Sep 3, 2008

Mechanical Harvester Fits Bush-Like Trees

Horticultural researchers at Michigan State University (MSU) began to evaluate the potential of over-the-row harvesting of tart cherries this summer.

The goal is to harvest tart cherries more efficiently and with quality high enough that they can be distributed in both the fresh and processing markets (even though the market for tart cherries is mostly for processing.) In such a harvesting system, tart cherry trees would be treated like bushes and harvested like blueberries.

The evaluation began in July using an off-the-shelf blueberry harvester, the Korvan 7420 from Oxbo International, on a mixture of seedling tart cherry trees of all sizes, shapes and descriptions. The trees are part of a large collection at the Clarksville Horticultural Research Station, bred and planted on more than 25 acres by MSU tart cherry breeder Amy Iezzoni. Some are nearly ready for release as named varieties. The candidates used for this test were seedlings in her breeding program that were naturally small in stature and canopy size. Larger, traditional-size trees were not harvested.

The U.S. tart cherry industry today is dominated by one variety – the light-red-skinned, white-fleshed, 400-year-old French variety Montmorency – grown to moderate size on either of two rootstocks, Mazzard and Mahaleb. They are harvested primarily by trunk shakers that drop cherries onto inclined-plane catching frames. Iezzoni has worked to develop new varieties that are resistant to cherry leaf spot and to introduce different European influences – having a less tart, deeper red flesh than Montmorency.

The test trees contained different kinds of fruiting wood – some pendant and willowy, some short and spurry – and were of different styles, from weepy to upright, some having single trunks and others multiple, some bush-like, some tree-like and varying in size from very dwarfed to full size. They also varied in maturity dates, so pull force for removal varied greatly as well.

To be included in the test, the trees had to fit into the harvester, the throat of which is 5 feet wide and 8 feet tall. Selections were chosen that are naturally small. Existing trunk shakers need space between the ground and canopy, so with this machine the canopy can extend down to about a foot above the ground and still provide space for catching plates below the shaking mechanism. Trees can be shortened in the trunk, saving canopy volume.

An early demonstration drew lots of researchers and media coverage, which made organizer Ron Perry nervous. This is just a first test, a time to generate ideas and questions, he kept reminding everyone.

On the other hand, Perry, after chairing the MSU Horticulture Department for several years, has left administration and returned to academe as the department’s orchard and vineyard technology specialist. With new funding for mechanical harvesting research, authorized in the Specialty Crops Initiative in the recently passed Farm Bill, Perry is organizing the MSU team to compete for part of the action.

Perry Hedin, from the Cherry Industry Administrative Board, was there to watch as well. The Michigan Cherry Committee put up funds to transport and test the harvester.

The great virtue of the $150,000-plus Korvan harvester is that it exists and is in commercial production. It is used to harvest blueberries and raspberries for both fresh and processing markets. Researchers like Perry and his MSU colleague Jim Flore agree that fruit trees are malleable and can be fitted to machines more easily than machines can be invented to fit trees. Given a machine, they can use genetics, rootstocks, hedging, pruning, training – lots of tools to shape a tree to it.

Two years ago, Flore brought to the same test orchard a machine designed for citrus harvest by USDA agricultural engineer Don Peterson, now retired, who developed it at the USDA research station in Kearneysville, W.Va.

While its key harvesting mechanism looked a lot like the key part of the Korvan machine, Peterson’s “spiked drum” did a much more complete job of removing fruit in the Clarksville test. The spiked drum looks like a hedgehog of plastic fingers that, when vibrated properly, effectively shake cherries from the trees.

The Korvan machine has a similar set of steel spikes – 20 of them on spindles 22 rows deep. But the gentler motion that shakes just enough to remove ripe blueberries, leaving green ones behind for future harvests, did not remove all the cherries, although the researchers said the Korvan did less canopy damage. Use of growth regulators like Ethephon to reduce pull force would also have a place in the system, Perry said.

The folding plastic plates that slide around the tree trunks provided a good catching surface. And the rest of the machine’s features – the elevator that moves fruit away from the catching surface and then elevates it to a platform above the trees – were effective. The trash removal system – fans that blow away leaves and debris – worked.

The machine moved continuously at about 1 mile an hour, a good speed in Perry’s assessment.

The Korvan 7420 has two sorting stations on its top platform where the operator sits. Blueberries or raspberries pass by two people who remove defective berries, and good berries then drop into flats for stacking at the rear of the platform. The concept, according to Josh Culp, the Korvan service manager who ran the machine along with service technician Adam Wallace, is to create a product ready for fresh market or cold storage without further handling.

While this machine wasn’t yet ready for commercial cherry orchards -– and properly designed orchards are not yet ready for the machine – MSU sweet cherry expert Greg Lang noted the difficulty in finding specialized machine manufacturers willing to build for an undeveloped market. He, too, agreed that Peterson’s machine performed well. It does one side of the tree at a time; the Korvan does both sides. Still, no manufacturer has picked up the Peterson machine. Efforts to harvest fresh-market quality sweet cherries, under way at the University of Washington, are focusing more on a limb shaker and catching frame.

Perry, however, is anxious to move ahead. His team is considering identifying elite candidates in Iezzoni’s breeding program that have potential and then getting test orchards in place. The work will take at least five years, he said. At the same time, he encouraged the Korvan representatives to take their observations back to the Oxbo engineers to fine-tune the shaking mechanism and lengthen the catching area.

Meanwhile, Flore was making calculations on a napkin during lunch, figuring how many trees per acre would be needed and what the spacing could be to fit the Korvan machine. Planting trees 9 x 3 or 9 x 4 (1,200 to 1,600 trees per acre) with canopies 8 feet tall would generate 100,000 cubic feet of fruiting area per acre, about 25 percent more than what growers currently achieve using trunk shakers that demand more open space and taller trunks.

That sounded pretty good to him.