PopBiology

Mastering biology through fascinating, accessible discoveries

Family First, Then the Flock: How Greylag Geese Decide When to Fly

Anyone who has watched a flock of geese lift off a field has seen something that looks like chaos resolving into order. One moment the birds are grazing. The next, a few are standing upright and shaking their heads, a rising racket of cackling fills the air, and a whole group is aloft, together. How do they decide when to go, and who goes with whom?

A study of free-living greylag geese (Anser anser) in Austria offers a detailed answer. The decision is not made by a boss goose. It is made through sound, and the sound is weighted by who is making it.

The problem every group has to solve

Living in a group pays off. Many eyes spot predators sooner, parents share the work of protecting young, partners can guard each other, and members learn where food is. But there is a catch. At any moment one bird may be hungry and keen to keep feeding while another is tired and ready to move. If everyone followed only their own needs, the group would drift apart and lose the very benefits that made it worth joining.

So groups need a way to share their readiness and agree on a time to move. Biologists describe three broad styles, though real animals rarely fit them neatly:

  • Despotic: one dominant individual decides and the others follow, even when it costs them. This is typical of species with strict, linear rank orders.
  • Democratic: many members signal their willingness, and the group moves once enough have chimed in (a threshold called a quorum). African wild dogs, which use sneezes to build a quorum, are one example.
  • Combined: each individual weighs what others are signalling against its own needs, then decides for itself whether to join. This suits loose, shifting societies where small units merge into bigger groups and split again.

Greylag geese appear to belong to the third group, with an important twist.

The study site

The research took place at the Konrad Lorenz Research Centre in Grünau im Almtal, Austria. The flock dates to 1935, when Lorenz established it from hand-raised eggs. The geese are non-migratory, used to people, and individually marked with coloured leg rings, so researchers know each bird’s identity, sex and family. The birds are also fed grain and grass pellets twice a day, which makes their movements to and from the feeding site predictable.

The team observed the geese on 28 days between 15 September and 9 November 2021. Autumn is when families and pairs gather into large flocks and make frequent short flights between feeding and resting spots. The population numbered 133 marked birds: 66 males, 43 females and 24 juveniles of unknown sex. Data were collected at the feeding field at the research station and at a nearby pond in Cumberland Wildpark, where the flock gathers in the evening before flying back.

An “attempt” began when a goose spontaneously started the take-off routine: standing alert, facing the direction it wanted to go, shaking its head from side to side, and calling. It ended when one or more birds took off, or when all signalling had stopped for 30 seconds. In pilot trials, birds that went quiet for that long had returned to feeding or resting. Of 651 recorded attempts, 587 were usable. Most of those excluded involved birds on the water, whose leg rings could not be read. The usable attempts came from 87 different initiating birds (34 females and 53 males). In total, 483 attempts (82%) ended in a successful flight.

Two calls, two jobs

Before taking off, geese use two distinct calls. Both carry information about who is calling and whether the caller is male or female, which helps when birds cannot see each other. The Locomotion call is a steady cackle that starts the display and means “I want to move.” The Pre-flight call is louder and comes in the last seconds before take-off, meaning “I’m about to fly.” The table compares them.

Locomotion callPre-flight call
SoundRepeated cackle of similar short elements at a steady rhythmA single loud syllable repeated at longer intervals
LoudnessAbout 77 dB at 1 m (medium range)About 82 dB at 1 m (long range)
TimingComes first and continues through the routineA few seconds before take-off
Meaning“I’m motivated to move” (walking, swimming or flying)“My urge to fly is strong”
Present in100% of attempts86% of attempts

A caution on the 100% figure for Locomotion calls: the researchers defined an attempt as beginning when a goose started Locomotion calling, so every attempt includes it by design. What the data do show is that attempts with more callers were more likely to succeed and produced larger flying groups. The data cannot show that silent departures never happen, because those would not have been sampled.

Bonded pairs can interleave the two calls. In the example recording in the paper, the female delivered bouts of Locomotion calls while the male slotted louder Pre-flight calls in between.

Sound suits this job better than sight. Flocks are dense, vegetation and terrain get in the way, and during the display every bird faces the direction it intends to fly rather than looking at its neighbours. Calls travel around obstacles and can tell a listener who is calling without eye contact.

What the numbers show

The researchers used a statistical model that answers two questions separately: did the flight happen at all, and if so, how many birds went? It also accounted for the same initiators and observation sessions appearing more than once. For a typical attempt where the family did not all call back, the model estimated an 81% chance of take-off and a flying group of 5.6 birds.

Two notes on reading the numbers. Effects on success are given as changes in the odds of failure. Odds compare failures with successes: 81 successes to 19 failures is about one failure for every four successes. A 68% drop in those odds means about one failure for every thirteen successes, so the chance of success rises from 81% to about 93%. That is a gain of 12 percentage points, not 68. Effects on group size are plain percentage changes. Where a figure is “per standard deviation,” it means a typical step up in the number of callers: about 3.6 more Locomotion callers or 1.4 more Pre-flight callers. Each figure comes with a 95% confidence interval, the range of values that fits the data, calculated from the paper’s tables.

1. Family comes first. The strongest predictor of a successful take-off was whether the initiator’s whole family called back: its mate and, for parents, all its young. When they did, the odds of failure fell by 68%, the same drop used in the example above. This fits how closely the birds stay together: pairs without young stayed together in 97% of trials, and families with young in 93%. In plain words: an attempt is much less likely to succeed when part of the family stays silent.

