9. A plethora of longbilled larks

¹Percy FitzPatrick Institute, University of Cape Town,
Rondebosch, 7701, email: pryan@botzoo.uct.ac.za
²Department of Genetics, Pretoria University, Pretoria, 0002,
email: PBloomer@nsnper1.up.ac.za

Most birders will confirm that the Longbilled Lark Certhilauda curvirostris is one of the most variable birds in southern Africa. If you put skins of the small, unstreaked, rufous birds from the eastern grasslands next to the large, heavily streaked, grey birds from the Northern Cape coast, you could be forgiven for placing them in separate genera, let alone different species. When Austin Roberts produced the first edition of his landmark Birds of South Africa in 1940, he recognized four species of longbilled lark: the Cape Longbilled Lark from the west coast, the Karoo Longbilled Lark from the western interior, the Eastern Longbilled Lark from the eastern grasslands, and the Damara Longbilled Lark. It was only in subsequent editions of Roberts’ that these forms were lumped together into a single species. The recent, rapid development of molecular techniques such as gene sequencing has provided a powerful new tool to obtain information about the relationships and taxonomic status of geographically variable populations. Used in conjunction with more traditional approaches, direct genetic assessments are being used to reexamine many of the species level conundrums. It was a combination of molecular and more traditional evidence that led to the recognition of Barlow’s Lark Mirafra barlowi as a distinct taxon in the Karoo Lark M. albescens complex (Ryan et al. 1998). We used the same approach to explore species limits in the longbilled lark complex. This note is a summary of the main findings of that study, published recently in Auk (Ryan & Bloomer 1999). We collected tissue samples from longbilled larks throughout their range. Sampling was most comprehensive in South Africa, with only patchy collecting from Namibia and none collected from southern Angola. We sampled nine of the twelve valid subspecies, missing only algida (coastal grasslands around Grahamstown), damarensis (central Namibia) and benguelensis (southern Angola). The race daviesi from the midlands of KwaZuluNatal was not reliably distinct from semitorquata-transvaalensis, although we sampled birds from within its range. We then sequenced the cytochrome b gene carried in the mitochondrial DNA, and compared this data with differences in morphology and display song. In a nutshell, the genetic evidence clearly showed there were three main groupings within the complex: 1) birds from northern Namibia, 2) birds from southern Namibia and the Karoo, and 3) birds from the west coast, Agulhas Plain and eastern grasslands. These groups differed from each other to almost the same extent as any of them differed from the presumed sister species, the Shortclawed Lark C. chuana, and thus warrant recognition as distinct species. If you can believe estimates based on molecular ’clocks’, members of the three groups have not shared a female ancestor for at least 3 million years! The three groupings were rather surprising, however, in that the greatest morphological variation is found within a single group (#3 above). Within this group, three subgroups were apparent in the gene sequence data, corresponding to birds from the west coast, south coast and eastern grasslands. These subgroups were less differentiated, but still had not shared a female ancestor for more than 1 million years. Given the marked vocal and morphological differences (there is virtually no overlap in size between birds of the same sex) among these subgroups, we suggest that they also should be recognized as distinct species. This results in the recognition of five ’new’ species of longbilled larks. Four are the same as those recognized by Roberts (1940), with the addition of the birds from the Agulhas Plain (brevirostris), which was only collected and described after Roberts (1940) Birds of South Africa was published (Roberts 1941). The species are named: Cape Longbilled Lark C. curvirostris, Agulhas Longbilled Lark C. brevirostris, Eastern Longbilled Lark C. semitorquata, Karoo Longbilled Lark C. subcoronata and Benguela Longbilled Lark C. benguelensis. We have followed Roberts (1940) in the use of common names, with the exception of the name for the northern Namibian/Angolan species, where there is some confusion (see below). As far as we know, the five species have largely non-overlapping ranges (Fig. 1). In fact, if you examine the atlas map for Longbilled Lark (ASAB2), you can clearly see a line of grid cells at 24°30’E where the species was either absent or very rare, corresponding with the boundary between the Karoo and grassland species (Dean et al. 1997). Similar gaps are evident between the Karoo and west and south coast populations. The only place where two species are known to overlap is west of the Richtersveld, where both Karoo and Cape Longbilled Larks have been collected together. However, narrow contact zones, possibly with hybridization, may occur in places. Depending on the species concept used, such narrow hybrid zones are not necessarily incompatible with specific status for the different populations (cf. Ryan 1997). Indeed we now have compelling evidence of a narrow hybrid zone between Barlow’s and Karoo Larks north of Port Nolloth (Cohen 1998), despite these species not being each others’ closest relative (Ryan et al. 1998). Getting back to longbilled larks, the greatest outstanding problem is in Namibia, where the boundary between the Karoo and Benguela Longbilled Larks is not resolved. It is evident that there is a major genetic boundary, but collecting was too patchy to identify the boundary. We assumed that the gap in the atlas distribution at 21°30’S corresponds with the species boundary. Unfortunately, the central and northern Namibian populations are similar in appearance, so field observations provide few clues. Plumage differences provide some supporting evidence for our inferred boundary. Moving from south to north, there is a marked reduction in streaking and an increase in the amount of rufous in the upperpart plumage in Karoo Longbilled Larks. This clinal trend is disrupted north of 21°30’S, with kaokoensis being somewhat more streaked than damarensis. In the absence of other evidence, we assume that damarensis is the northern extreme of the Karoo Longbill complex, which precludes the use of Roberts’ (1940) name for the northern Namibian taxon. However, more collecting is required to resolve this complex issue. REFERENCES Cohen, C. 1998. A hybrid zone between two nonsister species in the Karoo Lark Certhilauda albescens complex. Unpubl. Hons. Project, University of Cape Town. Dean, W.R.J., Allan, D.G., Herremans, M. Ryan, P.G. & Cohen, C. 1997. Family Alaudidae: larks. In: The atlas of southern African birds. Vol. 2: Passerines. Harrison, J.A., Allan, D.G., Underhill, L.G., Herremans, M., Tree, A.J., Parker, V. & Brown, C.J (eds), pp. 4-47, 687. BirdLife South Africa, Johannesburg. Roberts, A. 1940. The birds of South Africa. H.F. & G. Witherby, London. Roberts, A. 1941. Notes on some birds of the Cape Province. Ostrich 11: 112-135. Ryan, P.G. 1997. Lumping and splitting: the complexities of bird taxonomy. Africa: Birds & Birding 2(6): 64-67. Ryan, P.G. & Bloomer, P. 1999. The LongBilled Lark complex: a species mosaic in southwestern Africa. Auk 116: 194-208. Ryan, P.G., Hood, I., Bloomer, P., Komen, J. & Crowe, T.M. 1998. Barlow’s Lark: a new species in the Karoo Lark complex of southwest Africa. Ibis 140:605-619. Fig. 1. Ranges of species in the longbilled lark complex (after Ryan & Bloomer 1999).

Office Avian Demography Unit
Enquiries/More Information: adu-info@uct.ac.za
Last Modified 10-Aug-1999