www.elsevier.com/locate/ympev
1055-7903/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2006.11.022
Evolution and dispersal of three closely related macaque species,
Macaca mulatta, M. cyclopis, and M. fuscata, in the eastern Asia
Jui-Hua Chu
a, Yao-Sung Lin
a, Hai-Yin Wu
b,¤a Institute of Ecology and Evolutionary Biology, National Taiwan University, Taipei 106, Taiwan, ROC b Institute of Natural Resources, National Dong Hwa University, Hualien 974, Taiwan, ROC
Received 7 April 2006; revised 15 November 2006; accepted 16 November 2006 Available online 5 December 2006
Abstract
Macaca mulatta, M. cyclopis and M. fuscata are three closely related species in the fascicularis species group. M. mulatta is wide-spread
in Asia, while M. cyclopis and M. fuscata are restricted to Taiwan and Japan, respectively. Both M. cyclopis and M. fuscata are thought to be derived from ancient ‘mulatta’ populations in the eastern Asia. In this study, we analyzed sequences of mitochondrial DNA control region to provide genetic evidence for the evolution and dispersal scenario of the three species proposed by Fooden and Albrecht [Foo-den, J., Albrecht, G.H. 1999. Tail-length evolution in fascicularis-group macaques (Cercopithecidae: Macaca). Int. J. Primatol. 20, 431– 440]. Our results indicated that several localities in the southern China and Vietnam harbored multiple divergent mtDNA lineages that may not have evolved sympatrically. These divergent mtDNA lineages may have originated from diVerent ancient northern populations that retreated into southern localities during glacial periods. However, the age of the southward retreat and the northward recolonization may be dated back to a more ancient past during late middle Pleistocene (0.12–0.18 mya) instead of during the LGM (0.018 mya). Times of gene divergence between M. mulatta and the two island species, estimated by mean nucleotide diVerence, suggest the ancestral popula-tions colonized Taiwan and Japan around 0.38–0.44 mya. In addition, a more recent age of mulatta–cyclopis–fuscata population diver-gence (when ancient populations were isolated), estimated to be 0.17 mya by net nucleotide diverdiver-gence, is suggested.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Macaca mulatta; M. cyclopis; M. fuscata; mtDNA control region; Phylogeography; Eastern Asia
1. Introduction
The 19 extant species of the genus Macaca ( Brandon-Jones et al., 2004) have been organized into three to six species groups according to morphological or genetic diVer-ences (Delson, 1980; Fooden, 1976; Fooden and Lanyon, 1989; Groves, 2001; Hayasaka et al., 1996; Li and Zhang, 2005; Zhang and Shi, 1993). Among these species groups, the evolutionary relationship of the fascicularis species group, composed of Macaca fascicularis, M. mulatta,
M. cyclopis and M. fuscata, is of particular interest because
paraphyly has been observed from genetic data (Hayasaka
et al., 1996; Melnick et al., 1993; Morales and Melnick, 1998; Tosi et al., 2000, 2002, 2003). Based on maternally, paternally and bi-parentally inherited gene sequences, evo-lutionary relationship of the fascicularis species group has been described by Melnick and colleagues (Melnick et al., 1993; Morales and Melnick, 1998; Tosi et al., 2000, 2002, 2003). According to their studies, M. fascicularis is the ancestral form among the four species. M. mulatta may have originated from a fascicularis-like ancestor 2.5 million years ago (mya) and became widely distributed within a rel-atively short period. The rapid expansion was followed by mitochondrial diVerentiation between the eastern and west-ern parts of the species range (about 0.75 mya). After the intraspeciWc divergence, populations of eastern M. mulatta (China–Burma) colonized Taiwan and Japan during glacial periods when land bridges were available. Mitochondrial * Corresponding author. Fax: +886 3 8633260.
diVerentiation between M. mulatta and M. cyclopis or M.
fuscata was estimated to be 0.25 and 0.50 mya, respectively.
Female philopatry of Macaca species maintains a geo-graphic structuring of the maternally inherited mitochondrial DNA (mtDNA) that reXects patterns of cladogenic events (Melnick et al., 1993). It explains the observation that mtDNA haplotypes of eastern M. mulatta (those from China) are more similar to haplotypes of the two insular spe-cies (M. cyclopis and M. fuscata) than to haplotypes of west-ern M. mulatta (those from India–Pakistan), or mtDNA paraphyly of M. mulatta (Melnick et al., 1993; Morales and Melnick, 1998). However, such pattern was not detected in Y-chromosome (Tosi et al., 2000). This can be attributed to male-mediated gene Xow between eastern and western
M. mulatta that homogenized nuclear genetic variation
across geographic range (Tosi et al., 2002, 2003). Moreover, a southward introgression by male M. mulatta to populations of M. fascicularis generated a paraphyletic Y-chromosome pattern in M. fascicularis (Tosi et al., 2002, 2003).
According to Allen’s rule, size of extremities such as ears and tails of mammals increases from colder to warmer cli-mates in the same or closely related species, especially for those inhabiting a large latitudinal range. The composite distribution range of members in the fascicularis species group encompasses a wide latitudinal zone from 10°S to 40°N. The relative tail length (RTL, tail length/head and body length) in the species group generally decreases with increasing latitudes just as predicted by Allen’s rule ( Foo-den, 1997). However, the correlation does not apply to M.
mulatta inhabiting the 15°N–25°N latitudinal zone, i.e. the
southern populations (Fooden and Albrecht, 1999). RTL of extant southern mulatta populations is less than that of M.
cyclopis living in similar latitudes, but is similar to that of
the conspeciWc northern populations. Fooden and Albrecht (1999) also in Fooden (2000) suggested that modern south-ern mulatta populations did not originate within their pres-ent range; instead, they may be descendpres-ents of the populations retreating from farther north. They proposed a hypothetical scenario for the evolution and dispersal of M.
mulatta based on variation of RTL in the fascicularis
spe-cies group as well as the evolutionary history of the mulatta group (including M. mulatta, M. cyclopis and M. fuscata), which will be brieXy described in the following.
