Feeding behaviour studies in injured kestrels

A sample of nine (5 F, 4 M) adult Kestrels (_Falco tinnunculus_), temporarily in captivity due to rehabilitation but ready for release, were tested to ascertain for variation in the predatory behaviour. Tests were carried individually on 4 different prey: two vertebrates (lizard and laboratory mouse) and two invertebrates (grasshopper and earthworm). The birds were offered such prey once daily, not necessarily in subsequent days, and in random sequence. The latency to attack the various prey was similar. In contrast to the predation tests with vertebrates, the Kestrels preying on invertebrates landed in most cases (P < 0.05 for grasshopper and P < 0.01 for earthworm) a few centimetres from the prey itself, grasping it after a few steps. While the Kestrel toes encircled the mice completely, but they simply kept both lizards and grasshoppers squeezed on the ground; earthworms were instead gripped almost invariably with the beak. After capture, the birds used their beak to bite the prey at highly variable rate (P < 0.001). The vertebrates were those mostly bitten, particularly on the head (mouse: 32.0 ± 6.9 biting per test; lizard: 15.8 ± 2.6), while the earthworm was never bitten. The latency to prey ingestion, i.e. the time elapsed from grasping to ingestion, varied greatly (P < 0.001), possibly in relation to time spent biting. The duration of captivity did not affect any behaviour pattern performance, but only the latency to attack (P < 0.05). The Kestrel then confirms to have a strong behaviour plasticity even in the predatory sequence, being able to greatly modify its behaviour patterns in relation to different prey and showing that the behaviour sequence is not really stereotyped. Moreover, such a plasticity does not fade away even after long captivity, but affects only the time necessary to attack