Stop and Smell the Pollen: The Role of Olfaction and Vision of the Oriental Honey Buzzard in Identifying Food

Abstract

The importance of olfaction for various avian behaviors has become increasingly evident. So far, the use of olfaction for food detection among raptors has only been demonstrated for Cathartes vultures. The Oriental honey buzzard (Pernis orientalis) is a resident and migrant in Taiwan and regularly forages in apiaries. One of its foods in apiaries is yellow pollen dough, a softball-sized mixture of pollen, soybeans, and sugar that beekeepers provide as a supplementary food for bees. Given that pollen dough is not similar to any naturally occurring food, we hypothesized that buzzards identify the dough's nutritious contents using olfaction, perhaps in combination with vision. Using a series of choice experiments in which individuals could choose between two doughs, we showed that (1) buzzards almost unerringly chose pollen-containing over pollen lacking doughs when otherwise the doughs were identical in size, shape, and yellow color; (2) buzzards always preferred yellow over black or green doughs if both doughs contained pollen; (3) buzzards still preferred pollen-containing over pollen-lacking doughs when both doughs were black, but at a lower rate than in (1). We statistically excluded the possible influences of the doughs' relative brightness or of repeat visits by the same individuals. Our experiments thus suggest the use of a 'multi-modal foraging strategy' among buzzards whereby olfaction and vision are likely to be both used in identifying food at close distances. We also estimated the olfactory receptor gene repertoire size in the buzzard's genome which is almost five times as large as that of three other raptor species. Therefore, olfaction is likely of far greater ecological importance to this species than to other raptor species. We suggest that olfaction should be considered in the design of behavioral and genetic studies to better understand the use of multiple senses in avian behaviors.

Fig 1. Distribution of eight apiaries in central and southern Taiwan where our 216 two-choice experiments were conducted.

See S1 File for names and coordinates of the apiaries.

Fig 2. Number of foraging events in experiment 1 where the treatment is presence or absence of (a) pollen, (b) soybean and (c) sugar in yellow doughs (Table 1).

Open bars: number of events feeding on only one dough; light grey bars: number of events feeding on both doughs, whereby the first chosen dough contained all three ingredients; dark grey bars: number of events feeding on both doughs, whereby the first chosen dough contained only two ingredients.

Fig 3. Number of foraging events in experiment 2 where the treatment is color (yellow versus black or green) in doughs containing pollen (Table 1).

Open and solid bars as in Fig 2.

Fig 4. Number of foraging events in experiment 3 where the treatment is presence or absence of pollen in black doughs (Table 1).

Open and dark grey bars as in Fig 2, whereby dark grey bars means that the first chosen dough contained only two ingredients.

Fig 5. Spectra of reflected light by doughs in two-choice experiments.

(a) Yellow dough with all three ingredients (black line); yellow dough without pollen (red line); yellow dough without soybean (blue line); yellow dough without sugar (green line). (b) Black doughs with all three ingredients (black line); black doughs without pollen (red line); green doughs with all three ingredients (green line).

Fig 6. The estimated OR gene repertoire sizes of 13 species of birds representing eight orders.

Open and grey bars represent the estimated total number of functional and non-functional OR genes, respectively. Data are from this study (OHB = Oriental honey buzzard), Warren et al. [41] (zebra finch), Zhan et al. [42] (falcon 1 = peregrine falcon, falcon 2 = saker falcon), Doyle et al. [43] (golden eagle) and Steiger et al. [38] (all other species). Note that Doyle et al. [43] did not distinguish between the number of functional and non-functional OR gene.