Photo: Marlon Ferreira de Souza

Published on 01/28/2025

By Roseli Andrion  |  Agência FAPESP – One of the most efficient techniques in agriculture to protect crops, improve productivity and reduce the application of agrochemicals is the use of plasticulture in different agricultural productions. For this reason, plastic materials have been widely used in agriculture for several decades. However, while this material boosts agricultural production, without proper management it poses a major environmental challenge.

This scenario led a group of researchers from the Plasticulture Engineering Center (CEP) to develop a remote sensing system to detect agricultural areas that use plasticulture in their management. The solution uses machine learning on time series of satellite images to detect areas with agricultural plastic with almost 100% accuracy.

The CEP is an Engineering Research Center (ERC) supported by FAPESP and Braskem, a multinational petrochemical company controlled by Petrobras and Novonor. The center is hosted by the Interdisciplinary Center for Energy Planning (NIPE) of the State University of Campinas (UNICAMP).

One of the main focuses of the research is the detection of mulching, a technique that uses polyethylene film to reduce weed growth and control soil moisture and temperature. This gives the crop what it needs in the right amounts: light, water and nutrients.

On the other hand, mulching, which needs to be replaced with each new harvest, can contribute to agricultural plastic pollution. The method uses less plastic than greenhouses but has a greater potential for environmental pollution. “The greenhouse, on the other hand, is maintained for four or five years. In addition, the mulching is in direct contact with the soil and can leave lumps if not handled properly,” explains Marlon Fernandes de Souza, a researcher at the CEP.

According to the environmental agricultural engineer, the materials used in agriculture are becoming a major problem in some places, mainly due to improper disposal after use. “Our project aims to determine the amount of waste produced and propose solutions for managing it in a sustainable way.”

Before the study, there was no accurate survey of the areas where mulching is used in Brazil – not even the industry has this data. “The first step was to find out where plastic is used and in what quantity,” he points out. “Until then, the information was fragmented. With satellite imagery, we were able to delineate these areas and obtain data with almost 100% accuracy.”

Research using satellite imagery to detect plastics is more common in marine areas. “After global warming, which we still haven’t solved, many believe that the biggest environmental problem today is plastic pollution,” recalls Souza. By focusing on agricultural environments, the project represents an opportunity to help change the way the sector deals with this type of material.

According to the researcher, the study, conducted by the CEP and published in the scientific journal Environmental Science and Pollution Research, represents an important step in promoting the circularity of plastics in Brazilian agriculture. “Since it provides a reliable methodology for mapping agricultural plastic waste, it can help minimize the impact of degradation on the environment.”

Reverse logistics

One of the biggest global challenges to the proper management of plastic used in agriculture is the lack of infrastructure for collecting and recycling the material, especially in remote regions. “In some locations, the nearest facility is more than a thousand kilometers away, making reverse logistics economically unfeasible.”

The CEP group does not collect the waste identified, but the survey conducted by the researchers makes it possible to analyze the feasibility of setting up reverse logistics systems. Based on the results, it is possible to assess whether it is feasible to set up regional recycling plants or whether it is better to establish a transportation system to take the waste to more distant recycling centers.

The lack of a structured system for collecting and recycling agricultural plastics has led many producers to improvise solutions. “Unfortunately, there are still many producers who have huge piles of waste because they don’t know what to do with it. In some cases, the material accumulates for years because no one can collect it.”

This is more common with small producers who don’t have the negotiating power or sufficient disposal volume to attract recyclers interested in collecting the material. “Most large producers are able to negotiate collection with recycling companies, but small producers don’t have this advantage. They then accumulate large quantities of plastic without knowing how to dispose of it properly.”

The research could contribute to the development of more efficient public policies and waste management systems. One option is to create a reverse logistics model similar to the Campo Limpo system, which collects empty pesticide containers. In this system, producers return the empty packaging when they buy new products.

It would therefore be ideal to create a similar system for agricultural plastic films. “Implementing this type of solution requires the creation of public policies and the involvement of the entire value chain. Our research aims to provide the necessary information to make this type of initiative viable,” he explains.

Lack of legislation

The lack of specific legislation on plastics management is a growing concern in the agricultural sector. The Food and Agriculture Organization of the United Nations (FAO) is discussing voluntary codes of conduct for plastics management in the sector. But the future is still uncertain. “Ten years ago, there was virtually no such research. Today, there are discussions about best practices and recommendations, but there are still no clear regulations for practice.”

Souza points out that the goal of the project is to find solutions so that the use of plastic in agriculture brings benefits without causing damage to the environment. “The goal is to find ways of using this material in a way that doesn’t cause environmental problems. This discussion is taking place now: there have been several United Nations conferences last year to specifically address this issue.”

The article “Remote sensing detection of plastic-mulched farmland using a temporal approach in machine learning: case study in tomato crops” can be read at: link.springer.com/article/10.1007/s11356-024-35026-7.

 

Source: https://agencia.fapesp.br/53786