Molecular Ecology (2004) 13, 1551–1565
doi: 10.1111/j.1365-294X.2004.02140.x
Mitochondrial DNA phylogeography of western lowland gorillas (Gorilla gorilla gorilla)
Blackwell Science, Ltd
S T E P H E N L . C L I F F O R D ,* N I C O L A M . A N T H O N Y ,† M I R E I L L E B A W E - J O H N S O N ,* K A T E A . A B E R N E T H Y ,*‡ C A R O L I N E E . G . T U T I N ,*‡ L E E J . T . W H I T E ,§ M A G D E L E N A B E R M E J O ,¶ M I C H E L L E L . G O L D S M I T H ,** K E L L E Y M C F A R L A N D ,†† K A T H R Y N J . J E F F E R Y ,† M I C H A E L W . B R U F O R D † and E . J E A N W I C K I N G S * *Centre International de Recherches Médicales, Franceville (CIRMF), BP 769, Franceville, Gabon, †School of Biosciences, Cardiff University, Cardiff CF10 3TL, UK, ‡Department of Molecular and Biological Sciences, Stirling, FK9 4LA, UK, §WCS, Bronx Zoo, 185th Street and Southern Boulevard, Bronx, NY10460-1099, USA, ¶ECOFAC, B.P. 15115, Libreville, Gabon, **Department of Environmental and Population Health, Tufts University School of Veterinary Medicine, 200 Westboro Road, North Grafton, MA 01536, USA, ††The Graduate Center, The City University of New York, 365 Fifth Avenue, New York, NY 10016-4309, USA
Abstract The geographical distribution of genetic variation within western lowland gorillas (Gorilla gorilla gorilla) was examined to clarify the population genetic structure and recent evolutionary history of this group. DNA was amplified from shed hair collected from sites across the range of the three traditionally recognized gorilla subspecies: western lowland (G. g. gorilla), eastern lowland (G. g. graueri) and mountain (G. g. beringei) gorillas. Nucleotide sequence variation was examined in the first hypervariable domain of the mitochondrial control region and was much higher in western lowland gorillas than in either of the other two subspecies. In addition to recapitulating the major evolutionary split between eastern and western lowland gorillas, phylogenetic analysis indicates a phylogeographical division within western lowland gorillas, one haplogroup comprising gorilla populations from eastern Nigeria through to southeast Cameroon and a second comprising all other western lowland gorillas. Within this second haplogroup, haplotypes appear to be partitioned geographically into three subgroups: (i) Equatorial Guinea, (ii) Central African Republic, and (iii) Gabon and adjacent Congo. There is also evidence of limited haplotype admixture in northeastern Gabon and southeast Cameroon. The phylogeographical patterns are broadly consistent with those predicted by current Pleistocene refuge hypotheses for the region and suggest that historical events have played an important role in shaping the population structure of this subspecies. Keywords: control region, Gorilla gorilla, mitochondria, phylogeography Received 25 August 2003; revision received 11 December 2003; accepted 6 January 2004
Introduction Cyclical climate fluctuations during the late Quaternary are believed to have had a considerable impact on the past distribution and range dynamics of many African taxa (Lanfranchi & Schwartz 1990; Maley 1996). The colder, drier periods experienced during glacial maxima are believed to have led to the retraction of tropical forests into refugia and fostered allopatric divergence between isolated populations (Haffer 1982; Kingdon 1990; Grubb 2001). Major montane Correspondence: Stephen L. Clifford. Fax: + 241 67 72 95; E-mail:
[email protected] Stephen Clifford and Nicola Anthony contributed equally to this work. © 2004 Blackwell Publishing Ltd
refugia have been identified in Cameroon and the highlands associated with the western rift of eastern Democratic Republic of Congo, Rwanda and Uganda ( Maley 1996; Haffer 1982; Kingdon 1990; Grubb 2001). Additional minor refugia have also been proposed, based on palynological, geological and biogeographical data (Sosef 1994; Rietkirk et al. 1995; Maley 1996; Grubb 2001). It has also been suggested that fluvial refugia may have existed along major watercourses in the Congo basin where forest cover persisted during the arid phases of the Pleistocene (Colyn 1991; Grubb 2001). Several predictions have been made about the genetic and evolutionary consequences of ice-age refugia (Haffer 1982; Hewitt 1996; Dynesius & Jansson 2000) and whilst molecular phylogeography has been used to test
1552 S . L . C L I F F O R D E T A L . some of these predictions, most studies have focused on temperate or arctic systems (e.g. Byun et al. 1997; Petit et al. 1997; Avise & Walker 1998; Avise et al. 1998; Arbogast 1999; Holder et al. 1999). In contrast, surprisingly few phylogeographical studies have been carried out in equatorial Africa with some notable exceptions (e.g. Morin et al. 1994; Goldberg & Ruvolo 1997a,b; Gonder et al. 1997; Smith et al. 2000; Flagstad et al. 2001; Jensen-Seaman & Kidd 2001; MulokoNtoutoume et al. 2001; Eggert et al. 2002). Gorillas are restricted to closed canopy forest and high-altitude montane forest and do not occur in forest– savannah mosaic habitats (Tutin et al. 1997). Thus, changes in forest cover during Pleistocene glaciations may have had a profound effect on the colonization patterns and regional distribution of gorillas. Based on cranial and dental morphology, three subspecies of gorilla have traditionally been recognized (Groves 1967, 1970), Gorilla gorilla gorilla (western lowland gorilla), G. g. graueri (eastern lowland gorilla) and G. g. beringei (mountain gorilla). Western lowland gorillas are the most numerous (Harcourt 1996) and are separated from eastern subspecies by more than 850 km. They are found in Gabon, Cameroon, Equatorial Guinea, Congo, Central African Republic (CAR), Cabinda and Nigeria (Tutin & Vedder 2001). Eastern lowland gorillas occur in fragmented populations of both lowland and highland habitat in the Democratic Republic of Congo (DRC) (Hall et al. 1998; Omari et al. 1999) whereas mountain gorillas persist in two small populations in Rwanda and Uganda (Harcourt 1996). A recent evaluation of the available data has lead to a reclassification of the gorilla into two species (Groves 2001), the western gorilla G. gorilla and the eastern gorilla G. beringei. Within western gorillas, two subspecies have been proposed: G. g. gorilla (western gorillas except those in the Cross River area between Nigeria and Cameroon) and G. g. diehli, comprising of no more than a few hundred individuals in and around the Cross River. Within eastern gorillas, three subspecies have been proposed: G. b. graueri (eastern lowland), G. b. beringei (Virunga mountain gorilla) and an as yet unnamed third taxonomic unit from the Bwindi forest, Uganda (Sarmiento et al. 1996; Groves 2001). Although gorillas exhibit substantial ecological and morphological differentiation (Groves 1967, 1970, 2001; Sarmiento et al. 1996; Doran & McNeilage 1998; Sarmiento & Oates 2000; and references therein), molecular studies have only recently begun to quantify levels of genetic variation within and between wild gorilla populations (Garner & Ryder 1996; Field et al. 1998; Saltonstall et al. 1998; Clifford et al. 1999, 2003; Jensen-Seaman & Kidd 2001; Oates et al. 2002). Information from mitochondrial DNA (mtDNA) sequences has revealed high levels of variability within western lowland gorillas (Ruvolo et al. 1994; Garner & Ryder 1996; Noda et al. 2001), greater than that observed within eastern or mountain gorillas (Garner & Ryder 1996). However, very little information is currently available on
the geographical distribution of genetic variation in gorilla populations of known origin. Several studies point to evidence of strong phylogeographical structure within gorillas. Jensen-Seaman & Kidd (2001) have examined mtDNA variation in several eastern lowland and mountain gorilla populations throughout eastern Africa and Oates et al. (1999, 2002) postulated a genetically distinct population of gorillas in the Cross River region corresponding to one of four geographically defined ‘demes’ identified by Groves (1967). This study therefore aims to expand on an earlier study (Clifford et al. 2003) and to explore the potential role that Pleistocene refugia may have played in shaping gorilla population genetic structure. The first hyper-variable region (HV1) of the mitochondrial genome was chosen for investigation since previous studies have shown this marker to be well suited to intraspecific studies in great apes (e.g. Morin et al. 1993, 1994; Garner & Ryder 1996; Goldberg & Ruvolo 1997a,b; Gonder et al. 1997; Gagneux et al. 1999, 2001).
