“Black tigers” in a small Indian reserve suggest random genetic drift

October 17, 2021 • 9:15 am

The two greatest forces changing the frequency of gene variants in a population are natural selection and genetic drift. You’d better be familiar with natural selection by now, but genetic drift isn’t widely appreciated by non-evolutionists. It’s simply the change in frequency of genetic variants due to chance alone: the random sorting and representation of variants from one generation to the next, due not to any inherent increment or detriment to reproduction conferred by the genes.

Teaching genetic drift to students often involves letting them represent a population by choosing marbles out of a sack. If you have ten marbles in a sack, five red and five blue (representing a population with equal frequencies of two genetic variants), and choose five to be the genes in the parents of the next generation (population size must be finite), then you might get three red ones and two blue ones. You then make a new sack with the new population’s frequencies—6 red marbles and 4 blue ones. The frequency of the red variant has risen from 50% to 60%. Lather, rinse, and repeat, and you’ll see the frequencies of the marbles change each generation due to chance alone. Given enough time, all the marbles will be the same color, and then no further change can occur (this is called “fixation”). Thus we see gene-frequency change (which most of us define as “evolution”) occurring, but there’s been no natural selection, no deliberate choosing of marbles of one color. I often gave my students examples of gene frequency change in one population and said “what would you do to determine if this is due to selection?” (Answer: set up replicate populations. Selection will always drive the same variant to high frequency, while with drift you will get diverse and opposite changes among the replicate populations.)

The smaller the population, the greater the changes in gene proportions will occur (i.e., the stronger the “genetic drift”). In fact, if the population is small enough, genetic drift can overcome natural selection, increasing variants that actually diminish reproduction.  When you see small populations with high frequencies of odd variants, or even deleterious ones, you might begin to suspect the action of drift. Inbreeding can be seen as a form of genetic drift in a small population of restricted size, which is why one sometimes sees high frequencies of genetic diseases or defects in small populations of humans (here are some examples in the Amish).

The paper below, from the latest Proceedings of the National Academy of Sciences, shows a likely case of genetic drift involving a gene variant that causes bigger and darker stripes in tigers in India. You can read it by clicking on the screenshot below, or get the pdf here (the full reference is at the bottom of post).

There’s also a PNAS commentary on the paper above if you want the short take. Click on screenshot below, or get the pdf here.

India is home to two-thirds of the world’s tigers, and natural populations are often fragmented because of habitat destruction, and can also be small because of past hunting. A sampling of Indian tigers from wildlife reserves and zoos showed that one area, the Similipal Tiger Reserve in Odisha, had a high proportion of darkly striped tigers called “black tigers”. This is not the same as the melanism we see in black leopards and jaguars—both called “black panthers” though they’re different species. Below is a black tiger (right) compared to a “normal” tiger (click all figures to enlarge them.)

Below is a map showing where the authors sampled tigers. Circles are natural populations, and squares are zoos or captive reserves. The size of the circles and squares represents the sample size of tigers. I’ve put an arrow pointing to the area of interest, the Similipal Tiger Reserve.

Black tigers are seen only in Similipal Tiger or in small reserves or zoos.  The pie charts also show the frequencies of individuals that have zero (yellow), one (orange) or two copies of the mutant gene causing the unusual pattern (black color). The diagram below shows that black tigers “m/m” are found only in the wild in Similipal, but are also seen in two zoos, where they’ve probably been selected for breeding because they’re unusual. Further, the black tigers in the zoos all were founded by at least one ancestral individual from Similipal.

For some reason that one small wild population has a high frequency of the black variant (“allele”). (There are a minimum of 12 adult tigers in Simlipal, which is a minimum estimate. But there can’t be many more than that, as the rangers can identify the tigers.)

(From paper): Fig. 2. Distribution of the genotyped individuals. A total of 428 individuals were genotyped at the Taqpep c.1360C > T mutation site. Wild tigers are shown with a circular marker, and captive tigers (NKB, AAC, and Mysore Zoo) are shown with a square marker. The size of the square/circle indicates the number of individuals genotyped from a given area. In addition to the 399 Bengal tigers shown on the map, we genotyped 12 Amur, 12 Malayan, and five Sumatran tigers from Armstrong et al. (40) These are not shown on the map to allow the figure to focus on sampling within India. The fraction of the three genotypes in samples from the three populations in which pseudomelanistic tigers are present is shown with the pie chart. Similipal is the only population of wild tigers to have pseudomelanistic tigers, and the other two populations are of captive tigers. All wild tigers were homozygous for the wild-type allele at Taqpep c.1360C > T site except for Similipal individuals.

