MOLECULAR PHYLOGENETICS AND EVOLUTION

Molecular Phylogenetics and Evolution 27 (2003) 131142 www.elsevier.com/locate/ympev

Molecules and morphology: evidence for cryptic hybridization in African Hyalomma (Acari: Ixodidae)
David J. Rees,a,* Maurizio Dioli,b and Lawrence R. Kirkendalla
a

Department of Zoology, University of Bergen, Alle gaten 41, 5007 Bergen, Norway b Royal Veterinary College, London, UK Received 24 May 2002; revised 7 September 2002

Abstract The role of natural hybridization and introgression as part of the evolutionary process is of increasing interest to zoologists, particularly as more examples of gene exchange among species are identied. We present mitochondrial and nuclear sequence data for Hyalomma dromedarii, Hyalomma truncatum, and Hyalomma marginatum rupes (Acari: Ixodidae) collected from one-humped camels in Ethiopia. These species are well dierentiated morphologically and genetically; sequence data from the mitochondrial DNA (mtDNA) cytochrome oxidase I gene indicates 1014% divergence between the species. However, incongruence between morphology and the mtDNA phylogeny was observed, with multiple individuals of H. dromedarii and H. truncatum present on the same mtDNA lineage as H. marginatum rupes. Thus, individuals with morphology of H. dromedarii and H. truncatum are indistinguishable from H. marginatum rupes on the basis of mtDNA. Multiple copies of ITS-2 were subsequently cloned and sequenced for a subset of individuals from the mtDNA phylogeny, representing both normal and putative hybrid individuals. Very low sequence divergence (0.3%) was observed within normal individuals of both H. dromedarii and H. truncatum relative to the putative hybrid individuals (6 and 2.7%, respectively). The pattern of intra-individual variation in ITS-2 within putative hybrid individuals, particularly in H. dromedarii, strongly suggests that gene ow has occurred among these Hyalomma species, but no indication of this is given by the morphology of the individuals. 2002 Elsevier Science (USA). All rights reserved.
Keywords: Hyalomma; Ixodidae; mtDNA; ITS-2; Hybridization; Phylogeny

1. Introduction The signicance of hybridization and introgression as part of the evolutionary process has long been recognized by botanists but has traditionally received less attention from zoologists (Dowling and Secor, 1997). Natural hybridization has been hypothesized as acting as an evolutionary stimulus, with blocks of genes transferred between dierentially adapted systems, resulting in potential for adaptive shifts and evolutionary diversication (Anderson and Stebbins, 1954). Reluctance to assign an important evolutionary role to hybridization and introgression has arisen in part from a perceived rarity of animal hybrids, although it is emerging that gene exchange among animal species has
* Corresponding author. Fax: +47-555-89675. E-mail address: david.rees@zoo.uib.no (D.J. Rees).

been more common than previously believed (Dowling and Secor, 1997). Two factors that may contribute to the apparent rarity of animal hybrids are: (a) the traditional view that hybridization disrupts coadapted gene complexes, resulting in less t ospring that are then selected against (but see Barton, 2001), and (b) the difculty of detecting natural hybrids, particularly as they may not often be looked for. The tick genus Hyalomma Koch is represented in Africa, southern Europe, and Asia and inhabits regions with a long dry season (Matthysse and Colbo, 1987). The genus currently comprises 30 species and subspecies and has a complex history of synonymization (Camicas et al., 1998; Hoogstraal, 1956; Matthysse and Colbo, 1987; Pegram and Higgins, 1992). Ticks have adapted remarkably well to human domestication of livestock (Hoogstraal, 1978; Hoogstraal and Aeschlimann, 1982) and are of considerable economic importance through-

1055-7903/02/$ - see front matter 2002 Elsevier Science (USA). All rights reserved. doi:10.1016/S1055-7903(02)00374-3

132

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

out Africa due to both the direct eects of feeding on livestock and tick-borne diseases such as Theileria equi, T. annulata, and Babesia caballi (De Kok et al., 1993; Hoogstraal, 1956; Pegram et al., 1981). Species of Hyalomma are also known as potential vectors of many viral, bacterial, and protozoan pathogens that are of direct concern to humans. For example, H. marginatum marginatum is known as a potential vector of CrimeanCongo haemorrhagic fever (CCHF), tick-borne encephalitis, and of the rikketsia, Coxiella burnetti (Qfever) (De Kok et al., 1993; Hoogstraal, 1956). Ten species of Hyalomma have been recorded in collections from livestock (cattle, camels, sheep, and goats) in Ethiopia (Pegram et al., 1981). While the majority of these species are localized, three are relatively widespread: Hyalomma dromedarii, H. truncatum, and H. marginatum rupes. For this study, these species were collected from a herd of one-humped camels (Camelus dromedarius) in south-east Ethiopia. H. dromedarii is considered mainly a North African and Asian tick with a small population in East Africa (Hoogstraal, 1956) and the distribution of this species coincides with that of the most common host, the one-humped camel (Hoogstraal, 1956; van Straten and Jongejan, 1993). H. dromedarii has been described as the most completely desert-adapted ixodid tick of Africa (Pegram et al., 1981). Hyalomma marginatum rupes and H. truncatum are most commonly found on domestic cattle but also utilize other domestic and wild hosts (Cumming, 1998; Hoogstraal, 1956). Both species are widespread throughout Africa (Matthysse and Colbo, 1987) but absent from Eurasia and Arabia (Pegram and Higgins, 1992). Several studies involving collections of ticks from camels have noted a predilection in H. dromedarii for attachment in the nostrils, a specialized attachment site not utilized by other species of Hyalomma (e.g., Dioli, 1992; Dolan et al., 1983; Rutagwenda, 1985). The most common attachment sites for H. marginatum rupes and H. truncatum on both camels and cattle are the perineal/ inguinal area and the base of the tail (Baker et al., 1989; Dioli, 1992; Dolan et al., 1983; Mohammed, 1977; Rutagwenda, 1985). The initial impetus for a molecular analysis of this group of Hyalomma species came from the observation of a considerable number of ticks, typical of H. dromedarii morphology, utilizing attachment sites considered unusual for this species, primarily the anogenital region (M. Dioli, pers. obs.). Our original goal was to ascertain whether the H. dromedarii-like individuals from the anus could represent a cryptic species utilizing very dierent attachment sites than those generally associated with H. dromedarii. Maternally inherited mitochondrial DNA (mtDNA) sequences for the cytochrome oxidase I (COI) gene have previously been used in phylogenetic studies and have clearly separated Hyalomma species (Murrell et al., 2000) and this

