What happens to the queens that don't make it

The nuptial flight gets written up as a beginning, on this blog as much as anywhere. Queens go up, mate, land, shed their wings, dig in, and a colony starts. Every clause of that is true, and the sequence still misleads, because it follows the queens we end up noticing. For nearly all of them the story stops partway through. Estimates of combined flight and founding mortality sit close to 99 percent (Gordon & Kulig 1996; Helms & Kaspari 2015).

You can feel the number without any literature. Watch one strong July nest for an afternoon and you'll see queens leave by the hundred. Come back the next summer and the neighbourhood does not hold a hundred new colonies. Somewhere between the launch and the first anniversary, almost the whole cohort disappears.

Here's where it goes.

The swarm is a buffet announced in advance. Gulls and starlings work the rising columns, swifts and swallows cut through them, and over North America purple martins do the same job thoroughly enough that researchers have used them as sampling instruments: a 2016 Oklahoma study fitted martins with altitude loggers and identified the prey they carried home, and found ants through the lower atmosphere up to about 160 m, roughly 520 feet (Helms et al. 2016).

Synchrony is the counter-move. Hölldobler & Wilson (1990) describe the logic as predator satiation: if every nest in the area releases on the same afternoon, the local predators are full long before the sky empties. No individual queen gets safer, but the cohort loses a bounded share rather than being picked off at leisure across a month. The strategy doesn't dodge the losses. It spends them.

A queen that lands has traded a fast predator problem for a slow one. She's flightless the moment she breaks her wings off, calorie-dense, and conspicuous against pavement or bare soil. Birds again, spiders, ground beetles, and above all other ants: foraging workers treat a wandering foreign queen as food, and in dense territory the ground she lands on already belongs to someone.

Then there's water. Johnson (2021) followed the desert harvester ant Veromessor pergandei in central Arizona and dug up founding queens after two weeks of dry conditions: they were measurably more dehydrated than queens still sitting in their natal nests. That's a desert species under desert stress, but the direction holds generally. A searching queen is spending water she has no way to replace, and every hour she fails to find a nest site raises the price.

Digging in ends the visible dangers and starts a quieter one. A fully claustral queen closes her chamber and never leaves to feed. Everything the first brood needs comes out of her own body: the flight muscles she'll never use again are broken down internally (shown in fire ants by Barker 1979), and her fat reserves run down as the larvae grow (Wheeler & Buck 1996). It's a one-shot conversion. If the brood fails, from flooding, fungus, or reserves that were slightly too small, there's no second attempt with what's left.

How that founding works, and which species do it differently, is covered in fully claustral or not.

The bleakest version of the number comes from the same Arizona monitoring. Across 20 years of watched plots, new V. pergandei colonies successfully established in only 4 of them. Rainfall from January to June predicted which years those would be in 17 of 20 cases (Johnson 2021). Read that again from a queen's point of view: in most years, the entire cohort was written off by weather that fell months before anyone flew.

Once you hold the 99, the strange features of the flight stop being strange.

The absurd overproduction makes sense: it's the only strategy that works when the per-queen odds are this bad. The synchrony makes sense, as above. And the patience makes sense too. A colony doesn't rear queens until it can afford to throw them away, which in the Florida harvester ant Pogonomyrmex badius means waiting until the workforce passes roughly 700 workers (Smith & Tschinkel 2006). Reproduction is a surplus behaviour, and why one colony flies twice in a summer runs on the same logic.

It also puts the keeper's test tube in an honest light. A caught queen skips the birds, the rival workers and the drought entirely, which is most of the list above; you caught a queen this week, now what covers her remaining risks, which are real but far smaller. She is, by a wide margin, one of the lucky ones.

Gordon, D.M. & Kulig, A.W. (1996) and Helms, J.A. & Kaspari, M. (2015) are the sources behind the "close to 99%" mortality estimate, as summarised in Giannetti, D., Schifani, E. & Grasso, D.A. (2025), An adhesive drone trap to study the flight altitude preferences of winged ants, Current Zoology 71(5): zoaf002.

Helms, J.A., Godfrey, A.P., Ames, T. & Bridge, E.S. (2016). Predator foraging altitudes reveal the structure of aerial insect communities. Scientific Reports 6: 28670. Source of the purple martin foraging data and the ant flight altitudes.

Hölldobler, B. & Wilson, E.O. (1990). The Ants. Harvard University Press. Source for predator satiation and founding strategy.

Johnson, R.A. (2021). Desiccation limits recruitment in the pleometrotic desert seed-harvester ant Veromessor pergandei. Ecology and Evolution 11: 294 to 308. Source of the 4-in-20-years recruitment record, the rainfall prediction and the dehydration measurements.

Barker, J.F. (1979). Flight-muscle histolysis in Solenopsis invicta, Experientia. Cited for the internal breakdown of flight muscle after mating.

Wheeler, D.E. & Buck, N.A. (1996). Depletion of reserves in ant queens during claustral colony founding. Insectes Sociaux 43(3): 297 to 302. Cited for the direction its title states.

Smith, C.R. & Tschinkel, W.R. (2006). The sociometry and sociogenesis of reproduction in the Florida harvester ant, Pogonomyrmex badius. Journal of Insect Science 6: 32. Source of the 700-worker threshold.