Mitochondrial DNA evolution and population history of the Tenerife skink Chalcides viridanus.
ABSTRACT Recent studies of island lizards have suggested that historical vicariance as a result of volcanism may have played an important role in shaping patterns of within-island genetic diversity. The skink, Chalcides viridanus, shows variation in morphology within the volcanic island of Tenerife. Two mitochondrial DNA (mtDNA) fragments (from the 12S and 16S rRNA regions) were sequenced in individuals from 17 sites to evaluate the relationship between current phylogeography and the geological history of the island. Three main clades were detected. The two most basal clades were restricted to areas representing the ancient precursor islands of Teno and Anaga in the northwest and northeast of Tenerife, respectively. The third clade showed a widespread geographical distribution and provided evidence of a recent rapid expansion after a bottleneck. Within-island cladogenesis appears to have taken place during a recent period of volcanic activity and long after the ancient islands had been united by the eruptions that led to the formation of the Canadas edifice. Evidence of similar biogeographical histories are found in other species in the Canary archipelago, supporting the volcanism scenario as a potentially widespread cause of within-island differentiation in reptiles.
- SourceAvailable from: ncbi.nlm.nih.gov[show abstract] [hide abstract]
ABSTRACT: We describe a method for co-estimating 4Nemu (four times the product of effective population size and neutral mutation rate) and population growth rate from sequence samples using Metropolis-Hastings sampling. Population growth (or decline) is assumed to be exponential. The estimates of growth rate are biased upwards, especially when 4Nemu is low; there is also a slight upwards bias in the estimate of 4Nemu itself due to correlation between the parameters. This bias cannot be attributed solely to Metropolis-Hastings sampling but appears to be an inherent property of the estimator and is expected to appear in any approach which estimates growth rate from genealogy structure. Sampling additional unlinked loci is much more effective in reducing the bias than increasing the number or length of sequences from the same locus.Genetics 06/1998; 149(1):429-34. · 4.39 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: A technique is presented for the partitioning of nucleotide diversity into within- and between-population components for the case in which multiple populations have been surveyed for restriction-site variation. This allows the estimation of an analogue of FST at the DNA level. Approximate expressions are given for the variance of these estimates resulting from nucleotide, individual, and population sampling. Application of the technique to existing studies on mitochondrial DNA in several animal species and on several nuclear genes in Drosophila indicates that the standard errors of genetic diversity estimates are usually quite large. Thus, comparative studies of nucleotide diversity need to be substantially larger than the current standards. Normally, only a very small fraction of the sampling variance is caused by sampling of individuals. Even when 20 or so restriction enzymes are employed, nucleotide sampling is a major source of error, and population sampling is often quite important. Generally, the degree of population subdivision at the nucleotide level is comparable with that at the haplotype level, but significant differences do arise as a result of inequalities in the genetic distances between haplotypes.Molecular Biology and Evolution 08/1990; 7(4):377-94. · 10.35 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The Antwerp database on large subunit ribosomal RNA now contains 607 complete or nearly complete aligned sequences. The alignment incorporates secondary structure information for each sequence. Other information about the sequences, such as literature references, accession numbers and taxonomic information is also available. Information from the database can be downloaded or searched on the rRNA WWW Server at URL http://rrna.uia.ac.be/Nucleic Acids Research 02/1999; 27(1):174-8. · 8.28 Impact Factor
Molecular Ecology (2000) 9, 1061–1067
© 2000 Blackwell Science Ltd
Blackwell Science, Ltd
Mitochondrial DNA evolution and population history of
the Tenerife skink Chalcides viridanus
R. P. BROWN,* R. CAMPOS-DELGADO† and J. PESTANO†
*School of Biological & Earth Sciences, Liverpool John Moores University, Byrom St., Liverpool L3 3AF, UK, †Departamento de
Genética, Facultad de Ciencias Medicas y de La Salud, Universidad de Las Palmas de Gran Canaria, 35080 Las Palmas, Spain
Recent studies of island lizards have suggested that historical vicariance as a result of vol-
canism may have played an important role in shaping patterns of within-island genetic
diversity. The skink, Chalcides viridanus, shows variation in morphology within the
volcanic island of Tenerife. Two mitochondrial DNA (mtDNA) fragments (from the 12S
and 16S rRNA regions) were sequenced in individuals from 17 sites to evaluate the
relationship between current phylogeography and the geological history of the island.
