Conservation Genetics

, Volume 10, Issue 4, pp 909–914

mtDNA indicates profound population structure in Indian tiger (Panthera tigris tigris)

Authors

  • Reeta Sharma
    • Unit of Evolutionary Biology/Systematic Zoology, Institute of Biochemistry and BiologyUniversity of Potsdam
    • Wildlife Institute of India
  • Heiko Stuckas
    • Unit of Evolutionary Biology/Systematic Zoology, Institute of Biochemistry and BiologyUniversity of Potsdam
  • Ranjana Bhaskar
    • Wildlife Institute of India
  • Sandeep Rajput
    • Wildlife Institute of India
  • Imran Khan
    • Wildlife Institute of India
  • Surendra Prakash Goyal
    • Wildlife Institute of India
    • Unit of Evolutionary Biology/Systematic Zoology, Institute of Biochemistry and BiologyUniversity of Potsdam
Short Communication

DOI: 10.1007/s10592-008-9568-3

Cite this article as:
Sharma, R., Stuckas, H., Bhaskar, R. et al. Conserv Genet (2009) 10: 909. doi:10.1007/s10592-008-9568-3

Abstract

We analyzed mtDNA polymorphisms (parts of control region, ND5, ND2, Cytb, 12S, together 902 bp) in 59 scat and 18 tissue samples from 13 Indian populations of the critically endangered Indian tiger (Panthera tigris tigris), along with zoo animals as reference. Northern tiger populations exhibit two unique haplotypes suggesting genetic isolation. Western populations from Sariska (extinct in 2004) and Ranthambore are genetically similar, such that Ranthambore could serve as a source for reintroduction in Sariska. Zoo populations maintain mitochondrial lineages that are rare or absent in the wild.

Keywords

ConservationIndian tigermtDNAPopulation geneticsScat

Introduction

The Indian tiger (Panthera tigris tigris) inhabits India, Nepal, Bhutan, Bangladesh, and Myanmar. The species is critically endangered, as populations declined from an estimated 20,000–40,000 in the early 20th century to around 2,500 individuals (Sankhala 2005), mainly due to habitat destruction and poaching. One dramatic recent example of poaching is the complete extinction of the Sariska population in 2004. In 1973, the Indian government launched the “Project Tiger” to ensure the maintenance of viable tiger populations in India. One area of particular importance is Rajaji/Corbett which comprises the most vital conservation area in Northern India for large mammals.

mtDNA markers are widely used in conservation genetics. Here, we assess mtDNA variability in wild Indian Tiger populations, with particular focus on Northern (Corbett, Rajaji) and Western populations (Sariska, Ranthambore). Including samples from the now extinct Sariska population, we provide a genetic basis for reintroduction measures. Specimens from Indian zoos serve as a reference. As blood or tissue samples from wild tiger specimens are difficult to obtain, we mainly use scat for DNA analysis.

Material and methods

We obtained tiger samples from 13 wild populations throughout India (Fig. 1). Scat samples were collected in the years 2002–2005 and preserved according to Wasser et al. (1997). We also opportunistically collected muscle, blood, and skin. Based on estimations of home ranges (Seidensticker et al. 1999), scat samples found more than 10 km apart from each other were considered originating from different individuals. Populations were assigned to the regions Northern, Northeastern, Western, Central, and Southern India (Fig. 1). Data from Chitwan (Indo-Nepal border) and Nagarhole (Southern India) (Luo et al. 2004) were included. Additional samples were obtained from captive tigers (Zoo Chandigarh, n = 9, source unknown; white tigers from Zoo Delhi, n = 2, originating from Rewa, Central India).
https://static-content.springer.com/image/art%3A10.1007%2Fs10592-008-9568-3/MediaObjects/10592_2008_9568_Fig1_HTML.gif
Fig. 1

Map of India showing sample sites and indicating assignment of sample sites to geographic areas. Samples taken from Luo et al. (2004) are labelled with an asterisk

