PATCH ASSESSMENT IN FORAGING FLOCKS OF EUROPEAN STARLINGS - EVIDENCE FOR THE USE OF PUBLIC INFORMATION

A field experiment was carried out to determine whether group-foraging starlings (Sturnus vulgaris) use public information to help them esti mate the quality of an artificial resource patch and depart accordingl y. Three kinds of information are potentially available in a group: pa tch-sample information, pre-harvest information, and public informatio n. These three types of information can be combined into four patch as sessment strategies: (1) patch-sample alone; (2) patch-sample and pre- harvest; (3) patch-sample and public; and (4) patch-sample, pre-harves t, and public. Depending on the foraging environment we presented to t he starlings, each assessment strategy made a unique set of prediction s concerning the patch departure decisions of pairs of birds based on differences in their foraging success. The environment was manipulated in two ways: by altering the variability in patch quality and by chan ging compatibility, the ease with which individual birds could simulta neously acquire both patch-sample and public information. Our observat ions on patch persistence and departure order demonstrate that the sta rlings used a combination of patch-sample and public information, but not pre-harvest information, to estimate the quality of the experiment al patch. Moreover, our results suggest that starlings use public info rmation only when it is easily available and ignore it under incompati ble conditions. This study provides the first evidence of public infor mation use in a patch assessment problem.