Short communication A Gulf of Guinea island endemic is a member of a Mediterranean-centred bird genus GARY VOELKER, 1 * MARTIM MELO 2 & RAURI C. K. BOWIE 3 1 Department of Wildlife and Fisheries Sciences and Texas Cooperative Wildlife Collections, Texas A&M University, College Station, TX 77843, USA 2 Percy FitzPatrick Institute, DST / NRF Centre of Excellence, University of Cape Town, Rondebosch 7701, South Africa 3 Museum of Vertebrate Zoology and Department of Integrative Biology, 3101Valley Life Science Building, University of California, Berkeley, CA 94720, USA Keywords: Dohrn’s Thrush-babbler, Horizorhinus, Sylvia. Dohrn’s Thrush-babbler Horizorhinus dohrni represents a monotypic genus endemic to the Gulf of Guinea island of Príncipe (Fig. 1). Despite being formally described nearly 150 years ago, issues regarding the validity of the genus name and the bird’s taxonomic position remain. Originally described under the name Cuphopterus by Hartlaub in 1866, this designation was subsequently abandoned due to the previous assignment (just months prior) of Cuphopterus to a new Hymenoptera genus (Deignan 1964). Although a new genus name (Cuphor- nis) was proposed for dohrni by Giebel in 1872, Horizo- rhinus as proposed by Oberholser has been employed since 1899 (Deignan 1964). Beyond the issue of the validity of the genus name, reliably placing Horizorhinus in a family has proven diffi- cult. Deignan (1964) placed the genus as incertae sedis after the Timaliinae (babblers) and before Panurinae (parrotbills) under Muscicapidae. Sibley and Monroe (1990; see also Watson et al. 1986) placed the genus inside Muscicapidae (Old World chats and flycatchers), while noting that it had been considered a babbler, a fly- catcher, a thrush (Turdidae) or a warbler (Sylviidae). Collar and Robson (2007) placed Horizorhinus in Timal- iidae, due to similarities in morphology and voice to several babbler genera. Most notably, Horizorhinus is close to Kupeornis in ethology and morphology (de Nau- rois 1994). However, Horizorhinus is also noted as shar- ing vocal similarities with the warbler genus Sylvia (Collar & Robson 2007). As part of a molecular systematic assessment of species and generic relationships in the family Muscicap- idae, we included Horizorhinus, according to its place- ment in that family in the Sibley and Monroe (1990) tapestry. Our initial assessments clearly indicated that Horizorhinus was not a muscicapid. Our goal in this paper was to finally place Horizorhinus unambiguously within a family, and to clarify its generic status. METHODS Our initial assessment that Horizorhinus was not in Muscicapidae was based on an analysis of over 200 mus- cicapid species and over 30 outgroup taxa (G. Voelker & R.C.K. Bowie unpubl. data). These outgroup taxa included members of the Monarchini, Vangini, Picathart- idae, Bombycillidae, Cinclidae, Sturnidae, Turdinae, Paridae, Sittidae, Polioptilinae, Parinae, Regulidae, Zoste- ropidae, Sylviinae, Passerinae, Motacillinae, Prunellinae, and several solely New World families. In this initial assessment, we relied on the mitochondrial ND2 gene to utilize additional sequence data from Timaliidae bab- bler taxa available on GenBank; Horizorhinus has been placed in that family recently (del Hoyo et al. 2006). Specifically, we included Illadopsis puveli (EU686320), Illadopsis cleaveri (EU686296), Illadopsis fulvescens (EU686315), Illadopsis pyrrhopterum (EU652715), Illad- opsis rufipennis (EU652716) and Turdoides plebejus (AY136603). Based on our initial assessment, we focused our sub- sequent analyses on Sylviinae (babblers, Wrentit Cha- maea fasciata and Sylvia) taxa, and relied on data from the ND2 and cytochrome b (cyt-b) genes. We sequenced these genes from Horizorhinus , Chamaea fasciata, and from all members of Sylvia except althaea and leucomel- aena (after del Hoyo et al. 2006). We used previously published protocols and primers (Outlaw et al. 2007) to generate sequence data. Products were purified using the Qiaquick PCR purification kit (Qiagen, Valencia, CA, USA). We performed 20-lL cycle sequencing reactions on purified PCR product using BigDye labelled termina- tors (ABI, Foster City, CA, USA). Reactions were purified and DNA precipitated using isopropanol. Cycle sequencing reactions were run out on an ABI377 auto- mated sequencer. Each sample was sequenced for both the light and heavy strands. All sequence data (1035 base pairs of ND2, 998 of cyt-b) were aligned using SEQUEN- CHER v. 4.5 (Gene Codes, Ann Arbor, MI, USA). In addi- tion to these taxa, we downloaded eight other babbler taxa from GenBank (Fig. 2). Sequences of Horizorhinus and Sylvia atricapilla have been deposited on GenBank *Corresponding author. Email: gvoelker@tamu.edu ª 2009 The Authors Journal compilation ª 2009 British Ornithologists’ Union Ibis (2009), 151, 580–583