What wild mixed-species parrot flocks can teach you at home

Deep in the Peruvian Amazon, dozens of parrot species gather at mineral-rich clay licks each morning, and every arrival requires a rapid decision about where to land and who feels safe to stand beside. In a 2023 behavioral study by Vanessa Ferdinand, Elle Pattenden, Donald J. Brightsmith, and Elizabeth A. Hobson, researchers analyzed 11 years of observations and found parrots weren’t primarily identifying one another by species. Instead, they relied on fast social shortcuts — things like size, shape, and visible features — to navigate mixed flocks efficiently. You can read the complete study here.

If you’ve ever watched parrots gather at a clay lick in Peru, it looks chaotic—dozens of birds, multiple species, constant motion, everyone making tiny choices about where to land and who to tolerate. The surprise from new research is that those choices may be driven by a few simple shortcuts rather than detailed identification of every species around them.

Researchers analyzed 11 years of observations of 12 parrot species feeding together at a large riverbank clay lick in the Peruvian Amazon. Every five minutes, observers recorded which species were present in each zone of the cliff. That gave the researchers a rare window into “live” social decision-making: who shows up together, and more importantly, who chooses to join a group that’s already on the wall.

Wild parrots use fast categories, not long checklists

Humans tend to label birds by species first. The parrots in this study didn’t seem to operate that way.

When the team tested different explanations for joining decisions, the strongest models weren’t “species A joins species B.” The strongest models were simpler: broad categories like body size, clade/shape, or visible color features (head or face color, for example). Across the board, these coarse categories beat a species-level explanation in the statistical model comparisons.

In plain terms: in a busy, risky environment, parrots may lean on quick signals they can process fast.

Size differences predicted friction

Another practical takeaway: pairs of species with big size gaps were more likely to avoid each other, while similarly sized species were more likely to affiliate. That pattern held even after the researchers accounted for the fact that some species prefer certain parts of the clay lick or certain times of day.

That aligns with something many owners notice at home: when there’s a big size imbalance, the smaller bird often has to spend more energy staying safe—physically and socially.

What this means for you at home

If you have more than one parrot, or you’re introducing a new bird, this research supports a simple mindset: your birds may be reacting to category cues before they ever get to “individual relationship.”

Here are a few ways to use that:

Treat first impressions as “pattern matching.” Early tension doesn’t always mean two birds are doomed socially. It can mean “this looks like a risky category” (much bigger body, more dominant posture, unfamiliar coloring, unfamiliar movement style). That can soften with slow, predictable exposure. (The paper can’t prove this for captive birds, but it’s consistent with the kind of shortcuts they detected in the wild.)
Be extra conservative with big size gaps. The study found size differences linked to avoidant associations in the wild. At home, that’s a reason to manage proximity carefully—separate stations, distance, and no forced sharing of “high value” spots like food bowls, favored perches, or tight play areas.
Design the room to reduce fast, reactive decisions. At the clay lick, birds are constantly deciding where to land and when to leave, under pressure. In your home, you can remove pressure: multiple perches at different heights, clear “escape routes,” and enough spacing that a smaller bird doesn’t have to squeeze past a larger bird to move around.
Watch joining behavior, not just co-existing. One of the smartest parts of this paper is the difference between “they were present together” and “who chooses to approach whom.” At home, a bird that repeatedly chooses to move closer (on their own) is giving you stronger information than two birds merely being out at the same time on opposite ends of the room.
Expect asymmetry. In the study, joining decisions were often one-directional: one species reliably joined another, without the reverse being true. At home, one bird may be the consistent “approacher” while the other is the “tolerator.” That dynamic can still be stable; it just means you should judge comfort by the less enthusiastic bird’s body language.

A useful way to think about it

Your parrots aren’t running a spreadsheet of pros and cons when they see another bird. They’re making fast calls with limited bandwidth. This research suggests those calls often rely on broad signals—size, shape, and visual features—because that’s efficient when decisions are constant and consequences can be real.

If you want, tell me what species you have at home (and whether they’re housed separately, free-flight together, or doing gradual intros), and I’ll translate these findings into a simple intro plan and “what to watch for” checklist tailored to your setup.

‍

Summary of the research

This paper asks a deceptively simple question: when wild parrots decide which other birds to feed alongside, how much “detail” do they actually use to make that decision? The team used 11 years of field scans from a famous clay lick in southeastern Peru (Colpa Colorado, Tambopata), focusing on 12 common parrot species that frequently gather to eat mineral-rich clay. Observers recorded which species were present in the same zone every 5 minutes, especially in early wet season mornings (Oct–Dec, before 9am).

They analyzed the data in two complementary ways. First, they looked at co-presence (which species tended to be on the same clay-lick zone at the same time), while controlling for confounds like time of day and zone preferences. That produced a network of affiliations (pairs that appeared together more than expected) and avoidances (pairs that appeared together less than expected). Second, they looked at joining decisions—the moment-to-moment “who joins whom” pattern (e.g., species A arrives in a zone where species B was already present across adjacent 5-minute scans).

Key findings:

These weren’t just random crowds around a resource. Even after controlling for time/zone usage, there were real, structured affiliations and avoidances among species. In other words, the identity of who was already there influenced who showed up.
Body size mattered for who tolerated whom. Species pairs with bigger size differences were more likely to show avoidant relationships; similarly sized species were more likely to affiliate. The authors saw this pattern in their main dataset and then replicated it in a held-out validation split.
Joining behavior was directional and more revealing than simple co-presence. Co-presence relationships were often reciprocal (if A affiliates with B, B affiliates with A). Joining decisions were much less reciprocal—more like “A tends to join B,” without the reverse necessarily being true—suggesting asymmetries in how species use social information.
The big headline: parrots did not seem to rely on species-level ID to decide who to join. The team tested nine possible “categorization schemas” (species, clade/shape, large-macaw vs not, several color features, and multiple size-based schemes). For every species, a simpler schema beat the full species-level model in their model comparison framework (AIC), implying parrots often use compressed rules—like “birds with this head color,” or “large vs not,” or “my clade”—rather than “that exact species.”
Different species used different simple schemas. Some fit best with size, others with clade/shape, others with head/face color, and the best-fitting schema was fairly consistent for many species across their data split.

The authors frame this as coarse-graining: animals may reduce a complicated social scene into a small number of useful categories that are “good enough” for fast decisions under risk (predators, aggression, limited feeding spots).