A kept seed that is never cargo at all — sealed inside its own parent’s stem by a flower that never opened, carried because the stem is stiff enough to build with, and planted only when the nest rots down around the stem it was built from
R885 closed on mud: earth a bird chose as material, with a seed bank hidden inside it by accident of geography. Its closing turn asked for the tighter case — the cargo and the wall made of the same plant. It exists, and it turns out to be the case the very study behind R885 was actually built around. In the mud, the seed was a stranger to its container. Here the container is the seed’s own body’s parent: a dead stem, gathered for its stiffness, with its seeds still sealed inside it — put there by flowers that never opened, never met a pollinator, and never shed anything.
Sealed before it was ever a seed. Many plants run two mating systems at once. Their open, chasmogamous flowers cross-pollinate and drop their seeds in the ordinary way. Their cleistogamous flowers never open at all: they fertilise themselves inside the closed bud, cheaply, with no petals, no nectar and no visitor. In many such plants — grasses above all — the selfed seeds do not fall. They stay inside the stem that made them. Japanese stiltgrass, the species the nest researchers targeted, shows the split plainly: its outcrossed flowers sit in racemes at the stem tips, while its cleistogamous flowers are tucked along the stem, concealed inside the upper leaf sheaths. One plant, two kinds of seed — the mixed ones set loose at the top, the selfed ones locked in the walls of the house they were born in.
The stem outlives the plant. Stiltgrass is an annual, dead by winter — but its stems are in no hurry to follow. Dead stems of cleistogamous grasses persist standing through the winter and into spring, still holding viable seed, and that persistence is exactly what puts them on a bird’s shopping list. Nest-building birds in spring gather precisely this kind of stubborn, dry, structural material. The bird is choosing a stem the way the swallow chose mud: for what it will do in a wall. But where the swallow’s mud merely contained a seed bank it happened to be scooped from, the stem is the seed’s first and only container. The seed has never been a free object. It was made inside the vehicle, and it will ride inside the vehicle, and nothing in its life so far has ever separated the two.
Counted still attached. The nest-box half of the germination study lets this be counted rather than asserted. Nine nests built inside experimental boxes were dissected seed by seed: about 211 seeds per nest on average, and significantly more of them — about three in five — were still attached to the plant material they arrived in than were loose in the nest or present in droppings. Across all 87 nests, open mating was the commoner system by species, yet closed-mating plants supplied the majority of seedlings by sheer abundance, and among the handful of species that turned up most often in nests, most were cleistogamous. Stiltgrass itself appeared in nests sitting between about one and a hundred metres from the nearest known patch of it — for a plant otherwise reckoned a stay-at-home, moved by gravity and water, with no equipment for wind or animal travel at all.
The homebody out-travels the traveller. Here is the inversion this page exists for. The standing assumption about sealed, selfed seeds was that they are the plant’s stay-at-home reserve: kept in the stem precisely so they remain beside the parent, in ground already proven, while the outcrossed seeds do the exploring. The nest record suggests the gradient can run backwards. The outcrossed seed drops from its spikelet at the parent’s feet. The sealed seed, never dropped, rides inside a harvested stem to a nest a hundred metres off — nearly the full span of a small bird’s material-gathering range. Less mixed, less adventurous by every intention, and the one that actually leaves. And its release is R885’s release with the delay doubled: the nest must fail and the stem must rot, the seed set free at last into the litter of the structure built out of its own parent — planted, quite literally, in the remains of the house and the body that carried it.
Why this keeper is different from the others in the run. The bur was picked up by contact, the wage was paid and discarded, the mud was scooped and cemented. Every one of those seeds existed on its own before its keeper took it. This seed never did. Keeper and kept are one organism’s two generations: the dead parent is the vehicle, the sealed child is the passenger, the bird is only the haulage — recruited, as with the mud, entirely by the material’s fitness for a wall. The keeping is not a container found around a seed. It is a seed that was never let out of its maker.
One stem: sealed in, kept standing, built with, rotted free. Schematic, compressed from the published record.
Walk it in order.
Schematic only: the four stages compress a mating-system literature and one nest study into a single stem’s story, and no study followed one sealed seed from a named stem into a named nest and out as a seedling. The dissected nests in boxes were built by a small set of cavity-nesting species, not by the full 23-species sample. Labelled arithmetic: three in five of about 211 seeds is roughly 128 still-attached seeds per box nest — an illustration from the study’s own figures, not a load any given nest carries. The hundred-metre stiltgrass figure rests on the few nests that germinated it, and the claim that sealed seeds may out-travel open ones is offered by the authors as a testable possibility, not a result. The claim this page rests on is narrower and directly documented: in many cleistogamous grasses the selfed seeds remain inside the parent stem, dead stems persist through winter still holding viable seed, birds gather such stems as nest material, most seeds counted in dissected nests were still attached to that material, and closed-mating plants supplied the majority of what germinated from the nests.