Shaping of hooks in New Caledonian crows.

Many animals use tools, but their understanding of physical forces or causal relations is unclear (1, 2). Primates are considered the most versatile and complex tool users, but observations of New Caledonian crows (Corvus moneduloides) (3–5) raise the possibility that these birds may rival nonhuman primates in tool-related cognitive capabilities. We report here an experiment inspired by the observation that a captive female spontaneously bent a piece of straight wire into a hook and successfully used it to lift a bucket containing food from a vertical pipe (Fig. 1A). This occurred on the fifth trial of an experiment in which the crows had to choose between a hooked and a straight wire and only after the hooked wire had been removed by the other subject (a male). The animals had prior experience with the apparatus, but their only previous experience with pliant material was 1 hour of free manipulation with flexible pipe-cleaners a year before this experiment, and they were not familiar with wire (6). To investigate the importance of this observation, we conducted several new trials in which we placed a single straight piece of garden wire (0.8 mm in diameter, 90 mm long) on top of the tube and did not intervene until either of the birds obtained the food (a valid trial) or dropped the wire irretrievably into the tube (an invalid trial). Out of 10 valid trials (interspersed with seven invalid ones), the female bent the wire and used it to retrieve the food nine times, and the male retrieved the food once with the straight wire (7). To bend the wire, she first wedged one end of it in sticky tape (available around the bottom of the tube and the side of the plastic tray containing the apparatus) or held it in her feet at a location 3 m from the food, where there was no tape. She then pulled the other end orthogonally with her beak (see Movie S1), resulting in a bend with an angle of 74 30° (mean SE) (see Fig. 1B for individual tool shapes). She started to bend the wire 35 8 s after the start of each trial and used the resulting hook 6 2 s later. In all cases but one, she tried with the straight wire (for 15 4 s) before starting to make the hook. In all valid trials, the birds retrieved the food within 2 min. Thus, at least one of our birds is capable of novel tool modification for a specific task. In the wild, New Caledonian crows make at least two sorts of hook tools using distinct techniques (3, 4), but the method used by our female crow is different from those previously reported and would be unlikely to be effective with natural materials. She had little exposure to and no prior training with pliant material, and we have never observed her to perform similar actions with either pliant or nonpliant objects. The behavior probably has a developmental history that includes experience with objects in their environment ( just as infant humans learn about ev-