Page 3 - Agostini_alii_2007

58 SHORT COMMUNICATIONS VOL. 41, NO.1

Figure 1. The study area and approximate paths used by European Honey-buzzards during spring migration across the
Central Mediterranean (dotted arrow: expected path; CB 5 Cap Bon Peninsula; P 5 Pantelleria; M 5 Marettimo; U 5
Ustica; L 5 Lipari Islands; SM 5 Straits of Messina).

corded in 444 hr of observation (265 hr at Pantelleria, Honey-buzzards were seen reaching the site from SW and
179 hours at Marettimo). A total of 6446 individuals mi- then leaving the island toward E. Prevailing winds, from N-
grated in flocks, 5397 at Pantelleria, 1049 at Marettimo, NNE, W-WNW-NW, and WSW-SW-SSW had a significant
with an average flock size of 25.0 6 4.0 (SE) and 9.9 6 effect on the migratory flow at the two sites (x 2 5 45.92,
1.3 (SE) birds, respectively and a maximum flock size of df 5 2, P , 0.01, Table 1). During N-NNE and W-WNW-
757 at Pantelleria. Among 142 (65.7%) flocks detected NW winds flocks passed mostly via Pantelleria; conversely,
over the sea while reaching Pantelleria, all were reported during south-westerly winds, flocks passed mostly via Mar-
from the WNW, apparently coming from the Cap Bon ettimo, although during such atmospheric conditions ob-
Peninsula (Fig. 1). A total of 74 (34.3%) flocks reached servations were made mostly at Pantelleria because of poor
the site undetected as they crossed the sea; of these, 53 visibility at Marettimo (x 2 5 13.96, df 5 2, P , 0.01, Ta-
moved along the western coast heading NE, six headed ble 1). Because of its geographic characteristics, Maret-
SW, and 15 were observed inland. European Honey-buz- timo not uncommonly experiences poor visibility (see also
zards left Pantelleria heading NE. At Marettimo, European Agostini and Logozzo 1998). The circadian pattern of mi-