Three main clades were detected. The two most basal clades were restricted to areas
representing the ancient precursor islands of Teno and Anaga in the northwest and
northeast of Tenerife, respectively. The third clade showed a widespread geographical
distribution and provided evidence of a recent rapid expansion after a bottleneck.
Within-island cladogenesis appears to have taken place during a recent period of volcanic
activity and long after the ancient islands had been united by the eruptions that led to the
formation of the Cañadas edifice. Evidence of similar biogeographical histories are found
in other species in the Canary archipelago, supporting the volcanism scenario as a potentially
widespread cause of within-island differentiation in reptiles.
Keywords: evolution, geographical variation, island, lizard, phylogeography, volcanism
Received 13 November 1999; revision received 21 February 2000; accepted 21 February 2000
A large number of studies have demonstrated patterns
of vertebrate population differentiation over continental
regions that reflect historical fragmentation of gene flow
(Avise 1992; Hewitt 1996; Taberlet 1998; Douglas et al.
1999). In terrestrial species, range contractions mediated
by global climatic changes appear to be one of the
primary causes of this fragmentation. More recently,
differentiation over much smaller geographical areas has
been highlighted by within-island studies of lizards (e.g.
Thorpe & Báez 1987; Brown et al. 1991; Malhotra &
Thorpe 1991; Schneider 1996; Thorpe & Malhotra 1996;
Báez & Brown 1997). Global climatic scenarios provide
unlikely phylogeographical explanations in these cases.
Instead, evidence is beginning to favour the hypothesis
that historical volcanism is an important and widespread
cause of within-island cladogenesis and played a major
role in shaping these patterns of genetic diversity. This
is supported by mitochondrial DNA (mtDNA) surveys
(Pestano & Brown 1999; Thorpe et al. 1996) and the fact
that within-island variation has been described on volcanic
islands, particularly on Tenerife and Gran Canaria in the
Canary archipelago (centre 28°N, 16°W) off northwest
Africa. The influence of this mode of differentiation on
insect diversity has already been established (Carson et al.
The ability of these phylogeographical patterns to explain
within-island polymorphism in phenotypic characteristics
is clearly of interest. Cladogenesis in the Tenerife lizard
Gallotia galloti is thought to have occurred ≈ 0.7 million
years ago (Ma), but the resultant mtDNA signature is
slightly discordant with the morphological variation
(Thorpe et al. 1996). Evidence of earlier cladogenesis
(≈ 3 Ma) is found in the Gran Canarian skink, Chalcides
sexlineatus, where morphological and phylogeographical
patterns are concordant (Pestano & Brown 1999). Parallel
geographical variation in morphology is present in the
Tenerife skink, C. viridanus (Brown et al. 1991) although it
Correspondence: R. P. Brown. Fax: +(44)-151-2073224; E-mail:
1062 R. P. BROWN, R. CAMPOS-DELGADO and J. PESTANO
© 2000 Blackwell Science Ltd, Molecular Ecology, 9, 1061–1067
appears to be incompatible with the volcanism scenario
put forward for G. galloti on the same island.
The present work evaluates phylogeographical and
morphological concordance in C. viridanus and putative
associations between the former and the volcanic history
of Tenerife. This is achieved though a detailed analysis
of among-population diversity in C. viridanus, based on
mtDNA sequences from within the 12S and 16S rRNA
regions. Most of the major volcanic eruption events within
Tenerife occurred over 0.6 Ma, which may have left clearly
detectable signatures in the studied fragments, given their
rates of evolution within the Canary Island skinks (Brown
& Pestano 1998).
Materials and methods
Study species and sample sites
Chalcides sexlineatus is a small, ground-dwelling species,
which appears to show low vagility. It is most abundant
in low-altitude habitats in the north, with densities decreas-
ing in arid southern areas and at altitudes above ≈ 1100 m
(see Brown et al. 1993).
Tail tips were obtained from C. viridanus from 16 selected
sites within Tenerife (Fig. 1) and immediately stored in
99% alcohol. Skink morphology has previously been
quantified at all these sample sites (Brown et al. 1993).
Four individuals were also available from a new site (site
3). The total number of individuals from which sequence
data was collected was 85 (n per site = 4–8, except for
sites 1 [n = 3] and 14 [n = 1]). Low levels of introgression
appear to lead to reduced within-site mtDNA diversity
and this favours the use of increased numbers of sample
sites (increasing the probability that all major lineages
have been detected), as opposed to larger sample sizes,
when describing within-island phylogeographical patterns
(Thorpe et al. 1996; Brown & Pestano 1998).