gene was therefore selected for initial use. Following the results of the mtDNA study, a nuclear gene (the internal transcribed spacer ITS-2) was chosen to investigate possible hybridization among H. dromedarii, H. marginatum rupes, and H. truncatum. ITS-2 has been used in a number of tick phylogenetic studies at a range of scales (e.g., McLain et al., 1995; Wesson et al., 1993; Zahler et al., 1997) and the relatively high level of variability associated with this marker (including intra-individual variation, e.g., Rich et al., 1997) led to its selection over other candidates such as 28S, 16S, and 12S rDNA.

2. Materials and methods 2.1. Sample collection Ticks were collected by the second author from a domestic herd of one-humped camels near the town of Gode in south-east Ethiopia (5550 N, 43340 E) in October 1999 and June 2000. Hyalomma were also collected from camels near Isiolo in eastern Kenya (0210 N, 37350 E) (May 1998). Only male specimens were used in phylogenetic analyses because females of species in this genus can be dicult to distinguish. Samples in this study consisted of H. dromedarii collected from the nose and anus, and H. marginatum rupes and H. truncatum collected from the anus only. Extensive collection from the nose yielded no specimens of H. marginatum rupes or H. truncatum. No other Hyalomma species were present in collections. All specimens were stored in 70% EtOH, examined under a stereomicroscope and identied using available taxonomic keys (Hoogstraal, 1956; Matthysse and Colbo, 1987). Our species determinations were conrmed, for a voucher set of individuals used in the study, by J.L. Camicas of the Laboratoire dEpidemiologie des Maladies a Vecteurs, Montpellier, France (pers. comm.). All specimens used in this study are retained in the collection of M. Dioli. 2.2. DNA extraction, PCR amplication, and sequencing of COI Prior to DNA extraction, each sample was immersed in liquid nitrogen for 20 min. Entire individuals were then digested for 4 h and DNA subsequently column puried using reagents from a QIAamp DNA Mini Kit, following manufacturers recommendations. Individual ticks remained morphologically intact at the end of this DNA extraction process; 12 individuals were subsequently used to prepare scanning electron micrographs (see Fig. 4). The primers used for amplication of COI were TY-J-1449 50 -AATTTACAGTTTATCGCCT-30 (Murrell et al., 2000) and C1-N-2312 50 -CATACAAT AAAGCCTAATA-30 (Murrell et al., 2000). Together

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

133

these primers amplify a fragment of 863 bp. Polymerase chain reactions (PCRs) were performed in 25-ll volumes including 0.75 ll of each 10 lM PCR primer, 0.5 U of AmpliTaq DNA polymerase (PerkinElmer) and with a MgCl2 concentration of 2 mM. Two microlitres of DNA extract were used for amplication. Each of 40 PCR cycles comprised denaturation at 94 C for 30 s, annealing at 45 C for 1 min and extension at 72 C for 1 min. For some samples, annealing temperatures between 42 C and 47 C were necessary to achieve optimal amplication results. PCR products were column puried using a QIAquick PCR Purication Kit following manufacturers recommendations. Sequencing reactions using the PCR primers and the PerkinElmer BigDye terminator reaction mix were run on a PerkinElmer ABI automated DNA sequencer. The COI fragment was sequenced in both directions for all individuals and sequences were aligned by eye against the GenBank sequence of Dermacentor variabilis (Murrell et al., 2000). 2.3. PCR amplication, cloning, and sequencing of ITS-2 On the basis of the results of the COI sequencing, a subset of the samples from the mtDNA phylogeny was included in a study utilizing the nuclear marker ITS-2. Amplication of a portion of ITS-2 was performed using the primers of Zahler et al. (1997); RIB-8 50 GTCGTAGTCCGCCGTC-30 and a modied RIB-11 50 -GAGTACGACGCCCTACC-30 (this primer was modied by removal of the XbaI linker present in the original). Both primers are located in a variable region at the 30 end of ITS-2 and are conserved between Dermacentor spp. and Rhipicephalus sanguineus (Zahler et al., 1997). Together, these primers amplify a fragment of approximately 300 bp. PCRs were again performed in a 25-ll volume including 0.75 ll of each 10 lM PCR primer, 2 ll of DNA and 2 mM MgCl2 . AmpliTaq Gold DNA polymerase (PerkinElmer) was used in place of AmpliTaq to allow Hot Start PCR with minimal modication of reaction conditions. The PCR prole consisted of an initial enzyme activation step at 95 C for 10 min, followed by 40 cycles of denaturation at 94 C for 30 s and annealing/extension at 60 C for 1 min, and a nal step of 72 C for 10 min. Fresh (less than 1-dayold), unpuried PCR products were used in cloning reactions and transformation of chemically competent Escherichia coli (TOPO TA Cloning Kit for Sequencing, Invitrogen) following manufacturers recommendations. Transformation mixtures were then plated overnight at 37 C on LB plates containing 100 lg ml1 ampicillin. Colonies were then picked and transferred to 5 ml LB containing 50 lg ml1 ampicillin for overnight culture. Plasmids were then column puried using the Promega Wizard Plus SV minipreps DNA purication system following manufacturers recommendations. Sequencing reactions were performed as described for COI, using

