98 The Wildfowl Trust Observations on the behaviour and relationships of the White-backed Duck and the Stiff-tailed Ducks PAUL A. JOHNSGARD1 Summary 1. A number of behavioural and anatomical sources of evidence indicate that the White- backed Duck (Thalassornis) should be placed in the tribe Dendrocygnini and be regarded as a whistling duck rather than an aberrant Stiff-tailed Duck. 2. Behavioural evidence supports earlier views that the Black-headed Duck (Heteronetta) is the most generalized of the true stiff-tails, and it is suggested that the numerous unusual aspects of sexual behaviour found in this tribe can be traced back to a reduction of pair- forming and pair-maintaining mechanisms which can already be detected in Heteronetta. 3. The distinctive aspects of morphology and behaviour found in the Musk Duck (Biziura) can be attributed to the predictable effects of intense sexual selection resulting from the breakdown of pair bonds and the establishment of a completely polygamous or promis­ cuous breeding system in Biziura. 4. There is still too litde behavioural information concerning the Masked Duck to advo­ cate its generic separation (Nomonyx), although it is suggested that this species is probably the most isolated of the typical stiff-tails and presumably represents a less specialised evolu­ tionary line than do the other stiff-tails of the genus Oxyura. 5. Recent behavioural evidence tends to support earlier views that the remaining five species of Oxyura fall into two broad evolutionary groups; including leucocephala and jamaicensis on the one hand, and maccoa, vittata and australis on the other. The Stiff-tailed Ducks have traditionally At the species level the classification of been a relatively undisputed and well- the stiff-tails has been hampered by the defined group of species in the family considerable plumage similarities of most Anatidae, most members of which may species of Oxyura, as well as the meagre be easily distinguished from all other information available on their ranges and waterfowl by their elongated and pointed comparative anatomy. The problem has rectrices with stiffened shafts and by been especially vexing in southern South their bills having recurved nails. When America, where the affinities of the forms Eyton (1838) produced the first compre­ vittata and ferruginea to one another hensive classification of the Anatidae he have been in doubt, and their possible separated the stiff-tails as a distinct sub­ relationship to the North American Ruddy family (Erismaturinae) containing three Duck O. j. jamaicensis has also been a genera, Erismatura (Oxyura), Biziura, and problematic matter. Delacour and Mayr Thalassornis, the last genus being erected (1945) initially regarded these two South for the newly discovered White-backed American forms and the African Maccoa Duck, which Eyton first formally des­ Duck O. maccoa as races of a single cribed. He regarded this species as a “con­ southern hemisphere species that also in­ necting link between the genera Clangula cluded the Australian Blue-billed Duck and Biziura, the structure of the tail being O. australis. Boetticher (1952) later took nearly like that of the former genus, while the more extreme view that all of the the bill is like that of the latter.” Later southern hemisphere populations were systematists accepted this classification subspecies of O. jamaicensis. However, almost without modification, the only Helmayr and Conover (1948) pointed out change being that the genus Nomonyx that both South American forms breed was erected by Ridgway (1880) to distin­ together in central Chile, thus two species guish the Masked Duck from the other must be recognised. Lehmann’s descrip­ typical stiff-tails, primarily on the basis tion, in 1946, of a Colombian race of of its less recurved bill nail. This arrange­ Ruddy Duck O. jamaicensis andina sup­ ment persisted until the classic revision ported the view that the North American of the family by Delacour and Mayr Ruddy Duck is thus geographically con­ (1945), who reduced the stiff-tails to tribal nected to the South American popula­ rank (Oxyurini), merged Nomonyx with tions, and that ferruginea rather than Oxyura, and transferred the Black-headed vittata represents the southernmost repre­ Duck Heteronetta atricapilla into the sentative of this geographic series. tribe from its former position in the dab­ In an earlier review of the behaviour bling duck group, largely on the basis of patterns of the Anatidae (Johnsgard, Wetmore’s (1926) anatomical observations. 