TL;DR: The data indicate that Amadina should be assigned to the estrildines rather than the lonchurines, and her relatively large, mannikin-like conical bill is a result of convergence, probably reflecting selection pressure on bill shape associated with diet specializations.
Abstract: There is disagreement as to the number of genera comprising the mannikins (Lonchurinae). Some authors recognize two to five of the following genera:Lonchura, Euodice, Padda, Spermestes andOdontospiza. Other authors have merged the last four taxa into the enlarged genusLonchura. We bring together data from downy plumes (neossoptiles), nestling palate markings, begging vocalizations, agonistic behaviour, courtship postures and starch gel electrophoresis of allozymes to solve these taxonomic problems. PAUP analysis of the combined data set indicates thatSpermestes andOdontospiza belong to a monophyletic clade separate fromEuodice, Padda andLonchura. We propose mergingOdontospiza withSpermestes, which suggests an African radiation of the group, and placing the generaPadda andEuodice in withLonchura, which represents a mostly Australasian radiation. The genusAmadina has been treated by some as a mannikin (Lonchurinae) and by others as a waxbill (Estrildinae). Our data indicate thatAmadina should be assigned to the estrildines rather than the lonchurines.Amadina's relatively large, mannikin-like conical bill is a result of convergence, probably reflecting selection pressure on bill shape associated with diet specializations.