Materials and methods Sample collection and DNA extraction Shed hair collected from night nests (Tutin et al. 1995) comprised all of our samples except a single fecal sample collected in Gabon and a single museum pelt specimen reportedly collected in the 1930s from Belar, Cameroon. Samples throughout the entire gorilla range were collected over a period of several years through collaborations with researchers from existing study sites. Hair samples were stored at room temperature in a plastic container with silica gel. The fecal sample was collected in RNAlater (Ambion, Austin, TX, USA) and subsequently stored at −20 °C. Table 1 lists these contributors and the populations sampled. Hair was extracted using either a modification of the Chelex-100 protocol of Walsh et al. (1991) or a method modified from the protocol described by Vigilant (1999) based on proteinase K digestion of hair roots in a polymerase chain reaction (PCR) compatible buffer. DNA from the museum pelt was extracted using the QIAamp® DNA kit extraction protocol (Qiagen). Fecal DNA was extracted using the Qiagen Stool Extraction Kit (Qiagen). The CITES numbers for exportation and importation of all hair extracts are 00174 2 and 16640/01, respectively.
Mitochondrial DNA amplification As DNA extracted from noninvasive samples is highly degraded, only short DNA fragments could be amplified reliably (Clifford et al. 2003). Nested primers were designed to amplify a 258-base-pair (bp) fragment of the mitochondrial HV1 using the first-round primers PDPF1 (5′CACCATCAGCACCCAAAGCTAATAT-3′) and PDPR2 (5′-TTGTGCGGGATATTGATTTCACGGA-3′) and secondround primers L91-115 and H402-27 (Garner & Ryder 1996). © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1553 Table 1 List of sample sites, country of origin, site code, site number, number of individuals analysed per site and collectors of samples from wild gorilla populations examined in this study. *The sample from Rabi (site 7) was from feces. All other samples were derived from shed hair. **CDP (site 11) is a captive individual from the Centre de Primatologie (CDP) at CIRMF, Gabon, who originated from central Gabon. ***Samples from Belar (site 22) and Uele (site 23) are from Museum specimens
Site
Country
Code
Site no.
Bwindi Karisoke Kahuzi-Biega
Uganda Rwanda DRC
BWD RWD KBG
1 2 3
4 2 6
Tshiaberimu Itombwe Conquati Rabi* Petit Loango Lopé
DRC DRC Congo Gabon Gabon Gabon
TSH ITW CQT RAB PLO LOP
4 5 6 7 8 9
1 3 1 1 1 3
Lastourville CDP**
Gabon Gabon
LAS CDP
10 11
1 1
Ipassa Belinga Itombe Lossi Bai Hokou
Gabon Gabon Gabon Congo CAR
IPS BEL ITO LOS CAR
12 13 14 15 16
2 2 1 1 8
Nouabalé-Ndoki Lobéké Dja Afi Mts./Cross River Mont Alen
Congo Cameroon Cameroon Nigeria Equatorial Guinea Cameroon DRC
NDK LBK DJA CRS EQG
17 18 19 20 21
1 5 4 4 5
(WCS). Sequences from Genbank (L76749, L76752) Sequences from Genbank (L76750, L76751) D. Bonny, K.P. Kiswele (CNRS), I. Omari, C. Sikubwabo (ICCN), L. White, J. Hall, I. Bila-Isia, H. Simons Morland, E. Williamson, K. Saltonstall, A. Vedder, K. Freeman, B. Curran (WCS) J. Yamagiwa (Kyoto). Sequences from Genbank (L76771, L76772, L76773, AF187549) Sequence from Genbank (AF50738) I. Omari, F. Bengana ( ICCN); J. Hart ( WCS) B. Goossens (UWC), A. Jamart (HELP) S. Lahm (IRET) J. Yamagiwa (Kyoto University) C. Tutin, K. Abernethy, E. Dimoto, J.T. Dinkagadissi, R. Parnell, P. Peignot, B. Fontaine (CIRMF), M.E. Rogers, L. White, B. Voysey, K. McDonald, (Edinburgh), R. Ham (Stirling), J.G. Emptaz-Collomb Y. Mihindou (WCS-MIKE) J. Wickings (CIRMF) (equivalent to sequence from Genbank L76764) S. Lahm, J. Okouyi (IRET) S. Lahm (IRET). Sequence from Genbank (L76763) P. Telfer (NYU) M. Bermejo, G. Illera, F. Maisels (ECOFAC) M. Goldsmith (Tufts University), L. White (WCS). Sequences from Genbank (AY079508, AY079509, AY079510, L76761) P. Walsh (WCS) L. White, L. Usongo (WCS) E. Williamson (ECOFAC), L. Usongo (WCS/ECOFAC) K. McFarland, J. Oates (CUNY, USA). E. Nwufoh (CRNP) M. Bermejo, G. Illera (ECOFAC)
CAM UEL
22 23
1 1
M. Harman (Powell-Cotton Museum) Sequences from Genbank (AJ422244)
Belar*** Uele***
No. individuals
Cycle conditions for both first- and second-round PCR were as follows: 94 °C for 3 min followed by 50 cycles of 94 °C for 30 s, 50 °C for 30 s and 72 °C for 30 s and a final step of 72 °C for 10 min. Both extraction blanks and reaction blanks containing only PCR reagents were also included in nested reactions to control for potential contamination. Extraction blanks or PCR blanks almost never yielded products. The exception to this was the occasional amplification of human DNA which could be readily identified owing to a large deletion (∼70 bp) in the gorilla HV1 sequence relative to human and chimpanzee sequences (see Garner & Ryder 1996). For fecal DNA extracts, PCR products were amplified over 35 cycles from 1 µL DNA using the L91/ H402 primer combination and an annealing temperature of 62 °C. Second-round PCR reactions were conducted using identical conditions. All reactions contained final concentrations of 1.5 mm Mg2+, 1× buffer (Invitrogen), © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
Collectors
200 µm dNTPs, 0.2 µm of each primer and 0.5 units of Taq polymerase in a 10-µL volume PCR reaction.
DNA sequencing and data analysis PCR products were purified using the Turbo-purification KitIII (Gene Clean) and were either cloned into the PCR2.1TOPO vector (Invitrogen) prior to sequencing with M13 Forward (−20) and M13 reverse primers or were directly sequenced with L91-115 and H402-427 primers. In all cases, DNA was sequenced using the Big Dye V2.0 sequencing kit (Applied Biosystems). Sequences derived from cloned PCR products were assembled from a consensus of a mean of 4.7 clones per PCR product. Several published haplotypes of known geographical provenance were also included in phylogenetic analyses. These sequences are available in GenBank under the following accession numbers: L76749,
1554 S . L . C L I F F O R D E T A L . AY530102 to AY530154). To test whether different PCR primers yielded equivalent product and subsequent sequences, a subset of seven samples (EQG1, EQG2, DJA2, CAR3, CAR4, LOP1, LOP2) were PCR amplified using the primers and conditions described by Oates et al. (2002) and the resulting sequences were found to match those generated using the nested PCR strategy described above. In addition, a previous study (Clifford et al. 2003) based on a smaller fragment of the mitochondrial control region, again used different primer sets, and also yielded equivalent sequences. Sequences were aligned in clustal X (Thompson et al. 1997) using the multiple alignment default parameters. A 26-bp region encompassing the poly C domain of HV1 was excluded from principal phylogenetic analyses because of difficulties in alignment and possible length and site heteroplasmy within this region (Bendall & Sykes 1995; Garner & Ryder 1996). However, phylogenetic analysis motifs within the poly C domain and diagnostic sites in the sequence flanking the poly C domain were used to classify putative Numt sequences into two categories (I and II). Table 2
