Characterization and Evolution of the Mitochondrial DNA Control Region in Hornbills (Bucerotiformes) Wayne Delport, 1,2 J. Willem H. Ferguson, 1 Paulette Bloomer 2 1 Department of Zoology and Entomology, University of Pretoria, Pretoria 0002, South Africa 2 Molecular Ecology and Evolution Program, Department of Genetics, University of Pretoria, Pretoria 0002, South Africa Received: 3 April 2001 / Accepted: 7 December 2001 Abstract. We determined the mitochondrial DNA control region sequences of six Bucerotiformes. Horn- bills have the typical avian gene order and their control region is similar to other avian control regions in that it is partitioned into three domains: two variable domains that flank a central conserved domain. Two characteris- tics of the hornbill control region sequence differ from that of other birds. First, domain I is AT rich as opposed to AC rich, and second, the control region is approxi- mately 500 bp longer than that of other birds. Both these deviations from typical avian control region sequence are explainable on the basis of repeat motifs in domain I of the hornbill control region. The repeat motifs probably originated from a duplication of CSB-1 as has been de- termined in chicken, quail, and snowgoose. Furthermore, the hornbill repeat motifs probably arose before the di- vergence of hornbills from each other but after the di- vergence of hornbills from other avian taxa. The mito- chondrial control region of hornbills is suitable for both phylogenetic and population studies, with domains I and II probably more suited to population and phylogenetic analyses, respectively. Key words: Mitochondrial DNA — Control region — Bucerotiformes — Characterization — Repeats — Sec- ondary structure — Phylogenetics — Population genet- ics Introduction The mitochondrial DNA genome has proven to be useful in many avian phylogenetic and population studies (re- viewed by Baker and Marshall 1997; Mindell et al. 1997; Moore and DeFilippis 1997). Advantages of working with mtDNA include maternal inheritance (Lansman et al. 1983), a higher rate of evolution than in single-copy nuclear DNA (Brown et al. 1979), and variable evolu- tionary rates of genes within the genome itself (Aquadro and Greenburg 1983; Cann et al. 1984). Therefore it is possible to address both phylogenetic and population- level questions using different genes of the same genome (Wenink et al. 1994). Traditionally coding genes of the mitochondrial genome have been reserved for phyloge- netic studies (Moore and DeFilippis 1997), whereas the control region has been considered more suitable for population level studies (Baker and Marshall 1997). The control region evolves three to five times more rapidly than the remainder of the mitochondrial genome (Aquadro and Greenburg 1983; Cann et al. 1984) and therefore can be difficult to amplify with universal ver- tebrate primers in a polymerase chain reaction (PCR). In addition, both nuclear integration (Sorenson and Fleischer 1996; Zhang and Hewitt 1996), although not restricted to the control region, and repeat motifs (Des- jardins and Morais 1990, 1991; Quinn and Wilson 1993; Ramirez et al. 1993; Randi and Lucchini 1998; Wenink et al. 1994) could render PCR amplification difficult. Part of the control region is thought to regulate tran- scription and replication of the vertebrate mtDNA ge- nome (Brown et al. 1986; Clayton 1991). These areas that may have function associated-secondary structure Correspondence to: Wayne Delport, Molecular Ecology and Evolution Program; email: wdelport@postino.up.ac.za J Mol Evol (2002) 54:794–806 DOI: 10.1007/s00239-001-0083-0 © Springer-Verlag New York Inc. 2002