By the observed trend between latitude and RTL in the
fascicularis species group, the ancient mulatta populations
originally inhabited between 15°N and 25°N were probably more similar to M. cyclopis in RTL (mean RTL D 0.80, the long-tailed populations), while those inhabiting farther north were more similar to M. fuscata in RTL (mean RTL D 0.15 at 40°N, the short-tailed populations). The lati-tudinal gradient in RTL was formed gradually during northward population expansion at least 0.04 mya. At this stage, the ancient mulatta populations colonized Taiwan and Japan when dispersing routes were available. Extant
M. fuscata and M. cyclopis are two relictual populations
descended from the ancient short-tailed populations and long-tailed populations, respectively. During the last glacial
maximum (LGM, about 0.018 mya), when climatic deterio-ration compressed suitable habitats southward, the short-tailed populations (mean RTL < 0.50) retreated from north in the eastern Asia and replaced the long-tailed popula-tions. During the southward range shift, a west–east gradi-ent of decreasing RTL had been established (mean RTL D 0.45 at 100°E, mean RTL D 0.30 at 120°E). How-ever, the factors responsible for the gradient are not clear. Subsequent to the LGM, i.e. during the Holocene expan-sion, both the eastern and western populations of M.
mul-atta dispersed northward.
If the southward dispersal of M. mulatta in the eastern Asia occurred during the LGM as described by Fooden and Albrecht (1999), the phylogenetic relationship of the extant local populations may not correspond to geographic aYnity and the dispersal pattern may be hard to trace. Fur-thermore, dispersal pattern of M. mulatta during Holocene expansion (about 0.01 mya) should be apparent from the geographic distribution of closely related mtDNA haplo-types. However, the low divergent level of mitochondrial ribosomal RNA genes that Morales and Melnick (1998) used may not be sensitive enough to reveal recent diVerenti-ation. A more polymorphic mtDNA marker is needed.
Mitochondrial DNA polymorphism is commonly used to reveal phylogenetic and phylogeographic relationships between populations of a species or closely related species (Avise et al., 1987). The control region (CR), a unique non-coding nucleotide sequence, is the most variable portion of mtDNA in mammals (Saccone et al., 1993). The sedentary nature of female macaques and the maternal inheritance of mtDNA make the molecule an ideal tool to elucidate popu-lation history (Melnick et al., 1993). Given the scenario hypothesized by Fooden and Albrecht (1999), objectives in this study are: (1) to Wnd genetic evidence that supports the southward retreat in M. mulatta; (2) to trace the possible dispersal patterns of northward range expansion of M.
mul-atta; and (3) to estimate the time of divergent events of
ancestral M. cyclopis and M. fuscata.
2. Materials and methods
2.1. Sample collection
In this study, mtDNA CR sequences of the four species in the fascicularis species group were analyzed. Samples of 148 M. cyclopis (107 fecal, 32 hair and 9 tissue samples) col-lected from Wve regional populations in Taiwan (GenBank accession nos. AY878873–AY878925 and DQ143984– DQ143987) and 5 M. fascicularis of unknown origin (Gen-Bank accession nos. AY884307–AY884311) were sequenced. Sequences of 98 M. mulatta sampled from 11 provinces of China and Vietnam (GenBank accession nos. AF135271–AF135368, unpublished data by Ding et al.) and 50 M. fuscata sampled from six local populations in Japan (GenBank accession nos. AJ419855–AJ419864, Marmi et al., 2004) were included. Localities of samples with known geographic origins are indicated in Fig. 1.
2.2. DNA extraction, PCR ampliWcation and nucleotide sequencing
Total genomic DNA was extracted by traditional phenol– chloroform procedure (Kocher et al., 1989). Some of the DNA samples extracted from feces were further eluted by a silica pellet method (Geneclean III, Bio101) when the Wrst PCR attempt failed. The mtDNA CR was ampliWed by PCR using a primer set of DL1 (forward): 5⬘-CCAGAAATGAA CACCCTTCCTAGGGC-3⬘ (Chu et al., 2005) and Saru5 (reverse): 5⬘-GCCA GGACCAAGCCTATTT-3⬘ (Hayasaka et al., 1991) that produced a 1.4Kb fragment composed of the whole CR region and the conservative peripheral regions. The long PCR product may reduce the possibility to amplify a nuclear CR pseudogene. The PCR mixture contained 10 pmol of each primer, 200M dNTP, 1.5mM MgCl2, 1£
PCR reaction buVer (20mM Tris–HCl, pH 8.4, 20mM KCl), 1 unit Taq polymerase (GIBCO BRL) and about 1–3L DNA extract in a total volume of 50L. A DNA thermal cycler (Biometra) was programmed to perform initial dena-turation at 95 °C for 10 min and 32 cycles of 1 min at 95 °C, 1 min at 58 °C, and 2 min at 72 °C. A 10 min Wnal extension at 72 °C was also included. PCR elutes were partially sequenced by ABI 377 or ABI 3100 automated DNA sequencer using DL1 (forward) and another internal primer DH1 (reverse):
5⬘-CGGAGATGGGGGTGGGGGGGTTTG-3⬘ (Chu
et al., 2005).
2.3. Analysis of mitochondrial DNA sequences
Sequences were aligned by Pileup program in GCG (Wis-consin Package Version 10.1, Genetics Computer Group, Madison, WI). A fragment of 386 nucleotide sites was avail-able for analyses after alignment. For a comparison of
genetic polymorphism of the four species, nucleotide diver-sity () and mean nucleotide diVerence within species (dx) were calculated using Arlequin 2.0 (Schneider et al., 2000).
Phylogenetic relationship of unique haplotypes was con-structed using neighbor-joining (NJ) method (Saitou and Nei, 1987) and Bayesian approach and was rooted by sequences of M. fascicularis. NJ tree was constructed by MEGA 3.1 (Kumar et al., 2004). When constructing phylo-genetic trees, distance estimated by a sophisticated model often generates a large variance and thus reduces the reso-lution of the tree topology when conducting bootstrap (Nei and Kumar, 2000b). The situation could be worse when the dataset is composed of a large number of taxa with short nucleotide sequence (Nei and Kumar, 2000a). To avoid such problem, a simple p-distance model was used to calcu-late genetic distance between distinct sequences. Statistical support of the tree topology was estimated by interior branch analysis with 1000 bootstrap replicates.
Bayesian analysis of the sequence data was performed using MrBayes 3.1 (Huelsenbeck and Ronquist, 2001; Ron-quist and Huelsenbeck, 2003), with starting trees generated randomly. A total of 5,000,000 generations were run and the likelihood scores stabilized after 500,000 generations. A tree was saved every 100 generations to give 50,000 trees, of which the Wrst 10,000 were rejected (corresponding to those obtained before the likelihood scores stabilized). Three Markov Chain Monte Carlo runs each having four simulta-neous chains were performed independently. The joint pos-terior probability distribution of each run was congruent, suggesting that the chains were run for a suYcient number of generations to sample the posterior probability land-scape adequately. Bootstrap values or likelihood probabili-ties of 90% and 70% were considered as thresholds of high and moderate supports for a clade, respectively.