L76750, L76751, L76752 (mountain gorillas); L76771, L76772, L76773, AF187549, AF050738 (eastern lowland gorillas); L76754, L76760, L76761, L76763, L76764, L76766, AY079508, AY079509, AYO79510, AF250888 (western lowland gorillas). Also included was the sequence (AJ422244) obtained from a museum specimen from Bondo, in the Uele Valley (D.R.C.) where gorillas do not occur (Hofreiter et al. 2003). A series of candidate nuclear copies of mtDNA (Numts) of the mtDNA control region available through GenBank (Accession Numbers AF240448 –AF240453, AF240455 – AF240458) were also included in phylogenetic analyses. Minor differences in the polycytosine (poly C) domain sequence were observed between clones and were coded by consensus. Occasionally base conflicts between plasmid clones outside the poly C domain were encountered and these changes were coded by majority. These base conflicts could arise through (i) PCR errors, (ii) heteroplasmy which has been diagnosed in hairs (Grzybowski 2000), or (iii) Numt allele polymorphisms. The sequences generated from this study have been deposited in GenBank (Accession Numbers
Table 2 Shared derived characters used to diagnose different Numt groups and consensus sequences from a 26 bp tract of the poly cytosine (Poly C) domain within HV1. The base pair position of each diagnostic sites and the start and end sites of the Poly C sequence are derived from a 258 bp alignment of all the sequences used in this study. Variable sites are coded by their respective degenerate code. Indels are represented by consensus. The symbol (-) indicates a deletion relative to other sequences in the full alignment. The symbol (?) indicates missing data. Sequence origin (Cloned, PCR product or Genbank) and haplogroup association/Numt class are also indicated
Sequence
0 7 9
0 8 9
0 9 2
0 9 3
1 1 2
1 3 8
1 5 7
1 6 3
Origin
RWD1_L76751 RWD2_L76750 BWD1 BWD2 BWD3_L76749 BWD4_L76752
C C C C C C
C C C C C C
C C C C C C
C C C C C C
CCCCTCACCCCCCATTCCCTGCTCAC CCCCTCACCCCCCATTCCCTGCTCAC CCCCTCACCCCCCATTCCCTGCTCAC CCCCTCACCCCCCATTCCCTGCTCAC CCCCTCACCCCCCATTCCCTGCTCAC CCCCTCACCCCCCATTCCCTGCTCAC
A A A A A A
T T T T T T
GENBANK GENBANK PCR PCR GENBANK GENBANK
TSH1_AF50738 KBG1 KBG2 KBG3_L76773 KBG4_AF187549 KBG5_L76771 KBG6_L76772 ITW1 ITW2 ITW3
C C C C C C C C C C
C C C C C C C C C C
T T T T T T T T T T
C C C C C C C C C C
CCCCTCACCCCCCATCCCTTGCCCAC CCCCTCACCCCCCATCCCTTGCCCAC CCCCTCACCCCCCATCCCTTGCCCAC CCCCTCACCCCCCATCCCTTGCCCAC CCCCTCACCCCC-ATCCCTTGCCCAC CCCCTCACCCCC-ATCCCTTGCCCAC CCCCTCACCCCC-ATCCCTTGCCCAC CCCCTCACCCCC-ATCCCTTGCCCAC CYCCTCACCCTC-ATCCCTTGCCCAC CCCCTCACCCCC-ATCCCTTGCCCAC
A A A A A A A A A A
C C C C C C C C C C
GENBANK PCR PCR GENBANK GENBANK GENBANK GENBANK PCR PCR PCR
CRS1 CRS4 CRS2 CRS3 LBK1 LBK2 LBK4 LBK5 DJA1 DJA4 DJA2 UEL1_AJ422244 IPS2 DJA3
C C C C C C C C C C C C C C
C C C C C C C C C C C C C C
T T T T T T T T T T C C C C
C C C C C C C C C C C C C C
CTCCC-CTTCCCCCCCCC-TCCTCCA CCCCCTCTTCCCCCCCCC-TCCTCCA CCCCCCCTTCCCCCCCCC-TCCTCCA CCCCCCCTTCCCCCCCCC-TCCTCTA YCCCYCCTTCCCCCCYCCCTTCTCCA CCCCYCCTTCCCCYCYCC-TCCTCCA CCCCCCCTTCCCCCCCCC-TTCTCCA CTCCCCCTTCCCCCCCCC-TCCTCCA CCCCCYCYTYCCCCCCCCCTCYTCYA CCCCCCCTTCTCCCCCCC-TCCTCCA CCCCCC-TT---------------CA YYYYYYYYYYYYYYYYYYYYYYYYYA CCCCCCCTT--CCCCCCCCCCCTTCA CCCCCYCTT-CCCCTCCCCCCCTCCA
A A A A A A A A A A A A A A
T T T T T T T T T T T T T T
PCR PCR PCR CLONE CLONE CLONE PCR CLONE CLONE CLONE PCR GENBANK CLONE CLONE
Poly C
Haplogroup/ Numt class
Haplogroup A
Haplogroup B
5 4 4 6 4 4 7
Haplogroup C
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1555 Table 2 Continued
Sequence
0 7 9
0 8 9
0 9 2
0 9 3
1 1 2
1 3 8
1 5 7
1 6 3
Origin
CAR1 NDK1 CAR7 CAR8_AY079509 CAR3 CAR2_L76761 CAM1 CAR4 CAR6_AY079510 CAR5_AY079508 LBK3 EQG1 EQG2 EQG3 EQG5 EQG4 RAB1 LOS1 LAS1 BEL1 LOP3 ITO1 BEL2_L76763 LOP2 CQT1 CDP1_L76764 IPS1 PLO1 LOP1
C C C C C C C C C C C C C C C C C C C C C C C C C C C C C
C C C C C C C C C C C C C C C C C C C C C C C C C C C C C
T T T T T T T T T T T C C C C C C C C C C C C C C C C C T
C C C C C C C C C C C C C C C C C C C C C C C C C C C C C
CCCCCCCTTCCCCCCCCCCCGCACTG CCCCCCCTTCCYCCCCCC-CGCACTG CCCCCCCTTTCCCCCCCCCCGCACTG CCCCCCCTTCNCCCCCCC-CGCACTG CCCCCCCTTCCCCCCCCC-CGCACTG CCCCCCCTTCCCCCCCCCCCGCACTG CCCCCCCTTCCCCCCCCC-CGCACTG CCCCCCCTTCCCCCCCCCCCGCACTG CCCCCCCTTCCCCCCCCC-CGCATTG CCCCCCCTTCNCCCCCCC-CGCACTG CCCCCCCTTCCCCCCCCCCTGCACTG CCCCCCCTTCCCCCTCCCCTCCTCTG CCCCCCCTTCCCCCTCCCCTCCTCTG YCCCCCCTTCCCCCCCCC-TCCTCTG CCCCCCCTTCCCCCCCCC-TCCTCTG CCCCCCCTTCCCCCCCCC-CGCTCTG CCCCCCCTTCCCCCCCCCCCGCTCTG CCCCCCCTTCCCCCCCCCCCGCTCTG CCCCCCCTT-CCCCCYCC-CCYTCTG CCCCCCCTTCCCCCCCCC--GCTCTG CCCCCCCTTCCCCCCTCC-CGCTCTG CCCCCCCTTNCCCCCCCC-CGCTCTG CCCCCC-TTCCCCCCCCC-CGCTCTG CCCCCCCTTCCCCCCCCC-CGCTCTG CCCCCCCTTCCCCCCCCCC-GCTCTG CCCCCCCTTCCCCCCCCC-CGCTCTG CCCCCCCTTCCCCCCCCC-CGCTCTG CCCCCCYTTCCYCCCCCC-CGYTYTG CCCCCCCTTCCCCCCCCC-CGCTCTG
A A A A A A A A A A A A A A A A ? A A A A A A A A A A A A
T T T T T T T T T T T T T T T T T T T T T T T T T T T T T
CLONE CLONE PCR GENBANK PCR GENBANK PCR PCR GENBANK GENBANK PCR PCR PCR CLONE PCR PCR CLONE CLONE CLONE CLONE CLONE CLONE GENBANK PCR CLONE GENBANK CLONE CLONE PCR
ITO1a AF240448 AF240456 BEL1a L76760
C T T T T
C C C C T
C C C C C
C C C C C
CCCCCCCCT--C-----ACTGCTCCA CCCCCCCCT--C-----ACTGCTCCA CCCCCCCCT--C-----ACTGCTCCA CCCCCCCCT--C-----ACTGCTCCA CCCCCCCTC--CCCCCTACTGCTCCA
A A A A A
T T T T T
CLONE GENBANK GENBANK CLONE GENBANK
AF240452 ITW4 AF240449 AF240457 L76754 DJA5b
T T T T T T
T T T T T T
A A A A A A
C C C C C C
TTCCCC----CCCTCCGC----TCCA TTCCCC----CCCTCCGC----TCCA TTCCCC----CCCTCCGC----TCCA TTCCCC----CCCTCCGC----TCCA TTCCCC----CCCCCCCC--GCTCCA TTCCCC----CCCCCCCCCSGYTCCA
G G G G G G
G G G G G G
GENBANK PCR GENBANK GENBANK GENBANK CLONE
LBK5a Rab1a L76766 AF240455 AF240453 AF250888 AF240451
T T T T T T T
C C C C C C C
A A A A A A A
T T T T T T T
CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA CCCCCC-----CCATCCCCTGCTCTA
G G G G G G G
-
CLONE CLONE GENBANK GENBANK GENBANK GENBANK GENBANK
DJA4a ITW5 LOP4 AF240450 AF240458 DJA5a
T T T T T T
C C C C C C
A A N A A A
C C C C C C
ACCCCCTC--CCCACTTCCTGCTCCA ACCCCCTC--CCCACTTCCTGCTCCA ACCCCCTC--CCCAYTTNCTGTTCCA ACCCCCTC--CCCACTTCCTGCTCCA ACCCCCTC--CCCACTTCCTGCTCCA ACCCCCTC--CCCACTTCCTGCTCCA
G G G G G G
-
CLONE PCR PCR GENBANK GENBANK CLONE
Poly C
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
Haplogroup/ Numt class
5 4 4 4 4 4 4 6 Haplogroup D 4 4 4 4 4 4 7
Numt Class I
5 4 4 4 4 6 4 4 4 4 7
Numt Class II
