Foraging: The Bay-headed Tanager spends most of its time foraging in the canopy. The bird tends to perch upright on twigs or branches to obtain small fruits; on few occasions it has been observed hovering or snatching berries while still in flight (Isler and Isler 1999). Fruit is usually swallowed whole; sometimes, pieces are pecked from larger fruits (ffrench 1991, Isler and Isler 1999). Frequently occurs at feeding aggregations of mixed species flocks (Isler and Isler 1999).
The Bay-headed Tanager forages for insects on the undersides of branches using the diagonal-lean method (Snow and Snow 1971, Isler and Isler 1999). Snow and Snow (1971) describe this behavior as follows: "the bird hops along a branch and every few feet leans down first on one side then on the other to examine the underside." The species prefers branches with living foliage; only a few records occurred on dead twigs or branches (Snow and Snow 1971). These branches are usually bare or lightly moss-covered and no greater than 5 cm in diameter (Snow and Snow 1971 and data from Steven Hilty, cited in Isler and Isler 1999). The bird also occasionally searches moss on branches, hanging moss, flower heads, hanging dead leaves, seed stalks, tree trunks, and vines in search of insect food (Isler and Isler 1999). Rarely, it may sally (Isler and Isler 1999).
Naoki (2003) quantified arthropod foraging behavior of Tangara gyrola at two sites, one in Costa Rica (subspecies bangsi) and one in Bolivia (subspecies catharinae). Overall, these results were similar to those found by Snow and Snow (1971) in Trinidad (Naoki 2003). In Bolivia, arthropod foraging was performed using mostly the reach-down (45.5%), hang-down (27.3%), and glean (15.2%) maneuvers (classification of foraging maneuvers follows Remsen and Robinson 1990). Individuals mostly searched bare branches (66.7%) and partially-moss-covered branches (36.4%). Perches used for insect foraging were mostly 0.5-3.0 cm (84.4% of observations). Most foraging occurred between 5-10 m above the ground (66.7% of observations). In Costa Rica, the same attack methods were preferred: reach-down (37.9%), hang-down (17.9%), and glean (20.2%); the same substrates preferred: bare branch (67.6%), partially-moss-covered branch (27.6%); the same branch size preferred: 0.5-3.0 cm (83.3% of observations); and 68.1% of the foraging observations were between 5-10 m above ground. The fact that arthropod foraging behavior was similar in Costa Rica and Bolivia, as well as in a less diverse site in Trinidad, led Naoki (2003) to conclude that there was no evidence for ecological release or niche expansion for the Bay-headed Tanager in Trinidad. Insect foraging behavior is stereotyped across the distribution and likely is under genetic control.
In addition to data on arthropod foraging, Naoki (2003) also collected data on fruit foraging. In Bolivia, Bay-headed Tanagers foraged for fruit mostly by gleaning (64.2%) and using the reach-down maneuver (11.9%). Fruits mostly were obtained between 5-10 m (40.4%), 10-15 m (25.0%), or less than 5 m (25.0%) above the ground. In Costa Rica, a greater diversity of maneuvers were used for fruit foraging: glean (29.6%), reach-up (11.9%), reach-out (19.0%), reach-down (13.8%), and hang-side (16.4%). Fruit-foraging occurred mostly between 5-10 m (54.0%), 10-15 m (17.1%), or less than 5 m(20.3%) above the ground. Fruit species eaten and their proportions also differed at the two sites (see Food). Differences in fruit foraging between the two subspecies likely reflect geographic variation in fruit availability, whereas arthropod foraging behaviors appear more stereotyped within Tangara species (Naoki 2003).