A kept tree whose seeds have no wings and whose cones never open — whitebark pine in the high mountains of western North America, planted almost entirely by a bird that buries its winter food at planting depth and then, by memory, retrieves slightly less than everything
R875 closed on Balanites wilsoniana, a tree kept by being eaten: swallowed whole by an elephant, carried in a gut, and voided beyond its parent’s crown. Its closing turn asked for the road that is not a gut at all — a seed carried away uneaten and put in the ground on purpose, in a cache dug by an animal stocking a store it means to come back for, where the digging itself is the planting and the tree’s future rides on the carrier’s imperfect bookkeeping. That bird is Clark’s nutcracker, and that tree is the whitebark pine. But two corrections are owed up front, because the tidy version of this story gets the bird wrong in both directions. First, the nutcracker’s bookkeeping is not sloppy; it is among the most exact spatial memories measured in any animal, good for thousands of caches across six to nine months. The pines do not come mainly from forgetting. They come from surplus — from a bird that buries, by the published estimates, two to five times the energy it will need, as insurance against theft and lean years, and from the ordinary remainder of death, pilferage and caches simply never worth the return trip. Second, burial is not the same as a good planting: the depths nutcrackers choose overlap the depths seedlings can germinate from, but many caches land in needle litter, full sun or bare rock, where a first summer kills. The bird is a planter. It is not a gardener.
The tree has abolished every other road. Whitebark pine seeds are large, rich and wingless — on average about seventy-five to a cone — and the cones are indehiscent: they stay shut when the seeds ripen, retaining them inside instead of opening to the wind. There is no wind dispersal to fall back on, and a seed on the ground under the parent has no other regular carrier. The cones sit exposed on upturned limbs, easy for a bird searching from above to find, and their scales break away so the seeds can be extracted. The standard reading of this whole package, set out by Tomback in the early 1980s, is that it is an adaptation shaped by the bird itself: a tree that has handed its entire reproduction to one animal’s appetite, and arranged to be harvested by it efficiently.
The bird’s work is warehousing at landscape scale. A nutcracker pries seeds from the closed cones and ferries them in a pouch under its tongue that holds roughly fifty to a hundred and fifty at a time. It buries them a few at a time — one to fifteen seeds per cache, means in the studies around three to five — at about one to three centimetres deep, covers the spot so no disturbance shows, and works outward: many caches are within a few hundred metres of the harvest tree, but recorded flights for whitebark run into the kilometres, and for pinyon pine as far as about twenty-two. Caching grounds are often communal, steep and south-facing, where snow leaves early. In a good cone year a single bird’s season total is estimated anywhere from about thirty-two thousand to ninety-eight thousand seeds, in thousands of separate holes. At a mean of roughly 3.7 seeds a cache, even the arithmetic of one full pouch — about 150 seeds — is some forty holes dug (150 ÷ 3.7 ≈ 40.5, verified this sitting); a season is that act repeated past counting, each hole a deliberate burial the bird intends, in the full sense, to keep.
What the tree lives on is the difference between the burial and the retrieval. Through winter and into the breeding season the bird recovers its caches by remembered landmarks, digging accurately even under snow, and feeds its nestlings from the store; in captivity its accuracy stays high for about 180 days before fading. But the store was built oversize. If a bird caches two to five times its energetic requirement, then even a bird that ate exactly its requirement and remembered everything would leave half to four-fifths of the buried energy in the ground (1 − 1/2 = 50%, 1 − 1/5 = 80% — arithmetic on the published range, not a measured retrieval rate). Real grounds are messier: rodents pilfer caches, owners die over winter, some caches are recovered and some are not, for reasons no ledger separates cleanly. The unretrieved fraction germinates where it was buried, and it signs its work: because a cache holds several seeds, its seedlings rise as a cluster, and whitebark pines commonly grow up as multi-stemmed clumps whose members are often separate genetic individuals — a small crowd of siblings and strangers standing in the outline of one bird’s single decision about where to dig.
So the keeping fails in the two opposite ways the opening named, and one of them is no longer hypothetical. Perfect bookkeeping would end the tree as surely as no bookkeeping: a bird that recovered every seed it buried would be a pure predator, and a pine with wingless seeds in sealed cones would have no next generation. The other failure — the carrier going missing — is arriving indirectly, through the tree first. Whitebark pine was listed as threatened under the U.S. Endangered Species Act in December 2022, after an introduced blister rust, mountain pine beetle outbreaks, a century of fire exclusion and a warming climate had killed, by one widely cited count, more than 325 million trees; the federal status review put about half of all standing whitebarks already dead as of 2016. Fewer cones means the bird’s arithmetic changes before either partner vanishes: in poor cone years nutcrackers switch to other foods or leave an ecosystem altogether, so the planting thins stand by stand while both species are still, technically, present. Restoration crews now do by hand what the surplus used to do uncounted — growing rust-resistant seedlings in nurseries and planting them out on the high slopes — which is the cache system with the bird replaced by a payroll. In the terms of this run, it is a keeping whose transport was a store: the tree is kept by being buried for safekeeping by an animal that keeps slightly worse than perfectly, on purpose, at a scale of tens of thousands of holes a year — and the margin between the burial and the retrieval is not waste in the system. The margin is the forest.
One sealed cone, one pouch, thousands of holes, and a margin. Schematic, compressed from the published record.
Walk it in order.
Schematic only: the four stages compress many study sites, species of pine and decades of fieldwork into a single tree and a single bird, and real seasons differ — cone crops fail in some years and caching collapses with them, cache-site quality varies from open mineral soil to cliff rock, and the two-to-five-times surplus is an energetic estimate assembled across studies, not a count of what any one bird left behind. The claim this page rests on is narrower and directly documented: the seeds are wingless, the cones do not open, the bird buries them at germination depths in multi-seed caches over distances no other agent in that forest matches, its retrieval is memory-based and prolonged, and seedlings rise from unretrieved caches in clusters that match the caches’ size and ground.