1556 S . L . C L I F F O R D E T A L . lists the poly C sequence and synapomorphic sites for each sequence. These Numt sequences were included in initial phylogenetic analyses but were removed from the minimum spanning network and subsequent population genetic analyses. Putative Numts were identified in both eastern and western gorilla samples from the following sites: two samples from Itombwe, DRC (site 5), two samples from Dja (site 19), one sample from Lobéké (site 18), one sample from Rabi, Southwestern Gabon (site 7), one sample from Lopé, Central Gabon (site 9), and one sample from Belinga (site 13) and Itombwe (site 14), both in Northeastern Gabon. For phylogenetic analyses, the two data sets (with and without Numts) were analysed with modeltest 3.04 (Posada & Crandall 1998) to determine the substitution model that best fitted the data according to a hierarchical likelihood ratio test. The K81uf + G model was selected in both cases. modeltest was also used to estimate among-site-rate heterogeneity by estimating the value of the α shape parameter of the gamma distribution. Estimates were derived separately for the complete data set including all putative Numt sequences (α = 0.39) as well as for unique mitochondrial haplotypes (α = 0.49). Sequences were then analysed by the neighbour-joining (Saitou & Nei 1987) and maximum likelihood (Cavalli-Sforza & Edwards 1967) methods implemented in paup 4.0b10 (Swofford 1998) using the appropriate model and empirical base frequencies. The programme arlequin version 2.0 (Schneider et al. 2000) was used to estimate nucleotide diversity within nominal subspecies (Nei 1987), calculate per cent sequence divergence estimates between mtDNA haplogroups, carry out a hierarchical analysis of molecular variance (amova: Excoffier et al. 1992) and construct a minimum spanning network of mtDNA haplotypes. For the amova, sequences were partitioned either by subspecies (sensu Groves 1970) or by the four principal haplogroups (A–D) recovered in our phylogenetic analysis. Populations within these major subdivisions were defined by their respective subgroups (C1, C2, D1–3). The demographic parameter τ was estimated for haplogroups and subgroups within haplogroups with sample sizes of 10 sequences or more using a general nonlinear least squares approach (Schneider & Excoffier 1999). Confidence intervals for these estimated parameters (α = 0.05) and the validity of a model of sudden demographic expansion within haplogroups was also assessed (Schneider & Excoffier 1999).
Results Sequence divergence patterns and population genetic structure In total, 83 sequences were analysed, 53 of which were generated in this study and 30 were retrieved from GenBank. All sequences were examined over a 232-bp region
equivalent to the HV1 domain that excluded a 26-bp region of the Poly C domain. Of these sequences, 59 of the 83 sequences were presumed mitochondrial (Fig. 1), of which 16 were derived from eastern gorillas (six mountain and 10 eastern lowland) and 43 from western lowland gorillas. Two of these western lowland gorilla mitochondrial sequences were amplified from museum specimens (CAM1 and UEL1). Of the 43 western gorilla sequences, 20 were derived from cloned PCR products, 16 were derived from direct sequencing of PCR products and seven were derived from GenBank. Overall, 36 unique haplotypes (Fig. 2) were identified across 23 sites throughout the range of western lowland gorillas. The number of individuals sequenced per site is illustrated in Table 1. The remaining 24 sequences, comprising seven cloned, three PCR and 14 GenBank derived sequences were characterized as Numts and following phylogenetic classification were subsequently removed from population genetic analysis. Phylogenetic analysis, using both neighbour-joining and maximum likelihood methods gave concordant tree topologies with or without Numts sequences. Figure 1 shows the neighbour-joining tree generated with all sequences (n = 83) included in the analysis and Fig. 2 the maximum likelihood tree with only unique mitochondrial haplotypes (n = 36). Both phylogenetic analyses (Figs 1 and 2) and a minimum spanning network of the data (Fig. 3a) recovered four principal haplogroups (A–D). Haplogroups A and B correspond to mountain and eastern lowland gorillas, respectively. Haplogroups C and D are restricted to western lowland gorillas, are largely nonoverlapping, and are both distributed over wide geographical areas. Figure 3(b) illustrates the spatial distribution of haplotypes within the four major haplogroups and their proportional representation. Haplogroup C (n = 14) spanned an area from the Cross River in Nigeria (site 20), through Dja (site 19) to Lobéké in Southeast Cameroon (site 18), Ipassa in Gabon (site 12) and a museum sample collected from the DRC (site 23). Haplogroup D (n = 29) comprised all other western gorillas sites from the CAR, Congo, Equatorial Guinea and Gabon (sites 6–18) and one museum sample from southern Cameroon (site 22). Whereas haplogroup C was the most diverse, there appeared to be little or no geographical pattern among phylogenetic subdivisions (C1, C2) within this group. Conversely, phylogenetic analysis of haplogroup D recovered three well-defined genetic subdivisions (D1–D3) that were each confined to the following areas: (i) Equatorial Guinea (D1), (ii) CAR (D2) and (iii) Gabon and adjacent Congo (D3). Limited haplotype mixing between major haplogroups C and D was evident in southeastern Cameroon and northeastern Gabon where one sample from Lobéké, Cameroon (otherwise C) appeared to be almost identical to haplotypes across the river in CAR (D2). Similarly, one sample from northeastern Gabon (otherwise D3) possessed a haplotype characteristic of haplogroup C. Finally, a museum specimen ostensibly © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1557
Fig. 1 Neighbour-joining bootstrap consensus tree of gorilla mitochondrial HV1 sequences alongside candidate Numt sequences. Threeletter taxa name with number corresponds to sample identification given in Table 1 and GenBank accession numbers are given for those samples retrieved from the database. Numbers above tree branches correspond to the percentage bootstrap replicates for that branch. Estimates of bootstrap support are based on 1000 replicates and the tree is unrooted. Haplogroups A to D and subgroups C1, C2, D1, D2 and D3 are indicated, and candidate Numts (classes I and II) are shaded in grey.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
1558 S . L . C L I F F O R D E T A L . Fig. 2 Maximum likelihood tree with branch lengths generated from unique mitochondrial HV1 haplotypes. Midpoint rooting is employed. Three-letter taxa name with number corresponds to sample identification given in Table 1 and GenBank accession numbers are given for those samples retrieved from the database. Taxa names labelled with an asterisk represent multiple individuals with shared common haplotype. These are BWD2/3 (BWD2, BWD3_L76749), KBG1/3 (KBG1, KBG3_L76773), ITW1/2/3 (ITW1, ITW2, ITW3), LBK1/2/5 (LBK1, LBK2, LBK5), EQG1/2/3/5 (EQG1, EQG2, EQG3, EQG5), CAR/LBK/NDK (CAR1, CAR7, CAR8_AY079509, LBK3, NDK1) and GAB/ CON (BEL1, BEL2_L76763, CQT1, ITO1, LOP2, LOP3, LAS1, LOS1, PLO1, RAB1). Haplogroups A to D and subgroups C1, C2, D1, D2 and D3 are indicated.