1960), I deplored the lack of behavioural 1 Contribution (No. 380) from the Department of Zoology and Physiology, University of Nebraska, Lincoln, Nebraska. Illustrated by the author. Stiff-tail Behaviour 99 information on the stiff-tails, and sug­ rushes or reeds over water and contain no gested that such knowledge might serve down. However, they are sometimes loca­ to evaluate the validity of including such ted at the edges of ponds as well (D’Eath, aberrant genera as Heteronetta and Thal- 1965). The clutch size has been variously assomis in the tribe, as well as to help reported as ranging between two and 14 establish species limits and taxonomic eggs, although the latter figure doubt­ groupings within the genus Oxyura. Since less represents “ dump-nests ” produced then it has been my good fortune to by more than one female. The eggs are observe all of the nine species of Oxyurini surprisingly large (Dr. Janet Kear in­ (Delacour, 1959) in life, and to observe formed me that the average weight of 21 sexual displays among six of them. eggs was 82.5 grams), and are of a dis­ Although it is still premature to believe tinctive pale rusty brown colouration. that an adequate knowlege of stiff-tail This colour and their very smooth surface behaviour is at hand, it is nonetheless distinguish them from typical stiff-tailed possible now to make some comparisons duck eggs. Each egg represents more than and conclusions that were impossible in ten per cent of the adult bird’s weight, 1960. Some such observations have since which ranges from 680 to 790 grams been published (Johnsgard, 1965a), but (Phillips, 1926). This surprisingly large others were obtained too late to permit size is roughly equivalent to the situation inclusion in my general survey of Anati­ in stiff-tails, since unincubated eggs of dae behaviour. Thus, special attention North American Ruddy Ducks average will be paid here to the re-evaluation and 74.7 grams (Janet Kear, pers, com.), and probable relationships of the various adults of this species range from 560 to genera included in the tribe, and to a to 680 grams (Phillips, 1926). It may be review of the available behavioural infor­ hypothesized that in both species this mation on the more typical stiff-tails large egg size is a functional adaptation (Oxyura), insofar as it may reflect and that permits the development of unusu­ further clarify their evolutionary relation­ ally precocial young that are able to forage ships. independently by diving shortly after Without question the White-backed hatching. Thus, Clark (1964) reports that Duck Thalassornis leuconotus is one of downy Maccoa Ducks may spend ten to the most inadequately studied species of fifteen seconds under water when forag­ waterfowl. The downy young were first ing. Interestingly, Johnson (1965) reports correctly described by Delacour and illus­ that unusually large eggs are also typical trated by Peter Scott in 1959, the of Torrent Ducks (Merganetta armata) tracheal anatomy remained undescribed which likewise have extremely precocial until 1961, and it also was not until then young that are able to dive and navigate that the reticulated tarsal surface condi­ swift currents with ease (Johnsgard, tion of the species was first noted (Johns­ 1966a). gard, 1961a). The unusual tracheal struc­ In 1965, the first nesting attempt by the ture, the associated whistling voices of White-backed Ducks at Slimbridge was both sexes, and the reticulated tarsal pat­ in a clump of willow herb Epilobium tern suggested to me that perhaps the and reeds within a foot or two of the White-backed Duck had been incorrectly water’s edge, perhaps because no emer­ placed in the Oxyurini, and that it might gent vegetation is present in their pond. actually be an aberrant whistling duck. Likewise, in 1966, a nest was built in a However, during my two years of study similar clump of willow herb and reeds at the Wildfowl Trust between 1959 and approximately 18 inches from the water, 1961, the species, although represented by but in this case it was also immediately nine individuals, failed to breed, and very adjacent to the gravelled public pathway few behavioural observations of taxono­ that is used by visitors to the Trust. In­ mic significance could be obtained (Johns­ deed, several large pebbles an inch or gard, 1965a). Then, during a return visit more in diameter were incorporated into to the Trust in July of 1966, I learned the nest scrape, which was lined with that, following a relatively unsuccessful grass but lacked any contour feathers or breeding attempt in 1965 (Johnstone, down. The nest was approximately one 1966), a second nesting had begun shortly foot above the level of the water, but no before my arrival. ramp led to it. Rather, the birds climbed Accounts in the literature (e.g., Mack- up on shore a foot or two to the side of worth-Praed and Grant, 1952) indicate the nest and entered it indirectly. At the that under natural conditions the nests of time I arrived on 4th July, six eggs had White-backed Ducks are usually built in already been laid at approximately daily 100 The Wildfowl Trust intervals. These eggs had been initially by the male was most surprising, since replaced with white wooden dummy eggs male stiff-tails normally take no interest until it was found that the adults were in the nest.