Keeping paused inside the shell — an annual killifish embryo waiting in the cracked mud of a vanished pool, both parents long dead, until rain remakes its world and only then lets it finish
R928 closed with a hatchling whose keeping was finished in advance: a loggerhead mother placed a clutch in warm sand, left for years, and the sand incubated and sexed it unattended for roughly two months. This page takes the turn it offered and stretches that unattended interval past a season. Here the egg does not hatch on any schedule the parents could have planned around. It is buried in the bottom of a pool that dries up and ceases to exist, its development is suspended partway through, and it waits — not for a keeper to return, because there is no keeper left alive — but for the habitat itself to return.
A pool with an expiry date. The turquoise killifish, Nothobranchius furzeri, is a small annual killifish of ephemeral pools in semi-arid Zimbabwe and Mozambique, filled by seasonal rains and gone again in the dry season. Adults live compressed lives: in monitored wild populations most disappeared within roughly three to ten weeks as their pools dried, and only a minority of populations lasted much longer; captive lifespans are likewise counted in months. While water remains, mature fish spawn repeatedly into the mud — one wild estimate put a female’s output in the range of a few dozen to over a hundred eggs a day, varying more between populations than within them. Then the pool desiccates and every adult dies. There is no migration out and no survivor to tend what was laid. The population that will exist next year is, at that moment, entirely eggs in dry sediment.
Three places to stop. Annual killifish embryos do not simply slow down; the literature describes up to three distinct arrests, called diapause I, II and III, and species differ in which they use. The first falls early, at a dispersed-cell stage peculiar to these fish: instead of staying in an organised sheet, the early cells scatter across the yolk and only later re-aggregate to form the embryo’s axis — work on the turquoise killifish reports that this axis is self-organised by the embryonic cells rather than laid down in advance by the mother. The second arrest, diapause II, falls roughly midway through development, at a stage when many organs are already formed: summaries of the work describe a developing brain and a heart that stops beating for the duration and starts again afterwards. The third is not a pause in development at all but a pause after it: a fully formed embryo, finished inside the shell, waiting for inundation to cue hatching. An egg in the mud may therefore be unfinished, half-finished, or finished and merely unhatched — three very different kinds of waiting filed under one dry pool bed.
How long the mud keeps it. The dry season in the turquoise killifish’s range is reported at roughly five to eleven months, and that is the interval a typical egg bank must bridge. Secondary summaries of the diapause research put the average pause at around eight months, with laboratory embryos surviving up to about two years — a span longer than the whole adult life of the fish that laid them, several times over. Crucially, one well-known laboratory finding is that the pause appears not to be spent life: embryos that had been held in diapause for longer than a normal adult lifespan went on, in that work, to grow, reproduce and live as long as fish that had never paused. The keeping is not a slow leak. Time in the mud, on that evidence, does not count against the fish in the way time anywhere else does. A close relative used in the dormancy literature, the Venezuelan annual killifish Austrofundulus limnaeus — named here as that species, not merged into the turquoise killifish’s record — carries the same arrangement to an extreme of oxygen tolerance: its diapause II embryos are reported as the most anoxia-tolerant vertebrate embryos known, surviving without oxygen for periods measured in weeks to months in the experiments, because a buried egg in drying, warming mud cannot count on air either.
Not all of them believe the first rain. When rain finally refills the pool, hatching is not unanimous. Accounts of annual killifish describe a spread of responses: many fully developed embryos hatch on inundation, but a share hold back through the first wetting, and some eggs deposited in one dry season remain dormant through the following rainy season and hatch a year or more later. The literature reads this as bet-hedging against a false start — a brief rain that wets the mud and then fails would kill any fish that hatched into it, while an embryo that stayed in the shell loses only time. So the egg bank is not a single cohort on a single timer. It is a population of different decisions made by eggs that cannot compare notes, laid by parents who are no longer alive to be right or wrong about the season. The pool’s next generation is whoever’s waiting happened to match the rain that actually came.
And then, no time at all. What hatches is a fish in a hurry that is the mirror image of the egg’s patience. Turquoise killifish are reported as reaching sexual maturity faster than any other vertebrate — the figures differ by source and conditions, from about two weeks in some accounts to a few weeks in others — and the whole adult career of spawning daily into the mud is fitted into the weeks or months before the water goes. The species spends most of its existence as an egg that does nothing, and the rest of it as an adult that can afford to waste no day at all. The keeping and the kept are the same animal at two speeds: the parents’ frantic season is only the brief interval in which the egg bank reloads itself.
One egg bank’s year, in order: adults spawning repeatedly into the mud of a rain-filled pool that will not last the season; the pool drying, every adult dying, and the eggs left alone in cracked sediment; development arrested — early among dispersed cells, or midway with organs formed and the heart stopped, or after completion with hatching alone withheld; a dry season of roughly five to eleven months in the mud, the pause costing, on the laboratory evidence, no later lifespan; rain remaking the pool, some embryos hatching on the first inundation and others holding back against a false start, a few for another year; and the hatchlings maturing in weeks, spawning daily, and dying with the next drying — the bank reloaded. Schematic, compressed from the published record.
Walk it in order.
Schematic only: the stages compress one egg bank, and real timings vary by pool, population and year. Labelled cautions: the dry-season length, wild adult survival and daily egg output are reported spreads from particular monitored populations, not species constants; the roughly eight-month average diapause and the up-to-two-years laboratory figure come from secondary summaries of the diapause research and must not be retold as a field measurement for any given pool; the three diapauses are not all used equally by every annual killifish — species differ, and in A. limnaeus the first is reported as rarely entered while the second and third are routine; the anoxia figures belong to A. limnaeus experiments and are not turquoise killifish measurements; the finding that a long diapause costs no later lifespan is one laboratory programme’s result, not a law for wild egg banks, whose survival across a year recent modelling work describes as low and variable; maturity figures conflict across sources — about two weeks versus a few weeks — and are kept here as that conflict, not averaged; and bet-hedging is the literature’s reading of the spread in hatching, supported by studies of hatching probability and timing, not a decision any single egg can be shown to have made.