Molecular Phylogenetics and Evolution
Vol. 14, No. 3, March, pp. 479–483, 2000
doi:10.1006/mpev.1999.0722, available online at http://www.idealibrary.com on
NOTE
Testing Hypotheses of Vicariance in the Agamid Lizard
Laudakia caucasia from Mountain Ranges
on the Northern Iranian Plateau
J. Robert Macey,*,1 James A. Schulte II,* Haji G. Kami,† Natalia B. Ananjeva,‡
Allan Larson,* and Theodore J. Papenfuss§
*Department of Biology, Campus Box 1137, Washington University, St. Louis, Missouri 63130-4899; †Department of Biology,
Faculty of Sciences, University of Agricultural Sciences and Natural Resources, P.O. Box 49165, Gorgan, Iran;
‡Zoological Institute, Russian Academy of Sciences, 1 Universitetskaya nab., 199034, St. Petersburg, Russia;
and §Museum of Vertebrate Zoology, University of California, Berkeley, California 94720
Received May 10, 1999; revised August 4, 1999
The Laudakia caucasia species group of the Iranian
Plateau is a model system for historical biogeography.
Tectonic activity caused by collision of the Indian and
Arabian plates with Eurasia has compressed intervening Iranian plates on which the L. caucasia species
group is endemic, producing mountain barriers that
geographically fragment these populations. L. caucasia
occurs on the northern margin of the Iranian Plateau,
which has been uplifting for the last 5 million years
(Sborshchikov et al., 1981), forming three major mountain chains (Fig. 1): (1) west of the Caspian Sea, the
Lesser Caucasus of the Iranian Plateau connects to the
Greater Caucasus; (2) the Elburz Range of Iran is
situated directly south of the Caspian Sea; (3) the
Kopet-Dagh and Balkhan mountains of Turkmenistan
and northeast Iran rise to the east of the Caspian Sea.
Macey et al. (1998) examined populations of L. caucasia, L. erythrogastra, and L. microlepis, the three
species in the group, and reported a fully resolved
phylogenetic estimate with every branch well supported (minimum bootstrap of 98% and a decay index of
at least 4). That study suggests an east-to-west pattern
of fragmentation of populations along the northern
margin of the Iranian Plateau, with a major break
separating populations occurring in the greater and
lesser Caucasus from Turkmen populations in the
Balkhan and Kopet-Dagh mountains. Here, we report
the phylogenetic position of L. caucasia from the Elburz
Range located between these two montane regions,
plus a new sample from the Lesser Caucasus of Georgia. The new samples are from Baladeh (36° 128 N 51°
488 E) road, 12 km E. of Jct with Karaj (35° 488 N 50°
1 To whom correspondence should be addressed. Fax: (314) 9354432. E-mail: [email protected].
598 E) to Chalus (36° 388 N 51° 268 E) road, Mazandaran Province, Iran (population 9 in Fig. 1, uncatalogued specimen of the sixth author being deposited at
the Museum of Vertebrate Zoology, MVZ-TP24781,
AF172704), and from just south of the botanical gardens, southern hills of Tbilisi (41° 428 N 44° 458 E),
Georgia (population 8 in Fig. 1, deposited at the Museum of Vertebrate Zoology, MVZ 218720, AF172705).
We investigate three hypotheses for historical fragmentation of L. caucasia from the three major mountain chains of the northern Iranian Plateau (Fig. 1): (1)
the first phylogenetic divergence occurs in the east and
separates Kopet-Dagh and Balkhan populations from
the more western populations in the Elburz and Caucasus; (2) the earliest divergence occurs in the west and
separates Caucasus populations from the more eastern
populations in the Elburz, Kopet-Dagh, and Balkhans;
and (3) the earliest divergence occurs in the south and
separates the Elburz population from the more northern populations in the Caucasus, Kopet-Dagh, and
Balkhans. The first hypothesis is compatible with our
original suggestion of an east-to-west pattern of fragmentation of populations and species (L. caucasia and
L. erythrogastra) in the mountain ranges of the northern Iranian Plateau. The second hypothesis indicates
successive periods of fragmentation first in the east and
then in the far west. The third hypothesis requires a
more recent dispersal event across the Caspian Sea
during periods of lower sea level following vicariant
separation of populations.
