The genus Phyllachora contains numerous obligate fungal parasites that produce raised, melanized structures called stromata on their plant hosts. Most members of this genus are not of significant economic concern, with the exception of P. maydis, the causal agent of tar spot of maize (Zea mays). Tar spot of maize has emerged as a major threat to maize production throughout the Americas and continues to spread throughout North America. To date, species designations for Phyllachora have been based on host associations and morphology, and the origin and diversity of the pathogen that causes tar spot is unknown. We assessed the sequence diversity of 186 single stroma isolates collected from 16 hosts representing 15 countries by amplification of the ITS and LSU gene regions. Samples included both herbarium and contemporary strains that covered a temporal range from 1905-2019. These 186 isolates were grouped into 5 distinct species with strong bootstrap support. We found three closely related, but genetically distinct groups of Phyllachora are capable of infecting maize in the United States, we refer to these as the P. maydis species complex. Based on herbarium species, we hypothesize that these three groups in the P. maydis species complex originated from Central America, Mexico and the Caribbean. Although two of these groups were only found on maize, the third and largest group contained contemporary strains found on maize and other grass hosts, as well as herbarium specimens from maize and other grasses that include 10 species of Phyllachora. The herbarium specimens were identified based on morphology and host association, but our data indicates there may be significant synonymy in the Phyllachora genus and additional work on species delineation and host specificity should be considered.

Crop pest detection and mitigation remains an extremely challenging task for the farmers. Majority of the pest classification and detection techniques rely on supervised deep learning frameworks that require significant human intervention in labeling the input data, thereby making the down-stream tasks tedious. Therefore, this study presents a self-supervised learning (SSL) approach to classifying 12 types of agricultural insect pests from 9549 RGB images, by leveraging the Bootstrap your own latent (BYOL) algorithm. SSL uses minimal labeling and is indifferent to data augmentations or distortions. Hence, latent representations from pretrained SSL networks could be generalized well for downstream tasks like classification or object detection. For desirable classification of the insect images, the greatest challenges observed were: i) large intra-class variation (the same insect was found with different colors and patterns), and ii) complex background with inconspicuous foreground. Hence, to overcome these issues and aid generalizability of the representations learned through BYOL, entropy-guided segmentation (segments based on texture not color), is proposed as input to the SSL network in this study. Both raw and segmented images were separately fed to two independent BYOL SSL networks, i.e., with ResNet18 and ResNet50 architectures as the backbone. The efficacy of the latent representations for downstream applications was assessed using linear evaluation, and subsequently compared with traditional transfer learning outcomes from ResNet18 and ResNet50. The results indicated that, while ResNet50 backbone intuitively performed better in all cases, SSL aided with entropy-based segmentation led to ~94% classification accuracy compared to raw images (with ~90% maximum accuracy).

The genus Phyllachora contains numerous obligate fungal parasites that produce raised, melanized structures called stromata on their plant hosts referred to as tar spot. Members of this genus are known to infect many grass species but generally do not cause significant damage or defoliation, with the exception of P. maydis which has emerged as an important pathogen of maize throughout the Americas, but the origin of this pathogen remains unknown. To date, species designations for Phyllachora have been based on host associations and morphology, and most species are assumed to be host specific. We assessed the sequence diversity of 186 single stroma isolates collected from 16 hosts representing 15 countries. Samples included both herbarium and contemporary strains that covered a temporal range from 1905-2019. These 186 isolates were grouped into 5 distinct species with strong bootstrap support. We found three closely related, but genetically distinct groups of Phyllachora are capable of infecting maize in the United States, we refer to these as the P. maydis species complex. Based on herbarium species, we hypothesize that these three groups in the P. maydis species complex originated from Central America, Mexico and the Caribbean. Although two of these groups were only found on maize, the third and largest group contained contemporary strains found on maize and other grass hosts, as well as herbarium specimens from maize and other grasses that include 10 species of Phyllachora. The herbarium specimens were identified based on morphology and host association, but our sequence data indicates some Phyllachora species are capable of infecting a broad range of host species and there may be significant synonymy in the Phyllachora genus and additional work on species delineation and host specificity should be considered.