2. More callers, better odds. Beyond family, the size of the chorus mattered. For every typical step up in Locomotion callers (about 3.6 more birds), the odds of failure fell by about 45%. For every typical step up in Pre-flight callers (about 1.4 more birds), they fell by about 34%. In plain words: the more birds join the calling, the more likely an attempt is to end in take-off.

3. More callers, bigger flocks. Each typical step up in Locomotion callers made the flying group about 39% larger. The Pre-flight effect was much smaller, about 7.5%, and it is the least certain estimate in the table below. So the Locomotion call appears to do the heavy lifting of recruiting the wider flock, while the Pre-flight call looks more like a signal that take-off is imminent. In plain words: more callers means a bigger group in the air.

4. Sex did not matter, and parenthood mattered only for group size. Whether the initiator was male or female made no clear difference to either outcome. Being a parent had no clear effect on the chance of getting airborne, but parents did fly in larger groups, as the next section explains. In plain words: it doesn’t matter who starts the attempt; what matters is who answers.

The table below gives each estimate with its 95% confidence interval (CI). A wide interval means a less certain estimate.

EffectEstimate95% CI
Whole family calls back: lower odds of failure68%46–81%
More Locomotion callers (per standard deviation): lower odds of failure45%24–61%
More Pre-flight callers (per standard deviation): lower odds of failure34%8–54%
More Locomotion callers (per standard deviation): larger flying group39%29–50%
More Pre-flight callers (per standard deviation): larger flying group7.5%0.4–16%
Parent initiator: larger flying group36%10–73%
More Locomotion callers (per standard deviation): more non-family joiners54%33–80%

Why parents fly in bigger groups

Parents took off in groups about 36% larger than non-parents. Did parents have some special pull over the rest of the flock?

Apparently not. The researchers counted only the non-family birds that joined each flight, and parental status made no detectable difference. The likely explanation is simple arithmetic. A parent’s family is bigger: when a parent recruited only its own family, the smallest group was three birds, versus two for a non-parent and its mate. The youngsters also add their voices, so the chorus is louder from the start. Of the factors tested, only the number of Locomotion callers predicted how many non-family birds joined, with each typical step up bringing in 54% more outsiders.

The authors interpret this as mimetism, a kind of positive feedback in which the more birds are already doing something, the more likely others are to join. This is a reasonable reading, and it fits the data, but the study did not test it directly. It is better treated as a plausible contributing mechanism than as a demonstrated chain of events. In plain words: parents’ flights are bigger because their families are bigger and louder, not because parents are more persuasive.

A two-step decision

Putting the findings together, a goose’s decision looks like two checks in sequence:

  1. A family check. Before committing, the bird appears to need its mate and young to join in. Because the calls carry individual signatures, it could in principle verify this by ear alone. A family member’s call seems to count for far more than a stranger’s.
  2. A flock check. With family on board, a growing number of callers from the wider flock pushes the decision over the threshold and shapes how many birds fly together.

The authors call this a combined decision with an identity-weighted quorum. No bird gives orders. Each goose blends its own state with the public information in the air. The weighting itself is inferred from the pattern of results rather than measured directly, and the authors point to playback experiments as the way to pin it down.

How this compares with other species

The geese are not alone. Closely related species such as Canada geese and several swans perform similar pre-flight displays with head movements and chorus calling. Domestic geese show a similar link between the number of vocalising birds and the size of departing groups. Green woodhoopoes in dense forest rely almost entirely on vocal coordination to move between feeding sites. And in wild dogs and several primates, not all “votes” count equally, since the participation of particular individuals changes how readily a group moves. What stands out in the greylags is how clearly the data point to family membership as the first filter.

Why it matters

The study is a clear example of coordinated group action without a leader. Simple individual rules, such as “wait for my family to call” and “join in more readily when many others are calling,” are enough to produce smooth, flock-wide movement. It also shows that in societies where small, close-knit units live inside larger, fluid groups, counting votes is not enough. Who is voting matters too.

Cautions

It is an observational study of one habituated, regularly fed flock during one autumn, so it describes patterns in this population rather than proving cause and effect in every goose society. Several further limits are worth keeping in mind:

  • Departures from land only. Attempts begun on water were excluded because the rings could not be read.
  • Few unpaired birds. Only 22 attempts involved unpaired birds, so little can be said about them.
  • Few large choruses. Most attempts had only a few callers (about 3.6 on average, and no more than 20), so the model is much less certain about what happens when many birds call.
  • Family silence is ambiguous. When a mate or youngster stays quiet, the data cannot say whether it is acting as a brake on the initiator or is simply not ready to move itself.

What comes next

The authors suggest playback experiments with recorded calls from mates, familiar flockmates and strangers, to find the actual quorum thresholds and test the identity weighting directly. They also propose examining whether the order and interleaving of male and female calls carries extra information, and linking who starts successful departures to the flock’s long-term social network. They note that individuals seemed equally able to start a flight but differed in how many birds they recruited, which makes that last question especially interesting.

The next time a flock of geese noisily takes off, it may be worth listening closely. Behind the din is a quick negotiation: first among family, then across the flock.

Source

Study: Greylag geese (Anser anser) initiate group flight based on the number and identities of callers
Authors: Dina A. A. Mostafa, Alan G. McElligott, Nicolas M. Adreani, Elodie F. Briefer (2026)
Read the full paper: https://www.biorxiv.org/content/10.64898/2026.10.26.761644v1

Leave a Reply

Your email address will not be published. Required fields are marked *