Sorting out tigers (Panthera tigris): mitochondrial sequences, nuclear inserts, systematics, and conservation genetics

Sequences from complete mitochondrial cytochrome b genes of 34 tigers support the hypothesis that Sumatran tigers are diagnostically distinct from mainland populations. None of the latter, including Bengals, Siberians, or Indochinese tigers, were found to have fixed diagnostic characters. Phylogenetic analysis of these sequences confirms these results. Within the framework of a phylogenetic species concept, current evidence thus supports the recognition of two distinct taxa, and within the context of this definition they could be ranked at the species level. This paper also documents a previously unrecognized nuclear insert of mitochondrial DNA that includes, minimally, mitochondrial homologues of a control region that lacks the feline mitochondrial repeat sequences, a complete cytochrome b gene, and complete tRNAThr and tRNAPro genes. In a phylogenetic analysis of the nuclear cytochrome b-like sequences and various feline mitochondrial sequences, the nuclear insert clusters with lion mitochondrial cytochrome b sequences, which suggests the insert is at least as old as the split between lions and tigers. The results of this study emphasize the importance of doing more to conserve Sumatran tigers. Because they are underrepresented in zoos relative to Bengals and Siberians, an effort should be made to increase captive breeding stocks of Sumatrans. That Sumatrans are a distinct taxonomic entity relative to mainland populations can be used in educational programs to increase conservation efforts within Indonesia. are diagnosably distinct. Regulating the illegal trade of tigers and identifying the geographic sources of that trade also depend on knowledge of diagnosable taxa and the genetic or morphological markers that identify them. Finally, countries have a vested interest in identifying the diagnosable taxa present within their borders for a number of reasons: discerning patterns of diversity and endemism, formulating conservation management plans, as well as establishing a basis for environmental education programs (Cracraft, 1997). We undertook this study to investigate whether any of the currently recognized subspecies of tigers are diagnosably distinct when using DNA sequencing techniques with sufficient power to resolve phylogenetic patterns in closely related organisms. An affirmative answer to that question then leads one to investigate the issue of the interrelationships of these taxa, their historical biogeography, and the tempo of branching events that may underlie current patterns of taxonomic and genetic distinctness. During the course of this study we discovered an heretofore unrecognized fragment of tiger mitochondrial-like sequences, containing a mitochondrial-like control region (D-loop) lacking the feline repeat sequences, a complete cytochrome b gene, and complete tRNAThr and tRNAPro genes. For reasons to be discussed, we infer these to have been inserted into the nuclear genome. Because the accuracy of any statement about patterns of genetic variation, and their relevance for conservation genetics, depends on distinguishing between mitochondrial and nuclear sequences, we characterize these mitochondrial-like sequences and then compare them to their mitochondrial counterparts. MATERIALS AND METHODS Samples and provenance Blood samples of a lion (Panthera leo) and 34 captive tigers (except for P. t. amoyensis which was not available to us), all of presumed known parentage, were provided by the following sources [information in parentheses after the laboratory code for individual specimens includes: sex, zoo or source, and an identifying number such as local zoo number (ID) and studbook number (SB), when available]: Panthera tigris tigris: B2 (F, Topeka Zoo ID 1182 via Brookfield Zoo ID 8-G-3,4; SB 8411), B3 (F, Atlanta Zoo ID 750275 via Brookfield Zoo ID13-I-4), B4 (M, Atlanta Zoo ID 750175 via Brookfield Zoo ID 13-J-4), B5 (F, Cincinnati Zoo, ID 189158, SB 360), B6 (M, wild caught Nagarhole Park, India, from S. J. O’Brien, ID Pti-102), B7 (F, wild caught Nagarhole Park, India, from S. J. O’Brien, ID Pti-103), B8 (M, wild caught Nagarhole Park, India, from S. J. O’Brien, ID Pti-104), and B9 (M, wild caught Nagarhole Park, India, from S. J. O’Brien, ID Pti-105). Panthera tigris altaica: S1 (F, Omaha Zoo, ID 6067, SB 2895), S2 (F, Philadelphia Zoo ID 101571 via Brookfield Zoo ID 26-F-3,4; SB 3110), S3 (F, Minnesota Zoo ID MN5467 via Brookfield Zoo ID 26-G-5; SB 3124), S4 (M, Minnesota Zoo ID MN5466 via Brookfield Zoo, ID 26-H-5,6; SB 3123), S5 (M, Wildlife Conservation Society ID 861078; SB 3009), S6 (M, Wildlife Conservation Society ID 861079; SB 3010), S7 (F, Wildlife Conservation Society ID 721344; SB 845), S8 (F, Omaha Zoo ID 5684; SB 2645), S10 (M, Omaha Zoo ID 4530; SB 2430), S11 (F, Omaha Zoo ID 3976; SB 2393), S12 (F, San Diego Zoo ID 173268 via S. J. O’Brien ID Pti-66; SB 762), S13 (F, S. J. O’Brien, ID Pti-111), S14 (F, Dallas Zoo ID 812821 via S. J. O’Brien, ID Pti-85; SB 1852), and S15 (M, S. J. O’Brien ID Pti-106). Panthera tigris sumatrae: Su1 (M, Omaha Zoo ID 5448; SB 527), Su2 (F, Omaha Zoo ID 5228; SB 380), Su3 (F, San Diego Zoo ID 589111; SB 753), Su4 (M, San Diego Zoo ID 103497; SB 312), Su5 (M, San Diego Zoo ID BK-8), Su6 (M, London Zoo SB 592), Su7 (M, San Diego Zoo ID 587362 via S. J. O’Brien ID Pti-95; SB 718), Su9 (M, National Zoo ID NZP110519 via S. J. O’Brien ID Pti-109), and Su10 (M, San Diego Zoo ID 687610 via S. J. O’Brien ID Pti-82). Panthera tigris corbetti: C1 (F, Cincinnati Zoo ID 190014; SB 6), C2 (F, Cincinnati Zoo ID 190015; SB 18), and C3 (F, San Diego Zoo ID 592048; SB 28). Panthera leo: Brookfield Zoo ID 34-L-8. Historically, breeding records for zoo animals have varied in quality, but modern-day studbooks of tigers are considered to have high reliability as far as breeding histories are concerned (R. Tilson, pers. comm.). All samples used in this study are assumed to be ‘pure bred’ representatives of their subspecies and no individual was used if there was the slightest evidence it might be of hybrid origin (for example, large numbers of ‘Bengal’ tigers in zoos are descendants of hybrids). Because mitochondrial DNA was being assayed, care was taken not to use samples that were known to share the same maternal lineage. For those individuals with studbook numbers it was possible to trace lineages back four or five generations, and thus the independence of many of the individuals could be reasonably assessed. For those specimens represented by studbook numbers, the majority could be traced back to the wild, but at most only to country of origin several generations ago. We discovered no example in which an individual’s ancestry involved an unexpected geographic origin. The sample sizes used here are less than ideal, but for certain questions larger sample sizes are unnecessary (assuming correct identification of the subspecies). If two or more subspecies lack diagnostic variation relative to each other, increasing their sample sizes will not change that pattern of variation and identify them as being distinct. If, however, two or more subspecies are distinct using given sample sizes, then it may be that an increase in sample size for any of them will reveal the existence of polymorphism at the diagnostic sites, thus falsifying the hypothesis of their taxonomic distinctness. The current geographic distribution of tigers is ‘artificial’, in the sense that they have been eradicated from many areas, thus except possibly through the use of 140 J. CRACRAFT ET AL.