10. CORRESPONDENCE

Taxonomic classification of birds based exclusively on morphology and plumage traits has often been found to be inconsistent with true evolutionary history when tested with molecular phylogenies based on neutrally evolving markers. Here we present cytochrome-b gene sequences for the poorly known Little Brown Bustard Eupodotis humilis and analyse it in a phylogenetic context together with all other bustard species in the family Otididae. Our results suggest that this species is more closely related to the Little Bustard Tetrax tetrax than to other members of the Eupodotis genus. This finding is consistent with previous results suggesting polyphyly in the genus Eupodotis and with the fact that many of the phenotypic traits used to classify members of the family Otiidae are not phylogenetically informative.     

8. CORRESPONDENCE

This paper investigates, through experiments using surrogate predators, differences in intraspecific alarm calls between familiar and unfamiliar Bronze Mannikin Spermestes cucculatus groups. Four groups of mannikins were captured with mist nets from four areas in Durban (i.e. original groups) and randomly mixed (i.e. assorted groups). These groups were exposed to latex terrestrial snakes and mounted aerial raptors, and their alarm calls and predator response behaviours recorded. The Bronze Mannikins were able to discriminate between predators of different sizes, and increased their calling rate and decreased the end frequency of the alarm call in response to larger predators. This perhaps signalled increased threat, while simultaneously decreasing the conspicuousness of the flock. When the alarm call structure of the original and assorted groups in response to both raptors and snakes was compared, birds in original groups called more often, but paused longer between calls. Anti-predator behaviour differed in that assorted groups were less vigilant and aggressive toward the predators and panicked more frequently. In these groups, a failure to transfer the predation threat information might have caused the group to stop behaving cohesively and reliably. The manipulated experiments carried out in this study indicated that Bronze Mannikins were able to communicate predator size risk to conspecifics, but not as successfully to unfamiliar group members, showing that the investment, probability through altruistic payback, is greater in stable groups.     

CORRESPONDENCE

J. Cao and C. Z. He              877
Les populations de nombreuses espèces d'oiseaux Nord-américains montrent des signes de déclin. La collecte d'informations sur le succès reproductif permet de détecter rapidement les populations fragiles et aide au développement de pratiques de gestion de l'habitat appropriées. Dans cet article, nous étudions la performance du modèle Bayésien (He, 2003) appliquéà des taux de survie au nid dépendant de l'âge dans le cas de visites irrégulières. Nous montrons que les estimations sont satisfaisantes, mise à part celle de la survie à l'âge 1. Plus le nombre de jours entre deux visites est grand, plus cette survie risque d'être sous-estimée. Nous avons étudié ce problème et développé trois approches pour contrôler ce biais. Les simulations montrent que les trois approches permettent d'améliorer de façon significative l'estimation de la survie de la classe d'âge 1.