Fig. 1. Sampling localities of Macaca mulatta, M. fuscata and M. cyclopis. (1) Anhui, (2) Zhejiang, (3) Fujian, (4) Henan, (5) Hubei, (6) Hunan, (7) Guiz-hou, (8) Sichun, (9) Yunnan, (10) Guangxi, (11) Hainan, (12) Vietnam, (13) Hakusan, (14) Takahama, (15) Awajishima, (16) Takasakiyama, (17) Koshima, (18) Yakushima, (19) North Taiwan, (20) Central Taiwan, (21) South Taiwan, (22) East Taiwan and (23) Northeast Taiwan. Geographic distribution of the three phylogenetic clades of M. mulatta, haplogroups I, II and III, is also shown on the map. Arrows indicate that M. cyclopis and M. fuscata may be genetically more related to M. mulatta of haplogroups I and II, respectively.
2.4. Demographic inferences
Mismatch distribution analysis (MDA) was conducted to infer possible population expansion of a species. MDA assumes the pairwise nucleotide diVerences of haplotypes drawn from a population to have a unimodal distribution if it has experienced a demographic expansion event; other-wise, the distribution is multimodal reXecting a stochastic shape of gene trees (Rogers and Harpending, 1992). Good-ness of Wt, i.e. the sum of square deviations (SSD) between the observed and the expected mismatch (Schneider and ExcoYer, 1999), and Harpending’s raggedness index ( Har-pending, 1994) were calculated to examine the statistical support of the expansion event. Fu’s Fs test (Fu, 1997) was conducted to further conWrm the demographic expansion. Because insuYcient sampling tends to bias the analysis, each unique haplotype was used once regardless of its fre-quency of occurrence in samples by assuming the increase of haplotypes refers to a population expansion event ( Har-pending et al., 1993). The approach may ignore recent expansions but can detect relatively ancient ones.
Time of a conWrmed expansion was estimated by
D 2ute, where is estimated by MDA, u is the sum of per
nucleotide mutation rate in the DNA region under study and te is time since expansion. Mutation rates of the
mtDNA CR estimated from various phylogenetic studies (reviewed by Parsons et al., 1997) were in a range of 0.025– 0.26 site¡1myr¡1. In this study, we adopted the medium of
the reported values, 0.14 site¡1myr¡1, which corresponds to
a divergence rate of 28% per million years in the mtDNA CR. Hence, time since expansion was estimated by te (myr) D/(2 £ 0.14 site¡1myr¡1£386 sites). A 95% con
W-dence interval for was obtained by a parametric bootstrap approach (Schneider and ExcoYer, 1999).
Initial times of the mtDNA CR divergence between the four species in the fascicularis species group were estimated by dxyD2utd, where dxy is the mean nucleotide diVerence between haplotypes of two species, td is time since diver-gence and u is the same as described above. Because ances-tral populations of M. cyclopis and M. fuscata may have originated from diVerent local populations of ancient
M. mulatta, genetic diversity in each of the two species is
supposed to be a subset of genetic diversity of M. mulatta in the eastern Asia. Thus, samples of M. mulatta were divided into three groups as revealed from the phylogenetic analy-sis for further comparison. The initial periods of gene diver-gence between M. cyclopis or M. fuscata and each of the three groups of M. mulatta were also estimated by
dxyD2utd. To correct the discrepancy between “gene diver-gence” and “population diverdiver-gence” due to the presence of ancestral polymorphism in populations (Edwards and Beerli, 2000), times of population divergence (i.e., when the two ancestral populations were isolated to each other) between species or populations in the mulatta group were estimated by net nucleotide divergence, dAD2ut (Nei and Li, 1979). The most conservative time of population diver-gence was estimated from the minimum value of dA. Here
dADdxy¡1/2(dx+ dy), where dxy is the mean nucleotide diVerence between haplotypes of species (or populations) x and y, dx and dy are the mean nucleotide diVerences between haplotypes within species (or populations) x and y, respectively.
Since the distribution of the haplotypes from eastern China in the phylogenetic tree was not in accordance to their geographic aYnity, pairwise nucleotide diVerences among haplotypes sympatric in Guangxi (14 haplotypes), Zhejiang (eight haplotypes) and Fujian (six haplotypes) provinces were examined for the possible ages of diver-gence. Distribution of pairwise nucleotide diVerences was calibrated on a time scale of Pleistocene events (Jenkins, 2001) by dxyD2utd.
3. Results
3.1. Mitochondrial DNA CR variation
Among the total sequences examined (N D 301), 124 dis-tinct haplotypes were identiWed. Basic genetic polymor-phism parameters of the four species are summarized in Table 1. In total, 188 variable sites were found, including 169 transitions, 44 transversions, and 20 indels. Nucleotide diversity was the highest in M. fascicularis (0.105 § 0.065,
§ SD), followed by M. mulatta (0.081 § 0.039), and was
smaller in M. fuscata (0.062 § 0.031) and in M. cyclopis (0.061 § 0.030). The trend of genetic diversity agrees with the phylogenetic relationship of the four species that ances-tral taxon tends to retain higher genetic diversity than derived taxa.
