INTRODUCTION

The subfamily Lygosominae contains over 600 species (Greer, 1970a; Matsui, 1992; Zug, 1993). Within this subfamily, three evolutionary lineages (i.e., Eugongylus, Mabuya and Sphenomorphus groups) are recognized on the morphological, karyological and immunogenetic grounds (e.g., King, 1973, 1990; Greer, 1979, 1989; Hardy, 1979; Baverstock and Donnellan, 1990; Donnellan, 1991a, b; Ota et al., 1988, 1991, 1995, 1996). Of these, the Mabuya group is mainly distributed in temperate and tropical Asia, central and southern Africa, and Australia. Mabuya, the largest genus of this group with broadest range, also occurs in Madagascar and South America including the West Indian Islands, but is not distributed in Australia (Boulenger, 1887; Matsui, 1992; Nussbaum and Raxworthy, 1994).

Three arboreal genera (Apterygodon, Dasia [sensu stricto] and Lamprolepis) and one terrestrial or semi-fossorial genus (Lygosoma [sensu Greer, 1977]) have been assigned to the Mabuya group together with Mabuya and a few other African and Australian genera. Of these, the former three taxa had been grouped together as the genus Dasia sensu lato (Smith, 1937; Mittleman, 1952), when Greer (1970b) proposed the current generic arrangements on the basis of morphological characters. He also argued that the Apterygodon–Dasia lineage and the Lamprolepis lineage had evolved independently from a Mabuya-like stock in Southeast Asia. With an extension of this view, Greer (1977) considered that, besides the genus Mabuya, those two arboreal lineages, Lygosoma, Australian members of the Mabuya group, the Eugongylus group, and the Sphenomorphus group constitute six phylogenetic lineages independently derived from the Asian Mabuya-like stock (he argued for the subsequent derivations of the African endemic genera of the Mabuya group from the Mabuya-like stock within this continent). However, the chronological order of these divergences was not hypothesized in that work. Later, Australian members of the Mabuya group, the Eugongylus group and the Sphenomorphus group were attributed to divergences earlier than that in Asian and African members of the Mabuya group (Greer, 1979, 1989). The remaining three lineages, Apterygodon–Dasia, Lamprolepis and Lygosoma, as well as Mabuya, are still considered as derived from the Mabuya-like stock in Asia (Greer, 1977), although their detailed relationships remain uncertain.

The genus Mabuya seems to have first emerged in South or Southeast Asia and then dispersed through Africa onto Madagascar and South America, because a few species from South and Southeast Asia exhibit most primitive states of characters among the extant Mabuya species (Greer, 1977). Although some authors (e.g., Greer, 1977) pointed out the possible non-monophyly of this genus due to its wide distribution and great morphological diversity, no comprehensive phylogenetic analyses have ever been made for the genus and its relatives to verify this prediction.

There have been a number of debates regarding the phylogenetic relationships and classification of lygosomine skinks, and most of relevant arguments have depended on morphological evidence (e.g., Mittleman, 1952; Greer, 1970a, 1974, 1979; Horton, 1972, 1973). However, due to the scarcity of informative characters, it is not easy to formulate a sufficiently reliable phylogenetic hypothesis for this group solely on the morphological ground. Phylogenetic analyses on the basis of molecular data are, therefore, expected to much contribute to the solution of this problem.

We sequenced a part of mitochondrial DNA for representatives of Asian and African Mabuya, and the three other lineages supposedly derived from the Mabuya-like stock in Asia (see above), and analyzed resultant data phylogenetically. The purpose of this study is to reveal the pattern and process in the early evolution of the widespread and apparently substantially diverged Mabuya group in Asia and Africa.

MATERIALS AND METHODS

Tissues were obtained from eight Southeast Asian species belonging to five genera of the Mabuya group (Apterygodon vittatus, Dasia gricea, D. olivacea, Lamprolepis smaragdina, Lygosoma bowringii, Mabuya longicaudata, M. multifasciata and M. rudis), and two African Mabuya (M. quiquetaeniata and M. striata) (Table 1, see Appendix for further detail). We selected Eumeces latiscutatus of the subfamily Scincinae, a possible closest relative of Lygosominae (Greer, 1970a), as an outgroup for which tissues were available to us.

Table 1

Distribution of the genera of the Mabuya group. Asterisk (*) indicates taxonomic and/or geographic groups studied in the present analysis. (a) including western Oceanian islands; (b) including Madagascar; (c) including West Indies Islands. See Appendix for detailed localities.

