Studying fruit fly through machine learning

In the Western Cape, Ceratitis capitata, commonly known as the Mediterranean fruit fly or Medfly, is one of the most economically harmful pests in orchards, causing fruit damage and posing a phytosanitary risk.

Four years ago, Gulu Bekker, an intrepid Stellenbosch researcher, wondered what would happen if he employed techniques that are normally used to analyse “big data” in finance, medicine, and marketing, to study fruit fly distribution patterns. Bekker, who recently received his PhD for his efforts, gave integrated pest management (IPM) the lead role in his novel study, in which he explored the distribution of fruit flies in the Western Cape. He used machine learning techniques, combined with geospatial analyses in his research. Machine learning is an umbrella term for techniques that allow the user to train a computer to make certain decisions based on what it has learned. Geospatial analyses describe the statistical analyses of data that have a spatial component.

How did he do it?

Bekker used trap catch data to visualise the distribution range of fruit flies in a pome-producing area in the province. Approximately 400 traps were selected. A geographic information system (GIS) was then used to map the mean trap catch per trap location of four consecutive production seasons. He combined this information to identify monthly and seasonal long-term hot and cold spots in the regions of Elgin, Grabouw, Villiersdorp and Vyeboom. Hot spots are defined as areas with relatively high trap catches and cold spots areas with low trap catches, relative to the rest of the study area.

Findings

The results showed that the distribution of fruit fly hot and cold spots in his study area is driven by long-term climate variables. The hot spots were concentrated in the hotter and drier areas, and the cold spots in the colder and wetter areas. The study indicated that rainfall was the most important driver of these hot and cold spots. However, these drivers are area-specific. “This means that the same drivers cannot be used to model fruit fly distribution in areas with different climates and geographic characteristics,” Bekker explains. This finding motivates for finer-scale monitoring of climate in different regions, he says. It will allow for more accurate and area-specific models.

Although the drivers of fruit fly spatial distribution were different between the regions, the distance from orchards to urban areas in the early fruiting season emerged as a strong driver in all the areas. This was the case in the regions of Elgin and Grabouw, Vyeboom, Villiersdorp, as well as the Warm Bokkeveld. “Again, this brings up the question of the impact of unman-aged home gardens in urban areas on the fruit fly populations in surrounding commercial orchards,” Bekker emphasises. Bekker also zoomed in on orchard level to understand fruit fly populations at this scale. This is part of efforts to understand why orchards with low numbers of fruit flies in traps often experience lots of damage related to fruit fly infestation. “Many producers are baffled by the fact that there is often not a clear correlation between trap data and the damage caused in orchards by these insects,” he says. “Trap catch and fruit damage is a spatial problem and needs to be looked at in this way. We know that fruit fly distributions in orchards are spatially variable. It just makes sense that the damage would also be spatially variable. The monitoring method for both the fly and the fruit damage must be conducted on the same scale. This notion is, however, still unexplored and more research is needed.”

Why is the study relevant?

The research is relevant to the integrated management of fruit flies and potentially other insect pest species on a local and regional scale. “This study visualised the patterns of fruit fly distribution that we always believed were true,” says Bekker. “This is the first time we could see fruit fly distribution on a piece of paper and actually make sense of it on an area-wide scale.” Bekker says the results could be used to improve area-wide integrated fruit fly management programmes through more precise spatial planning of management actions, such as the sterile insect technique and the bait application technique. Based on his results, it is recommended that previously unman-aged fruit trees in home gardens should be managed, either by fruit stripping or applying protein baits. This is needed to prevent fruit fly population build-up in these habitats, which could later invade high-value export fruit crops in surrounding commercial orchards. “In terms of area-wide management, area-wide programme managers should employ spatial analyses techniques to see where hot and cold spots occur, and then adapt their practices accordingly,” Bekker says. “Precision management is the way forward. If you are not doing it yet, perhaps you should start thinking of it.” The industry also needs to take a long-term view on integrated pest management, he says. “That is our best insurance policy to ensure resilient farms and sustained production.”

More weather stations that are efficiently managed, as well as sustained monitoring can play a significant role in managing fruit flies in the region. “It is important to monitor fruit fly populations, but of equal importance is well organised and filed trap monitor-ing data,” Bekker indicates. There must also be a willingness to share this data with research institutions, he believes. This will help ensure that improvements are made to monitoring systems and control strategies.“Living in an age of ‘big data’ and machine learning, the possibilities of what can be done become endless,” Bekker says. “Every-one must build a culture of data sharing. In the long-term, both farmers and research institutions would hugely benefit from it.”

Who is Gulu Bekker?

Bekker grew up on a farm in the Eastern Cape between Hofmeyr and Steynsburg. His dissertation was challenging in more ways than one. Bekker got married a month before he started his PhD and became a dad during the final year of his study. “My wife
(Grethe) and the twins were my inspiration and motivation for getting it done.”
Before enrolling for his PhD, Bekker worked as a research assistant with the Integrated Pest Management Group in the Department of Conservation Ecology and Entomology and as an environmental consultant.
His PhD study, titled Spatiotemporal analyses of fruit fly populations in selected areas of the Western Cape, was funded by Hortgro, the International Atomic Energy Agency and the National Research Foundation. Bekker’s work was supervised by Dr Pia Addison and Matthew Addison, both from the Department of Conservation Ecology and Entomology, and Prof Adriaan van Niekerk, from Stellenbosch University’s Department of Geography and Environmental Studies.
His research was recently published in the journals Computers and Electronics in Agriculture and African Entomology.
Bekker is currently a post-doctoral fellow in the Department of Conservation Ecology and Entomology.
He is looking at automatic identification techniques to identify the most economically important fruit fly species in SA. “Hopefully, this research will lead to developing a tool that can be used at import and export ports (like harbours) to identify not only adult flies but also immature stages (of the insect),” he says.

Bekker is also applying machine learning techniques to under-stand the distribution patterns of codling moth populations in the Western Cape. Like fruit flies, these insects pose a major phytosanitary risk to the industry.