Feeding Adaptations in Whistling Ducks (Dendrocygna)

Abstract

WATErFOWl. have radiated into a variety of niches by evolving such diverse feeding structures as serrated bills for capturing fish (Mergus) and lamellated bills for sifting food items (Anas and especially Spatula). Goodman and Fisher (1962) thus described and analyzed the feeding apparatus in a number of representative anatids, placing particular emphasis on the bones and musculature of the head. More recently Kear and Burton (1971) described in detail the feeding adaptations of the New Zealand Blue Duck (Hymenolaimus malacorhynchos) in an effort to understand the function of this species' bill flap. We recently reported structural adaptations in the locomotor behavior (Rylander and Bolen MS) and the foot structure (Bolen and Rylander 1974) in four species of whistling ducks: the Black-bellied Whistling Duck (Dendrocygna autumnalis), the Fulvous Whistling Duck (D. bicolor), the Plumed Whistling Duck (D. eytoni), and the Wandering Whistling Duck (D. arcuata). The distinctive adaptations for terrestrial and aquatic locomotion in these species and their phylogenetic implications prompted us to examine their feeding apparatuses for evidence of parallel evolution between the two species representing sympatric populations in North America (D. autumnalis and D. bicolor) and the two species that are sympatric in Australia (D. eytoni and D. arcuata). MATERIALS AND METHODS Our data were obtained from at least two adult specimens of each of the four species. Measurements were made with dial calipers, or in some cases with an ocular micrometer incorporated in a binocular dissecting microscope. To measure the length of the semicircular canal, we carefully removed the roedial part of the periotic bone with a steel cutting burr on a rotary electric tool. Once the semi-circular canal was opened at each end, a human hair was inserted, cut with iridectomy scissors to a length equal to the canal, then removed from the canal with forceps and measured. To measure the height of a lainella (i.e. the greatest distance that it projects into the maxillary cavity), one of the lateral edges of the maxilla was shaved with a scalpel, parallel with the axis of the maxilla, until a cross section of the lamellae at their maximum heights was exposed. The height of each lainella was then easily measured with an ocular micrometer. Brain sections could not be prepared adequately because of deterioration, so we estimated the size of the optic and vestibulocochlear lobes by measuring the length and width of the depressions in the skull accommodating these lobes. We prepared an outline of the profile of the upper mandibular cavity by tracing a photograph of a clay cast of this cavity. 86 The Auk 91: 86-94. January 1974