the M13 forward and reverse primers supplied with the TOPO TA Cloning Kit (Invitrogen). 2.4. Phylogenetic analyses Published COI sequences for several species of Hyalomma were obtained from GenBank and included in our dataset. These were H. dromedarii (laboratory strain, Egypt), H. truncatum (South Africa), H. marginatum rupes (Zimbabwe), and H. aegyptium (laboratory strain, Belgium) (all from Murrell et al., 2000). Aligned COI sequence data were analyzed using PAUP* (Swoord, 1998). Maximum-parsimony (MP) analysis was performed using D. variabilis as an outgroup. ITS-2 sequence alignment was performed with the Multalin package (Corpet, 1988; available online via http:// prodes.toulouse.inra.fr/multalin/multalin.html) using the DNA comparison table, a gap weight of 5 and a gap length weight of 0. Haplotype networks were constructed from the aligned ITS-2 sequence data using the program TCS (Clement et al., 2000). This program allows the estimation of genealogical relationships from DNA sequences using the statistical parsimony method of Templeton et al. (1992). This method involves calculation of a pairwise distance matrix for haplotypes and calculation of the probability of parsimony for all pairwise comparisons until a cut-o point of 0.95 is exceeded. The maximum number of mutational dierences allowed between pairs of sequences under the parsimony criterion is determined by the number of dierences associated with the probability just before the 95% cut-o. This is the basis for connections made between haplotypes and the resulting graphical output represents the plausible connections made between haplotypes, including missing intermediates. DNA sequences have been deposited in the EMBL Nucleotide Sequence Database under Accession Nos. AJ437061AJ437101 (COI) and AJ437360AJ437402 (ITS-2).

3. Results 3.1. mtDNA analysis Aligned COI sequences, including those of Murrell et al. (2000) obtained from GenBank, consisted of 793 bp with variability at 29.9% of sites. Of the variable sites, 71.3% were parsimony-informative. Maximumparsimony analysis resulted in nine equally parsimonious trees that diered only in minor rearrangements of terminal branches. The maximum-parsimony phylogeny (Fig. 1) consists of three major mtDNA lineages. One lineage (A) comprises H. dromedarii from Kenya, the Egyptian laboratory strain, and several Ethiopian samples from both nose and anus attachment sites.

134

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

Fig. 1. One of the nine equally parsimonious trees for species of Hyalomma using mitochondrial COI sequence data. Branch lengths are proportional to substitutional change. Bootstrap values indicate nodes gaining more than 70% support (heuristic search, 500 replicates). Major clades are indicated by the letters A, B, and C. For Ethiopian H. dromedarii, samples collected from camels noses and anogenital region are indicated by N and A, respectively. Individuals subsequently used for cloning and sequencing of ITS-2 are indicated by an asterisk (*).

Maximum divergence within this clade is 0.5% (uncorrected). A second lineage (B) contains samples of H. truncatum from Ethiopia and the South African GenBank sequence. Divergence among the Ethiopian samples is 1.1% while between Ethiopia and South Africa, the maximum divergence is 10.3%. A third mtDNA lineage (C) comprises all H. marginatum rupes (Ethiopia plus the GenBank sequence from Zimbabwe) and also three individuals of H. truncatum and six H. dromedarii collected from the anus. Identical haplotypes are shared by H. marginatum rupes, H. dromedarii, and

H. truncatum. Maximum divergence within this clade is 3.8%. Minimum COI divergences between these lineages are 14.3% (AB), 12.2% (AC), and 10.0% (BC). 3.2. ITS-2 data On the basis of the mtDNA phylogeny, both normal (morphology + mtDNA) and putative hybrid (normal morphology but H. marginatum rupes-type mtDNA) individuals of H. dromedarii and H. truncatum were sequenced for multiple clones of ITS-2 along with indi-

