical interest to plant biologists, pathologists, and breeders. This review takes stock of the available evidence related to the MDR hypothesis. Questions about MDR are considered primarily through the lens of forward genetics, starting at the organismal level and proceeding to the locus level and, finally, to the gene level. At the organismal level, MDR may be controlled by clusters of R genes that evolve under diversifying selection, by dispersed, pathogen-specific genes, and/or by individual genes providing MDR. Based on the few MDR loci that are well-understood, MDR is conditioned by diverse mechanisms at the locus and gene levels.

Development of microsatellite loci in Artocarpus altilis (Moraceae) and cross-amplification in congeneric species.

PREMISE OF THE STUDY: Microsatellite loci were isolated and characterized from enriched genomic libraries of Artocarpus altilis (breadfruit) and tested in four Artocarpus species and one hybrid. The microsatellite markers provide new tools for further studies in Artocarpus. • METHODS AND RESULTS: A total of 25 microsatellite loci were evaluated across four Artocarpus species and one hybrid. Twenty-one microsatellite loci were evaluated on A. altilis (241), A. camansi (34), A. mariannensis (15), and A. altilis × mariannensis (64) samples. Nine of those loci plus four additional loci were evaluated on A. heterophyllus (jackfruit, 426) samples. All loci are polymorphic for at least one species. The average number of alleles ranges from two to nine within taxa. • CONCLUSIONS: These microsatellite primers will facilitate further studies on the genetic structure and evolutionary and domestication history of Artocarpus species. They will aid in cultivar identification and establishing germplasm conservation strategies for breadfruit and jackfruit.