THE RELATIONSHIP BETWEEN CLIMATE AND ANNUAL CYCLES IN THE COTINGIDAE

This survey is based primarily or the state of moult of over 4000 specimens of cotingas from all parts of the neotropical region. The seasonality of moult thus revealed is combined with existing knowledge of breeding seasons and seasonal environmental changes in an attempt to work out the broad pattern of annual cycles and their relation to climate. Within any local population the date of onset of moult may vary according to sex and age. In genera in which both sexes participate in nesting, males and females begin to moult at about the same time, or the males slightly in advance of the females. In genera with marked sexual dimorphism, in which only the female attends the nest, males may begin to moult well before females, at about the time that the latter begin egg-laying. The former group includes the genera Pachyramphus and Tityra, comprising species that are largely insectivorous, and the latter group includes the more specialized frugivorous genera. In all areas with well-marked seasonality, the ‘frugivorous group’ moults on average before the ‘Pachyramphus group’. It appears to be a general rule for first-year birds to moult earlier than older birds. A regional survey embracing all parts of the neotropical region shows that the peak of onset of moult occurs towards the end of the dry season (frugivorous group) or early in the wet season (Pachyramphus group). The changing moult seasons, strikingly in conformity with the geographical changes in the period of heaviest rainfall, are traced along a number of transects from Mexico in the north to Paraguay and Bolivia in the south. Such evidence as there is suggests that the main period of onset of moult in the frugivorous and Pachyramphus groups coincides with the period when their food is approaching or at its seasonal peak of abundance. It seems that both breeding and moult, which are almost entirely mutually exclusive, are as far as possible timed to coincide with this most favourable period; but whereas the moult takes a more or less fixed length of time the period when breeding is possible varies greatly in different species. Widely different patterns of annual cycle may result from the interaction of the two processes. Examples are given both from the cotingas and from species of other families with similar ecology. The proximate factors controlling the timing of the moult are briefly considered. It is suggested that increasing food availability is the main environmental controlling factor, and that an endogenous circannual cycle of moult must also be involved.