Crow research videos

Alex Taylor was the first to show in 2007 that NC crows can spontaneously use a tool as a metatool. Alex’s experiment suggested that metatool use came from crows recognising that a tool can be used in a general way to get another object, not only food. The video shows the first trial by Gypsy. The work was reported in Taylor et al. (2007) (see Publications). Footage taken by Alex Taylor.

Click the image to start the video.

The manufacture of a multi-step pandanus tool by a NC crow at Pic Ningua. Multi-stepped tools are only made on Grande Terre and are the most complex pandanus tool design. They take longer to make than the simpler wide pandanus tools and are associated with a high degree of laterality in manufacture. See our Publications for more details. Footage taken by Gavin Hunt.

Click the image to start the video.

The manufacture of a wide pandanus tool by a NC crow (Pandora) on the island of Maré. Crows on Maré only make the simpler wide pandanus design. We reported this behaviour in Hunt and Gray (2003, 2006) (see Publications). Footage taken by Gavin Hunt.

Click the image to start the video.

A young crow manufacturing a hooked tool in Parc Rivière Bleue from a live, forked twig. It first removes and discards one side of the fork (not shown), then removes the tool twig with a stump where the discarded fork was broken off. The crow then refines the tool by removing the leaves and ‘sharpening’ the stump so it is finer and more pointed. NC crows are the only nonhumans that make hook tools. The 3-D crafting process to make hooked-twig tools is the most complex example of tool shaping in animals. We reported this behaviour in Hunt and Gray (2004) (see Publications). Footage taken by Gavin Hunt.

Click the image to start the video.

An adult NC crow at Sarraméa using a fork in a branch to ‘aim’ a candle nut onto a rock below to try and break it and extract the nut kernel. NC crows are the only bird species know to drop food from a substrate rather than when flying. The use of the fork means a high degree of predictability about where the nut will land and may increase the chance of the nut breaking. We reported this behaviour in Hunt et al. (2002) (see Publications). Footage taken by Gavin Hunt.

Click the image to start the video.

A NC crow folding a wide pandanus tool before using it to extract food from a hole. This behaviour was the first evidence of innovative tool use in the wild, which confirmed behavioural flexibility in crows’ tool skills. We reported this behaviour in Hunt et al. (2007) (see Publications). Footage taken by Michael Anderson.

Click the image to start the video.

Alex Taylor tested NC crows with the modified two-trap-tube to investigate their ability for reasoning about physical problems. He found that 3 of 6 crows learnt to avoid a hole (the trap) and that they did so by the use of a causal rather than arbitrary rule (i.e. always move meat away from a hole). The video shows Obo solving the initial trap-tube task. Alex reported this behaviour in Taylor et al. (2009a, 2009b) (see Publications). Footage taken by Alex Taylor.

Click the image to start the video.

Alex used a simple modification to his first metatool paper to confirm that crows were not innovating metatool use by only simple associative learning. To do this, he required crows to obtain the metatool by pulling up a length of string hung from a perch. Crows naïve at metatool use solved this problem rapidly, pulling up the metatool even though there only experience with it had been a negative one. This showed that they had probably used cognition more complex than associative learning when innovating metatool use. This work was reported in Taylor et al. (2010) (see Publications). Footage taken by Alex Taylor.

Click the image to start the video.