A century of orchard systems

The trees were sprayed using 400-gallon wooden tanks and enormous Hardie air-blast machines powered by their own six-cylinder Perkins diesel engines. Dall remembers seeing bulldozers brought in to pull spray rigs when conditions were too muddy for tractors.

Open-vase trees capture more light than untrained trees, but light interception and distribution tend to be limited to the outer canopy, while the shaded portions of the tree are unproductive. The open-vase trees also take many years to come into full production.

"In the old days, growers had time to wait for their trees to fill their space," says horticultural adviser Christo Strydom.

"Nowadays, we don't have time – we need to harvest."

Tripling tree densities

The 1970s were an exciting time for South African pome-fruit growers. They were gradually transitioning from seedling rootstocks to M.793 and taking their first tentative steps toward higher-density plantings.

"I started at Oak Valley in 1970, and that was the year Don Heinicke came from the United States and convinced us that our trees were too big," says Dall.

Heinicke was one of the pioneers in studying light distribution in tree canopies. Their work launched a transition to simpler canopies that continues to this day.

In SA this shift began with the three-leader closed vase devised by Prof Daan Strydom of the Department of Horticulture at Stellenbosch University (SU).

"The three-leader closed vase was the rage in the 1970s," says Dall. "We headed the tree and grew three upright leaders with side branches. It was an improvement, but getting the leaders equally strong was very difficult, and even three leaders created too much shade."

Closed-vase orchards were typically planted at 5 x 3 m to give a density of around 660 trees p/ha – triple that of open-vase orchards. But establishment costs were still relatively low as trellising was not used.

"Before the closed vase, an Italian consultant convinced the apple industry that the palmette system was the best," recounts Dall.

"It was highly intensive and expensive and was not really going to work on seedling rootstocks." Palmettes came and went within a decade but left growers averse to trellising for far longer.

Closed-vase trees were a natural progression from open-vase trees and allowed growers a first taste of higher densities. The downside was that production costs remained high,
explains Prof Wiehann Steyn, general manager at Hortgro Science. "Spraying, pruning, and harvesting all remained inefficient."

The rise of central leaders

The work of New Zealand horticulturist Dr Don McKenzie gave rise to the early central leaders. Trees were planted at 5 x 3 m and were intended to have very structured scaffold branches spaced about 1 m apart. In reality, says Dall, the trees were less formal because of branching limitations.

"We did a lot of notching to get the breaks – it was actually a crazy system," he adds. "So, we moved to a more informal system, and then solaxe came."

There were many variations on the central-leader theme, but they were all a nightmare to prune. Steyn recollects the process on the farm where he grew up. Light distribution was addressed by removing large branches each autumn before any detailed pruning could begin.

"And the detailed pruning was a massive job," says Steyn. "It took the whole winter."

Horticultural adviser Dr Nigel Cook learnt about central leaders as an SU student under Prof Daan Strydom in the late 1980s. "We were taught that we were going to central-leader trees for better light distribution, but we had a much bigger branching problem than anyone else due to a lack of winter chill."

The role of mild winters in delayed foliation and basal dominance has long been known. South African researchers had been trying to find solutions since at least the 1950s, and rest-breaking treatments were already in use by 1960.

Cook explains that the combination of insufficient chill and apical dominance reduces lateral branching in central-leader trees – a challenge that growers face to this day. However, the original central leaders evolved into the solaxe and spindle trees that populate most contemporary South African orchards.

Solaxe

In the late 1970s a Frenchman called Jean-Marie Lespinasse observed that apple-tree branches produced less vegetative and more reproductive growth when bent below horizontal. This had profound implications for the apple industry.

Lespinasse and colleague Dr Pierre-Éric Lauri went on to develop the solaxe system in which branch angles are manipulated to manage tree growth and fruit production. It soon gained traction worldwide.

Cook tells how he adapted the solaxe for South African conditions after visiting Lauri. "I joked that we should call the South African version the SA-laxe."

Solaxe training gave growers the option to control vigour without heading cuts. "The solaxe relies almost exclusively on thinning cuts," says Cook, "because we don't want to invigorate the tree and reduce yields by excessive pruning."

South African apple growers began adopting the solaxe in the late 1990s, nearly doubling yields p/ha.

"A branch below horizontal is so much more productive and so much less vigorous," says Dall. "With solaxe, we weren't growing wood – we were growing fruit. We even converted old trees to solaxe."