Tissue samples were homogenized in a digest buffer and
whole genomic DNA was obtained using an organic
extraction protocol (Gross-Bellard et al. 1973). The L1091
and H1478 primers (Kocher et al. 1989) were used to
amplify a fragment of ≈ 384 bp from the 12S rRNA region
of the mtDNA. The polymerase chain reaction (PCR) was
performed as follows in the presence of 2 mm MgCl2: 35
cycles of denaturation at 94 °C for 30 s, annealing at 48 °C
for 30 s and extension at 72 °C for 30 s. 16SaR-L and 16Sd-
H primers (Reeder 1995) were used to amplify a fragment
of ≈ 535 bp from the 16S rRNA region of the mtDNA,
according to the following PCR profile and in the presence
of 1.5 mm MgCl2: 30 cycles of denaturation at 94 °C for
30 s, annealing at 50 °C for 45 s and extension at 72 °C for
30 s. The light strands were sequenced for all 85 individuals
using an ABI 373 sequencer (University of Liverpool sequen-
cing service). Complementary sequences were obtained for
56 and 19 individuals for 16S and 12S rRNA, respectively.
Fig. 1 The 17 samples site, the three proposed
ancient islands (shaded areas) (Ancochea et al.
1990) and altitudes on the island of Tenerife
(surface area 2034 km2).
POPULATION HISTORY OF THE TENERIFE SKINK 1063
© 2000 Blackwell Science Ltd, Molecular Ecology, 9, 1061–1067
Sequence analysis and haplotype relationships
The 16S and 12S rRNA sequences were aligned separately
using clustal w (Thompson et al. 1994). Secondary
structures were inferred after alignment with models
for Sceloporus undulatus (De Rijk et al. 1999; Van de Peer
et al. 1999). Haplotypes were clustered using neighbour–
joining (NJ) on Kimura 2-parameter distances with a
transition : transversion (TS : TV) weighting of 1 : 10 (see
Brown & Pestano 1998) using phylip (Felsenstein 1993).
Empirical TV : TS ratios were generally lower than this ow-
ing to low levels of sequence divergence, so the weighting
was based on data from a cross-species study of Chalcides
(Brown & Pestano 1998). C. viridanus from La Gomera and
El Hierro were used as outgroups (see Brown & Pestano
1998). Maximum parsimony analyses (MP) were carried
out using paup (version 4b2a) (Swofford 1998). Bootstrap
majority-rule consensus trees (500 replications), based on
heuristic searches, were computed for these data sets
(branches collapsed to give polytomies where support < 50%).
Among-site nucleotide diversity
NST statistics and a between-site mtDNA dissimilarity
matrix based on the number of nucleotide substitutions
between randomly selected pairs of individuals from
different sites were computed using the program haplo2
(Lynch & Crease 1990). Only one sample site showed
zero nucleotide diversity, so significance of geographical
structuring was tested using the program nucleodi
(Holsinger & Mason-Gamer 1996). This calculates a hier-
archical site structure based on between-site nucleotide
diversity, with the (null) hypothesis of no differentiation
between daughter nodes tested at each node on the tree.
Negative or zero branches were collapsed on the tree.
Two different methods were used to test for neutrality of
mutations, representing alternative rationales. The test
of Tajima (1989) is based on differences between estimates
of the product of the mutation rate and the effective
population size obtained from: (i) the number of segregating
sites; and (ii) the mean number of nucleotide differences
between sequences. The test of Fu & Li (1993) is based on
comparison of mutations in new ‘external’ (leading to tips)
and old ‘internal’ (connecting internal nodes) branches of
a genealogy. Both of these tests were applied to the data
(where haplotype diversity was sufficient).
Changes in population size were estimated using a
method based on a Metropolis-Hastings genealogy sampler
(Kuhner et al. 1998; program: fluctuate). This allowed
estimation of Θ = 2Nfµ, where Nf is the female population
size and µ represents the mutation rate per site, and also
the population growth rate (denoted as g). Again, this
method was only applied when there were more than
two haplotypes in the clade to be tested.