3.2. Phylogeny reconstruction
Branching patterns and the statistic supports in the NJ and the Bayesian trees were similar except for some tip clus-ters. Hence, only the NJ tree is illustrated in Fig. 2 to show the relationship between haplotypes. The NJ tree comprised Wve phylogenetic clades with high bootstrap support (>90%) and a single lineage (Zhejiang4 of M. mulatta) outward the major clades (Fig. 2A). M. cyclopis and M. fuscata each formed a monophyletic clade, while M. mulatta comprised three haplogroups (Haplogroup I to III) and the single line-age. The clustering patterns of the haplotypes within the two island species clades corresponded to their geographic aYn-ity (Fig. 2B), while the patterns within the Haplogroups I and
Table 1
Sample size (N), number of haplotypes (H), number of polymorphic sites (P), and nucleotide diversity () in the 5⬘ end partial segment of the mtDNA CR of the four Macaca species in the fascicularis species group
Species N H P (SD) M. fascicularis 5 5 86 0.105 (0.065) M. mulatta 98 57 138 0.081 (0.039) M. cyclopis 148 53 107 0.061 (0.030) M. fuscata 50 9 61 0.062 (0.031) Total 301 124 188
II of M. mulatta corresponded to geographic aYnity as well. The Haplogroup I consisted of haplotypes from Hunan, Yunnan and Sichuan in China, while the Haplogroup II comprised of haplotypes from Guangxi in China and Viet-nam. As for the Haplogroup III, it contained haplotypes from a wide range from Henan, Hubei, Anhui, Zhejiang, Fuj-ian, Guangxi to Hainan in China and Vietnam (Fig. 1). In
this haplogroup, closely related haplotypes were distributed in several provinces of China or Vietnam, and some sympat-ric haplotypes in Henan, Zhejiang, Fujian, Guangxi, and Vietnam were too divergent to be clustered together (Fig. 2A). In addition, since bootstrap supports among the Wve phylogenetic clades were less than 90%, the relationship of these clades could not be concluded. However, under the
Fig. 2. Neighbor-joining tree illustrating the phylogenetic relationship among the 124 mtDNA CR haplotypes of the four Macaca species in the fascicu-laris species group. The percentage bootstrap support (1000 replicates) greater than 70 is listed for interior branches. The haplogroups of M. cyclopis (53 haplotypes) and M. fuscata (nine haplotypes) are compressed in (A) and the haplogroups of M. mulatta are compressed in (B) in order to simplify the tree topology. M. fascicularis (Wve haplotypes) is the outgroup to root the tree. Haplotype codes of M. mulatta follow Ding et al. published in the Genbank.
criterion of moderate support (bootstrap value > 70%), M.
fuscata was grouped with Haplogroup II into a cluster
(boot-strap value D 81%), which belonged to a larger assemblage with M. cyclopis and the Haplogroup I included (bootstrap value D 79%).
3.3. Demographic inferences
Statistical parameters of MDA did not reject a smooth, unimodal distribution of pairwise nucleotide diVerences between unique haplotypes in each of the four species in the
fascicularis species group (Table 2). It indicated that popula-tion expansion occurred in all of the four species. Only the
ages of expansion in M. mulatta and M. cyclopis were described because samples of M. fascicularis (Wve haplo-types) and M. fuscata (nine haplohaplo-types) were under-repre-sented. Mismatch distributions of M. mulatta and M.
cyclopis are illustrated in Fig. 3. Expansion events of the two species were further conWrmed by the large negative values of Fu’s Fs (Table 2). The values of were 33.5 in M. mulatta and 23.9 in M. cyclopis, which gave the estimated ages of expan-sion to be 0.31 mya (95% CI of teD0.24–0.41 mya) and 0.22 mya (95% CI of teD0.15–0.37 mya) for the two species, respectively.
The mean number of nucleotide diVerences (dx) within spe-cies showed the same decreasing trend from the basal taxon
Fig. 2 (continued)
Table 2
Population expansion parameters and Fu’s Fs test of the four Macaca species in the fascicularis species group
a The time unit of t is in myr.
b SSD: The sum of square deviations between the observed and the expected mismatch (Schneider and ExcoYer, 1999). c RI: Harpending’s raggedness index (Harpending, 1994).
Species (95% CI) t a(95%CI) SSDb(p) RIc(p) Fu’s Fs (p)
M. fascicularis 49.5 (35.4–62.1) 0.46 (0.33–0.57) 0.075 (p D 0.75) 0.200 (p D 0.53) 1.27 (p D 0.47) M. mulatta 33.5 (26.2–44.8) 0.31 (0.24–0.41) 0.002 (p D 0.83) 0.001 (p D 1.00) ¡24.13 (p < 0.01) M. cyclopis 23.9 (16.6–40.2) 0.22 (0.15–0.37) 0.004 (p D 0.34) 0.003 (p D 0.70) ¡24.12 (p < 0.01) M. fuscata 30.6 (23.1–36.0) 0.28 (0.21–0.33) 0.035 (p D 0.25) 0.099 (p D 0.14) ¡1.08 (p D 0.18)
(M. fascicularis, dxD40.5) to the tip taxa (M. cyclopis and M.
fuscata, dxD26.2 and 24.7, respectively) in this species group (Table 3). The values of dxy between M. cyclopis and each of the three haplogroups of M. mulatta ranged from 42.7 to 47.1 (Table 3), which corresponded to an initial gene divergent time of 0.40–0.44 mya. Similarly, the values of dxy between M.
fuscata and each of the three haplogroups of M. mulatta
ranged from 40.8 to 45.4, which corresponded to an initial gene divergent time of 0.38–0.42 mya. The minimum values of
dA between M. cyclopis or M. fuscata and M. mulatta were 18 and 19.5 (Table 4), which indicated the most conservative time of population divergence to be 0.17 and 0.18 mya, respectively. In addition, various divergent levels were found between
hap-lotypes sympatric in Guangxi, Zhejiang and Fujian. The pair-wise nucleotide diVerences of these sympatric haplotypes showed coincident modes around 0.01, 0.06–0.12 and 0.18– 0.32 mya (Fig. 4), the timing of which approximated to vari-ous interglacial periods in the Pleistocene.
4. Discussion
Previous authors have reported mtDNA paraphyly among mulatta lineages from India and China (Hayasaka et al., 1996; Melnick et al., 1993; Morales and Melnick, 1998; Tosi et al., 2003). We also found that M. mulatta is not a mitochondrially monophyletic taxon. However,
Fig. 3. Observed and expected mismatch (model frequency) distributions given the null distribution (unimodal, bell-shaped) corresponding to population expansion for (A) M. mulatta, and (B) M. cyclopis. Arrows indicate similar modes of population expansion.