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

135

viduals of H. m. rupes. For both H. dromedarii and H. truncatum, two normal individuals were cloned and sequenced, one for eight clones, and a second individual for two clones. Two individuals of H. marginatum rupes were sequenced, one for ve clones, and a second for two. One putative hybrid H. dromedarii was sequenced for seven clones, and two putative hybrid H. truncatum were sequenced; one for seven clones and a second for two clones. The aligned ITS-2 sequences are shown in Fig. 2. Construction of a haplotype network for the ITS2 sequence data using TCS resulted in two distinct networks (Fig. 3). These two networks dier by a minimum of 11 substitutions, two one-base and one three-base insertion/deletion (see Figs. 2 and 3). One network (a)

comprised only H. dromedarii clones. The sequences obtained from seven clones from one normal H. dromedarii (HY01) were identical (termed H. dromedarii type in Fig. 3 for simplicity), while the eighth diered from the majority by only one substitution. Two clones from a second normal individual (HY43) were also identical to this majority type. Of the seven clones sequenced for the putative hybrid H. dromedarii individual (HY37), two were identical to the majority H. dromedarii sequence type; the remainder were not connected to this network. The second network (b) generated by the TCS analysis consisted of all clones from H. truncatum and H. marginatum rupes individuals and also the remaining

Fig. 2. Alignment of polymorphic sites among 43 ITS-2 clones for Hyalomma individuals. Nucleotide positions relate to those following sequence alignment using Multalin. Sequence labels indicate individual code, clone number, and species, respectively (D, H. dromedarii; M, H. marginatum rupes; T, H. truncatum). Dashes represent alignment gaps and dots indicate sequence identity to H. dromedarii individual HY01, clone 1.

136

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

Fig. 3. Haplotype network resulting from the analysis of ITS-2 sequence data using TCS. Branches connecting haplotypes represent one-step mutations and small circles indicate missing haplotypes. Haplotypes with the highest outgroup probability are displayed as squares and other haplotypes are displayed as ovals. Where a haplotype comprises more than one sequence, the morphological species, individual code, and number of clones involved are listed (connected to the haplotype by a broken line). For unique haplotypes, the individual and clone number is shown within the oval (e.g., clone 2 from individual HY57 appears as HY57-2). Clones from individuals considered normal and putative hybrid on the basis of mtDNA are indicated by lled circles and crosses, respectively.

ve clones from the putative hybrid H. dromedarii (HY37). Two groupings are apparent in this second network, relating to the majority sequence types seen in H. marginatum rupes and H. truncatum. These two groups are dierentiated by two substitutions (Figs. 2 and 3). The sequences for three clones from one individual H. marginatum rupes (HY28) were identical and matched by the two clones from a second individual (HY52) (this sequence type has been termed H. marginatum rupes type in Fig. 3). Of the two remaining H. marginatum rupes clones, both diered from the majority type by two substitutions, but one (HY28 clone 10) was of identical sequence to the ITS-2 sequence observed in the majority of normal H. truncatum clones. Four of seven clones from the putative hybrid

H. dromedarii (HY37) were of identical sequence to the majority H. marginatum rupes type. Greater sequence variation was observed among H. truncatum clones than in H. dromedarii or H. marginatum rupes. For normal H. truncatum, seven clones had identical sequences and the eighth diered by one substitution (HY58). One clone from a second individual (HY33) was again identical to this majority H. truncatum type ITS-2 sequence, and a second diered by one substitution. One clone from the putative hybrid H. dromedarii (HY37) was identical to the majority ITS-2 sequence type observed in H. truncatum. For the putative hybrid H. truncatum, of seven clones from one individual (HY57), three were identical to the majority H. truncatum type, three diered from this by two

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

137

substitutions, and one was closer to the H. marginatum rupes majority type (diering by two substitutions). One clone from a second putative hybrid H. truncatum individual (HY54) was identical to the majority H. truncatum type, and a second clone diered from the H. marginatum rupes ITS-2 type by one substitution.

4. Discussion 4.1. Incongruence between morphology and mtDNA On the basis of mtDNA sequence data, none of the three Hyalomma species included in this study form a monophyletic group (Fig. 1). Multiple samples of both H. truncatum and H. dromedarii collected from the anus are indistinguishable from H. marginatum rupes in terms of mtDNA sequence. In the mtDNA phylogeny, the main H. dromedarii lineage (Fig. 1, lineage A) comprises the GenBank reference sequence (Murrell et al., 2000) and samples from Kenya, as well as all Ethiopian nose samples and multiple samples collected from the anus. The H. truncatum lineage (Fig. 1, lineage B) comprises the GenBank reference sequence from South Africa (Murrell et al., 2000) and two of ve individuals from Ethiopian camel herds, with signicant COI divergence between our Ethiopian samples and the published GenBank sequence for this species (10.3%). The H. marginatum rupes lineage (Fig. 1, lineage C) contains the GenBank reference sample from Zimbabwe (Murrell et al., 2000) and all Ethiopian H. marginatum rupes samples. In addition, this lineage contains multiple individuals of both H. dromedarii (anus only none from the nose) and H. truncatum. These individuals are morphologically typical of their respective species but are indistinguishable from H. marginatum rupes on the basis of mtDNA. Given that many dierent gene trees comprise a species tree (see, e.g., Avise, 2000; Doyle, 1997; Page and Charleston, 1997; Page and Holmes, 1998), several processes may be invoked to explain apparent paraphyly of a species in a phylogeny constructed from a single locus. These include: (i) convergent evolution, (ii) retention of ancestral genetic diversity, (iii) amplication of pseudogenes, and (iv) introgressive hybridization. The presence of individuals with H. dromedarii and H. truncatum morphology on the H. marginatum rupes mtDNA lineage is unlikely to be explained by convergent evolution of these two morphotypes. Specimens of H. dromedarii and H. truncatum possessing the H. marginatum rupes mtDNA type (those on lineage C in Fig. 1) do not just tend towards these species morphologically, they are indistinguishable from normal specimens (see Fig. 4). Persistence of an ancestral allelic polymorphism within a species can lead to incongruence between gene and species trees, particularly if the di-

