At 5:47 on the morning of May 12, 2026, a chestnut-sided warbler was singing from the top of a staghorn sumac that had pushed itself through a fracture in the concrete pad of the old LTV Coke Works in Hazelwood, on the south bank of the Monongahela River in Pittsburgh.
The bird had been singing for eleven minutes. Edith Crale, who had driven down from Ithaca the day before, marked the song on a Cornell Lab data sheet she was filling in pencil. She wrote: 5:36 first song, persistent, full phrase.
The pad is roughly the size of a city block. It was poured in 1947 and last used for coke ovens in 1998. Since then, the site has been the subject of four redevelopment proposals, all stalled, and one quietly successful eight-year effort by the volunteer group Three Rivers Birding Club to record what came back.
By 6:10 a yellow warbler joined from the willow scrub at the river edge, then a common yellowthroat from a wet patch where the slab had cracked enough to hold runoff. By 6:40 the count for the brownfield was nine warbler species across about four hectares of unintended habitat.
This is not the way a birder is taught to think about productive land. The places that produce warblers in the Allegheny region are supposed to be deciduous canyons and the protected uplands of state forests. Brownfields are not even supposed to count.
And yet the Pittsburgh data, gathered patiently across two decades by volunteers and now formally analyzed by a small team at Carnegie Mellon's Department of Biological Sciences, shows that several of the city's post-industrial sites carry warbler species counts during May migration that rival or exceed those of nearby protected woodland.
The leading hypothesis is structural. Brownfields, when they are left alone, tend to develop a mosaic that mature forest does not have: open ground, low forb cover, mid-height shrub layer in the form of sumac and dogwood and Russian olive, and an emerging canopy of black locust and tree of heaven. Warblers, particularly during migration, want exactly that gradient.
Hazelwood is not unique. Crale's working list of productive Pittsburgh brownfields includes Nine Mile Run before its restoration, the upper reach of Saw Mill Run, a former rail yard near the Strip District, and two parcels along the Allegheny in Aspinwall where the bank stabilization has been allowed to revert to willow.
The complication is that the brownfields are also, by definition, contaminated. The Hazelwood pad sits on soils with elevated arsenic, lead, and polycyclic aromatic hydrocarbons. The state has classified it as Act 2 cleanup status, and there is no public access without a permit.
Whether this matters to the warblers is an open question. The chestnut-sided is here for ten days at most, eating caterpillars off the early sumac leaves and the willow catkins. The yellow warbler will stay to breed if the willow scrub holds. The common yellowthroat may already be on territory.
What the Carnegie Mellon team is studying now, with a small grant from the Heinz Endowments, is whether the breeding birds carry detectable heavy-metal loads in feather samples collected non-lethally at banding. Preliminary results from 2025 suggested elevated lead in two of seventeen yellow warbler samples, but the sample size is not yet useful.
The redevelopment question is the loud one. The Almono partnership, which owns most of the former LTV site, has been working since 2002 toward a mixed-use plan called Hazelwood Green. Construction has been slow but real. A robotics building opened in 2023. A pedestrian bridge to the river has been graded.
The plan, in its current form, preserves about thirty percent of the parcel as open space, including a continuous riparian strip along the Monongahela. The brownfield pad where the chestnut-sided was singing is scheduled, in the long-range concept, to become a research campus.
It is possible to make a case for both outcomes. The neighborhood wants jobs. Pittsburgh needs tax base. The pad cannot stay a pad forever. And the warblers, in the migration biology that matters here, are not depending on this one site so much as on a network of such sites distributed across the eastern flyway.
What can be lost is the structural mosaic, the staggered shrub-to-young-tree gradient that a mature park lawn will not produce and that mature forest does not produce either. That mosaic is what Pittsburgh has been growing, accidentally, on its industrial scars for forty years.
Pell Murphy, who joined Crale at Hazelwood on the morning of May 12, made the case in a 2024 piece for this magazine that the city should explicitly designate a portion of its brownfield acreage as managed early-successional habitat. He proposed three pilot sites. None have been adopted.
By 7:30 the warbler activity at the Hazelwood pad had quieted. A Baltimore oriole was working the locust at the parcel's eastern edge. A gray catbird, which is not a warbler but acts like one in May, was scolding from a dogwood thicket. The sun was on the slab.
Crale wrote in the margin of her sheet: nine warbler species, four hectares, ninety minutes. She underlined the four hectares.
On the drive back across the Hot Metal Bridge she said, mostly to herself, that the difficulty with brownfield birding is that it asks the field to revise its own categories. The places that produce the species are not the places the field has trained itself to value.
The chestnut-sided was almost certainly gone by the next morning. The willow scrub will still be there in June, and the yellow warblers in it. What happens to the slab will be decided by people who are not, on the whole, listening at 5:36 a.m.