Table 3
Mean number of nucleotide diVerences among (dxy, above diagonal), within (dx, diagonal) and net nucleotide diVerences among (dA, below diagonal) the four Macaca species in the fascicularis group in a segment of 386 nucleotide sites of the mtDNA CR
M. fascicularis M. mulatta M. cyclopis M. fuscata
M. fascicularis 40.5 59.8 62.4 66.2
M. mulatta 24.2 30.8 43.2 44.1
M. cyclopis 29.0 14.7 26.2 45.6
although we did not have the sequences from India–Paki-stan to re-examine the reported pattern, our result revealed a diVerent paraphyletic pattern in mtDNA among mulatta lineages unreported elsewhere. By applying a highly poly-morphic mtDNA marker and a larger panel of mtDNA sequences than previous studies (Deinard and Smith, 2001; Hayasaka et al., 1996; Morales and Melnick, 1998; Tosi et al., 2000; 2003; Zhang and Shi, 1993), we found that western (Yunnan–Sichuan) and eastern Chinese lineages were separated from each other in the phylogenetic tree with respect to M. cyclopis and M. fuscata (Fig. 2A). The result supports the recent divergence of M. cyclopis and M.
fuscata from Chinese M. mulatta and the retention of
genetic polymorphism in the Chinese M. mulatta as described elsewhere (Hoelzer, 1997; Melnick et al., 1993; Morales and Melnick, 1998). Moreover, the monophyly of
M. cyclopis and M. fuscata suggests postglacial vicariance
after the ancestral mulatta populations dispersed into Tai-wan and Japan, respectively.
4.1. Retreat-recolonization history of M. mulatta
In this study, no distinct geographic pattern was found in haplotypes of M. mulatta distributed in the eastern
Table 4
Mean number of nucleotide diVerences among (dxy, above diagonal), within (dx, diagonal) and net nucleotide diVerences among (dA, below diagonal) the Wve Macaca populations in the mulatta group in a segment of 386 nucleotide sites of the mtDNA CR
Minimum values of dA between M. cyclopis or M. fuscata to each of the three haplogroups of M. mulatta are underlined. a Haplogroup I (Yunnan-Sichun).
b Haplogroup II (Guangxi-Vietnam). c Haplogroup III (Eastern China).
M. cyclopis M. mulatta I M. mulatta II M. mulatta III M. fuscata
M. cyclopis 26.2 42.7 47.1 42.8 45.8
M. mulatta Ia 18.0 23.2 44.7 38.5 43.4
M. mulatta IIb 30.2 29.2 7.7 34.7 40.8
M. mulatta IIIc 19.3 16.5 20.4 20.8 45.4
M. fuscata 20.4 19.5 24.6 22.7 24.7
Fig. 4. Distribution of pairwise nucleotide diVerences of M. mulatta haplotypes in Guangxi, Zhejiang, and Fujian provinces of China aligned with the time frame of glacial and interglacial periods in the Pleistocene. Dotted line indicates the pairwise nucleotide diVerence of 13, corresponding to divergence around 0.12 mya. Sympatric haplotypes diVering by more than 13 nucleotides may have evolved allopatrically. Asterisks indicate possible expansion modes during interglacial periods in the Pleistocene.
China (Fig. 2A). Multiple divergent mtDNA lineages (espe-cially those diVering by more than 20 nucleotides) were found sympatric in Henan, Zhejiang, Fujian and Guangxi provinces of China and Vietnam, but no haplotype closely related to these lineages in nearby localities. It shows a crepant phylogeographic pattern against isolation by dis-tance (IBD) model (Wright, 1943). By assuming mtDNA haplotypes accumulate during population expansion and a mutation rate of 0.14 site¡1myr¡1 in the mtDNA CR, possi-ble nucleotide diVerences between sympatric haplotypes evolved from a single mtDNA matriline during diVerent temporal ranges can be estimated. Applying the time frame of glacial and interglacial events described by Jenkins (2001), the divergent time of haplotypes diVering by about two nucleotides is less than 0.01 my (i.e. during Holocene). Accordingly, diVerences of 6–13 and more than 20 nucleo-tides between haplotypes correspond to divergence since the last and the second last interglacials (0.06–0.12 mya and >0.2 mya), respectively (Fig. 4).
However, it is less likely for multiple mtDNA lineages to coexist for such a long time under the process of lineage sorting (Avise et al., 1984; Hoelzer et al., 1998; Wallman et al., 1996) when eVective population size decreased dra-matically during glacial periods (e.g. 0.01–0.06 mya and 0.12–0.18 mya, etc.). Moreover, even if these divergent lin-eages evolved sympatrically from the same mtDNA matri-line, phylogenetic relationship between them should still follow the IBD model not observed in this study. Hence, it is less likely that these divergent mtDNA haplotypes evolved sympatrically.
Two other scenarios may lead to sympatry of multiple divergent mtDNA lineages observed in this study. It may have resulted either from southward retreat that caused a stochastic collection of divergent lineages in a locality (i.e. a refugium) during glacial period, or from population expan-sion that brought divergent mtDNA lineages into contact during interglacial period. In the former case, it infers “forced dispersal” for populations characterized by female philopatry. A great scale of mtDNA lineage extinction and range shift may have occurred during the retreat process due to competition or simply deteriorated environments. A corollary to it is a great gap of nucleotide diVerence between sympatric haplotypes and the lack of closely related haplotypes nearby. In the latter case, it infers sec-ondary admixture due to population expansion. Since pop-ulation expansion tends to maintain mtDNA lineages, haplotypes closely related to these divergent mtDNA lin-eages can be found in nearby populations.
Results in this study indicate that phylogenetic relation-ship of haplotypes distributed in the eastern China is against the IBD model. Multiple divergent mtDNA lin-eages were found coexisting in several localities without closely related lineages nearby. This suggests a complex amalgamation of distant mtDNA lineages due to the south-ward retreat in the eastern China. Thus, the genetic data supports that there have been a southward retreat as pro-posed by Fooden and Albrecht (1999). In addition, it also
indicates that possible glacial refugia may include regions in Zhejiang, Fujian and Guangxi provinces that harbor divergent mtDNA lineages in China (probably in the mountain regions such as Wuyi-Shan and Nan Ling).
Fooden and Albrecht (1999) proposed that the south-ward retreat and the subsequent recolonization of M.
mul-atta in the eastern Asia occurred during and after the LGM
(0.018 mya), respectively. If ages of the two events were as proposed, the dispersal pattern of retreat may be hard to trace, while that of recolonization should be quite apparent. The resulting pattern should be some closely related mtDNA haplotypes (diVered by one or two nucleotides) overlaying in a wide range (e.g. Chen et al., 2005; Conroy and Cook, 2000; Fu et al., 2005; Hatase et al., 2002; Ritchie et al., 2004; Ruedi and Castella, 2003). However, we did not detect any such pattern. Instead, geographic aYnity observed in most sub-clades consisting of haplotypes diVer-ing by less than 13 nucleotides suggests to date the recoloni-zation event back to a more ancient age. In addition, nucleotide diVerence between those sympatric but divergent mtDNA lineages (>20 nucleotides) also suggests a more ancient time of the retreat event than during LGM. Hence, the scenario presented by Fooden and Albrecht (1999) was supported by our genetic data, however, the chronology of the scenario is arguable. In the following, we propose plau-sible ages for the southward retreat and the northward recolonization of M. mulatta, based on the divergent level of mtDNA CR haplotypes.