The researchers got samples of captive tiger DNA easily, but getting wild tiger DNA is hard. That involved tracking the tigers and collecting their feces, saliva from prey, or shed hairs. Sequencing can tell you immediately whether you have tiger DNA or something else. I’m not quite clear about how they managed to distinguish the tracks or prey of individual tigers in the wild from that other tigers, but differences in the DNA from different samples would tell you how many tigers you’re dealing with.

If it is indeed a single gene causing blackness, it behaves as a recessive; that is, you have to have two copies of the mutant form to be a black tiger. With no copies or only one copy paired with the “normal” allele, you have the normal tiger pattern. Here’s a genealogy of color from breeding records of captive tigers. Orange represents normal-patterned tigers, while black are “black tigers.” Circles represent females and squares males.

You see that two orange tigers can produce a black one; in these cases the orange tigers each carried one copy of the recessive “black” allele; they were “heterozygotes”.  This doesn’t absolutely establish that it’s a single recessive gene; it would strengthen the case if they mated two black tigers together and got all black offspring, which is what you predict from a recessive gene.

From paper: (From paper): (B) The pedigree of the captive tigers sampled for this study. The individual labels shown in red are for the tigers whose genome was sequenced for this study (NKB17 is not shown in the pedigree). The genotype values are indicated for the individuals sampled and successfully genotyped at the mutation site (+/+ for wild-type homozygote, +/m for heterozygote, m/m for mutant homozygote, and x/x for missing genotype). Squares represent males, and circles represent females. Pseudomelanistic phenotype is represented in solid black shapes. The dashed line shows the presence of the same individual at two spots in the pedigree.

But how do they know that the black pattern is caused by a single gene? The authors’ whole-genome sequencing found one gene whose variants comported completely with the color: if you had two copies of the mutant, which has a DNA sequence that eliminates formation of the protein coded by that gene, you were black, but if you had no copies or only one, you were normally colored.  This gene is called Taqpep, which has been implicated in making dark variants in other cat species (see below). The full name is “transmembrane amino-peptidase Q”, and the mutant form, which doesn’t function at all, is called Taqpep pH454Y. We’re not sure how the “normal” gene works in pattern formation: the enzyme is involved in degrading other proteins, and also helps form the placenta in humans!

What we do know is that other mutant felids with darker and broader stripes also have mutations in the Taqpep gene. Below is a figure from the commentary paper showing homozygous mutations in that gene in the tiger as well as in the domestic cat and in the cheetah, where it produces cheetahs with dark blotches instead of spots (see below). Each of the three Taqpep mutations is different, so here we have an example of “convergent evolution,” independent species arriving at similar appearances via independent mutations. These mutations must have occurred since the common ancestors of the three cats, which lived 11.5 million years ago for all three, and 8.8 milion years ago since the ancestor of the domestic cat split from the ancestor of the cheetah.

(From paper): Fig. 1. Convergent evolution of broadened stripes/spots in cat species. The phenotype has arisen independently in the domestic cat (Felis catus), cheetah (Acinonyx jubatus), and tiger (Panthera tigris). (A) The phylogeny on the left depicts the relationships among the three species; numbers above branches indicate the divergence times (in million years ago) among their respective lineages; a timescale is shown at the bottom (tree and node dates are from ref. 17). In each of these species, the phenotype is caused by unique mutations in the Taqpep gene, whose positions in the encoded protein are indicated below the respective branch. Coat pattern images are modified from the photos provided in the original articles: ref. 10 for domestic cat and cheetah; ref. 8 for tiger. (B) Schematic of the Taqpep protein indicating the positions of the five pattern-altering mutations shown in A (color coded per species).

Below, a “king” cheetah (right) next to a normal cheetah:

Why the dark tigers in Similipal? Given that the gene is rare elsewhere except in zoos, and that the Similipal population is small, genetic drift is a likely explanation. The mutation could be “neutral” (i.e., conferring neither a reproductive advantage or disadvantage compared to “normal tigers”, or it could even be slightly harmful. If the dark form were selectively advantageous, you’d likely see it in many Indian populations as it increased in frequency. (Further genome analysis shows no sign that the gene has risen in frequency due to selection, but we can’t say that with absolute assurance.)

