Whether it’s calculating your risk of catching the new coronavirus or gauging the chance of rain on your upcoming beach vacation, you use a mix of statistical, physical, and social information to make a decision. So do New Zealand parrots known as keas, scientists report today. It’s the first time this cognitive ability has been demonstrated outside of apes, and it may have implications for understanding how intelligence evolved.
“It’s a neat study,” says Karl Berg, an ornithologist and parrot expert at the University of Texas Rio Grande Valley, Brownsville, who was not involved with this research.
Keas already had a reputation in New Zealand—and it wasn’t a great one. The olive-brown, crow-size birds can wield their curved beaks like knives—and did so on early settlers’ sheep, slicing through wool and muscle to reach the fat along their spines. These days, they’re notorious for slashing through backpacks for food and ripping windshield wipers off cars.
To see whether keas’ intelligence extended beyond being mischievous, Amalia Bastos, a doctoral candidate in comparative psychology at the University of Auckland, and colleagues turned to six captive keas at a wildlife reserve near Christchurch, New Zealand. The researchers taught the birds that a black token always led to a tasty food pellet, whereas an orange one never did. When the scientists placed two transparent jars containing a mix of tokens next to the keas and removed a token with a closed hand, the birds were more likely to pick hands dipped into jars that contained more black than orange tokens, even if the ratio was as close as 63 to 57.
That experiment combined with other tests “provide conclusive evidence” that keas are capable of “true statistical inference,” the scientists report in today’s issue of Nature Communications.
The researchers also showed the keas two jars that each contained an equal number of black and orange tokens. But the experimenter could only reach the tokens located above a solid barrier. Most of the kea correctly chose hands that had reached into the jar with the greatest ratio of black tokens above that divider, showing that they based their predictions solely on physical information—the number and relative quantities of tokens above the barrier.