vergence time between species is short (Pamilo and Nei, 1988). Several studies have invoked incomplete sorting of alleles to explain the presence of a species on multiple mtDNA lineages (e.g., Juan et al., 1996; Rees et al., 2001; Sperling et al., 1999; Vogler and DeSalle, 1993). For Hyalomma, however, incomplete sorting of mtDNA alleles is not a compelling explanation due to the depth of divergence between these Hyalomma species as a whole (1014.3% COI) indicating that these species are not of recent origin. Nuclear copies of mitochondrial sequences have been detected in a variety of organisms (reviewed by Bensasson et al., 2001; http://www.pseudogene.net). These nuclear mitochondrial pseudogenes (Numts) may be inadvertently amplied by PCR and may lead to incorrect phylogenetic reconstruction. Other than the unexpected placement of H. dromedarii and H. truncatum individuals on the H. marginatum rupes mtDNA lineage, no symptoms associated with Numts (Bensasson et al., 2001) were observed in Hyalomma. An alternative explanation that would account for the observed variation in both the mtDNA and ITS-2 data involves introgressive hybridization, that is, the incorporation of the genes of one species into the gene pool of another. Past or ongoing hybridization between males of H. dromedarii and H. truncatum and female H. marginatum rupes could account for the observed mtDNA phylogeny. This possibility was further investigated by cloning and sequencing multiple copies of an additional nuclear marker (ITS-2) for both normal and putative hybrid individuals. 4.2. Intra-individual and interspecic variation in ITS-2 Signicant intra-individual sequence variation among cloned copies of ITS-2 (which, being located in ribosomal DNA, occurs in multiple copies in eukaryote genomes) has previously been reported for ticks of the genus Ixodes (Rich et al., 1997). In that case, eight copies of ITS-2 (300 bp) were cloned and sequenced from each of two individuals, revealing 3.54.5% nucleotide polymorphism within individuals. This contrasts with other organisms where lower intra-individual ITS-2 variability has been found (e.g., mosquitoes, less than 2%: Wesson et al., 1992; Onyabe and Conn, 1999). In the Hyalomma individuals in this study, within-individual heterogeneity of ITS-2 copies is clearly not universal. In normal individuals of both H. dromedarii (HY01) and H. truncatum (HY58) the level of nucleotide polymorphism was 0.3% (1 substitution in 302 bp, eight clones from each individual). Far greater levels of sequence diversity were observed in the putative hybrid individuals; 6% in H. dromedarii (HY37; 18 nucleotide polymorphisms, seven clones) and 2.7% in H. truncatum (HY57; eight polymorphisms, seven clones). The pattern of intra-individual variation in ITS-2 sequence appears

138

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

Fig. 4. Scanning electron micrographs of Hyalomma individuals included in the molecular study (prepared after DNA extraction). Dorsal view, ventral view, and spiracular plates, respectively, are shown for normal (morphology and mtDNA) individuals of H. dromedarii (a, b, c: all from individual HY09), H. truncatum (d, e, f: all from HY33), and H. marginatum rupes (g, h, i: all from HY52). Morphology of putative hybrid (normal morphology but H. marginatum rupes type mtDNA) individuals is also shown: H. dromedarii (j, k, l: HY42 dorsal, ventral; HY44 spiracular plate) and H. truncatum (m, n, o: HY54 dorsal; HY57 ventral, spiracular plate).

to support the existence of gene ow among Hyalomma species. Clones from normal H. dromedarii show very little intra-individual sequence variation (one substitution in a total of 10 clones). In contrast, clones from the putative hybrid H. dromedarii individual (HY37) displayed signicant sequence variation, with clones identical to the majority types found within H. marginatum rupes and H. truncatum as well as normal H. dromedarii (involving 1012 substitutions, one three-base and three one-base indels between clones; Fig. 2). Although other factors could account for the presence of multiple, divergent copies of ITS-2 within an individual (e.g., retention of ancestral polymorphic copies despite concerted evolution of rDNA), the fact that intra-individual variation was only found in the putative hybrid individual, and that the divergent copies were identical to

those found in other species, makes interspecic hybridization a more plausible explanation. The pattern of intraspecic variation in ITS-2 observed in H. dromedarii was also evident in H. truncatum. For normal H. truncatum, low intra-individual divergence was observed (one substitution in eight clones from one individual, and two substitutions in one of two clones from a second individual). For putative hybrid H. truncatum, three clones from one individual (HY57) matched the sequence predominant in normal H. truncatum, while three others diered from this sequence by two substitutions. A seventh clone diered from the majority sequence of H. marginatum rupes by two substitutions (and by four from H. truncatum). For two clones from a second putative hybrid H. truncatum individual (HY54), one clone matched the majority

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

139

Fig. 4. (continued)

