HABITAT USE AND FORAGING SUCCESS OF ROSEATE AND COMMON TERNS FEEDING IN FLOCKS IN THE AZORES David MONTICELLI* 1 , Jaime A. RAMOS** and José PEREIRA*** SUMMARY.—Habitat use and foraging success of roseate and common terns feeding in flocks in the Azores. Aims: To study the foraging ecology of roseate terns and common terns in the Azores sub-tropical wa- ters where the former species is relatively abundant compared to congeners. In particular, we asked whether foraging behaviour differed between both species, which factors determined individual success at forag- ing-flocks, and how did our findings compare with previous studies of Atlantic populations in North Amer- ica (temperate), and Caribbean (tropical). Location: Ponta das Contendas (38°39’N, 27°05’W), Terceira Island, Azores. Methods: Foraging-flocks were characterized and compared in terms of species (single- or mixed- species), size (no. of individuals), type of habitat (inshore, exposed coast, deep blue water), and presence- absence of biotic effects. We used a generalized linear model (GLM) approach to study the effect of se- lected factors (species, habitat, biotic effect, cloud cover, wind speed, and their interactions) on individual foraging parameters at these flocks (number of dives min -1 , number of aborted dives min -1 , number of prey caught min -1 ). Results: Both tern species were most often observed in mixed-flocks in the exposed coast habitat, but roseate terns were also prone to feed in mono-specific flocks over blue water. Only one quarter of the to- tal number of flocks recorded was in the presence of biotic effects, a situation where common terns, but not roseate terns, were twice more numerous. The GLM results suggested that species and habitat alone were good predictors of prey caught min -1 : common terns achieved the highest success in inshore bays (1.2 fish min -1 ) while roseate terns maximized their success on exposed coasts (0.8 fish min -1 ). In the presence of biotic effects, both species dove at a higher frequency, and aborted fewer dives per min, but there was no marked effect on capture rates. Increasing wind speed negatively affected the rate at which both species aborted dives, but again, there was no effect on capture success. Conclusions: Foraging success did not increase in the presence of biotic associations that should nor- mally facilitate prey capture. This suggests that interspecies competition may arise at mixed-flocks, as found in North American studies. Overall, our results paralleled previous findings that the common tern is adapted to feed in calm, inshore bays. Because this latter habitat was scarce in the study area, we hy- pothesized that this contributes towards explaining why common terns do not breed in high numbers at this colony site (ten percent of the Azores population). Conversely, the large area occupied by the marine coastal (exposed coast) and oceanic (blue water) environments in the study area may, at least in part, jus- tify the large number of roseate terns found there (one quarter of the Azores population) in 2001. * Laboratoire de Foresterie tropicale et sub-tropicale, Faculté universitaire des Sciences agronomiques de Gembloux, 5030 Gembloux, and Royal Belgian Institute for Natural Sciences, 1000 Brussels, Belgium. ** Institute of Marine Research (IMAR), Departamento de Zoologia, Universidade de Coimbra, 3004- 517 Coimbra, Portugal. *** Direcção Regional do Ambiente, Direcção de Serviços de Conservação da Natureza, 9930-210 La- jes do Pico, Azores, Portugal. 1 Corresponding author: david.monticelli@sciencesnaturelles.be Ardeola 53(2), 2006, 293-306