First, the major southward retreat, which generated a complex pattern of CR haplotypes in the eastern China, occurred during the second last glacial (0.12–0.18 mya or earlier, corresponding to nucleotide diVerence >13 between sympatric haplotypes) instead of the LGM (0.018 mya). The retreat apparently displaced the native populations (the long-tailed mulatta populations) inhabiting the 15°N– 25°N latitudinal zone and resulted in a discrepant phyloge-netic relationship between M. cyclopis and M. mulatta. That is, M. cyclopis is not mitochondrially more related to
M. mulatta in the eastern China (such as populations in
Fujian and in Zhejiang). Instead, it is more related to M.
mulatta in the southwest China (Yunnan, Haplogroup I)
(Figs. 1 and 2). However, the connection between M.
cyclo-pis and the Haplogroup I is not well supported
(bootstrap D 28%) probably due to the extinction of the “transitional populations”. The extant populations of M.
mulatta in Yunnan and M. cyclopis may be two relictual
groups descended from ancestors more similar to the long-tailed mulatta populations. The mountain area in the south-western China proposed as a refugium for various animals and plants in China (Harrison et al., 2002) has protected Yunnan populations from displacement. Nonetheless, a decreasing west–east gradient of RTL from 90°E to 115°E (Fooden, 2000) and a more or less homogeneous pattern of allozymes (Fooden and Lanyon, 1989) and microsatellites (Morin et al., 1997) in M. mulatta indicate that male-medi-ated gene Xow from the eastern China population to the Yunnan population was common. Besides, postglacial
vicariance may prevent gene Xow between China and Tai-wan and thus retained the morphological character of a longer tail in M. cyclopis.
Subsequently, during the last interglacial (0.06–0.12 mya, corresponding to nucleotide diVerences about 6–13 between haplotypes), the major northward recolonization occurred that covered the current range of M. mulatta in China. Population expansion during this period can be inferred from a similar level of lineage divergence within provinces of Yunnan and Fujian, as well as between Guan-gxi, Fujian, Zhejiang and Vietnam. During the last glacial, populations may have been isolated in multiple refugia located in Henan, Zhejiang, Fujiang and Guangxi, instead of performing another large-scaled retreat. Similar loca-tions of glacial refugia in the southern China have also been suggested for a conifer species, Cunninghamia konishii (Lu et al., 2001).
Afterward, the Holocene expansion was mostly conWned locally. It was inferred from the observation that the same or related haplotypes were found within provinces of China, e.g. haplotypes of Haplogroup II in Guangxi, haplo-types in Zhejiang, Henan, and Yunnan (Fig. 2A). However, the occurrence of the same or related haplotypes in distant localities provides some exceptions (detailed data not showed). For example, haplotype Sichuan1 occurs in Sic-hun and Hainen, while haplotype Yunnana5 occurs in Yunnan and Hunan, as well as two closely related haplo-types, Zhejiang9 and Vietnam6, occur in two distant locali-ties Zhejiang and Vietam. The exceptions need to be re-examined for the possibility of anthropogenic introduction.
4.2. Divergence and demographic history of ancient mulatta populations
Mean nucleotide divergences between M. mulatta and each of the two island species, M. cyclopis and M. fuscata (dxy, Table 3), suggest that ancestral mulatta populations colonized Taiwan and Japan in approximately the same time period around 0.38–0.44 mya during the early stage of the third last glacial (0.32–0.44 mya). The result is also sup-ported by fossil records that suggest the existence of M.
cyclopis in Taiwan and M. fuscata in Japan at least 0.3 mya
(Fooden and Wu, 2001) and 0.4 mya (Kamei, 1969), respec-tively. Considering ancestral polymorphism, the most con-servative estimates of the population divergent time between M. mulatta and each of the two species are around 0.17 and 0.18 mya, respectively. The estimates indicate the times of population (or species) isolation were during the end of the second last interglacial (0.18–0.32 mya) and the beginning of the second last glacial (0.12–0.18 mya). The observed monophyly in M. cyclopis and M. fuscata sug-gests that ancestral population of each species originated from diVerent regional population in mainland at around 0.4 mya, and, during 0.17–0.4 mya, dispersed from main-land to Taiwan or Japan when dispersal routes were avail-able. Furthermore, the divergent level between the three species also indicates that successful dispersal was unlikely
since the second last glacial. It is concordant with the paly-nological evidence that most of the exposed continental shelf connecting Taiwan and China was covered by savanna-like vegetation (Liew et al., 1998) that is not suit-able for macaques. The cool and arid climate even extended to a lower latitudinal zone of the Sunda region ( Gathorne-Hardy et al., 2002; Heaney, 1991).
The MDA did not reject a smooth, unimodal distribu-tion of nucleotide diVerences between haplotypes in M.
mulatta and M. cyclopis. It suggests that the population
expansion occurred around 0.22 and 0.31 mya for M.
cyclo-pis and M. mulatta, respectively (Table 2). Besides, there are multiple modes (arrows placed in Fig. 3) embedded in the empirical curves of both species. The similarity of the modal patterns in the two species indicates that they have experienced similar demographic history during Pleisto-cene. The expansion that dated back to a more ancient past (0.31 mya) also indicates earlier population colonization.
Molecular dating of any event (e.g. divergence of popu-lations or population expansion) is subject to the referred mutation rate and needs to be interpreted with caution (Hillis et al., 1996). Nevertheless, the estimated ages of pop-ulation expansion in the study are consistent with the gla-cial history, and the divergent times between M. cyclopis,
M. fuscata and M. mulatta are within the ranges reported
elsewhere (Hayasaka et al., 1996; Marmi et al., 2004; Morales and Melnick, 1998; Tosi et al., 2003). The hypo-thetic scenario suggested by Fooden and Albrecht (1999) is supported by our mtDNA CR data. However, the ages of the retreat and the recolonization of ancestral M. mulatta may be dated back to a more ancient time during the late middle Pleistocene (0.13–0.18 mya) and late Pleistocene (0.06–0.12 mya) instead of the LGM (0.018 mya). In addi-tion, by applying the net nucleotide divergence (dA, Nei and Li, 1979) and a more complete panel of mtDNA sequences than previous studies, a more recent population divergent time, 0.17 mya, for the three species of the mulatta group is suggested.
