Adrian J .F. K. Craig

Craig, A.J.F.K. 1999. Weaving a story: The relationships of the endemic Ploceidae of Madagascar. In: Adams, N.J. & Slotow, R.H. (eds) Proc. 22 Int. Ornithol. Congr., Durban: 3063-3070. Johannesburg: BirdLife South Africa.

Four Ploceidae occur naturally on Madagascar: Forest Fody Foudia omissa and Madagascar Fody F. madagascariensis, and Nelicourvi Weaver Ploceus nelicourvi and Sakalava Weaver P. sakalava. A preliminary phylogenetic analysis using morphological characters and features of their breeding biology supports the view that the genus Foudia may be closest to the African genera Quelea and Euplectes. The Madagascan Ploceus species may be linked to the solitary, insectivorous African weavers, or the Asian Baya Weaver P. philippinus, rather than derived from the colonial savanna weavers of Africa.

INTRODUCTION

The avifauna of Madagascar can be broadly divided into three categories: endemic groups such as the Mesitornithidae, representatives of Afrotropical families (differing often only at the subspecific level) like the Hamerkop Scopus umbretta, and species with Asian affinities such as the Madagascar Bulbul Hypsipetes madagascariensis (Langrand 1990). Since Crook's (1964) monograph on the weavers, the genus Foudia, with two species in Madagascar and three (or four) on other Indian Ocean islands, has been linked with the African genera Quelea and Euplectes, primarily on the basis of their nesting and courtship behaviour. Crook (1964) tentatively assigned the two endemic Ploceus species, Sakalava Weaver P. sakalava and Nelicourvi Weaver P. nelicourvi, to a group that includes both African and Asian ploceids, noting that little information was available for them.

Previous accounts of the biogeography of the Madagascar avifauna have neglected the Ploceidae; they were not mentioned at all by Griveaud (1967) nor by Keith (1980). Apart from incidental observations, little has been recorded concerning the biology of these species, which are also of minor conservation concern compared to other Malagasy endemics. Recent research has focussed on endangered Foudia species on Mauritius (Mauritius Fody F. rubra), Seychelles (Seychelles Fody F. sechellarum) and Rodriques (Rodrigues Fody F. flavicans) (Collar et al. 1994), and much information on the Madagascar Fody F. madagascariensis also comes from populations introduced on other Indian Ocean islands (e.g., Crook 1961; Safford 1997). In this paper I have used published information and data from museum specimens to compare the four Madagascar ploceids, F. madagascariensis, Forest Fody F. omissa/eminentissima, Ploceus nelicourvi and P. sakalava, to possible relatives in Africa and Asia. From Asia, I selected only Baya Weaver P. philippinus, the most widespread and best-known species. From Africa, I chose four Ploceus species with a wide distribution in southeastern Africa: two non-colonial insectivores Forest Weaver P. bicolor and Spectacled Weaver P. ocularis, and colonial granivores Spotted-backed Weaver P. cucullatus and Masked Weaver P. velatus; also all three members of the genus Quelea: Red-billed Quelea Q. quelea, Cardinal Quelea Q. cardinalis, and Red-headed Quelea Q. erythrops; and four representatives of the genus Euplectes: two bishops Golden Bishop E. afer and Red Bishop E. orix and two widows Red-shouldered Widow E. axillaris and Yellow-rumped Widow E. capensis. The Madagascar and Asian species I know only as study skins, but I have field experience with all the African species except Q. cardinalis.

METHODS

Plumage details and nesting data were taken both from examination of specimens and from descriptions in the literature, supplemented by my own field notes for the African species. All skeletal data were recorded by me personally from museum collections (Table 1). The main published sources for Asian and Madagascan birds were Benson et al. (1977), Crook (1960, 1961, 1963, 1964), Frith (1976), Langrand (1990), Newton (1959) and Rand (1936). For Foudia omissa, which is often treated as conspecific with Red-headed Fody F. eminentissimma (cf. Dowsett & Forbes-Watson 1990), I have used biological data from populations of F. eminentissimma on other islands. A North American icterid, the Red-winged Blackbird Agelaius phoeniceus, was used as the outgroup, and data for the House Sparrow Passer domesticus also were included. Trees were generated using both PAUP (Swofford 1993) and Hennig86 (Farris 1988).