DNA from scat and tissue samples was isolated using QIAamp DNA Stool Kit and DNeasy Kit (QIAGEN), respectively. Scat extractions were carried out in a separate laboratory dedicated to ancient and degraded samples. A blank extraction control was always included to check for cross contamination. Two PCR primer pairs (Table 1) were developed for the conserved part of the control region (CCR) based on a published tiger sequence (AF053055). Additional eight primer pairs were developed to detect known mtDNA polymorphisms (Luo et al. 2004). PCR was performed according to Tiedemann et al. (2005), with the following cycling parameters: 94°C for 10 min; 39 cycles of 94°C for 30 s, primer specific annealing temperature for 30 s, 72°C for 30 s; final 10 min at 72°C. Amplicons were analysed on an AB3100 or AB3130 automatic sequencer (Applied Biosystems).
Table 1

Primer pairs developed in this study targeting particular regions of the Panthera tigris tigris mitochondrial genome

Fragment

Primer

Primer sequence (5′ > 3′)

Product length (bp)

Annealing temperature (°C)

CCR (a)

PtigCR1F

ATTCATGATTTAGAACAGTTCTTTC

243

56.0

PtigCR1R

TAGTCATTAACCCATCGAGATG

CCR (b)

PtigCR2F

CAACGTGGGGGTGTCTATAGTGA

292

60.8

PtigCR3R

CCTTGTTTGTTTGTACGTGTGGAA

ND2 (a)

PtND2aF

TTCCCATCCTCATGAAAAAAT

118

55.2

PtND2aR

TGCAGCAGGTTGATGATAATT

ND2 (b)

PtND2bF

TTCCCCTCAAACAACTGTATAAA

112

59.9

PtND2bR

TTGGTGTGAGTGGGAGTAGTATG

ND5 (a)

PtND5aF

GCCATCAGCCATAGAAGGTCC

103

64.0

PtND5aR

TATGAGTGGGTGGAAGCGGAT

ND5 (b)

PtND5bF

TCCGCCCTACTCCATTCAAGC

120

64.0

PtND5bR

GATGGCCCCCAGGCATAGAGT

ND5 (c)

PtND5cF

CCCCGATTCAACTCCCTAAGCCC

123

64.0

PtND5cR

CGTTGGGGGGATGTTATGGGAGA

CytB

PtCytBF

GAATATACTACGGCTCCTACACC

172

59.9

PtCytBR

GTCCCAATATATGGGATTGCT

12S

Pt12SIIF

GCCATCTTCAGCAAACCCTAAA

206

64.0

Pt12SI&IIR

CCCGATTCAATCGAGCTCTCTAT

A 332 bp fragment of the CCR was sequenced for each specimen. The sequences of coding mitochondrial genes (ND2 (a): 54 bp; ND2 (b): 59 bp; ND5 (a): 60 bp; ND5 (b): 68 bp; ND5 (c): 57 bp; Cytb: 127 bp; 12S: 145 bp) yielded a composite haplotype of 570 bp for each individual. Specimens not yielding the entire 570 bp were excluded from further analyses. To verify mitochondrial origin, sequences were compared to nuclear mitochondrial insertions (Numts) known for tiger (AF053056, AF053053, DQ151551). We verified tiger origin of scat samples by comparing ND2 sequences to those of leopard (AY634383), which can occur sympatrically. Verified tiger mtDNA haplotypes were compared to haplotypes TIG1–6 from Luo et al. (2004). As our sequences are shorter, we add an asterisk to our haplotype names (e.g., TIG* 1 is identical at all positions analysed here to TIG1 in Luo et al. 2004). A haplotype network was constructed with TCS (Clement et al. 2000). We used Arlequin version 3.1 (Excoffier et al. 2005) to calculate haplotype/nucleotide diversity, to perform an analysis of molecular variance (AMOVA), to compute pairwise FST values, and to perform exact tests of population differentiation. Mismatch distributions were checked for signs of recent population expansion (Schneider and Excoffier 1999).

