Why one colony can fly twice in a summer

Somebody always writes in about this. They saw the big flight, the one that made the local news, ants pouring out of every third paving slab for an hour. Two, three weeks later, on an ordinary warm evening, they see it again. Same garden, same species by the look of it, a smaller crowd this time. Was that a different colony that was just running late? A late flight of some other species entirely? Or did they somehow miss the memo and this is what everyone else already saw weeks ago?

Often, none of those. Often it's the same colony, going up a second time.

The instinct is to think of a nuptial flight as something that happens to a species on a date, like a tide table. Warm enough, humid enough, calm enough, and the species flies. That's roughly true at the population level, which is why a forecast is worth building at all, but it skips a layer. The actual decision is made nest by nest, and a nest that has done well for itself has more room to make it twice.

The clearest statement of this is old and blunt. Hölldobler and Wilson, in The Ants, note that colonies can raise more than one batch of reproductives in a good season and fly them separately: an initial major flight, then one or more smaller flights afterward as the next batch matures. They frame it as ordinary colony behaviour when conditions allow it, not as some rare exception.

That's not an armchair claim either. It shows up, cited exactly that way, in the largest UK study of flying ants there is.

Hart and colleagues analysed 13,394 usable public reports of flying ants from the Royal Society of Biology's national survey, 88.5% of them the black garden ant, Lasius niger. Most of the paper is about weather: flights cluster above about 13 °C (55 °F), and essentially every day with a UK mean temperature over 25 °C (77 °F) had ants up somewhere in the country. Wind above 6.3 m/s, 14 mph, shuts it down hard.

But weather only explains when conditions allow a flight, not why one nest goes and its identical-looking neighbour doesn't, and the paper says so directly in its conclusions. Colonies don't all reach "flight ready" at once. Many are there by July, some earlier, and variation caused by colony-specific differences and local geography means others aren't ready until August. Colonies in warmer microhabitats, urban plots and spots near heat-retaining structures such as walls or paving, tended to have their winged ants ready to go about three days earlier on average than colonies in cooler settings.

Put a colony that reached readiness early together with a good enough season, and you get exactly the pattern Hölldobler and Wilson describe: it fires off its main batch, keeps building, and has a second, smaller one ready to send up weeks later. The paper cites that mechanism directly for why some nests produce more than one flight in a single favourable year.

Boomsma and Leusink's classic study of Dutch flight weather, run across three summers on a Dutch island, adds a piece that fits alongside this. They found no significant difference in the mean air temperature at which Lasius niger, L. flavus and two Myrmica species flew, which is a strange result on its face, since these are ants of very different sizes living in very different microhabitats. Their explanation was that a queen doesn't fly at a fixed air temperature so much as when her body temperature roughly matches the nest's soil temperature, a rough rule of thumb rather than a precise one, since a heavier queen basking in the open gets there by a different route than a light one in shaded grass.

That's a species-level idea, not a colony-repeat-flight idea, but it points at the same underlying fact: a flight is triggered by a local, physical threshold being crossed inside and around one nest, not by a calendar entry the whole species reads at once. Once you accept that, a second threshold-crossing weeks later, in the same nest, in a good year, stops being surprising.

A 2026 paper took a genuinely odd angle on this. Garcia-Garin and colleagues went through 72 swift carcasses recovered from bird strikes at Barcelona airport, collected across three separate years, and found ant remains from 16 different species in 31 of them, evidence of nuptial flights the swifts had flown straight into while foraging. When they compared the weather on days that produced ant strikes against days that didn't, temperature, humidity, wind speed and rainfall didn't differ significantly between the two. Their conclusion was that colony-level factors, not the day's weather alone, were doing a lot of the work in setting exactly which day a given flight happened on.

That's a different dataset, a different country, a different method entirely, airport bird-strike records rather than a garden survey, and it lands on the same idea from a completely separate direction: weather sets the window, but something happening inside the colony decides which day within that window, and for some colonies, how many days.

There's a lower bound on all this, and it comes from an unrelated harvester ant study that happens to nail down the number precisely. Smith and Tschinkel spent years tracking the sociometry of Pogonomyrmex badius colonies in Florida and found that a colony doesn't start producing reproductives at all until its worker population passes roughly 700 individuals. Below that, every calorie goes into workers. Above it, a colony has the surplus to gamble on queens and males, whose only job is to leave and probably die trying.

P. badius isn't the ant in most people's garden, and the 700 figure doesn't transfer directly to Lasius niger, whose colonies and biology are different in scale. What transfers is the logic: reproduction is a surplus behaviour, gated by colony size and resources, not an inevitability that switches on for every nest the moment the calendar and the weather agree. A large, well-fed niger colony crossing that threshold early, with a good summer still ahead of it, is exactly the kind of colony with room to fly twice.

Mostly, nothing changes about how you look. If you're checking a forecast for a species you already saw fly once this season, a second, smaller window opening a few weeks later isn't a bug in the data or a coincidence, it's the ordinary shape of a good year for that colony. The RSB's own survey data shows why a single "flying ant day" was always a myth in the first place: across the three years it ran, ants were reported flying somewhere in the UK on 96% of days between the start of June and the start of September. A season with that much spread has plenty of room in it for one nest to visit twice.

None of this means every late flight is a repeat, and it's worth being honest that you generally can't tell from the garden which explanation is which. A second small flight from your own lawn could be the same colony's second batch, or it could be a different, later species that happens to look similar from a distance, and short of marking individual queens there's no easy way to be sure which. What both explanations have in common is that they're normal, and a "missed" flying ant day is far more often a flight you haven't seen yet than one that will never come again.

Dealate scores conditions hour by hour through the season for the species active near you, rather than trying to name a single day, because the biology underneath it doesn't work that way either. It's a reason to look up again, not a guarantee that this evening is the one. If you want the mechanics of a single flight, what is a nuptial flight covers the event itself, and why your neighbour's ants flew and yours didn't covers the same colony-level unevenness from the street rather than the season. The second wave goes through the British species that are only just getting started once Lasius niger winds down.

Hart, A.G., Hesselberg, T., Nesbit, R. & Goodenough, A.E. (2018). The spatial distribution and environmental triggers of ant mating flights: using citizen-science data to reveal national patterns. Ecography 41(6): 877–888. DOI 10.1111/ecog.03140. Open access. Source of the 13,394-record dataset, the temperature and wind thresholds, and the "flight ready at different times" conclusion.

Hölldobler, B. & Wilson, E.O. (1990). The Ants. Harvard University Press. The textbook source, cited within Hart et al., for a single colony raising multiple batches of reproductives and flying an initial major flight followed by smaller ones.

Boomsma, J.J. & Leusink, A. (1981). Weather conditions during nuptial flights of four European ant species. Oecologia 50(2): 236–241. Source of the shared flight-temperature window across species and the idea that a queen flies once her body warms to roughly the nest's soil temperature.

Garcia-Garin, O., Espadaler, X., Navàs, F., Castells-Morral, M., Ruiz-Sagalés, M., Abril, S., Gómez, M. & Ferrer, X. (2026). Opportunistic airport-collision samples from swifts reveal ant nuptial flight phenology. Scientific Reports 16: 17194. DOI 10.1038/s41598-026-48191-1. Open access. Source of the swift-collision dataset and the finding that weather didn't separate flight days from non-flight days as cleanly as colony-level factors did.

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 roughly 700-worker threshold before a colony becomes reproductively active, used here for the general logic rather than as a figure for Lasius niger.