Precision pollination

The next generation of artificial pollination systems is designed to support crop-load optimisation. By Anna Mouton.

Prof. Matthew Whiting is an applied whole-tree physiologist in the Department of Horticulture at Washington State University. He spoke to Fresh Quarterly about artificial pollination and the potential of greater precision enabled by new technologies.

Q. What is artificial pollination?

We consider artificial pollination as anything different to introducing honeybee colonies — renting hives — to facilitate pollen transfer.

But if you think about tree-fruit production systems and how flowers are pollinated, at least in our area, it’s not natural because we don’t have much in the way of native pollinators. The honeybees we use are European.

Artificial pollination is not a new concept at all. Supplementing natural pollination in various agricultural systems goes back many decades.

Q. Why develop artificial pollination systems?

There are several key reasons, but crop-load management is one of the main reasons I began to work on artificial systems. I collaborate a lot with sweet-cherry growers, and they were struggling with overcropping and fruit quality on new dwarfing rootstocks.

We initially concluded that we need to thin better. I went down that path of looking at different candidates for thinning but then I realised that thinning is related to fruit set — to excessive fertilisation rates. We needed to take control of pollination rates. And you can do that with artificial pollination systems.

I believe we need to get away from the polliniser-pollinator model because both pollinisers and pollinators are fraught with annual variability in response to climate and other factors.

Colony collapse disorder has also been a significant concern in the United States. Our agricultural production systems need to be resilient to this phenomenon. Artificial pollination could, in some ways, represent the future of food in a very broad sense through yield security and climate resilience.

Q. Is artificial pollination used commercially in tree fruit?

We’re at a point now where we’re quite successful with broadcast applications of pollen suspensions for supplemental pollination. That’s usually for growers whose crop is too light for whatever reason: poor pollinator activity, poor bloom overlap, not enough or incompatible pollinisers—the list goes on.

We recommend that the grower still bring in beehives, but we’ll add one or two passes with the pollen spray. This typically requires a pass when there’s a decent population of open and receptive flowers, probably at about 40%–50% bloom and again at about 80% bloom.

There are companies in the United States that have been collecting pollen for a long time. We worked with them to develop the pollen suspension, and we used electrostatic sprayers that positively charge the pollen grains so they are attracted to the negatively charged stigmatic surface of the flowers.

The pollen suspension media keeps the pollen alive long enough to spray and allows it to form a consistent suspension. We generated the intellectual property for this in my lab, and I started a company licencing it from the university. So, I’m involved in this commercially as a pollen and suspension media provider.

The pollen suspension media is used mostly in cherries and almonds but works equally well in apples and pears.

I believe artificial pollination will go to replacement pollination — total artificial pollination — in the long run. In future orchards, you won’t have to worry about bloom timing, compatibility, quantities, or positions of pollinisers. And you won’t have to rent hives or worry about how many or how to distribute them.

Q. Tell us about your prototype robotic pollinator.

I find this idea really interesting: strategically pollinating a percentage of flowers on an apple tree to produce the desired crop.

Our prototype uses an off-the-shelf robotic arm with three joints and six degrees of freedom. It’s easy to manage and control, and it’s all-purpose, so you can mount whatever you want on it. We mounted a camera, a vision system, and a modified spray nozzle and connected them to a spray tank and pump. It’s on a platform with batteries to power it.

We know that apples have a king flower in the terminal position that opens first, surrounded by lateral flowers that typically open a day or more later. The computer vision system scans the tree to find clusters in which only the king flower is open, and the electrostatic nozzle sprays a little pollen on that flower.

The system doesn’t target the stigmatic surface directly, but we avoid clusters with open lateral flowers. So, we use the phenology of flowering to target pollination.

In our trials, we bag the branches just before the buds open to prevent pollination. Then, we withdraw the bags at night when there are no native pollinators or bee flights, test our system, and put the bags back on. This past year, we had over 90% pollination success.

We’re hoping to run more trials this spring, including netting a larger section of an orchard to keep out insects and removing pollinisers, if there are any, in that section of trees.

Q. What is the future of robotic pollination?

Right now, we’re at the one-arm, one-robot stage. But we’re working on other projects, such as with commercial companies developing multi-armed apple harvesters.

So, we imagine there will be options for multi-armed platforms that drive down the alleyway and access both sides of the row. In fact, we have submitted funding proposals to develop a multi-purpose robotic platform.

You could use the platform for harvesting, and then, when you want to pollinate, you take out the harvest module and put in the pollination one. Or put in the thinning or pruning one. In the future, we could have multi-armed autonomous robots. It’s kind of exciting — the possibilities and interest in robotics for tree fruit are impressive.

When I talk to farmers, they understand the concept of artificial pollination, of simplifying their thought processes and farming operations, and the benefits of being a little more precise.

For more information, contact Prof. Matthew Whiting at mdwhiting@wsu.edu.