Results

The 332 bp CCR fragment was identical in all 23 wild specimens typed, representing eight populations from Northern (Corbett, Rajaji), Northeastern (Twai), Western (Ranthambore, Sariska), and Central India (Panna, Kanha, Bandhavgarh) (Genbank no. EU527875). A single polymorphism was detected in Chandigarh zoo animals (EU527876). Fifty-nine out of 117 scat samples (50%) and all 18 tissue samples were successfully genotyped at all seven target regions (ND5, ND2, Cytb, and 12S), yielding 570 bp. A further 31 scat samples (26%) were partially genotyped (data not included in analysis). Six composite mitochondrial haplotypes were found (geographical occurrence in Table 2), four of them new (TIG* 7–TIG* 10; EU527859–EU527862 for 12S; EU527863–EU527866 for Cytb; EU527867–EU527870 for ND2; EU527871–EU527874 for ND5), while two haplotypes were identical to TIG5 and TIG6 from Luo et al. (2004). TIG* 9 was exclusively found in captive white tiger specimens from Rewa, Central India, while Zoo Chandigarh samples all showed TIG* 5. The parsimony network has a star-like topology (Fig. 2). The observed mismatch distribution (all populations pooled) did not significantly differ from expectation after a recent sudden expansion (P = 0.6).
Table 2

Analyzed wild tiger populations, relative haplotype frequencies, and genetic diversity measures (standard errors in parentheses)

Regions

Populations

Samples and sample size

Composite coding region mt haplotypes

Haplotype diversity

Nucleotide diversity

Scat samples

Tissue

N

Total

Analysed

TIG* 2

TIG* 3

TIG* 4

TIG* 5

TIG* 6

TIG* 7

TIG* 8

TIG* 10

Himalayan Indo-Nepal border

Chitwan NPa

   

12a

  

0.08

0.50

0.42

   

0.62 (0.09)

0.0012 (0.0010)

Northern India

Corbett TR

21

9

7

16

     

1

  

0.16 (0.09)

0.0003 (0.0004)

Rajaji NP

26

6

1

7

     

0.71

0.29

 

Nazibabad

  

1

1

     

1

  

Northeastern India

Twai WLS

1

1

 

1

       

1

n.d.

n.d.

Western India

Ranthambore TR

25

19

3

22

   

0.86

   

0.14

0.37 (0.09)

0.0019 (0.0015)

Sariska TR

4

4

 

4

   

0.50

   

0.50

Central India

Panna TR

7

4

 

4

   

0.75

0.25

   

0.13 (0.11)

0.00023 (0.0004)

Nasik

3

1

 

1

   

1

    

Kanha TR

7

3

3

6

   

1

    

Bandhavgarh TR

1

1

1

2

   

1

    

Umaria

  

2

2

   

1

    

Southern India

Kalakad-Mundanthurai TR

12

5

 

5

   

1

    

0.29 (0.16)

0.0008 (0.0008)

Nagarhole NPa

   

2a

0.50

0.50

      

Periyar TR

10

6

 

6

   

1

    

Total

117

59

18

77 (+14a)

0.01

0.01

0.01

0.56

0.07

0.24

0.02

0.08

0.62 (0.04)

0.0016 (0.0010)

aData from Luo et al. (2004)

NP = National Park; TR = Tiger reserve; WLS = Wildlife sanctuary; n.d. = Not determined

https://static-content.springer.com/image/art%3A10.1007%2Fs10592-008-9568-3/MediaObjects/10592_2008_9568_Fig2_HTML.gif
Fig. 2

Parsimony network showing the phylogenetic relationship between the mt haplotypes and their frequency (circle size) in wild tiger populations

Haplotype and nucleotide diversity was highest in Chitwan and Western India, intermediate in Southern India, and particularly low in Central and Northern India (Table 2). The AMOVA revealed a significant genetic structure among geographical regions (FST = 0.53; P < 0.001). Pairwise FST values among regions are significant between Northern India and all other regions (FST values between 0.64 and 0.88), as well as between Chitwan and Southern India (FST = 0.23), while all other FST were non-significant. Exact permutation tests for population differentiation were significant in all comparisons among regions, except between Central and Southern India.

Discussion

Our PCR amplification success rate on tiger scat samples was 50% (strict sense; specimens fully genotyped) or 76% (including partly genotyped specimens). This is lower than in a comparable study in more temperate regions (Russelo et al. 2004; 86.3%), possibly due to different climatic conditions which can influence DNA quality in scat samples (Nsubuga et al. 2004).