H. truncatum sequence and the second diered from the H. marginatum rupes sequence by one substitution. No clones from putative hybrid H. truncatum were identical to the majority H. marginatum rupes type. This may reect sequence divergence since a past hybridization event between these species, or alternatively, it may reect the presence of greater diversity of ITS-2 types within H. marginatum rupes. Among the H. marginatum rupes clones, ve are identical, one diers from this sequence by two substitutions, and one is identical to the majority H. truncatum sequence (Figs. 2 and 3). The presence of H. truncatum-type ITS within individual H. marginatum rupes could be accounted for in two ways, (a) retention of ancestral ITS within H. marginatum rupes or (b) introgression of H. truncatum nuclear genes into the H. marginatum rupes genome as a result of hybridization. Either of these is possible but the latter is perhaps more compelling, given the other indications of hybridization among these Hyalomma species. 4.3. Natural hybridization in Hyalomma Because of the absence of individuals of intermediate morphology, hybridization among the Hyalomma species in this study was not initially suspected, however, the genetic data suggest that natural hybridization has indeed occurred. Potential for hybridization among

species of Hyalomma has been noted by previous workers (e.g., Matthysse and Colbo, 1987), and experimental interspecic crosses occasionally produce juvenile or adult ospring in Hyalomma (Cwilich and Hadani, 1963; Pervomaisky, 1954) and in the closely related genus Rhipicephalus (Pegram et al., 1987; Zivkovic et al., 1986). Signicantly, in two such studies (Pervomaisky, 1954; Zivkovic et al., 1986), F1 progeny usually inherited the morphology of one of the parental species (rather than showing intermediate morphology). Fig. 5 illustrates two possible pathways to hybrid individuals that account for both the observed morphology and genetic data in this study. We assume that the majority-type ITS-2 sequence observed in H. truncatum and H. marginatum rupes indicates gene ow between these species and does not merely represent the retention of ancestral polymorphism. We attempt to account for the presence of the following classes of individuals revealed by this study: (a) H. truncatum morphology, H. marginatum rupes mtDNA, ITS-2 from both species, (b) H. dromedarii morphology, H. marginatum rupes mtDNA, ITS-2 from H. dromedarii, H. marginatum rupes, and H. truncatum, (c) H. marginatum rupes morphology, H. marginatum rupes mtDNA and ITS-2 from both H. marginatum rupes and H. truncatum. The mtDNA (matrilineal) phylogeny indicates the involvement of males of H. dromedarii and

140 D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142 Fig. 5. Simplest routes to hybrid individuals observed in this study. The three types of hybrid individuals suggested by mtDNA and ITS-2 sequence data are accounted by direct inheritance of paternal morphology in hybrid ospring (1) and by hybridization followed by backcrossing (2).

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142

141

H. truncatum and female H. marginatum rupes. In the case of direct inheritance of paternal morphology (Fig. 5.1), the classes of individuals (a)(c) described above can therefore be accounted for by the following sequence of events: an initial hybridization event between male H. truncatum and female H. marginatum rupes would result in male hybrid ospring with H. truncatum morphology, H. marginatum rupes mtDNA, and ITS-2 characteristic of both species (a). Secondary hybridization events between female progeny from this initial event and males of either H. dromedarii or H. marginatum rupes would result in individuals with (b) H. dromedarii morphology, H. marginatum rupes mtDNA, and ITS-2 from H. dromedarii, H. marginatumrupes, and H. truncatum, and (c) H. marginatum rupes morphology, H. marginatum rupes mtDNA and ITS-2 from both H. marginatum rupes and H. truncatum. An alternative pathway to the observed individuals, involving backcrossing, is presented in Fig. 5.2. In this scenario, an initial hybridization event between male H. truncatum and female H. marginatum rupes is followed by repeated backcrossing of female progeny to males of each of the three species, for as many generations as necessary to re-establish typical morphology. Problems exist with both of these scenarios, and at the present time we are unable to determine which process, if either, has led to the existence of the individuals we have observed in this study. Further eld and laboratory studies are required to determine the extent and frequency of hybridization in Hyalomma as well as the mechanisms leading to maintenance or re-establishment of typical morphology. 4.4. Concluding remarks If intermediate individuals are observed in the eld, then the possibility of interspecic hybridization is evident. However, the indications from experimental crosses (Pervomaisky, 1954; Zivkovic et al., 1986) and our study involving a natural population are that, in some cases, morphology alone may not indicate the hybrid origin of individuals. In the case of Hyalomma, hybridization was not suspected among the three species and hybrid individuals were only identied by molecular analysis involving multiple, independent markers. It is disconcerting to consider that the pattern evident in these Hyalomma may be repeated in other species and genera. Inferences from gene trees concerning relationships among populations or species may be incorrect if the gene tree does not accurately reect organismal phylogeny. The eects of introgression on levels of genetic variability within individuals and populations, and among populations and species, may be extremely difcult to assess, and may go unnoticed if a study involves only one species among several which are or have been hybridizing. Concerning the importance of identifying

introgression, as Moore (1995) reminds us, if hybridization is apparent, its eects can be taken into account and its occurrence incorporated into the historical narrative of the groupas it should be. As a cautionary measure, particularly in cases where hybridization is a possibility, we emphasize that molecular phylogenetic studies should not rely on sequence data from a single individual as being representative of a species. Additionally, in analyses of nuclear genetic diversity among populations or species, it is wise to make an initial assessment of the level of intra-individual variability in several individuals. Clearly, dierent interpretations would have resulted from assessment of either normal (0.3% ITS-2 variability) or putative hybrid H. dromedarii (6% ITS-2 variability), although both individuals would be treated the same on the basis of morphology. A number of questions are raised by our ndings. Are Hyalomma hybrids fertile? If so, what are the tness consequences of hybridization? By what mechanism is species-specic morphology maintained? Does hybridization aect the role of these species as disease vectors? Are other species of Hyalomma involved? How widespread and how frequent is hybridization in Hyalomma and does it occur in other genera? Is hybridization linked to host use? These and further challenges await.

