Problem solving Goldfinches

I realised only recently that Goldfinch manipulation abilities are superb. They hold large seeds with one foot while manoeuvring them into position to feed, or fold stems back to hold berries or seeds securely, or hand upside down to reach the thinnest branches. Yesterday I watched a small flock feeding on the catkins of a silver birch. The closest bird to me held a catkin while feeding from it and even changed the foot it used to hold it.
 It happens that Goldfinch manipulation abilities have been known for a long time (Plinius and Gesner mention it). Golfinches used to feature in animal shows, as seen in this old British Pathe film demonstrating the string pulling ability of a Goldfinch. The bird alternatively pulls and holds a string to reach a container with water or food.

String pulling is a common experimental setup to investigate the problem solving ability in birds. Parrots, corvids parids and finches can pass the test. In a series of experiments on captive-reared Goldfinches and Siskins, Uta Seibt and Wolfgang Wickler set to explore the problem-solving abilities of individual birds and what factors influence success. The birds were reared in cages with no access to material with which they could experiment their manipulative abilities and were tested when less than 5 months old using a 7-cm string holding a thimble with food at the end.

The researchers found that there were three groups of bird. 'Inventors' (23% of goldfinches and 62% of siskins) solved the string-pulling test by themselves, some times in the first experimental session, whereas ‘imitators’ (25% of goldfinches; 10% of siskins) were successful after watching conspecifics solving the problem. The remaining birds failed the test even when exposed to an experimented model.
 Something the experiment highlighted was the remarkable behavioural plasticity of Goldfinches and Siskins as shown by the diverse repertoire of behaviours the birds used to solve the problem (getting the food in the thimble):

Several individuals of both species hovered in flight near the thimble. They never succeeded in taking food, but still tried to reach food this way when they were already able to eat via string pulling and holding. A few goldfinches and siskins stretched from the cage fence towards the thimble but without successfully touching it. Two siskins and one goldfinch reached the thimble while hanging down from the perch head down like bats. Several siskins climbed down the string and ate from the thimble. Three other siskins shed some food from the thimble by heavily shaking the string and pecked the grain from the floor. (Goldfinches never moved to the cage floor.) Eight goldfinches and seven siskins invented a special method to flick the string around the perch, thus lifting the thimble indirectly. Some individuals used several of these techniques.

In addition, some individuals showed a preference for one or the other foot to hold the string.
Additional experiments tested the effect of being reared with access to branchlets. These seemed to facilitate success, but they were not necessary in several individuals. Similarly, the presence of a model facilitates success for a large proportion of the birds (the 'imitators'). Finally, the presence of food in the thimble was also necessary for the goldfinches to lift the string.
Bird species which instinctively used their feet to clamp food or branches while foraging - like the Goldfinch - should be able to solve the test, but the experiments show that birds still vary in their predisposition to experiment with the string and in their success. What proportion of wild Goldfinches use foot-bill coordination when foraging is hard to assess, but one would expect that given that natural conditions this would be higher than in the experiments. The wild birds are exposed to a much richer rearing environment, as they have very diverse diets and have to extract seeds from an array of plants from small perennials with soft pliable stems to trees with variably flexible branchlets (dandelion, teasel, lavender birch, ash, alder, rowan to give some examples). In addition, Goldfinches are highly social birds that feed together in flocks, so the learning opportunities would be plentiful.

Goldfinch holding birch catkin with left foot. Note that the catkin is still attached to the branch and the foot helps immobilise it, facilitating food retrieval.

More information
Seibt, U., & Wickler, W. (2006). Individuality in Problem Solving: String Pulling in Two Carduelis Species (Aves: Passeriformes) Ethology, 112 (5), 493-502 DOI: 10.1111/j.1439-0310.2005.01172.x . Here.