In fact, the authors did a simulation assuming that the Similipal population was isolated from other populations 10-50 tiger generations ago, and concluded that the population was likely founded by only a couple of tigers: two or three. In Similipal the frequency of the “dark” gene form is about 58%, while the light gene form is at about 42%. If there were random mating, you’d thus expect (0.58)² dark tigers there, or about 34% of all tigers. As you can see for the Similipal pie chart above, that is pretty close to what you get.

This would, then, be a good example to use when teaching about genetic drift, which is a difficult concept to teach well, involving mathematics that students don’t like. But when teaching you always need examples, and we can demonstrate drift in the lab using sacks of marbles or computer simulations. But it’s better to have examples from nature, and this is one that I’d use when teaching, as it satisfies the conditions for drift, there appears to be no selection favoring the black gene, and the population is known to be small and isolated.

The only other question is that of conservation. The Similipal population is endangered, and could be increased by bringing in other tigers. That would reduce the frequency of the black gene and of black tigers. It all depends on what you want to save: the tiger itself or the pattern? I’d go for the tiger, as the pattern genes will always be around in low frequency in the gene pool, but the tigers may disappear.

______________

Sagar, V. Christopher B. KaelinMeghana NateshP. Anuradha ReddyRajesh K. MohapatraHimanshu ChhattaniPrachi ThatteSrinivas VaidyanathanSuvankar BiswasSupriya BhattShashi PaulYadavendradev V. JhalaMayank, M. Verma Bivash PandavSamrat MondolGregory S. BarshDebabrata Swain, and Uma Ramakrishnan. 2021. High frequency of an otherwise rare phenotype in a small and isolated tiger population

11 thoughts on ““Black tigers” in a small Indian reserve suggest random genetic drift

  1. Very cool! I am lifting some of this to use in my classes.
    I have read that its thought the black jaguar has some selective advantage in dense forests, and so it is a bit more frequent there than elsewhere. But I have not checked further on that. That one looks to be a single mutation involving the Melanocortin receptor signaling pathway.
    Another example, somewhat a classic, are the white “spirit bears” which are white-ish black bears found in some areas of Canada. Their numbers are highest on some islands, and that suggests the result of the founder effect. It seems possible they have advantages when hunting in the winter, but I have not learned more on that.

    1. I’ll bet rosettes are better – otherwise why would nearly all be that pattern? Surely dark coloured ones would have occurred before & spread if they gave significant advantage?

      PS do the spirit bears not hibernate? I would say there cannot be that much advantage?

  2. Fascinating, and the explanation has a clarity that is to be envied – a loss to Chicago’s students that PCC is now PCC(E).

  3. I don’t think I really understood genetic drift until you explained it with the marbles. AHA! I’m saving this post for when I need a refresher course.

    And I’m with you, save the species, not the pattern. (Could that also be called the phenotype?)

    I’ve never seen a black tiger or a king cheetah; I find them both remarkable and beautiful.

  4. Based on this discussion and the one on “New Scientists” evolution articles in September of last year, my understanding is that genetic drift is important in changing gene frequency and for molecular clocks. Is not of much significance in speciation or morphological/physiologic changes, and in small populations can overpower adaptations that are helpful.
    Is that a fair description and is genetic drift important in any other way?
    (just wondering for my own education)

    1. In my book “Speciation” with Allen Orr, we have a whole chapter on drift vs. selection, and conclude that drift is not important in speciation: i.e., in creating reproductive isolation between populations. It is of course the basis of molecular clocks. As for morphologica/physiological changes, it’s hard to tell. Surely some observable characters are the result of drift, but how do we distinguish that from weak selection impossible to detect experimentally? Or, some traits that look “neutral” may simply be pleiotropic byproducts of traits that were fixed by natural selection.

      There’s an early book by Robson and Richards purporting to give many examples of traits resulting from drift, but I wouldn’t trust their examples, as it was a long time ago and there were, as I recall, speculations rather than experiments.

  5. I love hearing how you taught “genetic drift”. Where were you when I was struggling to learn genetics (1996)?
    I really adore this blog. We are so lucky that you take so much time and care and are so generous in your sharing. It is one of the highlights of my day.
    Thank you, Jerry
    Debi Z in Tucson

  6. Thank you for this really helpful explanation! It made sense even to a layperson like me. It probably helped that it used cats — one of my favorite topics — as the example.

Leave a Reply