from southwestern Cameroon possessed a haplotype characteristic of gorillas from the Sangha river region. Mean absolute pairwise sequence differences between eastern (haplogroup A and B) and western lowland gorillas (haplogroups C and D) was high (37.79/232; 16.29%) and recapitulates the large genetic distance previously identified between eastern and western lowland gorillas. By comparison, mean percentage pairwise estimates between mountain (haplogroup A) and eastern lowland gorillas (haplogroup B) were relatively low (14.85/232; 6.40%). However, the most striking divergence was between haplogroups C and D within western gorillas which at 21.59/232 (9.31%), is higher than the average divergence between the two eastern subspecies. Within haplogroups A–D, pairwise divergences were lower in the eastern haplogroups A (1.93/232; 0.83%) and B (2.00/232; 0.86%), than within the two western haplogroups, C (7.43/232; 3.20%) and D (7.12/232; 3.07%). In addition, pairwise sequence differences were lowest within subgroups from haplogroup D (and equivalent to values obtained for the eastern
subspecies), ranging from 0.59/232 (0.25%) in D1 to 2.00/ 232 (0.86%) in D2 and 2.11/232 (0.91%) in D3. In contrast, pairwise sequence differences were higher in the two subgroups C1 (3.24.43/232; 1.4%) and C2 (3.67/232; 1.6%). Estimates of nucleotide diversity (Nei & Li 1979) were also highest in western lowland gorillas overall (0.062 ± 0.031) and were comparable with previous published estimates (Garner & Ryder 1996; Jensen-Seaman & Kidd 2001). This value is over twice as large as that estimated for published human sequences (Vigilant et al. 1991) and approximately six times greater than that observed for either eastern lowland (0.009 ± 0.006) or mountain (0.008 ± 0.006) gorillas. Jensen-Seaman & Kidd (2001) also report similar estimates of nucleotide diversity in their study of eastern gorillas. Within western lowland gorillas, nucleotide diversity was equivalent in haplogroups C (0.032 ± 0.018) and D (0.031 ± 0.017), and was equally distributed between haplogroup C subdivisions C1 (0.014 ± 0.009) and C2 (0.016 ± 0.012), respectively, but not across subdivisions within haplogroup D (0.003 ± 0.003, 0.011 ± 0.007 and 0.009 ± 0.007 © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1559 Fig. 3 (a) Minimum spanning network constructed using the programme arlequin (Schneider et al. 2000) of pairwise absolute differences between gorilla mitochondrial DNA haplotypes. Three-letter taxa name with number corresponds to sample identification given in Table 1. The area of each circle is proportional to the number of sequences in each. Branch lengths are also proportional and hash marks for closely related haplotypes indicate individual mutational steps. Haplogroups A to D are colour coded and subgroups C1, C2, D1, D2 and D3 are indicated. (b) Geographic distribution and haplogroup designation (A–D) of sequences sampled from sites 1–23. The area of each circle is proportional to the number of sequences analysed at each site. The present day geographical distribution of gorillas is shaded in grey. Within the western gorilla range, subgroups within major halogroups C and D are as follows, C1 (sites 18–20), C2 (sites 12, 19, 23), D1 (site 21), D2 (sites 16 –18, 22) and D3 (sites 6–15). Gorillas from site 12 ( IPS — Ipassa, Gabon) and 19 ( LBK — Lobéké, Cameroon) exhibit haplotypes from both major haplogroups (C and D) and this is reflected in circle coloration.
in D1, D2 and D3, respectively). Of note are the values obtained for subgroup D3 which not only covers a large geographical area ( Fig. 3b) containing over 50% of extant gorillas ( Harcourt 1996) but also possesses the lowest genetic variability of any of the subgroups examined. An amova, where genetic variation was partitioned among the three traditional gorilla subspecies (mountain, eastern lowland and western lowland) indicated that over half of the total molecular variance (50.98%) was attributable to the differences among subspecies. A substantial proportion of the total variance (43.47%) was also attributable © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
to differences among subdivisions (C1, C2, D1–D3) within haplogroups. When western gorillas were subdivided into two groups corresponding to haplogroups C and D in our phylogenetic analysis, the among group variance rose to 65.04% of the total molecular variance. In the western gorillas, mismatch distribution patterns within haplogroups C and D were bimodal and multimodal respectively, indicating population structure and complex evolutionary histories within these two haplogroups. In contrast, mismatch distributions in the eastern haplogroup B were unimodal, and were consistent with a model of sudden demographic
1560 S . L . C L I F F O R D E T A L . expansion with a τ-value of 2.115, range 0.506 –3.187. Haplogroups A and D1 had too few individuals to be included in this analysis. When mismatch distributions within haplogroup D were analysed by subgroups with 10 or more sequences (D2, D3), the pattern of pairwise differences was also consistent with a model of sudden demographic expansion with τ-values ranging from 2.699 (1.081–5.56) in subgroup D2, to 1.105 (0.000 –3.202) in subgroup D3. These values are comparable with previous estimates from eastern lowland and mountain gorillas (Jensen-Seaman & Kidd 2001).
Characterization of putative Numt groups Nuclear transfers of mitochondrial DNA copies have been reported in both humans (Zischler et al. 1995; Mourier et al. 2001) and nonhuman primates (Collura & Stewart 1995; van der Kuyl et al. 1995; Zischler et al. 1998; Mundy et al. 2000). Several presumed Numts of the gorilla HV1 region were identified by phylogenetic analysis and are highlighted in grey boxes in the phylogeny depicted in Fig. 1. The classification of sequences into one of two Numt classes (labelled I–II) was based on synapomorphic sites within HV1 and distinct motifs within the poly C (Table 2). Numt class I was made up of divergent sequences from two cloned PCR product sequences from western gorillas (BEL1, ITO1) as well as two putative Numt sequences from GenBank (AF240448, AF240456). Within Numt class II, five cloned PCR products from four western gorillas (RAB1, DJA4, DJA5 and LBK5) yielded highly divergent sequences that clustered with putative Numt sequences (AF240449– AF240453, AF240455, AF240457, AF240458). In addition, direct PCR sequencing of samples from two eastern (ITW4, ITW5) and one western (LOP4) gorilla yielded only putative Numt Class II sequences. Finally, four western lowland gorilla sequences previously submitted to GenBank clustered with sequences from either Numt Class I (L76760) or II (L76754, L76766, AF250888). It is possible that all of the sequences presented as western gorilla mitochondrial DNA in haplogroups C and D (including mitochondrial sequences retrieved from GenBank) are Numts. This is a highly unlikely scenario, as similar mitochondrial-like western gorilla control region sequences have been amplified by independent research groups (Horai et al. 1995; Garner & Ryder 1996; Xu & Arnason 1996; Oates et al. 2002; Clifford et al. 2003) using multiple combinations of different primer pairs that all yielded equivalent sequences. In addition, the two complete gorilla mitochondrial genomes previously published by Horai et al. (1995) and Xu & Arnason (1996) contain HV1 sequences similar to the mitochondrial sequences published in this paper, clustering strongly with haplogroups C and D, respectively. The Xu & Arnason (1996) mitochondrial genome, in particular, is highly unlikely to be of nuclear origin because it was obtained from DNA enriched with
respect to mtDNA. Finally, our own analysis, involved multiple lines of evidence including the separation of mitochondrial and putative Numts by PCR product cloning, PCR amplification of HV1 using different PCR primer sets and the identification of Numt group synapomorphic sites and poly C motifs (see Table 2).
Phylogenetic relationships Figure 1 depicts the phylogenetic relationships between the 83 gorilla-derived sequences examined in this study whereas Fig. 2 depicts the relationship between unique gorilla mitochondrial haplotypes, excluding all Numts. As reported previously (Ruvolo et al. 1994; Garner & Ryder 1996), phylogenetic analyses reveal a major evolutionary split between eastern and western gorillas. Mountain gorillas (haplogroup A) are clearly differentiated from all eastern lowland gorillas (haplogroup B). The most striking aspect of this analysis occurs in the western gorilla, with a deep phylogenetic break apparent within western lowland gorillas (haplogroups C and D) that is strongly supported in both neighbour-joining and maximum likelihood analysis. The two haplogroups are predominantly monophyletic with evidence of limited mtDNA exchange at locations where the two haplogroups potentially come into contact. This haplotype exchange is shown by the presence of a Southeast Cameroon sequence (LBK3) in haplogroup D2 and a Gabonese sequence (IPS2) in haplogroup C2. Of note, is a single museum specimen, collected in southwestern Cameroon (CAM1) in the 1930s, that clusters with other sequences from haplogroup D2, despite being more than 700 km west of any D2 gorilla examined in this study. A second museum-derived sequence (AJ422244) from DRC, where gorillas do not occur and which is closer to the current eastern gorilla range, clusters in haplogroup C2 with sequences from Nigeria, Cameroon and Northern Gabon. The unexpected placement of these two sequences does however, raise questions about the reliability of using museum specimens when examining phylogeography in current populations (Hofreiter et al. 2003).
