Preference and motivation for different diet forms and their effect on motivation for a foraging enrichment in captive Orange-winged Amazon parrots (Amazona amazonica)

Motivation tests were conducted to assess preference strength for diet form and a cage enrichment device by Orange-winged Amazon parrots (Amazona amazonica) held in individual cages, N = 10. Each cage was equipped with two trough-type feeders, one of which had a hinged lid that required lifting to access feeder content; cost of lifting the lid could be increased by the addition of up to 480 g upon it, ～1.5 times the mass of an Amazon parrot. Motivation tests were conducted using three different diet forms of pellets (regular-sized, 0.16 g/pellet (cylindrical-shaped); large-sized, 3.4 g/pellet (cylindrical); and over-sized, 3–5 g/pellet (cuboid-to-spheroid)) manufactured from the same diet formulation. When regular pellets were concomitantly freely available, birds were still highly motivated to gain access to either large-sized or over-sized pellets, with 7 of 10 birds lifting the maximum weight of 480 g. In motivation tests comparing over-sized vs. large-sized pellets, birds worked more for over-sized pellets (when large-sized pellets were freely available) than vice versa: motivation for over-sized pellets exceeded that for large-sized pellets by approximately 242.67 ± 64.4 g (F = 14.2, P = 0.0055; Sign Test, P = 0.0078). Additional tests assessed motivation to access 2.5 cm/side wooden cube enrichment devices when regular or over-sized pellets were freely available. Birds removed more cubes when fed regular pellets (Sign Test, P = 0.0078) and lifted an average of 221.33 ± 64.62 g more to access them (F = 11.73, P = 0.009; Sign Test, P = 0.0063), than when over-sized pellets were freely available, suggesting that enrichment devices may act as foraging substitutes. Likewise, both wooden cubes and over-sized pellets elicited comparable podomandibulation (handling with beak and foot) behavior, and podomandibulation was reduced when parrots were fed regular as opposed to over-sized pellets (58 ± 10 s mean ± SE] vs. 4.27 min ± 31 s; P < 0.0001). Finally, there was no evidence of contrafreeloading: in control experiments, birds only removed pellets which were freely available, when the same pellets were concomitantly available in weighted-lid feeders (F = 120.20, P < 0.0001). In summary, these results show that captive Orange-winged Amazon parrots strongly prefer pellet forms that are substantially larger than what is commercially recommended for them; preferred forms more closely resemble that of native wild foods, such as palm fruits.