Acknowledgments We express our thanks to Dr. M.T. Fox, Department of Pathology and Infectious Diseases, Royal Veterinary College, London and Paul Hylliard and Janet Beccaloni of the Natural History Museum, London, for advice on Hyalomma and loan of type specimens. Thanks go to Prof. J.L. Camicas, Laboratoire dEpidemiologie des Maladies a Vecteurs, Montpellier, France for conrmation identication of samples. We thank Dr. Sileshi Mekonnen, Dr. Ibrahim Hussein, Dr. Nigist Mekonnen, and Mr. Abebe Mekonnen of the Acarology and Entomology Team at the Sabeta National Animal Health Center, Ethiopia, for help and advice on collecting. We also thank Eigil Erichsen, Laboratory for Electron Microscopy, University of Bergen for invaluable assistance with the scanning microscope. Helpful comments on the manuscript were provided by Brent Emerson and Endre Willassen, Rob DeSalle, and an anonymous referee. Funding was provided by the Norwegian Research Council (NFR), Project No. 128388/420, Applications of Molecular Techniques in Systematic Biology. References
Anderson, E., Stebbins Jr., G.L., 1954. Hybridization as an evolutionary stimulus. Evolution 8, 378388. Avise, J.C., 2000. Phylogeography: The History and Formation of Species. Harvard University Press, Cambridge, MA.

142

D.J. Rees et al. / Molecular Phylogenetics and Evolution 27 (2003) 131142 Page, R.D.M., Charleston, M.A., 1997. From gene tree to organismal phylogeny: reconciled trees and the gene tree/species tree problem. Mol. Phylogenet. Evol. 7, 231240. Pamilo, P., Nei, M., 1988. Relationships between gene trees and species trees. Mol. Biol. Evol. 5, 568583. Pegram, R.G., Higgins, A.J., 1992. Camel ectoparasites: a review. In: Allen, W.R., Higgins, A.J., Mayhew, I.G., Snow, D.H., Wade, J.F. (Eds.), Proc. 1st Int. Camel Conf., pp. 6978. Pegram, R.G., Hoogstraal, H., Wassef, H.Y., 1981. Ticks (Acari: Ixodoidea) of Ethiopia. I. Distribution, ecology, and host relationships of species infesting livestock. Bull. Ent. Res. 71, 339359. Pegram, R.G., Walker, J.B., Cliord, C.M., Keirans, J.E., 1987. Comparison of populations of the Rhipicephalus simus group: R. simus, R. praetextatus, and R. muhsamae (Acari: Ixodidae). J. Med. Entomol. 24 (6), 665682. Pervomaisky, G.S., 1954. Variation in pasture ticks (Acarina, Ixodidae) and its signicance for systematics. Trud. Vsesoyuz. Ent. Obshch. 44, 62201. Rees, D.J., Emerson, B.C., Orom i, P., Hewitt, G.M., 2001. Reconciling gene trees with organism history: the mtDNA phylogeography of three Nesotes species (Coleoptera: Tenebrionidae) on the western Canary Islands. J. Evol. Biol. 14, 139147. Rich, S.M., Rosenthal, B.M., Telford III, S.R., Spielman, A., Hartl, D.L., Ayala, F.J., 1997. Heterogeneity of the internal transcribed spacer (ITS-2) region within individual deer ticks. Insect Mol. Biol. 6, 123129. Rutagwenda, T., 1985. The control of important camel diseases in the IPAL study area. In: IPAL Technical Report No. E-7. UNESCO, Nairobi, Kenya, pp. 1075. Sperling, F.A.H., Raske, A.G., Otvos, I.S., 1999. Mitochondrial DNA sequence variation among populations and host races of Lambdina scellaria (Gn.) (Lepidoptera: Geometridae). Insect Mol. Biol. 8, 97106. Straten, M. van, Jongejan, F., 1993. Ticks (Acari: Ixodidae) infesting the Arabian Camel (Camelus dromedarius) in the Sinai, Egypt with a note on the acaricidal ecacy of Ivermectin. Exp. Appl. Acarol. 17, 605616. Swoord, D.L., 1998. PAUP*. Phylogenetic Analysis Using Parsimony (* and Other Methods), Version 4. Sinauer Associates, Sunderland, MA. Templeton, A.R., Crandall, K.A., Singh, C.F., 1992. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogram estimation. Genetics 132, 619633. Vogler, A.P., DeSalle, R., 1993. Mitochondrial DNA evolution and the application of the phylogenetic species concept in the Cicindella dorsalis complex (Coleoptera: Cicindelidae). In: Desender, K. (Ed.), Carabid Beetles: Ecology and Evolution. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 7985. Wesson, D.M., McLain, D.K., Oliver, J.H., Piesman, J., Collins, F.H., 1993. Investigation of the validity of species status of Ixodes dammini (Acari: Ixodidae) using rDNA. Proc. Natl. Acad. Sci. USA 90, 1022110225. Wesson, D.M., Porter, C.H., Collins, F.H., 1992. Sequence and secondary structure comparisons of ITS rDNA in mosquitoes (Diptera: Culicidae). Mol. Phylogenet. Evol. 1, 253269. Zahler, M., Fillipova, N.A., Morel, P.C., Gothe, R., Rinder, H., 1997. Relationships between species of the Rhipicephalus sanguineus group: a molecular approach. J. Parasitol. 83, 302306. Zivkovic, D., Pegram, R.G., Jongejan, F., Mwase, E.T., 1986. Biology of Rhipicephalus appendiculatus and R. zambeziensis and production of a fertile hybrid under laboratory condtions. Exp. Appl. Acarol. 2, 285298.

