For 21 years now, Dr. Ian Merwin has tended a 320-tree apple orchard on the eastern shore of Cayuga Lake near Cornell ­University’s Ithaca campus. He’s been studying the long-term effects of four different orchard floor management systems.

A key question is, are there better arrangements than trees in herbicide-treated strips separated by alleys of grass, the system used by most apple growers?

Merwin studies the effects on tree growth and yield, runoff and leaching of nutrients, and soil quality.

The results have been surprising. In a nutshell, you can grow apples almost equally well over the long term in all four systems. But a simple postemergence herbicide program—two applications of glyphosate each year in May and July—generated the highest fruit yield, providing good orchard floor conditions for working and, surprisingly, good protection of the soil from erosion and runoff.

The system that turned out to be the most different—and interesting—was one based on deep applications of bark mulch to keep tree strips weed-free.

Merwin reviewed the results of this long-term study with growers during Southwest Michigan Horticulture Days in Benton Harbor, making a mixed-media presentation using PowerPoint and a speaker phone linked to his office in Ithaca, New York.

The study began in 1992 when Royal Empire apples were planted on M.9/M.111 rootstock on a 10- by 20-foot spacing and trained to a vertical axe. Microsprinkler irrigation was used as needed. The soil is a heavy, silty clay loam, and high in organic matter (4 percent).

Drainage tile was installed under the orchard so that both subsurface and surface runoff water could be collected and analyzed.

Four systems

Merwin has studied four orchard floor management systems:

—Mowed red fescue turfgrass with no herbicides
—Hardwood bark mulch, renewed every two years as it decomposed during the first ten years, and then every four years to the present.  For the first five years, the mulch suppressed weeds, but over time, a glyphosate spray was necessary each May to control perennial weeds that became established in the mulch.
—Glyphosate applied in May and July each year (the postemergence program)
—Karmex, Solicam, and glyphosate herbicides applied in May each year (the preemergence program)

Measurements are taken to see what effects the systems have on soil physical properties, tree physiology and yields, agrichemical runoff and leaching, and nutrient availability and recycling.

DNA/RNA fingerprint tests are conducted on the fungi and bacteria in the soil to determine what is happening within the rhizosphere (the area directly around the roots) under the different management systems. Recently, studies have been done to determine potential apple replant disease problems.

Mowed grass

While vegetation is strongly competitive in peach orchards and in young apple orchards, the mowed grass in this irrigated and fertilized apple orchard only slightly ­suppressed yield over the long term.

“The trees were smaller and less productive early, but they adapted to the ­competition,” Merwin said.

Over two decades of observations, yields in the mowed-grass treatment were lower than from the glyphosate- or bark mulch-treated soil, but about equal to those from trees in bare strips treated with preemergence herbicides.

Glyphosate strips

After years of twice-annual treatment with glyphosate, Merwin said, mosses grew on the plots, creating a protective surface blanket that is almost ideal as a ground cover. It protects the soil from erosion, but does not compete with the trees.

While the surface tends to be weedy late in the season and during winter, those weeds are like a free cover crop. They are not competitive, and have no significant impact, he said.


After 15 years had elapsed, the striking result was that the treatment differences became statistically less significant over time. The soil in the preemergence herbicide strips had less phosphorus, less potassium, less manganese, less organic matter, and less cation exchange capacity, but differences were not statistically significant.

The preemergence herbicide-treated plots were the most vegetation-free of all the treatments.

In the drainage outflow, more nitrate leached from the bare soil than in any of the other systems.

Bark mulch

Trees in the bark mulch strip treatment became very different as the experiment developed. “After ten years, the trees got larger but not more productive of fruit. The fertility level of the soil rose—maybe too much for apples,” Merwin said, adding that it might be an option for growers working on soils with low fertility.

That kind of mulch—made of an abundant local resource from hardwood sawmills in northern New York State—created a truly different soil, which contained two to three times the levels of phosphorus and twice the calcium of the others, and almost twice the organic matter. Organic matter content rose to 8.6 percent. The pH was higher—7.2 compared to 6.4 in the other plots—and the cation exchange capacity was 25 percent higher.

“Even though the bark mulch was not mixed into the soil, the decomposed mulch at the soil interface was gradually incorporated into the soil and doubled soil organic matter content over the years of treatments,” ­Merwin said.

The preemergence herbicide-treated plots had lower levels of bacteria in the soil, the other three not being significantly different.  When levels of fungi in the soils were measured, the grass treatment had higher levels than the others.

The bark-mulched soil, in respiration tests, emitted significantly more carbon dioxide, a measure showing elevated microbial activity in the soil.

As the experiments go on, Merwin becomes more fascinated with the bark mulch system. It is more expensive to maintain, but the soil becomes more fertile. That can be an advantage, but also a disadvantage if growers need to control tree vigor.

“In Europe, they pay growers to sequester carbon,” he said. “If they ever do that here, bark mulch is the way to do it.”