Are predatory birds effective secondary seed dispersers?

We have studied the unusual phenomenon of secondary seed dispersal of Lycium intricatum seeds on a small oceanic Atlantic island (Alegranza, Canarian Archipelago) in which a small frugivorous lizard ( Gallotia atlantica ) and two different predatory birds participate, a shrike ( Lanius excubitor ) and a kestrel ( Falco tinnunculus ). Endemic lizards that are common prey of both bird species consume Lycium fruits. Lizard remains were significantly matched with the presence of Lycium fruits in the regurgitation pellets of the two predatory birds. Seeds were found in 7.3% of the lizard droppings, 31.0% of kestrel pellets and 55.7% of shrike regurgitations. The mean number of seeds per dropping or pellet was 4.8 ±4 in lizard, 20.2 ±34.5 in shrike and 6.7 ±8.1 in kestrel. The percentage of viable seeds showed significant differences among all four treatments, decreasing in the following direction: seeds collected directly from plants (98.0%), shrikes (88.0%), lizards (72.3%), and kestrels (31.7%). Seeds from Lycium fresh fruits and shrike pellets showed significantly higher germination rates than those from lizard droppings and kestrel pellets. While lizards and shrikes are effective seed dispersers, kestrel gut treatment decreases seed viability. Seed viability is always higher than seed germination in each of the four treatments. In this island environment, Lycium seeds are under an important random influence during the seed dispersal process. Secondary seed dispersal seems to acquire a relevant dimension in small and remote insular environments or isolated continental systems where interactions among the different elements involved are intense, all of them are abundant native residents, and they have been coexisting for a long time. © 2002 The Linnean Society of London, Biological Journal of the Linnean Society , 2002, 75 , 345–352.