Discussion Gorilla phylogeography In Africa, the lower temperatures and greater aridity experienced during Pleistocene glacial maxima are thought to have led to the fragmentation of rainforest taxa into forest refugia (Maley 1996). Contraction of species ranges into these refugia may have fostered divergence between fragmented populations whereas rapid colonization during periods of climate amelioration may have led to large geographical areas being occupied by relatively few haplotypes. Genetic exchange between adjacent phylogroups © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1561 could result from zones of secondary contact between expanding refugial populations. Genetic signatures of such past historical events may be apparent in contemporary population genetic structure and provide supporting evidence for the influence of such refugia on past population genetic processes (Hewitt 1996; Jensen-Seaman & Kidd 2001). The most important finding of this study is the exceptional geographical substructuring that exists within western lowland gorillas. The large genetic distances between two major haplogroups within the western gorillas is greater than the molecular divergence observed between eastern gorilla subspecies and implies a relatively long evolutionary separation. The results are broadly consistent with previous morphological analysis (Groves 1967), where western gorillas were portioned into four ‘demes’ designated, Nigeria, Plateau, Coastal and Sangha. Two of these, Nigeria and Plateau, coincide with the geographical distribution of haplogroup C in Nigeria and Cameroon, whereas the distribution of subgroups D2 (CAR), D1 (Equatorial Guinea) and D3 (Gabon and Congo) are largely equivalent to the location of the Sangha (D2) and Coastal demes (D1, D3), respectively. The geographical distribution of haplogroups C and D and their respective subgroups may therefore coincide, in part, with the locations of several major forest refugia in Cameroon, Equatorial Guinea, Gabon and Congo (Sosef 1994; Rietkirk et al. 1995; Maley 1996). All these areas could have potentially harboured forest remnants during dry phases of the Pleistocene and consequently led to the allopatric separation and divergence of western lowland gorillas. However, owing to the paucity of information on the precise location and nature of these refugia (Livingstone 1982; White 2001), and the lack of other phylogeographical data covering this region, further sampling and additional studies of other rain-forest-associated taxa will help elucidate the relationship between Pleistocene forest/fluvial refuges and gorilla genetic structure. The haplotype diversity and bimodal mismatch distributions in haplogroup C, reflect a complex population history and/or population genetic structure. Gorillas in this haplogroup, despite being substantially less abundant than their counterparts in haplogroup D are responsible for over half of the nucleotide diversity found within western lowland gorillas. Previous morphological analysis had placed the Cross River gorillas, from Nigeria (which belong to subgroup C1), as the most distinct of the four proposed ‘demes’ (Groves 1967, 1970). However, haplotypes within this subgroup have a much wider geographical distribution and are similar to other haplotypes in Cameroon and northeast Gabon. Further sampling in haplogroup C is therefore required to elucidate the complex patterns of genetic variation within this haplogroup more clearly. In contrast, despite considerably greater sampling within haplogroup D, the unimodal mismatch distributions and evidence of recent population expansion of two © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
out of the three subgroups (D2 and D3) are consistent with a refugial expansion scenario (Hewitt 1996; Jensen-Seaman & Kidd 2001). This is particularly true of subgroup D3, where identical haplotypes are found across the entire southern range of western gorillas in Gabon and Congo. The present location of subgroup D2, restricted in southern CAR and adjacent Congo, may be the result of a fluvial refuge postulated to have persisted during the arid periods of the Pleistocene (Colyn 1991). The identification of this novel phylogeographical division in gorillas is also supported by previous morphological analyses (Groves 1967, 1970) that designate gorillas in this region as belonging to the ‘Sangha’ deme. The importance of fluvial refugia in Africa, however, has yet to be evaluated. Overall, low levels of mtDNA diversity exist within subgroups in haplogroup D, the haplotype difference between subgroups being largely responsible for the relatively higher levels of diversity within this haplogroup as a whole. Geographic boundaries of the distinct haplogroups do not appear to coincide with any major rivers, suggesting that present river courses do not constitute a major barrier to gene flow in gorillas. The Sanaga river has previously been reported as a taxonomic boundary for several groups, including mandrills, drills, forest duikers and chimpanzee subspecies (Gonder et al. 1997; Grubb 2001). This appears not to be the case for gorillas, as populations on both sides of the river belong to the same haplogroup. The apparent phylogeographical break between adjacent haplogroups C and D across the Sangha river may be simply coincident with the expansion of two well-defined haplogroups. The presence of a haplogroup D sequence in Lobéké (LBK3) suggests limited haplotype exchange between adjacent groups C and D one side of the Sangha River. Similarly the association of one of the Northern Gabon sequences from Ipassa (IPS2) with sequences from Cameroon and Nigeria suggest haplotype intermixing. Additional sampling in potential zones of secondary contact should help elucidate the geographical boundaries of these haplogroups more clearly as well as quantify the degree of haplotype exchange more accurately. The existence of two distinct groups of presumed Numts in the present study suggest that interpretation of mtDNA phylogenies should proceed with caution. Incorporation of Numts into mitochondrial analyses can potentially confound interpretation of derived phylogenies and lead to erroneous conclusions (Sorenson & Fleischer 1996; Mourier et al. 2001). The presence of near identical eastern and western lowland gorilla sequences within the putative Numt clusters is consistent with the slower rates of molecular evolution in the nuclear genome (Brown et al. 1982) and if homologous, suggest that these insertions may have occurred prior to the divergence of eastern and western gorillas. Alternatively, the presence of identical alleles in eastern and western populations could be the
1562 S . L . C L I F F O R D E T A L . result of recent gene flow or incomplete lineage sorting between regional groups. We note that the identification of several distinct Numt groups in this and other studies (Garner & Ryder 1996; Jensen-Seaman & Kidd 2001) suggests that multiple independent transpositions and/ or gene duplication events have occurred at different times during gorilla history.
Conservation implications/conclusions Phylogeographical analyses of intraspecific genetic variation can provide valuable information on how genetic variation is partitioned within species and aid in the identification of evolutionarily significant units (ESUs) for conservation (Moritz 1994; 1995). The geographical distribution and almost complete reciprocal monophyly between distinct haplogroups within western lowland gorillas strongly supports the recognition of at least four distinct evolutionary lineages comprising mountain, eastern lowland and western lowland haplogroups C and D, although there is evidence of limited gene flow between neighbouring populations in Northeastern Gabon, Cameroon and the Dzanga-Sangha region of CAR. This observation is not inconsistent with genetic patterns predicted under Pleistocene refuge theory and does not preclude recognition of haplogroups C and D as distinct evolutionary entities. According to the model advanced by Moritz (1994), ESUs are designated on the basis of reciprocal monophyly at mitochondrial markers and significant divergence in nuclear loci. Our result in this respect is then preliminary as we present only data derived from maternal patterns of genetic structure. Among recently diverged taxa, conflicts between mitochondrial and nuclear loci (Shaw 2002) as well as between loci in the nuclear genome (Machado & Hey 2003) are to be expected. Therefore caution needs to be exercised when defining evolutionary lineages derived only from mitochondrial control region sequences. In addition to conflicts between data sets, the criteria used to establish significant evolutionary units remains contentious (e.g. Crandall et al. 2000; Fraser & Bernatchez 2001; Moritz 2002; and references therein). Specifically, criticisms have been made about (i) the over-reliance of molecular data in identifying units for conservation, (ii) the fact that ESUs do not necessarily represent adaptive units of significance and (iii) that reciprocal monophyly is an overly restrictive assumption given what is known of the gene dynamics of speciation. Crandall et al. (2000) have called for a distinction between historical and recent phenomena and proposed that both ecological and genetic tests of nonexchangeability should be considered when defining ESUs. In western gorilla populations, whilst there is evidence of correspondence between patterns of genetic diversity in the present study and morphological differences observed between distinct population groups (cf. Groves 1967, 1970),
there is also evidence of disparity (Sarmiento et al. 1996; Sarmiento & Oates 2000). For example, Nigerian gorillas are morphologically distinct but belong to a large mitochondrial haplogroup that encompasses populations as far as Southeastern Cameroon. One potential explanation for this discrepancy is that morphological and ecological traits under selection may evolve much more rapidly than the neutral molecular markers used here so that disparities may arise in recently diverged taxa. Moreover, regional differences in potentially adaptive traits are difficult to assess because of insufficient knowledge of the ecology and behaviour of free-ranging gorilla populations. Nevertheless, the mitochondrial data presented in this study significantly advances our understanding of the molecular phylogeography of this endangered species of great ape and point to regionally distinct mitochondrial lineages that reflect an appreciable history of isolation from one another. Within the context of the biogeography of this group, our results also indicate that changes in the distribution of forest vegetation during the Pleistocene could have fostered divergence between regional populations and that conservation efforts should take these regional differences in genetic diversity into account. Future studies should seek to unequivocally identify nuclear insertions, assess genetic structure using nuclear loci and collect further ecological and behavioural data in order to better understand the conservation status and evolutionary significance of the major lineages identified in this study.