Acknowledgments
We would like to acknowledge S.W. Chang, K.F. Lin, S.M. Chen, M.W. Fan, Wildlife labs of NDHU and NTU for their assistance of sample collection. GCG program ser-vice is kindly provided by National Healthy Research Insti-tutes (NHRI) in Taiwan. The research was supported by National Science Council, Taiwan (NSC-89-2311-B-002-082) and the fund of the Institute of Zoology, Academia Sinica to H.Y. Wu.
References
Avise, J.C., Neigel, J.E., Arnold, J., 1984. Demographic inXuences on mito-chondrial DNA lineage survivorship in animal populations. J. Mol. Evol. 20, 99–105.
Avise, J.C., Arnold, J., Ball, R.M., Bermingham, E., Lamb, T., Neigel, J.E., Reeb, C.A., Saunders, N.C., 1987. IntraspeciWc phylogeography: the
mitochondrial DNA bridge between population genetics and systemat-ics. Annu. Rev. Ecol. Syst. 18, 489–522.
Brandon-Jones, D., Eudey, A.A., Geissmann, T., Groves, C.P., Melnick, D.J., Morales, J.C., Shekelle, M., Stewart, C.B., 2004. Asian primate classiWcation. Int. J. Primatol. 25, 97–164.
Chen, S.Y., Su, Y.H., Wu, S.F., Sha, T., Zhang, Y.P., 2005. Mitochondrial diversity and phylogeographic structure of Chinese domestic goats. Mol. Phylogenet. Evol., 37, 804–814.
Chu, J.H., Wu, H.Y., Lin, Y.S., 2005. Mitochondrial DNA diversity in two populations of Taiwanese macaque (Macaca cyclopis). Conserv. Genet. 6, 101–109.
Conroy, C.J., Cook, J.A., 2000. Phylogeography of a post-glacialcolonizer: Microtus longicaudus (Rodentia: Muridae). Mol. Ecol. 9, 165–175. Deinard, A., Smith, D.G., 2001. Phylogenetic relationships among the macaques:
evidence from the nuclear locus NRAMP1. J. Hum. Evol. 41, 45–59. Delson, E., 1980. Fossil macaques, phyletic relationships and a scenario of
deployment. In: Lindburg, D.G. (Ed.), The Macaques: Studies in Ecol-ogy, Behavior, and Evolution. Van Nostrand Reinhold, New York, pp. 10–30.
Edwards, S.V., Beerli, P., 2000. Perspective: gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies. Evolution 54, 1839–1854.
Fooden, J., 1976. Provisional classiWcation and key to living species of macaques (Primates Macaca). Folia Primatol. 25, 225–236.
Fooden, J., 1997. Tail length variation in Macaca fascicularis and M mul-atta. Primates 38, 221–231.
Fooden, J., 2000. Systematic Review of the Rhesus Macaque, Macaca mul-atta (Zimmermann, 1780). Field Museum of Natural History, Chicago. Fooden, J., Albrecht, G.H., 1999. Tail-length evolution in fascicularis-group
macaques (Cercopithecidae: Macaca). Int. J. Primatol. 20, 431–440. Fooden, J., Lanyon, S.M., 1989. Blood–protein allele frequencies and
phylo-genetic-relationships in Macaca—a review. Am. J. Primatol. 17, 209–241. Fooden, J., Wu, H.-Y., 2001. Systematic review of the Taiwanese macaque, Macaca cyclopis Swinhoe, 1863. Field Museum of Natural History, Chicago.
Fu, Y.X., 1997. Statistical tests of neutrality of mutations against popula-tion growth, hitchhiking and background selecpopula-tion. Genetics 147, 915– 925.
Fu, J., Weadick, C.J., Zeng, X., Wang, Y., Liu, Z., Zheng, Y., Li, C., Hu, Y., 2005. Phylogeographic analysis of the Bufo gargarizans species com-plex: A revisit. Mol. Phylogenet. Evol. 37, 202–213.
Gathorne-Hardy, F.J., Syaukani, Davies, R.G., Eggleton, P., Jones, D.T., 2002. Quaternary rainforest refugia in south-east Asia: using termites (Isoptera) as indicators. Biol. J. Linn. Soc. Lond. 75, 453–466. Groves, C.P., 2001. Primate Taxonomy. Smithsonian Institution Press,
Washington, DC.
Hayasaka, K., Fujii, K., Horai, S., 1996. Molecular phylogeny of macaques: Implications of nucleotide sequences from an 896-base pair region of mitochondrial DNA. Mol. Biol. Evol. 13, 1044–1053. Harpending, H.C., 1994. Signature of ancient population-growth in a
low-resolution mitochondrial-DNA mismatch distribution. Hum. Biol. 66, 591–600.
Harpending, H.C., Sherry, S.T., Rogers, A.R., Stoneking, M., 1993. The genetic-structure of ancient human-populations. Curr. Anthropol. 34, 483–496.
Harrison, T., Ji, X.P., Su, D., 2002. On the systematic status of the late Neogene hominoids from Yunnan Province, China. J. Hum. Evol. 43, 207–227. Hatase, H., Kinoshita, M., Bando, T., Kamezaki, N., Sato, K., Matsuzawa,
Y., Goto, K., Omuta, K., Nakashima, Y., Takeshita, H., Sakamoto, W., 2002. Population structure of loggerhead turtles, Caretta caretta, nest-ing in Japan: bottlenecks on the PaciWc population. Mar. Biol. 141, 299–305.
Hayasaka, K., Ishida, T., Horai, S., 1991. Heteroplasmy and polymor-phism in the major noncoding region of mitochondrial-DNA in Japa-nese monkeys—association with tandemly repeated sequences. Mol. Biol. Evol. 8, 399–415.
Heaney, L.R., 1991. A synopsis of climatic and vegetational change in Southeast-Asia. Clim. Change 19, 53–61.
Hillis, D.M., Moritz, C., Mable, B.K., 1996. Molecular Systematics. Sinauer Associates, Sunderland, MA.
Hoelzer, G.A., 1997. Inferring phylogenies from mtDNA variation: Mitochon-drial-gene trees versus nuclear-gene trees revisited. Evolution 51, 622–626. Hoelzer, G.A., Wallman, J., Melnick, D.J., 1998. The eVects of social
struc-ture, geographical strucstruc-ture, and population size on the evolution of mitochondrial DNA: II. Molecular clocks and the lineage sorting period. J. Mol. Evol. 47, 21–31.
