This study evaluated the potential of using machine vision in combination with deep learning (DL) to identify the early blight disease in real-time for potato production systems. Four fields were selected to collect images (n = 5199) of healthy and diseased potato plants under variable lights and shadow effects. A database was constructed using DL to identify the disease infestation at different stages throughout the growing season. Three convolutional neural networks (CNNs), namely GoogleNet, VGGNet, and EfficientNet, were trained using the PyTorch framework. The disease images were classified into three classes (2-class, 4-class, and 6-class) for accurate disease identification at different growth stages. Results of 2-class CNNs for disease identification revealed the significantly better performance of EfficientNet and VGGNet when compared with the GoogleNet (FScore range: 0.84–0.98). Results of 4-Class CNNs indicated better performance of EfficientNet when compared with other CNNs (FScore range: 0.79–0.94). Results of 6-class CNNs showed similar results as 4-class, with EfficientNet performing the best. GoogleNet, VGGNet, and EfficientNet inference time values ranged from 6.8–8.3, 2.1–2.5, 5.95–6.53 frames per second, respectively, on a Dell Latitude 5580 using graphical processing unit (GPU) mode. Overall, the CNNs and DL frameworks used in this study accurately classified the early blight disease at different stages. Site-specific application of fungicides by accurately identifying the early blight infected plants has a strong potential to reduce agrochemicals use, improve the profitability of potato growers, and lower environmental risks (runoff of fungicides to water bodies).

The uniform application (UA) of agrochemicals results in the over-application of harmful chemicals, increases crop input costs, and deteriorates the environment when compared with variable rate application (VA). A smart variable rate sprayer (SVRS) was designed, developed, and tested using deep learning (DL) for VA application of agrochemicals. Real-time testing of the SVRS took place for detecting and spraying and/or skipping lambsquarters weed and early blight infected and healthy potato plants. About 24,000 images were collected from potato fields in Prince Edward Island and New Brunswick under varying sunny, cloudy, and partly cloudy conditions and processed/trained using YOLOv3 and tiny-YOLOv3 models. Due to faster performance, the tiny-YOLOv3 was chosen to deploy in SVRS. A laboratory experiment was designed under factorial arrangements, where the two spraying techniques (UA and VA) and the three weather conditions (cloudy, partly cloudy, and sunny) were the two independent variables with spray volume consumption as a response variable. The experimental treatments had six repetitions in a 2 × 3 factorial design. Results of the two-way ANOVA showed a significant effect of spraying application techniques on volume consumption of spraying liquid (p-value < 0.05). There was no significant effect of weather conditions and interactions between the two independent variables on volume consumption during weeds and simulated diseased plant detection experiments (p-value > 0.05). The SVRS was able to save 42 and 43% spraying liquid during weeds and simulated diseased plant detection experiments, respectively. Water sensitive papers’ analysis showed the applicability of SVRS for VA with >40% savings of spraying liquid by SVRS when compared with UA. Field applications of this technique would reduce the crop input costs and the environmental risks in conditions (weed and disease) like experimental testing.

Non-marketable crops are increasingly being used as a tool to promote agroecosystem services and sustainable agriculture. Nevertheless, crops vary greatly in the traits by which they capture resources and influence the local ecosystem. Here we report on the traits and associated soil microbial communities that relate to aboveground biomass production, nutrient capture, weed suppression, erosion control and building particulate organic matter of 22 different full-season cover crops. All agroecosystem services were positively correlated with maximum canopy height and leaf area. Rooting density was positively associated with indices of bacterial diversity. While some legumes produced the greatest standing N and P in aboveground biomass, they were also poor at capturing soil nitrate and promoted high levels of potential plant fungal pathogens. Conversely, Brassicaceae crops had the lowest levels of potential plant fungal pathogens, but also suppressed saprophytic fungi and rhizobia. Thus, not all crops are equal in their ability to promote all agroecosystem services, and while some crops may be ideal for promoting a specific agroecosystem service, this could result in a trade-off with another. Nonetheless, our study demonstrates that plant functional traits are informative for the selection of crops for promoting agroecosystem services.

Cover crops are increasingly being included in crop rotations as a mechanism to promote diversity and provide agroecosystem services, including weed suppression. Recently, cover crop mixtures have increased in popularity in an attempt to provide a greater diversity in ecological services as compared with monocultures. Several recent studies, however, have failed to detect a positive effect of cover crop diversity on biomass production or weed suppression. Here we assessed biomass productivity and weed suppression in 19 cover crops seeded as monocultures and 19 mixtures of varying species composition and functional richness (two- and three-species mixtures) of full-season cover crops in Atlantic Canada. Cover crop biomass production and weed suppression varied by species identity, functional diversity, and species richness. As cover crop biomass increased regardless of diversity, weed biomass declined. Highly productive forbs and grasses provided the greatest weed suppression in monoculture. In line with previous observations, mixtures were not more productive or weed suppressive on average than the most productive monocultures. We observed that the inclusion of the highly productive species buckwheat (Fagopyrum esculentum Moench) and sorghum–sudangrass [Sorghum × drummondi (Nees ex Steud.) Millsp. & Chase] in a mixture increased stand evenness, productivity, weed suppression, and spatiotemporal stability. Taken together, our results suggest that effects of diversity on mixture productivity and weed suppression are species specific. This further demonstrates the importance of species selection in cover crop mixture design.

Potato is the third most important staple food crop globally following rice and wheat. In the United States, potato is grown on approximately 410,000 ha with a farm-gate value of US$1,032 million. In Canada, potato is grown on approximately 134,000 ha with a farm-gate value of US$235 million. The objective of this manuscript, compiled by the Weed Science Society of America Weed Loss Committee, was to estimate potato yield loss caused by weed interference. Potato yield data from weedy and weed-free plots (or plots with >95% weed control) was obtained from researchers working on weed management in potato in the United States and Canada or from published manuscripts from 2000 to 2018. Potato yield loss from weed interference was 12% to 61% when no weed management tactics were implemented. The average yield loss for all states/provinces (where data was obtained) due to weed interference was 44%. Weed interference would cause a farm-gate loss of approximately US$465 million and US$61 in the United States and Canada, respectively, if weeds are not controlled. These results indicate that weed management is critical for successful potato production, and that an ongoing need for research exists on weed management in this crop.

Potato producers in the Atlantic Canadian provinces of New Brunswick (NB) and Prince Edward Island (PE) rely on the photosystem II-inhibiting herbicide metribuzin for weed management. Recently, potato producers in the region have reported unacceptable common lambsquarters control following an application of metribuzin. Tissue and seed samples were collected from escaped common lambsquarters populations from across the potato producing regions of NB and PE and screened for the Ser264Gly mutation in psbA. Overall, 46% of sampled populations possessed the Ser264Gly mutation across the region. Cross-resistance testing to atrazine, metribuzin and linuron confirmed populations with the Ser264Gly were resistant to triazines and triazinones but remained susceptible to linuron. Dose response analysis determined a moderate level of resistance to metribuzin in common lambsquarters which would not be controlled in producers fields. A field experiment was conducted in Fredericton, NB and Harrington, PE, to determine if currently registered and unregistered products and tank-mixes would control PSII-inhibitor-resistant common lambsquarters in potato. All evaluated products, with the exception of S-metolachlor, provided control equivalent to the weed-free check without compromising potato yield or quality. This study demonstrates that PSII-inhibitor-resistant common lambsquarters are found in Atlantic Canadian potato production systems, but can be controlled with currently registered herbicides and rates with alternative modes of action.