Small amounts of livers, removed from anesthetized or dead specimens and stocked at −80°C, were homogenized in extraction buffer [150 mM NaCl, 10 mM Tris-HCl (pH 8.0), 10 mM EDTA, 1% sodium dodecyl sulfate]. After digesting samples with Proteinase K (100 μg/ml) at 50°C for three hours, DNA was extracted with phenol (three times) and 25:24:1 of phenol/chloroform/isoamyl-alcohol (once), and was precipitated in ethanol with one-tenth volume of 3.0 M sodium acetate (pH 5.2). Samples resuspended in TE buffer were further purified by RNase digestion (20 μg/ml) at 37°C for one hour, followed by ethanol precipitation. DNA amplification and sequencing are described in detail elsewhere (Honda et al., 1999). A part of mitochondrial 12S and 16S rRNA genes was amplified by the polymerase chain reaction (PCR) using primer L1091 (5′-AAACTGGGATTAGATACCCCACTAT-3′) and H1478 (5′-GAGGGTGACGGGCGGTGTGT-3′), and L2606 (5′-CTGACCGTGCAAAGGTAGCGTAATCACT-3′) and H3056 (5′-CTCCGGTCTGAACTCAGATCACGTAGG3′), respectively (Kocher et al., 1989; Hedges et al., 1993). The numbering system followed that for the human sequence (Anderson et al., 1981).

Alignments for DNA sequences were determined based on maximum nucleotide similarity. We prepared a pairwise matrix of distance by Kimura's (1980) two-parameter model. The neighbor-joining (NJ) method (Saitou and Nei, 1987) was applied to infer relationships among taxa on the basis of the distance matrix. The degree of supports for internal branches of each tree was assessed by 1,000 bootstrap replications (Felsenstein, 1985). These analyses were performed by use of Clustal W 1.4 (Thompson et al., 1994). Maximum parsimony analysis (MP) was also performed using PAUP 3.1.1 with heuristic option (Swofford, 1993). In this analysis, each nucleotide base was regarded as a character and four kinds of salt as different character states. No frequency bias was assumed for transition and transversion. The confidence was assessed by 1,000 bootstrap resamplings (Felsenstein, 1985). In both analyses, gap sites were excluded.

RESULTS

Aligned sequences from two mitochondrial genes are presented in Fig. 1. The 12S rRNA fragment consisted of 389 total sites, 157 of which were variable. In the 16S rRNA fragment, there were 436 total aligned sites, 144 of which were variable. Intergeneric nucleotide replacements between five lygosomine genera varied from 70 base pairs (bp) (Apterygodon vittatus vs Dasia gricea) to 152 bp (Lamprolepis smaragdina vs Mabuya longicaudata). Nucleotide replacements between congeneric species of Dasia and Mabuya were observed in 73 and 74 bp (D. grecea vs D. olivacea from Borneo and Malay Peninsula, respectively), and from 82 bp (M. multifasciata vs M. rudis) to 121 bp (M. longicaudata vs M. quiquetaeniata or M. striata), respectively. Intraspecific nucleotide replacements of D. olivacea involved 18 bp (Malay Peninsula vs Borneo), whereas there were no replacements between two samples of L. smaragdina (Guam vs Saipan).

Fig. 1

Aligned sequences of a 825 bp segment of the 12S and 16S rRNA genes. The initial 389 bp in each row correspond to 12S rRNA gene sequence. The 16S rRNA gene sequence begins at the asterisk. Dot indicates an identity with the first sequence; dash denotes a gap.

The NJ dendrogram derived from mitochondrial DNA distance matrix (not given) is shown in Fig. 2A. The ingroup portion of this dendrogram was divided into two major clusters, of which one, consisting of Lamprolepis and Lygosoma, was completely supported in bootstrap iterations (100%). The other major cluster, supported in 94% of bootstrap iterations, contained Apterygodon, Dasia and Mabuya. The latter cluster was further split into two subclusters consisting of African Mabuya (99%), and Asian Mabuya, Apterygodon and Dasia (71%), respectively. Within the latter, Apterygodon and Dasia (86%), and three Asian Mabuya examined (93%) constituted lower subclusters. Conspecific samples exclusively constituted lowest clusters in all iterations (100%).

Fig. 2

(A) Neighbor-joining (NJ) dendrogram derived from distance matrix from 12S and 16S rRNA sequence data. Numbers at branch indicate bootstrap proportions in 1,000 bootstrap pseudoreplications. Branches without BP values were not supported in ≥50% of the replicates. Bar equals 0.1 Kimura's two-parameter distance. “Asia” includes the western Oceanian islands. (B) Maximum parsimony (MP) cladogram using heuristic bootstrapping analysis (691 steps, 211 bp informative under the condition of parsimony, consistency index=0.56). Branches without BP values were not supported in ≥50% of the replicates.

