City Verra

Street trees: choosing them, planting them, and keeping them alive between the curb and the sidewalk.

Roadway Tree lawn Sidewalk TICO Tilia cordata30 ft crown GYDI Gymnocladus dioicus40 ft crown ULAM Ulmus americana55 ft crown CEOC Celtis occidentalis35 ft crown 0 10 20 ft Roadway Sidewalk TICO Tilia cordata30 ft crown GYDI Gymnocladus dioicus40 ft crown ULAM Ulmus americana55 ft crown CEOC Celtis occidentalis35 ft crown 0 10 20 ft
The 400 block, surveyed in September. Crowns drawn to measured spread: TICO little-leaf linden, 30 ft; GYDI Kentucky coffeetree, 40 ft; ULAM American elm, 55 ft; CEOC hackberry, 35 ft.

Planting,

Why street trees die in their third summer, and the watering schedule that prevents it

Most failed plantings look fine for two seasons and then collapse in a dry August. The cause is almost always a root ball that never grew into the surrounding soil, and the fix is a weekly watering routine that lasts longer than most programs budget for.

by Marguerite Osei, 9 minutes

When a street tree dies in its first summer, everyone notices and someone usually gets a phone call. When it dies in its third, it tends to go quietly: a thinning crown in July, early color in August, and a dead leader the following spring. By then the planting contractor’s one-year warranty has lapsed, the watering crew has moved on to newer blocks, and the tree gets logged in the inventory as ‘declining, cause unknown.’ On the 412 trees our crew planted along the Milwaukee Avenue corridor in 2022 and 2023, almost all of our losses so far have fallen in that third growing season, and when we dug them up the root systems told the same story every time.

A nursery tree arrives with a root ball that is a small fraction of the root system it had in the field. ANSI Z60.1 sizes balled-and-burlapped stock at roughly ten to twelve inches of ball diameter per inch of trunk caliper, so a 2-inch linden comes with a ball about 24 inches across, and the digging has cut away most of its absorbing roots. For the next several years the tree lives almost entirely on the water held in that ball. The ball is often a heavier clay or field loam than the sandy, compacted fill in a typical tree lawn, and the textural difference creates an interface water does not cross easily. You can soak the surrounding pit and the ball itself will still be dry.

The working rule in temperate cities is that establishment takes about one growing season per inch of caliper, and longer in hot, paved sites. That means a 2-inch tree needs dependable water through at least its second full summer, and a 3-inch tree through its third. Most municipal programs fund one or two seasons. The third-summer collapse is what happens when watering stops before new roots have pushed far enough past the original ball to find water on their own. In the trees we excavated, new roots had spread 6 to 10 inches beyond the ball, about a third of the distance we would expect in a well-established tree.

Our schedule now is simple and dull, which is the point. Each tree gets 15 to 20 gallons once a week from May through September for three full seasons, applied slowly enough to soak the ball rather than run off it. For crews this means slow-release bags of the 20-gallon kind, filled on a fixed weekday route and pulled in October so they do not trap moisture against the bark over winter. For block stewards with a hose, it means a trickle at the base of the trunk for 30 to 40 minutes, not a five-minute spray across the mulch. In a week with an inch or more of rain, we skip. In a heat wave above 90°F for more than three days, we add a second fill.

We check rather than guess. Twice a season, a crew member pushes a soil probe or a long screwdriver into the root ball itself, about six inches from the trunk, and pulls a core with a soil auger on any tree that looks off. The ball should feel like a wrung-out sponge. Bone dry means the bag is draining down the side of the ball, which usually happens because the tree was planted with the ball sitting higher on one side or a mulch ring was piled against the trunk and sheds water outward. Saturated and sour-smelling means the pit has no drainage, a common problem in tree lawns over old building rubble, and that needs a different repair than more water.

None of this is expensive per tree, but it adds up across a program. At our crew rates, a third season of weekly watering costs about $85 per tree, against roughly $1,100 to remove a dead tree, grind the stump, and replant. The harder cost is administrative: the third-year trees are scattered across older planting areas, so they fall off routes built around this year’s work. We solved that by tagging every new tree’s inventory record with its planting date and generating the watering route from records younger than 36 months, rather than from the current year’s planting list.