The significance of the association between among-site
mtDNA diversities in C. viridanus and Galliotia galloti
was computed using matrix correlation with significance
determined by randomization (Mantel 1967). A haplotype
dissimilarity matrix was calculated for the G. galloti sequences
presented by Thorpe et al. (1996). Within-site nucleotide
diversity was assumed to be zero for this latter species.
Mantel’s test was also used to compare between-site
nucleotide diversity with between-site morphological
differences in C. viridanus colour pattern (four characters;
see Brown et al. 1991), scalation (six characters) and body
dimensions (nine characters; see Brown et al. 1993).
The morphology matrices were either Mahalanobis D2
distances between site centroids (scalation or body dimen-
sions), or Euclidean distances calculated from site means
The GenBank accession numbers for sequences presented
in this paper are AF232602–AF232667.
The 12S fragment provided 359 bp of unambiguous
sequence containing 18 variable sites, eight of which were
parsimony informative within Tenerife alone. All but one
of the variable sites corresponded to a loop region of the
rRNA (although note that the exception coincided with
an area of ambiguous alignment). Maximum sequence
divergence was 8 bp.
The 16S fragment provided 466 bp of homologous
unambiguous sequence, containing 20 variable sites (13
of which were parsimony informative). Fourteen variable
sites did not appear to be involved in secondary-structure
bp interactions. Maximum pairwise divergence was 12 bp.
The combined fragments yielded 33 haplotypes.
The MP and NJ trees provided similar topologies that
supported three ancient cladogenetic events. Two early
events first gave rise to a lineage currently confined to
sites 8 and 17 in the northwest, and then to a lineage con-
fined to site 10 in the northeast (Fig. 2). These accounted
for only three and two of the 33 haplotypes, respectively.
A subsequent event led to a relatively recent clade
comprising all remaining haplotypes (‘central’ haplotypes).
These were geographically widespread, being found
everywhere except at sites 10 and 17.
The majority of the total nucleotide variation was between
populations: NST = 0.768. There was considerable differ-
entiation between sites (Fig. 3). The phylogeographical
patterns were discordant with the three morphological
systems tested in the same species (Table 1). The pattern
was also discordant with the phylogeography of Galliotia
galloti from the same island (between-matrix correlation,
r = 0.031, P = 0.240 [2000 randomizations]).
1064 R. P. BROWN, R. CAMPOS-DELGADO and J. PESTANO
© 2000 Blackwell Science Ltd, Molecular Ecology, 9, 1061–1067
When applied to all Chalcides viridanus Tenerife haplo-
types, neither Tajima’s (1989) (D = –1.418, P > 0.10)
nor Fu & Li’s (1993) (D* = –1.148, P > 0.10, F* = 1.464,
P > 0.10) tests of neutrality were significant. These results
changed slightly when the central haplotypes alone
were analysed (Tajima’s D = –1.875, P < 0.05; Fu & Li’s
D* = –1.950, P > 0.10; Fu & Li’s F* = –2.262, P < 0.10),
suggesting possible deviations from expectations under
neutrality. Estimations based on central haplotypes, using
Metropolis-Hastings simulations, suggested a large value
of 2Nfµ, i.e. Θ = 0.410 and rapid population growth:
g = 3147.8 (log-likelihood 0.0049), respectively.
Is there an association between the geological history
of Tenerife and the genesis of the northeast, northwest
and central lineages? Potassium-Argon (K-Ar) dating of
rocks in Tenerife indicates three ancient areas of the
island: Teno in the northwest, Anaga in the northeast and
Roque del Conde in the southwest. The ages of these
areas are estimated at 7.4 million years (Myr), 6.5 Myr
and 11.6 Myr, respectively (Ancochea et al. 1990). Volcanic
eruptions in the central Cañadas edifice united these
ancient islands (although Teno and Roque del Conde may
represent parts of the same island). These central eruptive
Fig. 2 Maximum parsimony trees displaying
relationships, between haplotypes and
phylogeography. Values on internal nodes
are bootstrap values (branches collapsed
to give polytomies where bootstrap
support < 50%). A neighbour joining tree
also provided support for the central (C),
northwestern (NW) and northeastern (NE)
clades shown here.