RESULTS

Table 2 shows the character states for the species in Fig. 1. The uninformative characters, which showed the primitive condition in all taxa, were disregarded in the analysis. With regard to skeletal characters, these are remarkably uniform in the three genera Foudia, Euplectes and Quelea. I have data for eight other Euplectes species, and all members of this genus share the same character states for these nine skeletal features. In contrast, the 17 species of Ploceus examined so far represent seven different combinations of skeletal characters. This implies that the current genus, which includes more than 60 species in most checklists, is likely to be paraphyletic.

The genus Ploceus

There is a long tradition of including many species of weavers in the genus Ploceus. Reichenow (1886) recognised 38 taxa in six subgenera. He included P. sakalava in the subgenus Ploceus alongside the Asian species Streaked Weaver P. manyar, Black-breasted Weaver P. benghalensis and Asian Golden Weaver P. hypoxanthus; P. velatus was placed in the subgenus Sitagra; and P. cucullatus in Hyphantornis. Ploceus ocularis and P. bicolor were assigned to the genus Symplectes (Reichenow 1887). In the most recent comprehensive Afrotropical checklist, Dowsett & Forbes-Watson (1990) recognise 56 species of Ploceus. Of the current world checklists, only Wolters (1979) has broken up this large genus. He restricted Ploceus to the five Asian species; placed sakalava and nelicourvi as the only two species making up Nelicurvius; ocularis in Hyphanturgus; bicolor in Symplectes; and velatus and cucullatus in Textor. Unfortunately Wolters did not publish any reasons for these decisions, but Morlion (1980) noted differences in pterolyosis within the genus Ploceus that could support such subdivisions.

Other data on the taxa investigated here

Karyotyping has not been used widely in bird studies. Christidis (1986) commented that F. madagascariensis appeared to have a karyotype identical to that of P. velatus and P. philippinus. Thus this character may not provide useful information on relationships within the sub-family. DNA studies have yet to include any Madagascar ploceids. DNA-hybridization showed Q. quelea and Q. erythrops as close relatives, with Euplectes their sister taxon, and Ploceus species, including P. cucullatus, forming part of the same branch (Sibley & Ahlquist 1990).

CONCLUSIONS

While molecular techniques currently enjoy a high profile, the most cost-effective approach seems to be to frame a phylogenetic hypothesis using other characters, which can then be tested and refuted or supported by molecular data. Comparative behaviour studies, as demonstrated by Crook (1964), remain an accessible option. A recent analysis concluded that behavioural characters are just as reliable as morphological characters in phylogenetic studies (De Queirez & Wimberger 1993). Clearly we are not yet able to identify the closest relatives of the Madagascar ploceids with any confidence, but the fodies most likely are African in origin, whereas for the two Ploceus species an Asian ancestry also is plausible. I hope that fieldworkers in Madagascar will contribute new data to these questions.

ACKNOWLEDGEMENTS

My original field studies of ploceids in South Africa were stimulated and guided by the late Profs. G. J. Broekhuysen and K. Immelmann, and by Prof. G. L. Maclean and Dr. C. J. Skead. For access to museum specimens, I am grateful to D. Allan, T. Cassidy, K. Garrett, J. Hinshaw, A. C. Kemp, W. E. Lanyon, M. Louette, R. B. Payne, and D. Willard. Research funding has been provided by Rhodes University, and the Foundation for Research Development. N. Barker assisted with the systematic analysis, and P. E. Hulley and C. J. Whittington-Jones commented on a draft of the paper.

Wolters, H.E. 1979. Die Vogelarten der Erde, Lieferung 4. Hamburg & Berlin; Paul Parey.