Ecophysiology and technology in Italy

An ecophysiologist discusses how new technology can allow growers to communicate with their trees. By Anna Mouton.

Prof. Luigi Manfrini is an ecophysiologist in the Department of Agricultural and Food Sciences at the University of Bologna. His research focuses on precision management of fruit production, and he teaches a postgraduate course in precision orchard management. He also chairs the Mechanisation, Digitisation, Sensing, and Robotics Working Group of the International Society for Horticultural Science.

Manfrini spoke with Fresh Quarterly about the intersection between ecophysiology and technology and why growers should remain curious.

Q. How could Italian apple growers benefit from precision orchard management?

A. Fruit size is the main variable determining the return to the apple grower in Italy. Of course, you need to have colour if it’s a red fruit, but the cooperatives never look at the internal quality of the fruit.

Growers always need to balance production in terms of yield and fruit size. If the fruit numbers are very high but the fruit are too small, revenues will be very low. It’s much better to have a lower number of fruit but of a bigger size.

Growers also have the problem of biennial bearing — one tree is full of fruit one year, and the following year it has nothing. Properly managing the crop load by avoiding excessive yields ensures balanced production across the years.

My group is working on computer vision systems using cheap cameras that can be attached to a tractor or rover. Our latest results show that we are quite precise in predicting fruit count and size at the end of the season.

We can predict the crop load and fruit size distribution for apples within a couple of millimetres, and we’re also working on kiwifruit, peaches, apricots, and plums.

The idea is to allow growers to exploit the information from the orchard to manage individual trees best. They can apply management practices like changing the irrigation schedule, doing some thinning, or adjusting the fertilisation.

Q. Why aren’t growers already collecting fruit measurements to inform management?

A. As part of my PhD research, I developed a protocol to predict an orchard’s average fruit size and size class distribution. The technology was simple — go into the orchard with a calliper and measure the fruit growth three to five times during the season.

After finishing my PhD, I built a startup based on my research. The methodology worked well, and the results were good, but growers weren’t ready to use it — not because it was difficult but because it required their time.

During the season, growers are always rushing to follow the crop, and if you ask them to do something more, they mostly say it takes too much time, even if it’s only one hour per week.

If I had been using the computer vision system 13 years ago when I built the startup, it would have overcome the problem of going to the orchard to measure fruit with a calliper.

I also asked the growers to collect very small samples to be quicker. Even if it doesn’t see all the fruit, a computer vision system will measure a thousand times more fruit than people would measure. And a bigger sample will make us more precise.

Q. How could ecophysiology be applied in orchards?

A. I always try to teach my students that with physiology, you can speak to the plants and understand what they are saying to you. And sensors are the translators of what the plants are doing — sensors give us numbers that can tell us if the plant is stressed or not.

However, maybe a soil sensor tells us there isn’t enough water in the soil for the plant, but then we also need feedback from the plant. Is that water content in the soil really stressing the plant?

That’s why I want to check different parts of the orchard and the relative humidity and air temperature at the same time as having the plant’s feedback. I need to speak with the plant.

At the moment, growers do not use plant-based sensors, but different companies are releasing micro-potentiometers that measure plant water potential or sensors that measure transpiration. These must be calibrated in the orchard every year, and a third party must interpret the data.

My group is trying to create a platform to translate all the information from the orchard — from plant-based or environmental sensors — for growers. The idea is to have sensors that tell you if the plant is in a good condition or not.

The next step is to use the sensor to control, for example, the irrigation schedule. So the grower can look if everything is fine, but the plant directly requests the rest.

Q. What are some of your other projects?

A. At the moment, half of my group consists of engineers developing techniques and tools for counting fruit and reconstructing the canopy structure. We are also trying to create a big database of physiological information from sensors in the orchard.

We want to create a digital twin that can tell us what will happen in the following years in terms of carbohydrate and nutrient storage and what the plants will do. Precision fruit growing could be called prediction fruit growing because I would like to have an idea of what will happen in the future according to what happened in the past.

The topic is pretty difficult because there are many interactions, and we need to consider that every orchard is unique. In a single orchard, there are specific environmental conditions and management practices, such as protective nets that change the light environment or different pruning techniques.

One of the biggest problems with fruit trees is that they remember what happened in the past. It’s a perennial crop — trees remember if they used resources during the winter or because they had problems in previous years. So one year you find a very low number of flowers just because the year or two before you had a strange season.

Because of climate change, growers can’t really understand what their trees are doing. In Emilia-Romagna, we’ve had floods for two years in a row, and after that, we had very dry seasons. The long-term effect on plant productivity is not predictable.

Q. Are growers becoming more open to adopting technology?

A. According to the literature, only about 5% of growers use new technologies in precision agriculture in general. Precision orchard management is a specific part of precision agriculture, and I think the figure is even lower for orchards.

Agriculture is a slow world compared to automation or other sectors investing in new technologies. This is natural because we always need to wait to see the plant’s response and production.

But since I started researching, the growers’ minds are changing. In the beginning, they said, “My world is different from yours,” but now they are using mobile phones and apps. They’re more curious — many come to university and get at least a bachelor’s degree, so they start understanding how plants work.

Every year, growers are challenged by a new problem in the orchard environment — new pathologies, insects, climate change, or stress conditions. But I’m sure there is always a solution, and my suggestion to growers is to stay curious and be prepared to try sensors and technologies.