Resultant cladogram of MP (Fig. 2B) showed no substantial inconsistency with the NJ dendrogram in terms of branching topology, although Apterygodon, Dasia and the Asian members of Mabuya did not constituted an exclusive cluster.

DISCUSSION

On the basis of differences in skull and external morphology, Greer (1970b) thought that Apterygodon and Dasia (sensu stricto) are monophyletic among the three arboreal genera formally assigned to Dasia (sensu lato), whereas Lamprolepis emerged independently from the Asian Mabuya-like stock. Later, he emphasized this view by arguing that the Apterygodon–Dasia lineage, Lamprolepis, and the terrestrial/semi-fossorial Lygosoma constitute the three distinct phylogenetic lines independently derived from the Asian Mabuya-like stock (Greer, 1977). Karyological data (Ota et al., 1996) also offered a circumstantial support to Greer's (1977) view by indicating closer chromosomal similarities of the three arboreal genera with Asian species of Mabuya than with African congeners or other lygosomine groups. However, phylogenetic relationships inferred from DNA sequences in the present study do not support Greer's (1977) view with respect to the independent origins of Lamprolepis and Lygosoma, because these two genera exclusively constituted a cluster. Moreover, our results strongly suggest that the collective divergence of these two genera have occurred prior to the separation between the African Mabuya and the Asian Mabuya–Apterygodon–Dasia clade. These may contradict with Greer's (1977) view, which seemingly assumed that Lamprolepis and Lygosoma have derived from the Mabuya-like stock within Asia.

Based on the morphological character, Greer (1976, 1977) assumed that the African endemic genera of the Mabuya group and African species of Mabuya were derived from the Mabuya-like stock through in situ continental radiation rather than from multiple colonizations from outside. Relationships depicted in Fig. 2A and 2B do not contradict with the postulated monophyly of African members of the Mabuya group, although the number and size of samples examined are too small to draw any definite conclusion by this result alone.

Considering our results, the Asian and African members of the Mabuya group are likely to constitute two major evolutionary lineages, which may be referred to as the Lygosoma and Mabuya subgroups. Ecological similarity, involving morphological specialization to arboreal habits (e.g., Greer, 1970b), between the Apterygodon–Dasia clade of the Mabuya subgroup and Lamprolepis of the Lygosoma subgroup thus seem to be attributable to the convergence rather than to the recent common ancestry. Morphological and karyological similarities among the Apterygodon–Dasia clade, Lamprolepis, Lygosoma and Asian species of Mabuya (Greer, 1970b, Ota et al., 1996) are supposed to be symplesiomorphy, although a few species of Lygosoma seems to have differentiated karyotypes (de Smet, 1981).

Greer (1970b, 1977) thought that Apterygodon and Dasia sensu stricto are monophyletic, constituting one of the distinct phylogenetic lineages derived from the Asian Mabuya-like stock. This view was confirmed by the present results. Apterygodon differs from Dasia and Mabuya in having an ectopterygoid process and a karyotype of 2N=28 format, and in lacking pterygoid teeth: both of the latter have basically 2N=32 format karyotypes and pterygoid teeth, and lack the ectopterygoid process (Greer, 1970b; Ota et al., 1996). Relationships illustrated by our analysis also favor views of the previous authors that those character states unique to Apterygodon have evolved from states of corresponding characters in Dasia (Greer, 1970b; Ota et al., 1996).

The nucleotide replacements between species of Mabuya were larger than those between some combinations of different genera. Moreover, Asian Mabuya were not exclusively clustered with African congeners in NJ analysis (Fig. 2A), although this relationship was not support in enough bootstrap proportion in MP analysis (Fig. 2B). These suggest the genetic heterogeneity and the non-monophyly of the genus. Further analysis for more species of Mabuya, including those from Madagascar and South America, are strongly desired to revise its systematics.

Recently Vietnascinsus was described from Vietnam as another genus of arboreal skinks monotypic with V. rugosus (Darevsky and Orlov, 1994). We have had no chance to directly examine this skink, but judging from the original description, it may be closest to Lamprolepis because both genera share a medial separation of palatal rami of pterygoids (Greer, 1970b; Darevsky and Orlov, 1994). We thus suspect that Vietnascinsus belongs to the Lygosoma subgroup of the Mabuya group. This view definitely needs further verifications.