Mitochondrial DNA sequences [1722 aligned bases
from the genes encoding ND1 (subunit one of NADH
dehydrogenase), tRNAGln, tRNAIle, tRNAMet, ND2,
tRNATrp, tRNAAla, tRNAAsn, tRNACys, tRNATyr, and COI
(subunit I of cytochrome c oxidase)], were obtained
479
1055-7903/00 $35.00
Copyright r 2000 by Academic Press
All rights of reproduction in any form reserved.
480
MACEY ET AL.
FIG. 1. Map of the Middle East showing localities from which Laudakia populations are sampled. The two outgroup taxa, L. himalayana
(asterisk) and L. lehmanni (open circle), are from Tajikistan in the Pamir Mountains. L. microlepis (closed triangle) is endemic to the Zagros of
southern Iran. L. erythrogastra and L. caucasia populations are numbered as in Macey et al. (1998). The L. erythrogastra (closed circle)
population 1 is from the Badkyz Plateau and population 2 is from the eastern Kopet-Dagh of Turkmenistan. The L. caucasia (closed square)
populations 1 and 2 are from the central and western Kopet-Dagh of Turkmenistan, respectively. Population 3 is from the Caspian Sea flood
plain in Turkmenistan. Populations 4 and 5 are from the little and big Balkhan mountains in Turkmenistan, respectively. Population 6 is from
Armenia and 7 is from Dagestan, Russia. The new samples are numbered 8, representing the Lesser Caucasus of Georgia, and 9, representing
the Elburz of Iran. Three alternative hypotheses for fragmentation of L. caucasia populations on the northern margin of the Iranian Plateau
are depicted below the map. (1) The earliest vicariant split occurs in the east and separates Balkhan and Kopet-Dagh populations from the
more western Elburz and Caucasian populations. (2) The earliest vicariant split occurs in the west and separates Caucasian populations from
the more eastern Elburz, Balkhan, and Kopet-Dagh populations. (3) The earliest vicariant split occurs in the south and separates the Elburz
population from the more northern Caucasian, Balkhan, and Kopet-Dagh populations.
following Macey et al. (1998) and aligned to our earlier
sequences. The same two unalignable regions totaling
14 positions were excluded from analysis (positions
1388–1394 and 1613–1619). The sequence for L. caucasia from Georgia is identical in length to the sequences
for L. caucasia from Armenia, and Dagestan, Russia.
The sequence for L. caucasia from the Elburz of Iran is
one base shorter than the sequence for L. caucasia from
the Balkhans and central Kopet-Dagh, with a single
gap placed at alignment position 185.
Phylogenetic trees were estimated using PAUP*
(Swofford, 1998) with branch-and-bound searches. Bootstrap resampling was applied to assess support for
individual nodes with 1000 bootstrap replicates using
branch-and-bound searches. Decay indices (⫽‘‘branch
support’’ of Bremer, 1994) were calculated for all inter-
VICARIANCE IN Laudakia caucasia
nal branches using branch-and-bound searches that
retain suboptimal trees. The Wilcoxon signed-ranks
test (Templeton, 1983), conducted as a one-tailed test,
was applied to test the shortest trees relative to alternative hypotheses using PAUP* (Swofford, 1998), which
corrects for tied ranks. One-tailed probabilities are
close to the exact probabilities for this test but are not
always conservative, whereas the two-tailed test is
always conservative (Felsenstein, 1985). Two-tailed
probabilities are double the one-tailed probabilities.
Divergence times were estimated using the rate of
0.65% (a possible range of 0.61–0.70%) change per
lineage per million years (1.3% for pairwise comparisons). This calibration (Macey et al., 1998) is based on
four well-dated geologic events that produced vicariant
separation of Laudakia populations (Abdrakhmatov et
al., 1996; Girdler, 1984; Sborshchikov et al., 1981).
A single most-parsimonious tree results from analysis of the 1708 aligned positions (211 phylogenetically
481
informative). This topology is compatible with that of
our previously published tree (Fig. 2). Strong support
occurs for all branches, including positions of the new
sequences. Within L. caucasia, Caucasian populations
form a monophyletic sister group to all other L. caucasia populations sampled. The Georgian population is
the sister lineage to the Armenian population, and both
occur in the Lesser Caucasus. The Elburz population of
Iran is the sister lineage to all L. caucasia populations
from Turkmenistan (Balkhans and Kopet-Dagh).
