Researchers use AI to combat Asian rust

Professor Marcelo Giovanetti Canteri, from the Department of Agronomy at the State University of Londrina (UEL), is the director of the Center for Artificial Intelligence in Agriculture (Ciagro), which brings together researchers from several institutions in Paraná focused on developing technologies for the field. He is currently coordinating a project in partnership with the institution's Department of Computing (CCE) for the application of AI in agriculture, specifically in monitoring and combating soybean rust, within the Rust Alert.

Soybean rust (or Asian rust, its continent of origin) is a disease caused by the fungus Phakopsora pachyrhizi and attacks the leaves of the plant. It is the main disease of soybeans and has caused significant crop losses in the past.

In addition to UEL, IDR-PR, Adapar, UTFPR, Brazilian Agricultural Research Corporation, Faep and ABC Foundation are also working together to monitor more than 200 collectors spread across soybean plantations. It is an area that begins in Assis (SP), passes through Londrina and Campo Mourão, reaches Cascavel, and goes to Pato Branco and Guarapuava.

The collectors are “traps” mounted on plants that capture the fungus with the wind. And while before it was necessary for the researcher to go to the trap in situ to see the result and then analyze it with the help of a microscope, with AI he can monitor the presence (or not) of spores without leaving the laboratory, as the technology allows the collected spores to be automatically identified. IDR-PR implemented the network of traps

According to the professor, rust is a pest in its truest sense. "Even with all the measures to combat it, such as the annual sanitary break, research into biological control and the development of more resistant soybean species, it returns every year and requires the application of fungicides," he says. On average, these products need to be used annually in an area of ​​40 million hectares.

The state coordinator of the Sustainable Grains project of IDR-Paraná, Edivan Possamai, highlights that Paraná, through IDR-PR, was a pioneer in organizing a network of collectors of Asian rust spores, especially in the last eight years. On average, 160 collectors are installed annually and help farmers in their decision-making. "With this work, it is possible to reduce fungicide applications by 35%, with lower costs for the producer and less environmental impact", he states.

The project aims to reduce costs for producers through more effective monitoring, because other factors also interfere and can be acted upon. Three aspects must be taken into account: the plant, the climate and the fungus itself. Ciagro is automating the observation of the three factors, anticipating the arrival of data and ensuring faster decisions.

Monitoring tests for the 2023/2024 harvest analyzed two of the three aspects: climate and pathogen (fungus). For the next harvest (24/25), planting monitoring will be included. Professor Marcelo notes that, over time, the fungus has been increasing its resistance to fungicides, hence the urgent need to develop technology to combat the pest. New fungicides are also being created, but the use of AI can indicate the best time to use them, and even whether or how much it is worth doing so, depending on the characteristics of each region.

Climate data is an important component of the research. Previously, it took up to two weeks to determine the data, but now it is available in two days. This makes it easier to use, for example, the virtual alert issued by Embrapa's Anti-Rust Consortium, which provides information on the spread of rust throughout the country.

Each region has its own rainfall and humidity levels, which are observed individually. Pato Branco and Guarapuava, for example, are known for their humid climate and late planting. And the time of planting also matters: if it is done late, the fungus arrives with the plant less developed, increasing the risk of damage, that is, increasing the need for fungicide application, which means more expenses.

Despite being a research project, there are extension actions, since it directly involves rural producers. Embrapa and IDR are involved in this aspect. In addition, the project has been disseminated through publications, participation in scientific events and social networks. Papers have already been published in journals in the area of ​​Agronomy and presented at Computer Science events. Ciagro has already promoted two workshops on the use of AI in agriculture and the third will be at the end of November.

Technical Forum

On November 27, Adapar will hold a technical forum to define the Sanitary Vacancy periods and the soybean sowing calendar for Paraná in the 2025/26 harvest. This proposal and discussion with the production sectors aims to establish the dates to be forwarded to the Ministry of Agriculture and Livestock (Mapa), so that the analysis can be carried out in harmony with the other Federation Units.  

"These legislative measures are important for the Management of Asian Soybean Rust, because, once adopted, they are the beginning of reducing the pressure of disease inocula. In addition, the more restricted sowing window can provide a lower application of fungicides for rust control, which allows for less pressure on the active ingredients used to manage the disease in the field", explains the head of the Pest Surveillance and Prevention division in Agricultural and Forestry Crops at Adapar, Marcílio Araújo.