Acknowledgements This work was funded by the Leverhulme Trust London and the Darwin Initiative. We would like to thank all the collectors listed in Table 1 for their sampling efforts, M. Jensen-Seaman and B. Bradley for communicating data on Numts, A. J. Tosi and P. Walsh for useful discussions and M. Kazmierczak for logistical help.
References Arbogast BS (1999) Mitochondrial DNA phylogeography of the New World flying squirrels (Glaucomys): implications for Pleistocene biogeography. Journal of Mammology, 80, 142–155. Avise JC, Walker D (1998) Pleistocene phylogeographic effects on avian populations and the speciation process. Proceedings of the Royal Society of London, Series B, 265, 457–463. Avise JC, Walker D, Johns GC (1998) Speciation durations and Pleistocene effects on vertebrate phylogeography. Proceedings of the Royal Society of London, Series B, 265, 1707–1712. Bendall KE, Sykes BC (1995) Length heteroplasmy in the first hypervariable segment of the human mtDNA control region. American Journal of Human Genetics, 57, 248–256. Brown WM, Prager EM, Wang A, Wilson AC (1982) Mitochondrial DNA sequences of primates: tempo and mode of evolution. Journal of Molecular Evolution, 18, 225–239. © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1563 Byun SA, Koop BF, Reimchen TE (1997) North American black bear mtDNA phylogeography: implications for morphology and the Haida Gwaii glacial refugium controversy. Evolution, 51, 1647–1653. Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. American Journal of Human Genetics, 19, 233–257. Clifford SL, Jeffery K, Bruford MW, Wickings EJ (1999) Identification of polymorphic microsatellite loci in the gorilla (Gorilla gorilla gorilla) using human primers: application to non-invasively collected hair samples. Molecular Ecology, 8, 1556 –1558. Clifford SL, Abernethy K, White L, Tutin CEG, Bruford MW, Wickings EJ (2003) Genetic studies of western gorillas. In: Gorilla Biology: a Multidisciplinary Perspective (eds Taylor A, Goldsmith M), pp. 269–292. Cambridge University Press, Cambridge. Collura RV, Stewart C-B (1995) Insertions and duplications of mtDNA in the nuclear genomes of Old World monkeys and hominoids. Nature, 378, 485 –489. Colyn M (1991) L’importance zoogéographique du Bassin du fleuve Zaïre pour la spéciation. Annales Sciences Zoologiques, 264, 180–185. Crandall KA, Bininda-Emonds ORP, Mace GM, Wayne RK (2000) Considering evolutionary processes in conservation biology. Trends in Ecology and Evolution, 15, 290 –295. Doran D, McNeilage A (1998) Gorilla ecology and behaviour. Evolutionary Anthropology, 6, 120–131. Dynesius M, Jansson R (2000) Evolutionary consequences of changes in species’ geographical distributions driven by Milankovitch climate oscillations. Proceedings of the National Academy of Sciences of the USA, 97, 9115 –9120. Eggert SL, Rasner CA, Woodruff DS (2002) The evolution and phylogeography of the African elephant inferred from mitochondrial DNA sequence and nuclear microsatellite markers. Proceedings of the Royal Society of London, Series B, 269, 1993–2006. Excoffier L, Smouse P, Quattro J (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics, 131, 479–491. Field D, Chemnick L, Robbins M, Garner K, Ryder OA (1998) Paternity determination in captive lowland gorillas and orang utans and wild mountain gorillas by microsatellite analysis. Primates, 3, 199–209. Flagstad O, Syvertsen PO, Stenseth NC, Jakobsen KS (2001) Environmental change and rates of evolution: the phylogeographic pattern within the hartebeest complex as related to climatic variation. Proceedings of the Royal Society of London, Series B, 268, 667–677. Fraser DJ, Bernatchez L (2001) Adaptive evolutionary conservation: towards a unified concept for defining conservation units. Molecular Ecology, 10, 2741–2752. Gagneux P, Wills C, Gerloff U et al. (1999) Mitochondrial sequences show diverse evolutionary histories of African hominoids. Proceedings of the National Academy of Sciences of the USA, 96, 5077–5082. Gagneux P, Gonder MK, Goldberg TL, Morin PA (2001) Gene flow in wild chimpanzee populations: what genetic data tell us about chimpanzee movement over space and time. Philosophical Transactions of the Royal Society of London, Series B, 356, 889–897. Garner KJ, Ryder OA (1996) Mitochondrial DNA diversity in gorillas. Molecular Phylogenetics and Evolution, 6, 39– 48. Goldberg TL, Ruvolo M (1997a) Molecular phylogenetics and historical biogeography of east African chimpanzees. Biological Journal of the Linnean Society, 61, 301–324. © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
Goldberg TL, Ruvolo M (1997b) The geographic apportionment of mitochondrial genetic diversity in east African Chimpanzees, Pan troglodytes schweinfurthii. Molecular Biology and Evolution, 14, 976–984. Gonder MK, Oates JF, Disotell TR, Forster MRJ, Morales JC, Melnick DJ (1997) A new west African chimpanzee subspecies? Nature, 388, 337. Groves CP (1967) Ecology and taxonomy of the gorilla. Nature, 213, 890–893. Groves CP (1970) Population systematics of the gorilla. Journal of Zoology, 161, 287–300. Groves CP (2001) Primate Taxonomy, p. 350. Smithsonian Institute, Washington DC. Grubb P (2001) Endemism in African rain forest mammals. In: African Rainforest Ecology and Conservation (eds Weber W, White LJT, Vedder A, Naughton-Treves L), pp. 88–100. Yale University Press, New Haven and London. Grzybowski T (2000) Extremely high levels of human mitochondrial DNA heteroplasmy in single hair roots. Electrophoresis, 21, 548–553. Haffer J (1982) General aspects of the refuge theory. In: Biological Diversification in the Tropics (ed. Prance GT), pp. 641– 657. Columbia University Press, New York. Hall JS, Saltonstall K, Inogwabini B-I, Omari I (1998) Distribution, abundance and conservation status of Grauer’s gorilla. Oryx, 32, 122–130. Harcourt A (1996) Is the gorilla a threatened species? How should we judge? Biological Conservation, 75, 165–176. Hewitt GM (1996) Some genetic consequences of ices ages, and their role in divergence and speciation. Biological Journal of the Linnean Society, 58, 27–276. Hofreiter M, Siedel H, Van Neer W, Vigilant L (2003) Mitochondrial DNA sequence from an enigmatic gorilla population (Gorilla gorilla uellensis). American Journal of Physical Anthropology, 121, 361–368. Holder K, Montgomerie R, Friesen VL (1999) Glacial vicariance and historical biogeography of rock ptarmigan (Lagopus mutus) in the Bering region. Evolution, 53, 1936–1950. Horai S, Hayasaka K, Kondo R, Tsugane K, Takahata N (1995) Recent African origin of modern humans revealed by complete sequences of hominoid mitochondrial DNAs. Proceedings of the National Academy of Sciences of the USA, 92, 532–536. Jensen-Seaman MI, Kidd KK (2001) Mictochondrial DNA variation and biogeography of eastern gorillas. Molecular Ecology, 10, 2241–2247. Kingdon J (1990) Island Africa. Collins, London. van der Kuyl AC, Kuiken CL, Dekker JT, Perizonius WRK, Goudsmit J (1995) Nuclear counterparts of the cytoplasmic mitochondrial 12S rRNA gene: a problem of ancient DNA and molecular phylogenies. Journal of Molecular Evolution, 40, 652– 657. Lanfranchi R, Schwartz D (1990) Paysages Quaternaires de l’Afrique Centrale Atlantique. ORSTOM, Paris. Livingstone DA (1982) Quaternary geography of Africa and the refuge theory. In: Biological Diversification in the Tropics (ed. Prance GT), pp. 523–535. Columbia University Press, New York. Machado CA, Hey J (2003) The causes of phylogenetic conflict in a classic Drosophila species group. Proceedings of the Royal Society of London Series B-Biological Sciences, 270, 1193–1202. Maley J (1996) The African rain-forest — main characteristics of changes in vegetation and climate change from the Upper Cretaceous to the Quaternary. Proceedings of the Royal Society of Edinburgh, Series B, 104, 31–73.