The shortest estimate of phylogeny is compatible
only with hypothesis 2 in Fig. 1; the earliest phylogenetic split within L. caucasia occurred in the west.
When the overall shortest tree, which groups the
Elburz and Turkmen (Balkans and Kopet-Dagh) populations, is compared to the shortest alternative trees
(Appendix) compatible with hypotheses 1 and 3 of Fig.
1, these alternatives are rejected (n ⫽ 7, TS ⫽ 4,
P ⬍ 0.029; n ⫽ 6, TS ⫽ 0, P ⬍ 0.007; respectively).
FIG. 2. The single most-parsimonious tree found from a branch-and-bound search. The tree has a length of 565 steps and a consistency
index of 0.851. Bootstrap values are presented above branches and decay indices are shown in boldface below branches. Note that all internal
nodes are well supported. Populations are labeled with letters as in Fig. 1. A branching pattern is observed that is compatible only with
hypothesis 2 in Fig. 1, an initial vicariant split of L. caucasia populations in the western portion of the Iranian Plateau.
482
MACEY ET AL.
Our previous suggestion that populations of the L.
caucasia species group on the northern margin of the
Iranian Plateau experience vicariant splitting in a
strictly east-to-west pattern is not correct (hypothesis 1
in Fig. 1). Instead, the area cladogram indicates shifts
of tectonic activity on the Iranian Plateau through the
last 10 million years. First, north–south compression of
Iranian plates by the impinging Arabian and Indian
plates (Gondwanan) with Eurasia isolates L. microlepis
9 MYBP. Second, sharp uplifting of the eastern KopetDagh on the northeastern margin of the plateau isolates L. erythrogastra 3–4 MYBP. Third, uplifting in the
Lesser Caucasus and Elburz mountains separates Caucasian populations from eastern populations of L. caucasia 2–3 MYBP. The average pairwise sequence divergence between the two major clades of L. caucasia is
⬃2.77%, compatible with divergence at 2.1 MYBP
(Table 1).
Deep divergences on the northern margin of the
Iranian Plateau occur at the eastern and western
extremes, suggesting waves of tectonic activity affecting different regions at different times. Additional taxa
TABLE 1
Pairwise Comparisons of DNA Sequences among
Populations of the L. caucasia Species Group and
Outgroup Taxa
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
L. himalayana
L. lehmanni
L. microlepis
L. erythrogastra Badkyz Plateau (1)
L. erythrogastra Eastern KopetDagh (2)
L. caucasia Dagestan (7)
L. caucasia Armenia (6)
L. caucasia Caspian Sea Flood
Plain (3)
L. caucasia Western Kopet-Dagh (2)
L. caucasia Central Kopet-Dagh (1)
L. caucasia Little Balkhan (4)
L. caucasia Big Balkhan (5)
L. caucasia Georgia (8)
L. caucasia Elburz (9)
13.
L. caucasia
Georgia (8)
14.
L. caucasia
Elburz (9)
200, 11.8%
238, 14.0%
195, 11.5%
73, 4.3%
208, 12.3%
234, 13.8%
200, 11.8%
76, 4.5%
78, 4.6%
4, 0.2%
3, 0.2%
81, 4.8%
41, 2.4%
44, 2.6%
55,
55,
56,
41,
42,
40,
42,
41,
28,
29,
43,
3.2%
3.2%
3.3%
2.4%
2.5%
—
43, 2.5%
2.4%
2.5%
2.4%
1.6%
1.7%
2.5%
—
Note. Number of base substitutions between sequences is shown
first and percentage sequence divergence is shown second. Taxa and
populations 1–12 are labeled as in Table 3 of Macey et al. (1998) with
L. erythrogastra and L. caucasia population numbers corresponding
to Fig. 1 depicted in parentheses. Our previously published (Macey et
al., 1998) sequence for L. caucasia from Armenia (CAS194304,
AF028686) reported a G at alignment position 1481 (GenBank
position 1477) in the AA-stem of the tRNAAsn gene. The correct
nucleotide is a C, and the GenBank file has been updated. This error
artificially created an uninformative variable position. Hence, Table
2 in our previous paper has one less variable site in the tRNAAsn gene,
and Table 3 of that paper has one less substitution in pairwise
comparisons made to the L. caucasia population from Armenia.
that occur widely on the Iranian Plateau can be sampled
to test generality of this biogeographic model for faunal
barriers producing (1) north–south vicariance ⬃9 MYBP,
(2) isolation in the far east 3–4 MYBP, and (3) isolation
in the far west 2–3 MYBP.