Baker, M.K., Ducasse, F.B.W., Sutherst, R.W., Maywald, G.F., 1989. The seasonal tick populations on traditional and commercial cattle grazed at four altitudes in Natal. J. S. Afr. Vet. Assoc. 60, 95101. Barton, N.H., 2001. The role of hybridization in evolution. Mol. Ecol. 10, 551568. Bensasson, D., Zhang, D.-X., Hartl, D.L., Hewitt, G.M., 2001. Mitochondrial pseudogenes: evolutions misplaced witnesses. Trends Ecol. Evol. 16, 314321. Camicas, J.-L., Hervy, J.-P., Adam, F., Morel, P.C., 1998. The Ticks  ditions de lOrstom, Paris. of the World (Acarida, Ixodida). E Clement, M.D., Posada, D., Crandall, K.A., 2000. TCS: a computer program to estimate gene genealogies. Mol. Ecol. 9, 16571660. Corpet, F., 1988. Multiple sequence alignment with hierarchical clustering. Nucleic Acids Res. 16, 1088110890. Cumming, G.S., 1998. Host preference in African ticks (Acari: Ixodida): a quantitative data set. Bull. Ent. Res. 88, 379406. Cwilich, R., Hadani, A., 1963. Inter-specic hybridization of ticks of the genus Hyalomma. Acta Trop. 20, 178180. De Kok, J.B., dOliveira, C., Jongejan, F., 1993. Detection of the protozooan parasite Theileria annulata in Hyalomma ticks by the polymerase chain reaction. Exp. Appl. Acarol. 7, 846849. Dioli, M., 1992. A study of tick infestation in four nomadic camel herds in Northern Kenya. In: Schwartz, H.J., Dioli, M. (Eds.), The One-humped Camel (C. dromedarius) in Eastern Africa: A Pictorial Guide to Disease, Health Care, and Management. Margraf, Weikersheim, pp. 263267. Dolan, R., Wilson, A.J., Schwartz, H.J., Nelson, R.M., Field, C.R., 1983. Camel production in Kenya and its constraints. II. Tick infestation. Trop. Anim. Health Prod. 15, 179185. Dowling, T.E., Secor, C.E., 1997. The role of hybridization and introgression in the diversication of animals. Ann. Rev. Ecol. Syst. 28, 593619. Doyle, J.J., 1997. Trees within trees: genes and species, molecules and morphology. Syst. Biol. 46, 537553. Hoogstraal, H., 1956. African Ixodoidea. I. Ticks of the Sudan Bur. Med. Surg., Dept. Navy. US Gov. Print. O. (USGPO), Washington. Hoogstraal, H., 1978. Biology of ticks. In: Wilde, J.K.H. (Ed.), Tickborne Diseases and their Vectors. Centre for Tropical Veterinary Medicine, Edinburgh, pp. 314. Hoogstraal, H., Aeschlimann, A., 1982. Tick-host specicity. Bull. Soc. Ent. Suisse 55, 532. Juan, C., Ibrahim, K.M., Orom , P., Hewitt, G.M., 1996. Mitochondrial DNA sequence variation and phylogeography of Pimelia darkling beetles on the island of Tenerife (Canary Islands). Heredity 77, 589598. McLain, D.K., Wesson, D.M., Oliver Jr., J.H., Collins, F.H., 1995. Variation in ribosomal DNA internal transcribed spacers 1 among eastern populations of Ixodes scapularis (Acari: Ixodidae). J. Med. Entomol. 32, 353360. Matthysse, J.G., Colbo, M.H., 1987. The Ixodid Ticks of Uganda. Entomological Society of America, Maryland. Mohammed, A.N., 1977. The seasonal incidence of ixodid ticks of cattle in northern Nigeria. Bull. Anim. Health Prod. Afr. 25, 273293. Moore, W.S., 1995. Inferring phylogenies from mtDNA variation: mitochondrial-gene trees versus nuclear-gene trees. Evolution 49, 718726. Murrell, A., Campbell, N.J.H., Barker, S.C., 2000. Phylogenetic analyses of the Rhipicephaline ticks indicate that the genus is paraphyletic. Mol. Phylogenet. Evol. 16, 17. Onyabe, D.Y., Conn, J.E., 1999. Intragenomic heterogeneity of a ribosomal DNA spacer (ITS2) varies regionally in the neotropical malaria vector Anopheles nuneztovari (Diptera: Culicidae). Insect Mol. Biol. 8, 435442.