Why Fairfax County Plants Its Forests in Clumps

A forest planted in rows

For about fifty years, the standard way to replant a cleared site in the Mid-Atlantic has been the plantation: a short list of hardwood species set in rows ten to twelve feet apart, each sapling inside its own plastic tube to keep the deer off. It is fast, and a crew can put thousands of stems in the ground in a day. Five years later, many of those sites still look like what they are, a tree farm with grass between the trunks, thin shade, and almost no shrub layer.

Fairfax County's stormwater planners had a particular reason to want something better. Stream restoration in the county's parks often means taking out the trees along a bank to rebuild the channel, and the forest that grows back decides how long that stream goes without shade, leaf litter and fallen wood. In 2017 and 2018 they replanted four parks a different way.

What a superclump is

An old field in Virginia does not turn to forest evenly. A bird perches on a lone red cedar and drops the seeds it carried in; shade under the cedar knocks back the grass; seedlings take hold; the patch widens year by year until patches meet. Ecologists call this nucleation. Restoration biologists copy it on purpose and call it applied nucleation.

Fairfax County's version is the superclump. Each one is an oval about 20 feet across and up to 40 feet long, packed with roughly 65 native plants of 12 to 16 species, set two and a half to four feet apart. The planters fill four groups: shrubs such as blackhaw viburnum and ninebark; understory trees such as serviceberry and pawpaw; fast-growing canopy trees such as black locust, sycamore and river birch; and slow climax trees such as white oak, swamp white oak and black gum. Plants deer avoid go toward the outside, the thorny black locust and the strong-scented pawpaw, and a single four-foot wire fence circles the whole clump. Hardwood mulch covers the ground. The open land between clumps stays in meadow or mown grass, waiting for the clumps to spread into it.

A plantation puts one or two trees and up to four shrubs in the same 400 square feet that a superclump fills with 65 plants, about eleven times as many stems.

The claim

On cleared ground in the Virginia Piedmont, planting a site as dense, many-species clumps rebuilds the working parts of a forest (canopy, leaf litter and soil) faster than planting the same site in evenly spaced rows.

The evidence, five years in

In the fifth year after planting, county scientists and their partners laid out 161 research plots across eight Fairfax County reforestations and compared four conditions side by side: superclumps, plantation rows, ground left unplanted, and the nearest standing forest.

Three measures separated those conditions most cleanly, and all three are cheap to take: tree canopy cover, soil organic matter, and the share of ground covered by leaf litter. The superclumps reached 92 percent canopy cover, 6 percent soil organic matter and 65 percent leaf-litter cover, matching or beating the reference forest on the same sites. The rows and the unplanted ground trailed on all three.

The soil itself loosened. Soil bulk density, the mass of a set volume of dry soil, averaged 1.10 grams per cubic centimeter inside the clumps, close to the 1.08 of the standing forest, while the rows and the unplanted ground both sat at 1.20. Looser soil lets roots push through and lets rain soak in instead of running off toward the stream.

Why the clumps outran the rows

A superclump heals ground the way an old field does, on a faster schedule. Packing 65 plants into 400 square feet closes the canopy within a few seasons. Shade knocks back the grass and the sun-loving invasives that crowd a new planting, and the leaves that fall each autumn stay put under the mulch, rot, and build the soil's organic matter. Roots and soil life working through that litter loosen the ground; inside the clumps, bulk density moved toward the forest's number while the rows stayed packed.

Deer are part of the story. Northern Virginia's white-tailed deer browse young trees hard enough to stall a planting. The superclump answers in layers: stems too close together for a deer to walk through, thorny or strong-scented plants on the edge, and one four-foot fence around the whole clump in place of a plastic tube on every stem. Crowding helped beyond the deer. At Rock Hill, where rows and clumps were both fenced, sycamore survived 92 percent of the time in the clumps and 10 percent of the time in the rows.

The clumps also borrow from the forest next door. Fairfax usually plants them after stream restoration, beside the tree saves and buffers the crews left standing, so birds carry seed in from mature trees close by. A 2012 review of the method by Jeffrey Corbin and Karen Holl found that seed rain falls off sharply within 5 to 10 meters of a forest edge, and that larger patches near a seed source draw more new seedlings than small, isolated ones. In Costa Rica, Holl's team ran a test for 15 years at 15 sites and found that tree islands brought back most measures of forest recovery as well as full plantations did, at lower cost.

The method has limits. After five years, unplanted ground right beside a clump still looked like the rows and the bare plots, so the forest had not yet spread into the gaps; Fairfax now recommends 10 to 14 clumps per acre instead of the 5 to 8 it first planted. Where no forest stands nearby to supply seed, or where land has gone decades without trees, an even planting may serve better. Five years does not make a mature forest. It does put the working parts of one in place sooner, and in a county that clears streamside trees to repair its streams, sooner matters.