Analyzing altogether 902 bp of mtDNA (332 bp of Control Region and 570 bp of coding regions) in 77 wild Indian tigers representing 13 wild Indian populations from various regions of the Indian subcontinent and 11 specimens from two zoo populations yielded only six different composite haplotypes. Together with Luo et al. (2004), 10 Indian tiger haplotypes are now known, which differ from one another by only a few mutations. This low diversity is typical also for other tiger subspecies (Luo et al. 2004; Russelo et al. 2004). Yet, our composite mitochondrial haplotypes are informative regarding geographic population structure. The closely related haplotypes TIG* 1 and TIG* 9 (Fig. 2) were only found in captive zoo populations. Although our sampling from wild populations is far from exhaustive, our study indicates that these mitochondrial lineages might have disappeared from wild Indian tiger populations. The two closely related haplotypes TIG* 7 and TIG* 8 together had a frequency of 100% in Northern populations (i.e., all tigers in Rajaji/Corbett showed one of these two haplotypes), while it was absent in all other regions. Possibly, TIG* 7 and TIG* 8 arose locally from the ubiquitous TIG* 5 (Fig. 2; category V phylogeography sensu Avise 2000). This suggests that Rajaji/Corbett harbours a genetically distinct tiger population. Haplotype frequencies indicate further population structure: both the AMOVA and the pairwise analyses reveal significant differences among most tiger populations of different geographic areas across India. Genetic analysis of the recently exterminated population of Sariska (Western India) suggests potential connectivity to the extant Ranthambore population (co-occurrence of the otherwise rare TIG* 10 type), a scenario further corroborated by historical information on a forest connection among these two areas until 100 years ago. Thus, for reintroduction of tigers to Sariska, animals from Ranthambore should be considered, if regional genetic identity is to be maintained.

If we assume that (1) tigers have colonized the Indian subcontinent 250,000 to 100,000 years ago (Johnsingh and Goyal 2005) and (2) the mitochondrial lineages found originated from a common ancestor at the time of colonization, our overall nucleotide divergence (0.16%; Table 2) would nominally translate into a divergence rate between 0.64 and 1.6%/Myr for the composite sequence analyzed here. We do not argue that this algebraic exercise provides any precise estimate of a divergence rate in tiger mt genomes. Our rough estimates nevertheless fall into the range of divergence estimates for mitochondrial genes in Panthera (0.45–0.68%/Myr for 12S; 1.9–2.8%/Myr for ND2; Lopez et al. 1997), such that we conclude that the level of mtDNA diversity found in Indian tiger appears congruent with the aforementioned colonization scenario.

The simple star-like topology of the parsimony network (Fig. 2), the close phylogenetic relationship of haplotypes, the mismatch distribution, and the fact that the most common haplotype TIG* 5 is central in the network and distributed all over India (except in the Northern populations) is consistent with a sudden expansion of an ancestral population, presumably after colonization of the Indian subcontinent. However, such ancient events might be substantially overarched by the much more recent population decline, leading to a loss of genetic diversity. Such a loss was previously reported in wild Amur tiger (P. tigris altaica) populations, where mtDNA variation was only found in captive specimens, but not in the wild (Russelo et al. 2004). In our study, haplotype TIG* 9 was only found in white tiger specimens from Zoo Delhi representing the extinct population from Rewa (Central India). Possibly, animals carrying this haplotype were taken to captivity before this mitochondrial lineage became rare or even extinct in the wild. Similarly, the unique CCR haplotype in Zoo Chandigarh tigers could represent a mitochondrial lineage that is rare or extinct in the wild.

Acknowledgements

Financial support is acknowledged from University of Potsdam and DAAD. We are thankful to the Director and Dean of the Wildlife Institute of India, Dehradun, for providing necessary facilities and acknowledge the cooperation of Indian forest officials. Valerio Ketmaier and Christoph Bleidorn adviced in data analysis. Katja Moll, Sudhansu Mishra, and Udayan Borthakur provided technical assistance.

Copyright information

© Springer Science+Business Media B.V. 2008