Huelsenbeck, J.P., Ronquist, F., 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755.
Jenkins, I., 2001. The Pleistocene. In: Atlas of the Evolving Earth. Macmil-lan Reference USA, New York, pp. 70–93.
Kamei, T., 1969. Mammals of the glacial age in Japan-especially on Japa-nese monkey. Monkey 106, 5–12.
Kocher, T.D., Thomas, W.K., Meyer, A., Edwards, S.V., Paabo, S., Villa-blanca, F.X., Wilson, A.C., 1989. Dynamics of mitochondrial-DNA evolution in animals—ampliWcation and sequencing with conserved primers. Proc. Natl. Acad. Sci. USA 86, 6196–6200.
Kumar, S., Tamura, K., Nei, M., 2004. MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinformat. 5, 150–163.
Li, Q.Q., Zhang, Y.P., 2005. Phylogenetic relationships of the macaques (Cercopithecidae: Macaca), inferred from mitochondrial DNA sequences. Biochem. Genet. 43, 375–386.
Liew, P.M., Kuo, C.M., Huang, S.Y., Tseng, M.H., 1998. Vegetation change and terrestrial carbon storage in eastern Asia during the Last Glacial Maximum as indicated by a new pollen record from central Taiwan. Glob. Planet. Change 17, 85–94.
Lu, S.Y., Peng, C.I., Cheng, Y.P., Hong, K.H., Chiang, T.Y., 2001. Chloro-plast DNA phylogeography of Cunninghamia konishii (Cupressaceae), an endemic conifer of Taiwan. Genome 44, 797–807.
Marmi, J., Bertranpetit, J., Terradas, J., Takenaka, O., Domingo-Roura, X., 2004. Radiation and phylogeography in the Japanese macaque, Macaca fuscata. Mol. Phylogenet. Evol. 30, 676–685.
Melnick, D.J., Hoelzer, G.A., Absher, R., Ashley, M.V., 1993. MtDNA diversity in rhesus monkeys reveals overestimates of divergence time and paraphyly with neighboring species. Mol. Biol. Evol. 10, 282–295. Morales, J.C., Melnick, D.J., 1998. Phylogenetic relationships of the
macaques (Cercopithecidae: Macaca), as revealed by high resolution restriction site mapping of mitochondrial ribosomal genes. J. Hum. Evol. 34, 1–23.
Morin, P.A., Kanthaswamy, S., Smith, D.G., 1997. Simple sequence repeat (SSR) polymorphisms for colony management and population genetics in rhesus macaques (Macaca mulatta). Am. J. Primatol. 42, 199–213. Nei, M., Kumar, S., 2000a. Accuracies and statistical tests of phylogenetic
trees. In: Molecular Evolution and Phylogenetics. Oxford University Press, New York, pp. 165–186.
Nei, M., Kumar, S., 2000b. Evolutionary change of DNA sequences. In: Molecular Evolution and Phylogenetics. Oxford University Press, New York, pp. 33–50.
Nei, M., Li, W.H., 1979. Mathematical model for studying genetic varia-tion in terms of restricvaria-tion endonucleases. Proc. Natl. Acad. Sci. USA 76, 5269–5273.
Parsons, T.J., Muniec, D.S., Sullivan, K., Woodyatt, N., AllistonGreiner, R., Wilson, M.R., Berry, D.L., Holland, K.A., Weedn, V.W., Gill, P., Holland, M.M., 1997. A high observed substitution rate in the human mitochondrial DNA control region. Nat. Genet. 15, 363–368.
Ritchie, P.A., Millar, C.D., Gibb, G.C., Baroni, C., Lambert, D.M., 2004. Ancient DNA enables timing of the Pleistocene origin and Holocene expansion of two Adelie penguin lineages in Antarctica. Mol. Biol. Evol. 21, 240–248.
Rogers, A.R., Harpending, H., 1992. Population-growth makes waves in the distribution of pairwise genetic-diVerences. Mol. Biol. Evol. 9, 552–569. Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic
inference under mixed models. Bioinformatics 19, 1572–1574. Ruedi, M., Castella, V., 2003. Genetic consequences of the ice ages on
nurs-eries of the bat Myotis myotis: a mitochondrial and nuclear survey. Mol. Ecol. 12, 1527–1540.
Saccone, C., Lanave, C., Pesole, G., Sbisa, E., 1993. Peculiar features and evolution of mitochondrial genome in mammals. In: DiMauro, S., Wallace, D. (Eds.), Mitochondrial DNA in Human Pathology. Raven Press, New York, pp. 27–37.
Saitou, N., Nei, M., 1987. The neighbor-joining method—a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. Schneider, S., ExcoYer, L., 1999. Estimation of past demographic
parame-ters from the distribution of pairwise diVerences when the mutation rates very among sites: Application to human mitochondrial DNA. Genetics 152, 1079–1089.
Schneider, S., Roessli, D., ExcoYer, L., 2000. ARLEQUIN: A software for population genetics data analysis, version 2.0. Genetics and Biometry Laboratory, University of Geneva.
Tosi, A.J., Morales, J.C., Melnick, D.J., 2000. Comparison of Y chromo-some and mtDNA phylogenies leads to unique inferences of macaque evolutionary history. Mol. Phylogenet. Evol. 17, 133–144.
Tosi, A.J., Morales, J.C., Melnick, D.J., 2002. Y-chromosome and mito-chondrial markers in Macaca fascicularis indicate introgression with Indochinese M. mulatta and a biogeographic barrier in the Isthmus of Kra. Int. J. Primatol. 23, 161–178.
Tosi, A.J., Morales, J.C., Melnick, D.J., 2003. Paternal, maternal, and biparental molecular markers provide unique windows onto the evolutionary history of macaque monkeys. Evolution 57, 1419–1435.
Wallman, J., Hoelzer, G.A., Melnick, D.J., 1996. The eVects of social struc-ture, geographical strucstruc-ture, and population size on the evolution of mitochondrial DNA. 1. A simulation model. Comput. Appl. Biosci. 12, 481–489.
Wright, S., 1943. Isolation by distance. Genetics 28, 114–138.
Zhang, Y.P., Shi, L.M., 1993. Phylogenetic relationships of macaques as inferred from restriction endonuclease analysis of mitochondrial DNA. Folia Primatol. 60, 7–17.
