Carried for a wage

A kept seed whose carrier is recruited on purpose — a fat body grown on its own coat, carried home by an ant, stripped and eaten there, and the seed itself discarded intact, underground, in the nest’s waste ground

R881 closed on the dung beetle: recruited by accident, by a smell nobody meant to send it, burying a seed it never ate for a nursery of its own. Its closing turn asked for the opposite hiring — a carrier recruited on purpose, by a lure the seed itself carries, and a keeping that is not an accident but a wage the plant evolved to pay. That is myrmecochory, seed dispersal by ants, and the wage is the elaiosome: a small fleshy appendage on the seed, rich in lipids and protein. Two corrections first, because this page insists on them. The wage buys a very short ride, not a journey: in the northern hemisphere the average ant carry is only a little over a metre. And the wage is sometimes taken without the work being done at all — an ant that eats the elaiosome where it found it, and leaves the seed where it fell, has been paid and has not carried. The mutualism is real, widespread, and routinely cheated, on both sides.

What the wage is, and why it works as a handle. An elaiosome is not simply food tied to a seed. Its fatty chemistry resembles insect prey more than seed tissue — particular lipids in it are the sort ants associate with a dead insect — so a foraging ant treats the whole diaspore as a prey item to be seized and carried home, often gripping it by the elaiosome itself. The seed, in effect, has grown its own handle, and scented it as food. Inside the nest the elaiosome is fed to larvae or eaten by workers; the seed underneath it is too hard, or too unrewarding, to be worth opening in the cases that make this a dispersal system rather than predation, and it comes through intact. Then housekeeping does the planting: the spent seed is taken to an underground midden or ejected from the nest, into waste ground enriched with the colony’s refuse. The plant did not choose that ground. It bought a carrier whose ordinary domestic routine ends there, which is close enough to choosing it, over evolutionary time.

How far, and how many. Ant dispersal is estimated for roughly eleven thousand plant species worldwide, a striking convergence for so specific a trick. It is not, in most places, a long-distance system. A synthesis of northern-hemisphere records put the mean carry at about 1.16 metres, with a maximum of about 70 metres; other summaries put the most common carries between about half a metre and a metre and a half. For a tall tree that would be nothing. For the small herbs and shrubs that mostly use it, a metre is several body-lengths away from the parent, out of the densest shade and sibling crowding, and — crucially — underground or in a midden within hours, instead of on the surface where seed predators work. Longer Australian records exist and are kept apart on this page: they belong to different ants, different country, and they do not lift the northern mean.

What the metre buys. A major review of the evidence sorted the proposed advantages into three: escape from seed predators, dispersal for distance, and directed dispersal to favourable sites. In the studies that tested them, predator avoidance and distance were supported in most cases; directed dispersal was supported in about half. Most studies that tested more than one advantage found more than one, so the page does not pick a single reason. In mesic forest, the predator pressure is the vivid part: work summarised in that literature found predators taking a large share of uncollected seeds within days. An ant that removes a seed quickly, even by a metre, and puts it underground has changed its odds completely, out of proportion to the distance. In fire-prone country the same burial can matter for a different reason — depth is what a seed needs to survive a burn and germinate after it. The benefit, in other words, is context, not the ride itself.

Where it fails: the wage taken at the doorstep. This is the failure R881’s opening named in advance. In a Central European field study of three early-flowering plants — snowdrop, hollow root and European wild ginger, worked by a small acorn ant and a red ant — removal was low: about 10%, 26% and 34% of offered diaspores respectively. Many ants were seen eating the elaiosome on the spot, without carrying the seed anywhere. The authors judged that interaction, in that place with those ants, not clearly mutualistic at all: the plants paid and obtained no apparent benefit. A survey of Iberian myrmecochores makes the matching point from the other side: dozens of ant species interacted with the seeds, but only a small minority dispersed them to any significant degree; the rest were predators, or cheaters feeding in place. Plants cheat too, with rewards that are meagre, hard to remove, or chemically promised and not actually present. A wage, once evolved, invites everyone to find a way to pocket it for less work, and the plant to find a way to pay less for the same carry. That arms race, not a harmonious partnership, is the honest picture.

Why this keeper is different from the others in the run. The mallard kept by digesting incompletely; the predator kept by killing the first keeper; the beetle kept by accident, hired by a smell it was never sent. The ant is the first keeper in this sequence that the plant hired deliberately, with a structure grown for the hiring, priced in fat and protein, and shaped so the carrier picks the seed up by the payment itself. And its keeping has a shape none of the earlier ones had: the payment is consumed completely and the kept thing is not consumed at all. The wage is spent to nothing in the nest, and exactly because it is gone, the seed is thrown out with the rubbish — discarded, in enriched waste, in the dark, at shallow depth, which for many of these plants turns out to be the best ground within reach. Kept, in the end, by being thrown away by a carrier who was paid to do the carrying and never knew it had done the planting.

One seed, one wage, one short carry. Schematic, compressed from the published record.

Walk it in order.

Schematic only: the four stages compress separate records — a global estimate of how many species use ants, a northern-hemisphere synthesis of carry distances, a review of the advantages, and Central European and Iberian field work on cheating — into a single seed’s story, and no one study followed a seed from a parent plant through an ant’s nest to a seedling. The arithmetic illustration in stage two is labelled arithmetic, not a measurement: 26% of 100 offered hollow-root diaspores is 26 removed, computed from the published removal rate for that species in that study, where rates for the other two species were lower and higher respectively. The claim this page rests on is narrower and directly documented: many plants grow a lipid-rich appendage that induces ants to carry the seed to the nest and discard it intact after eating only the appendage; the typical carry is short, about a metre in northern records; and the arrangement fails, measurably, wherever ants eat the appendage in place instead of carrying.