Table 1 Mantel tests of association between mitochondrial DNA
(mtDNA) and morphology matrices (males [M] and females [F],
separated where appropriate)
Body dimensions (M)
Body dimensions (F)
r = 0.100,
r = 0.353,
r = 0.075,
r = 0.105,
r = –0.017,
P = 0.321
P = 0.061
P = 0.262
P = 0.246
P = 0.471
POPULATION HISTORY OF THE TENERIFE SKINK 1065
© 2000 Blackwell Science Ltd, Molecular Ecology, 9, 1061–1067
episodes occurred principally in three cycles between
3.5 and 2.7 Ma, 2.5–1.4 Ma and 1.1–0.2 Ma, each giving
rise to a new Cañadas edifice (Ancochea et al. 1999). The
ancient islands were probably united at the beginning of
the first cycle.
Broad estimates of the timing of cladogenesis are useful
in assessing whether this fits in with the pattern of diver-
gence in Chalcides viridanus. Independent of their reliability,
there is considerable heuristic value in applying molecular
clocks when distinguishing between putative causal events
that were well separated in time (Zamudio & Greene 1997).
A ‘ball-park’ figure of 1.5% pairwise sequence divergence
can be estimated based on the estimated date of coloniza-
tion of El Hierro from La Gomera (see Brown & Pestano
1998). This estimate is in agreement with published
sequence divergence rates (e.g. Mindell & Thacker 1996).
If this is correct then the central and northeastern lineages
diverged ≈ 0.9 Ma while the divergence of the north-
western lineage occurred 1.1 Ma, i.e. long before the ancient
islands were joined. Whether they involved vicariance
or colonization events is unknown. Colonization probably
took place ≈ 4 Ma (Brown & Pestano 1998) (close to the
time when the precursor islands were joined) so a con-
siderable time period subsequently appears to have
elapsed before the described within-island cladogenesis.
Molecular clocks are problematical (Hillis et al. 1996)
and so spatial correlations provide stronger evidence
of a relationship between cladogenesis and geology. The
northeastern lineage is located within the known extension
of the ancient island of Anaga, while the northwestern
lineage is located within Teno. Application of the hyper-
geometric distribution provides an insight into the prob-
ability of this pattern arising by chance. For our sampling
regime, if one lineage is confined to two of the 17 sample
sites (as for the northwestern lineage), and another to
one of 17 sites (as for the northeastern lineage), then the
probability that the former will comprise two of the four
Teno sites and the latter one of the five Anaga sites is
P = 0.0147.
The considerable postcolonization range expansion
by the central clade contrasts sharply with the spatially
limited northeastern and northwestern clades. Star-like rela-
tionships within the former, and a mean pairwise diver-
gence below 0.5%, suggest that this expansion may have
been recent following a bottleneck around 300 000 years
ago, i.e. towards the end of the last major eruptive cycle.
A recent selective sweep or bottleneck would be expected
to largely remove old mutations, leaving only new, low-
frequency mutations. Tajima’s neutrality test (Tajima 1989)
supports this hypothesis. Three lineages of the gecko
Tarentola delalandii on the same island also show evidence
of population growth following a relatively recent bottle-
neck, suggesting a common effect caused by a geological
event during the last eruptive cycle (T. Gübitz, personal
The Tenerife lacertid lizard Gallotia galloti shows a
discordant pattern of phylogeographical variation, but it
is still interpreted in terms of the ancient islands of Anaga
and Teno that form part of Tenerife (Thorpe et al. 1996).
Unlike C. viridanus, the ancient G. galloti lineages are
widespread, indicating considerable dispersal out of Anaga
and Teno. Levels of sequence divergence are comparable
between G. galloti and C. sexlineatus, raising the possibility
of a common causal event. Phylogeographical patterns in
Pimelia beetles also show similarities with C. viridanus
(Juan et al. 1996).
Why are phylogeographical patterns discordant with
geographical variation in generalized morphology? More
detailed analysis reveals that generalized patterns may
mask more subtle concordance. For example, site 10 indi-
viduals had northeastern haplotypes and (unlike other
individuals) possessed green tail pigment. However, they
showed little differentiation in other characteristics. Differ-
ential introgression between nuclear and mtDNA provides
an alternative explanation of this disparity, as does local
selection pressures for different morphologies, overriding
the historical effects that have shaped the mtDNA
variation (e.g. Juan et al. 1996; Thorpe et al. 1996). The
differential selection pressures component of this argument
Fig. 3 Hierarchical relationships among sites (Holsinger &
Mason-Gamer 1996). Polytomies arise where branch lengths ≤ 0
have been collapsed.