If you manage a planting program, the most useful single change is probably to move the watering commitment from ‘first two years’ to ‘until caliper in inches equals seasons watered,’ and to write that into the budget at planting time instead of finding the money later. If you are a neighbor with a young tree outside your door, find out when it went in. If it is in its second or third summer and looks healthy, it still needs the hose.

The ledger

Planting

  • Reading a nursery’s caliper tag before you sign for the truck

    How to check trunk caliper, root ball size, and girdling roots at delivery, and which defects are reason enough to send a tree back.

    by Dev Raman5 min

Soil & Pits

Species

  • Retiring the Bradford pear: replacement species for narrow tree lawns

    Five medium trees that tolerate a four-foot strip, overhead wires, and road salt without the Callery pear’s split crotches and invasive seedlings.

    by Hannah Kowalczyk7 min

  • Hackberry, Kentucky coffeetree, and Japanese pagoda tree under salt spray

    Three years of leaf-scorch notes from a plow route show which of these salt-tolerant species actually hold up within ten feet of a treated arterial.

    by Hannah Kowalczyk9 min

Pruning

Inventory

  • What a block-by-block inventory caught that a sample survey missed

    A complete street-tree count in one ward turned up 340 empty pits, a cluster of ash we thought were gone, and a planting backlog nobody had budgeted for.

    by Ruth Adeyemi8 min

Stewardship

  • Mulch volcanoes, and how to talk a neighbor out of building one

    Mulch piled against the trunk invites rot and girdling roots. A three-inch ring pulled back from the bark is better, and here is a polite way to say so.

    by Sam Feldkamp4 min

Field notes

Soil & Pits

Silva cells or structural soil: what each costs per tree over twenty years

Suspended pavement systems cost more up front, but the comparison changes once you count rooting volume per dollar and the trees you do not have to replace.

by Tomás Arrieta, , 11 minutes

Both approaches solve the same problem: a tree under pavement needs uncompacted soil, and a sidewalk needs a load-bearing base. Structural soil, typically the Cornell CU-Soil mix of crushed angular stone with about 20 percent loam by weight and a hydrogel binder, does both jobs in one material. Modular suspended pavement, such as Silva Cells or similar crate systems, carries the sidewalk on a plastic frame and fills the voids with ordinary, uncompacted planting soil.

The difference is in how much of that volume a root can actually use. In structural soil only the loam between the stones holds water and nutrients, so a cubic yard of CU-Soil offers roughly a fifth of the usable rooting medium of a cubic yard of loose loam. In a suspended system nearly all of the volume inside the frame is soil. On our last two downtown projects, suspended pavement came in at about $28 per cubic foot of soil installed against $11 for structural soil. Per cubic foot of usable soil, though, the suspended system was cheaper.

Over twenty years the bigger number is survival. We specify about 1,000 cubic feet of soil for a medium-canopy tree under pavement. Structural soil at that volume is achievable but rarely budgeted, so most projects install a third of it and get trees that stall at 6 to 8 inches in diameter. If your site can only afford a small volume, use suspended pavement for that volume. If you can excavate a large continuous trench, structural soil is the economical choice.

Pruning

Raising crowns over bike lanes without topping the tree

Protected bike lanes need about 10 feet of clearance. Here is how to get there over five years of structural pruning instead of one bad cut.

by Marguerite Osei, , 6 minutes

When a city converts a parking lane into a protected bike lane, the trees that used to overhang parked cars now overhang cyclists, and the request comes in to ‘clear them up’ before the lane opens. Most departments design for 8 feet of vertical clearance over sidewalks and 10 to 14 feet over the travel way, depending on what vehicles use it. Street sweepers and the small plows that clear separated bike lanes generally need the upper end of that range.

The temptation is to reach the target in one visit by taking the lowest scaffold limbs back to the trunk. On a young tree that can remove a third of the live crown at once, which goes well past the 25 percent most arboricultural guidance treats as a ceiling. It also leaves large wounds low on the trunk that compartmentalize poorly. The better approach is to identify the permanent lowest scaffold limb now, at or above the clearance line, and then shorten the temporary branches beneath it with reduction cuts each cycle rather than removing them.

Those temporary branches keep feeding the trunk and adding taper while the crown develops above them. On a three-year pruning cycle we usually take them off completely on the second or third visit, when they are under two inches in diameter and the wounds close within a couple of seasons. The bike lane gets its clearance a few years later than the opening date, so in the meantime we mark the low limbs with reflective tape and tell the transportation team the timeline up front.