APPENDIX
Trees used as alternative hypotheses in Wilcoxon
signed-ranks tests. Taxon names appear as in Fig. 2.
Lengths of trees and consistency indices (CI) (Swofford,
1998) are given in parentheses.
The most-parsimonious tree derived by constraining
western populations of L. caucasia to form a monophyletic group (eastern split, hypothesis 1 in Fig. 1) (570
steps, CI 0.844): (L. himalayana, (L. lehmanni, (L.
microlepis, ((L. erythrogastra-Badkyz Plateau, L. erythrogastra-eastern Kopet-Dagh), (((L. caucasia-Degestan, (L. caucasia-Armenia, L. caucasia-Georgia)), L.
caucasia-Elburz), ((L. caucasia-central Kopet-Dagh, (L.
caucasia-western Kopet-Dagh, L. caucasia-Caspian Sea
flood plain)), (L. caucasia-Little Balkhan, L. caucasiaBig Balkhan))))))).
The most-parsimonious tree derived by constraining
northern populations of L. caucasia to form a monophyletic group (southern split, hypothesis 3 in Fig. 1) (571
steps, CI 0.842): (L. himalayana, (L. lehmanni, (L.
microlepis, ((L. erythrogastra-Badkyz Plateau, L. erythrogastra-eastern Kopet-Dagh), (((L. caucasia-Degestan, (L. caucasia-Armenia, L. caucasia-Georgia)), ((L.
caucasia-central Kopet-Dagh, (L. caucasia-western
Kopet-Dagh, L. caucasia-Caspian Sea flood plain)), (L.
caucasia-Little Balkhan, L. caucasia-Big Balkhan))), L.
caucasia-Elburz))))).
ACKNOWLEDGMENTS
This work was supported by grants from the National Science
Foundation (DEB-9726064 to A.L., J.R.M., and T.J.P.), National
Geographic Society (4110-89 and 4872-93 to T.J.P. and J.R.M.), and
the California Academy of Sciences.
REFERENCES
Abdrakhmatov, K. Ye., Aldazhanov, S. A., Hager, B. H., Hamburger,
M. W., Herring, T. A., Kalabaev, K. B., Makarov, V. I., Molnar, P.,
Panasyuk, S. V., Prilepin, M. T., Reilinger, R. E., Sadybakasov, I. S.,
Souter, B. J., Trapeznikov, Yu. A., Tsurkov, V. Ye., and Zubovich,
A. V. (1996). Relatively recent construction of the Tien Shan
inferred from GPS measurements of present-day crustal deformation rates. Nature 384: 450–453.
Bremer, K. (1994). Branch support and tree stability. Cladistics 10:
295–304.
Felsenstein, J. (1985). Confidence limits on phylogenies with a
molecular clock. Syst. Zool. 34: 152–161.
Girdler, R. W. (1984). The evolution of the Gulf of Aden and Red Sea in
space and time. Deep-Sea Res. Part A, 31: 747–762.
VICARIANCE IN Laudakia caucasia
Macey, J. R., Schulte, J. A., II, Ananjeva, N. B., Larson, A., RastegarPouyani, N., Shammakov, S. M., and Papenfuss, T. J. (1998).
Phylogenetic relationships among agamid lizards of the Laudakia
caucasia species group: Testing hypotheses of biogeographic fragmentation and an area cladogram for the Iranian Plateau. Mol.
Phylogenet. Evol. 10: 118–131.
Sborshchikov, I. M., Savostin, L. A., and Zonenshan, L. P. (1981).
483
Present plate tectonics between Turkey and Tibet. Tectonophysics
79: 45–73.
Swofford, D. L. (1998). ‘‘PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods), Beta Version 4.0b1,’’ Sinauer, Sunderland, MA.
Templeton, A. R. (1983). Phylogenetic inference from restriction
endonuclease cleavage site maps with particular reference to the
evolution of humans and the apes. Evolution 37: 221–244.