Apples have been cultivated for millennia, so it seems strange that it took people so long to twig to solaxe. "I think that when we planted an orchard in the 60s and 70s, we left it for four or five years without doing anything to it," reflects Dall.

"It wasn't that expensive to establish an orchard, so the pressure to get into production early and get high yields was not really there."

Bending down all the branches and the top of the leader breaks apical dominance and promotes bourse-over-bourse bearing. Solaxe trees usually had permanent branches and were labour-intensive to prune and pick. They bore more fruit but not necessarily better-quality fruit.

Progress with pears

By the late 1980s most pear trees were central leaders – pears do not respond well to solaxe training – spaced 4.5 x 1.5 m. "Since then, not much has changed," says Strydom. "There are some variations, but we're still planting at more or less the same distance."

Most South African pear orchards are established on vigorous BP 1 rootstocks. Obtaining sufficient lateral branching on young trees can be a struggle, explains Strydom. "We want to do as little hard pruning as possible while still getting breaks so we can fill our canopy."

This usually comes down to heading young trees 3 – 4 weeks after green tip and applying rest-breaking agents and plant-growth regulators. Training young pear trees takes longer than training young apples, and pears are notoriously slow to come into production.

"Unfortunately, BP 1 doesn't give us the option of high-density plantings or nice manipulations," says Strydom.

He is excited about the newer quince rootstocks. Having achieved excellent results with Forelle on quince, he would like to see the development of dwarfing pear rootstocks for cultivars that are incompatible with quince.

"We need to start thinking more creatively about rootstock-cultivar combinations," says Strydom. "And we need to start experimenting with planting distances so that pears can compete with alternative crops in terms of cash flow."

Upwards and onwards

Research demonstrating how the relationship between tree density and light interception affects yields goes back to the 1970s. Studies with Golden Delicious found that yields peaked at 40 tonnes p/ha in low-density plantings of 660 – 1 100 trees p/ha compared to 70 tonnes in higher-density plantings.

Nowadays, most of our industry is based on tall spindles at 2 500 trees p/ha," says Steyn, "but new orchards are being planted at even higher densities."

Spindles are trees with narrow canopies. They come in many flavours, including slender and super, but all aim to capture as much light as possible throughout the canopy. Many other systems – like planar cordons and fruiting walls – have the same goal.

Higher densities became possible thanks to the adoption of dwarfing and semi-dwarfing rootstocks such as M.7 and M.9. Rootstocks also affect precocity – how early the tree starts bearing – which is critical to the economics of modern fruit farming.

Asked what he would plant today, Dall chooses G.778. "It would be an informal central spindle, almost bearing onto solaxe, because getting the branches below horizontal is necessary, but we would be renewing those branches regularly."

While researchers abroad are turning to formal multileader systems like the planar cordon, these rely on dependable breaks for success. With our already mild winters becoming even warmer, problems can be expected.

Cook thinks a system he calls an informal fruiting wall is a better solution for warmer areas. The trees are allowed to develop multiple informal leaders with weaker fruiting wood on the branches. They fill their space without needing as much pruning as central-leader trees.

"You're not fighting the weak leader that we encounter in warmer areas," explains Cook. "And you don't need to resort to aggressive pruning that reduces your yields."

Regardless of the system, modern orchards are not cheap to establish. More trees cost more money, and trellising is just one of many standard expenses. "Many of the considerations when deciding on a system have to do with costs," says Steyn. "But the biggest driver is maximising the quantity of first-class fruit."

When Dall started in the industry in the 1960s, crops were still measured in 44 lb-lug boxes, and 1 000 boxes p/ha was regarded as a good harvest. "I remember when anything above 50 tonnes a hectare on Royal Gala was phenomenal," he says. "Now, breaking even on Royal Gala requires almost 70 tonnes a hectare."

The bar keeps lifting, but Dall is confident that our industry is up to the challenge. "Fifteen years ago, 70 tonnes would have
been unbelievable. Now it's accepted that it will be achieved. And with the new rootstocks, I think everything will be over 100 tonnes – it's exciting!"

Featured Image: A banded-fruit-weevil spray trial in the 1980s. Tree densities are noticeably higher than in preceding decades. (B.N. BARNES. THE DECIDUOUS FRUIT GROWER 1980, 30(8), P280.).