1564 S . L . C L I F F O R D E T A L . Morin PA, Wallis J, Moore JJ, Chakraborthy R, Woodruff DS (1993) Non-invasive sampling and DNA amplification for paternity exclusion, community structure and phylogeography in wild chimpanzees. Primates, 34, 347–356. Morin PA, Moore JJ, Chakraborthy R, Jin L, Goodall J, Woodruff DS (1994) Kin selection, social structure, gene flow, and the evolution of chimpanzees. Science, 265, 1193 –1201. Moritz C (1994) Defining evolutionary significant units for conservation. Trends in Ecology and Evolution, 9, 373 –376. Moritz C (1995) Uses of molecular phylogenies for conservation. Philosophical Transactions of the Royal Society of London, Series B, 349, 113–118. Moritz C (2002) Strategies to protect biological diversity and the evolutionary processes that sustain it. Systematic Biology, 51, 238–254. Mourier T, Hansen AJ, Willerslev E, Arctander P (2001) The human genome project reveals a continuous transfer of large mitochondrial fragments to the nucleus. Molecular Biology and Evolution, 18, 1833 –1837. Muloko-Ntoutoume N, Petit RM, White L, Abernethy K (2001) Chloroplast DNA variation in a rainforest tree (Aucoumea klaineana, Burseraceae) in Gabon. Molecular Ecology, 9, 359 – 363. Mundy NI, Pissinatti A, Woodruff DS (2000) Multiple nuclear insertions of mitochondrial cytochrome b sequences in callitrichine primates. Molecular Biology and Evolution, 17, 1075 –1080. Nei M (1987) Molecular Evolutionary Genetics. Columbia University, New York. Nei M, Li W-H (1979) Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences of the USA, 76, 5269–5273. Noda R, Kim CG, Takenaka O et al. (2001) Mitochondrial 16S rRNA sequence diversity of hominoids. Journal of Heredity, 92, 490–496. Oates JF, McFarland KL, Stumpf RM et al. (1999) New findings on the distinctive gorillas of the Nigeria–Cameroon border region. American Journal of Physical Anthropology, 28, 213 –214. Oates JF, McFarland KL, Groves JL, Bergl RA, Linder JL, Disotell TR (2002) The Cross River gorilla: natural history and status of a neglected and critically endangered subspecies. In: Gorilla Biology: a Multidisciplinary Perspective (eds Taylor A, Goldsmith M), pp. 472–502. Cambridge University Press, Cambridge. Omari I, Hart JA, Butynski TM (1999) The Itombwe Massif, Democratic Republic of Congo: biological surveys and conservation, with an emphasis on Grauer’s gorilla and birds endemic to the Albertine Rift. Oryx, 33, 301–322. Petit R, Pineau E, Demesure B, Bacilieri R, Dacousso A, Kremer A (1997) Chloroplast DNA footprints of postglacial recolonization by oaks. Proceedings of the National Academy of Sciences of the USA, 94, 9996–10001. Posada D, Crandall KH (1998) Modeltest: testing the model of DNA substitution. Bioinformatics, 14, 817– 818. Rietkirk M, Ketner P, De Wilde JJFE (1995) Caesalpinoideae and the study of forest refuges in central Africa. In: The Biodiversity of African Plants (eds van der Maesen LJG, van der Burgt XM, van Medenbach de Rooy JM), pp. 608–623. Kluwer. Academic Publishers, the Netherlands. Ruvolo M, Pan D, Zehr S, Goldberg T, Disotell TR, von Dornum M (1994) Gene trees and hominoid phylogeny. Proceedings of the National Academy of Sciences of the USA, 91, 8900 – 8904. Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution, 14, 399 – 411.
Saltonstall KG, Amato G, Powell J (1998) Mitochondrial DNA variability in Grauer’s gorillas of Kahuzi-Biega National Park. Journal of Heredity, 69, 129–135. Sarmiento EE, Oates JF (2000) The Cross River gorillas: a distinct subspecies, Gorilla gorilla diehli Matschie 1904. American Museum Novitates, XX, 3304. Sarmiento EE, Butynski TM, Kalina J (1996) Gorillas of BwindiImpenetrable forest and the Virunga volcanoes: taxonomic implications of morphological and ecological differences. American Journal of Primatology, 40, 1–21. Schneider S, Excoffier L (1999) Estimation of past demographic parameters from the distribution of pairwise differences when the mutation rates vary among sites: application to human mitochondrial DNA. Genetics, 152, 1079–1089. Schneider S, Roessli D, Excoffier L (2000) Arlequin, a Software for Population Genetics Data Analysis, Version 2.000. Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, Geneva Switzerland. Shaw KL (2002) Conflict between nuclear and mitochondrial DNA phylogenies of a recent species radiation: what mtDNA reveals and conceals about modes of speciation in Hawaiian crickets. Proceedings of the National Academy of Sciences of the USA, 99, 16122–16127. Smith TB, Holder K, Girman D et al. (2000) Comparative avian phylogeography of Cameroon and Equatorial Guinea mountains: implications for conservation. Molecular Ecology, 9, 1505– 1516. Sorenson MD, Fleischer RC (1996) Multiple independent transpositions of mitochondrial DNA control region sequences to the nucleus. Proceedings of the National Academy of Sciences of the USA, 93, 15239–15243. Sosef MSM (1994) Studies in Begoniaceae. Agricultural University Papers, Wageningen, the Netherlands. Swofford DL (1998) PAUP*, Phylogenetic Analysis Using Parsimony V4. Sinauer Associates, Sunderland MA. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG (1997) The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acid Research, 25, 4876–4882. Tutin CEG, Vedder A (2001) Gorilla conservation and research in central Africa. In: African Rainforest Ecology and Conservation (eds Weber W, White LJT, Vedder A, Naughton-Treves L), pp. 429 – 448. Yale University Press, New Haven and London. Tutin CEG, Parnell RJ, White LJT, Fernandez M (1995) Nest building by lowland gorillas in the Lopé Reserve, Gabon: environmental influences and implications for censusing. International Journal of Primatology, 16, 53–76. Tutin CEG, White LJT, Mackanga-Missandzou A (1997) The use by rain forest mammals of natural forest fragments in an equatorial African savanna. Conservation Biology, 11, 1190 – 1203. Vigilant L (1999) An evaluation of techniques for the extraction and amplification of DNA from naturally shed hairs. Biological Chemistry, 380, 1329–1331. Vigilant L, Stoneking M, Harpending H, Hawkes K, Wilson AC (1991) African populations and the evolution of human mitochondrial DNA. Science, 253, 1503–1507. Walsh PS, Metzger DA, Higuchi R (1991) Chelex 100 as a medium for simple extraction of DNA for PCR-based typing from forensic material. Biotechniques, 10, 506–513. White LJT (2001) The African rain forest. In: African Rainforest Ecology and Conservation (eds Weber W, White LJT, Vedder A, © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
P H Y L O G E O G R A P H Y O F W E S T E R N L O W L A N D G O R I L L A S 1565 Naughton-Treves L), pp. 1–29. Yale University Press, New Haven and London. Xu X, Arnason U (1996) A complete sequence of the mitochondrial genome of the western lowland gorilla. Molecular Biology and Evolution, 13, 691–698. Zischler H, Geisert H, Castresana J (1998) A hominoid-specific nuclear insertion of the mitochondrial d-loop: implications for reconstructing ancestral mitochondrial sequences. Molecular Biology and Evolution, 15, 463 – 469. Zischler H, Geisert H, von Haeseler A, Pääbo S (1995) A nuclear fossil of the mitochondrial d loop and the origin of modern humans. Nature, 378, 489.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 1551–1565
Stephen Clifford is a postdoctoral researcher based in Gabon interested principally in primate population genetics and conservation issues pertaining to the region. Nicola Anthony is an assistant professor at University of New Orleans who is interested in the molecular ecology and phylogeography of tropical and temperate taxa. Jean Wickings is head of the Molecular Ecology Unit (UGENET) at CIRMF, Gabon, specializing in aspects of African biodiversity and biogeography, examining species ranging from primates and elephants to tropical flora.
