Post-fire recovery may start below the ash. Three years in, the number is missing.

The testable version of the belowground argument is narrower than the framing suggests, and it is worth stating precisely. Soil microbiomes recycle nutrients and support plant health, and by Jennifer Bhatnagar's account they help set how a forest responds to environmental change ^[6]. If that holds on a burn scar, then a hillside where bay and maple are coming back while oaks and redwoods are not ^[8] is reporting on what survived beneath the ash as much as on what anyone puts in the ground above it. Read that way, a program that counts seedlings planted is measuring its own effort rather than the site's capacity to keep them alive.

Nothing in this account measures either. There are no microbiome results, no seedling survival rates and no costs ^[3]. What there is: a first field season in the summer of 2023 ^[1] inside a program described as three years old ^[5], set against a question about what restoration will do for these landscapes decades from now ^[10]. Three post-fire growing seasons is a thin base for a decadal claim ^[2], and thinner still for the comparison a land manager actually needs, which is soil-treated planting against untreated planting on the same slope.

Scale compounds it. The Napa holding is 18.5 acres ^[4], about 7.5 hectares ^[1]. That is enough ground for a well-instrumented experiment and not enough for a landscape inference.

The species list is not the lab's own either. Residents wanted native oaks and redwoods rather than pines or firs, and favoured species that were fire-resistant and culturally meaningful; their priorities did not always match conventional restoration practice ^[12]. Bhatnagar says that forced a redesign, with theories of forest recovery tested using the species the community wanted ^[13]. Good practice for the neighbours, and a real limit on transfer: an agency replanting conifers will not find its trees in this test.

The provenance of the question deserves a plain sentence too. Her subject is microbes rather than trees ^[2], and wildfire had not been her research focus until these landowners got in touch ^[7]; the opening inquiry was whether truffles could be cultivated on land they owned ^[3]. Private money choosing which post-fire question gets three years of a soil lab is not a scandal, but it does mean the evidence arrives shaped by one property and the preferences attached to it.

What would move a restoration budget is not the mechanism, which is plausible enough, but a survival differential tied to a soil treatment and a price per surviving tree. Until that exists, the exportable finding from Napa is the cheap one: species return unevenly across a single burn ^[8], which a manager can record without a sequencer, and which says where to look before ordering stock.