The mean IC could be lowered significantly in the future while still selectively breeding for the white coat. A table of 698 possible family-line pairings shows that 78% could produce white cubs with IC less than the current overall mean (<0.2), and 39% with IC less than half the current mean (<0.1).
A closer look at the circumstances around the instances of inbreeding suggests a variety of causes unconnected to breeding for the white coat – convenience, animal safety concerns, lack of space, accident, and lack of accurate information to inform breeding choices.
1. The Data Set used
The data set records 396 tigers alive at around 1 December 2018, from the White Tiger Dataset – ie the current Bengal tiger population of zoos in India.
The 2018 National studbook classifies tigers as “Normal” or “White”, with Normal being orange tigers without the white gene, while White includes both white-colour tigers and orange heterozygous tigers, ie the product of white tiger breeding.  I will not use the term “Normal” as I consider white tigers to be normal and natural also, so have split them into:
- 193 White-gene-carriers = 123 white-colour tigers (ww) + 70 orange heterozygous tigers (Ow)
- 203 Non-whites = orange tigers assumed to not carry the white gene (OO and O?). 69 Wild-born tigers and 12 tigers with unknown parents are excluded from calculations, leaving 122 in this group.
2. Mean Inbreeding Coefficients (IC)
Mean inbreeding for the living population, including both white and orange tigers, is given as 0.2054 in the 2018 National studbook (data to 31 March 2018). In the conclusions, the highness of this figure is explained as a consequence of selective breeding for the white coat colour. However, there is no mean IC given for White-gene-carriers versus Non-whites to enable comparison.
Using the data set described above, I calculate:
- Overall mean IC = 0.2077
- Mean IC for the Non-whites = 0.1934
- Mean IC for White-gene-carriers = 0.2168
That’s not a lot of difference between the mean IC for non-whites and whites! So this does not support the conclusion that white tiger breeding is to blame for high mean IC.
A closer look at the circumstances around individual tigers and pairings suggests some more likely reasons for the high IC…
2.1 Inbreeding of Non-White tigersThere have been many instances of orange (non-white tigers) being inbred – IC’s range from 0.0 to 0.48, with a cluster of 32% in the range of 0.30 to 0.39. 
Sometimes the inbreeding happens by mistake, when a male cub is left with his mother or sister too long. (A zoo may not have the space to separate them, and zoos that keep tigers in safari-style enclosures tend to keep family groups that get along well, to avoid fighting.) This happened several years ago at one tiger safari where it was found that the father of a litter of cubs was the mother’s son from her first litter. As there had already been inbreeding in the history of this line, the IC of the cubs was very high, around 0.4800. (There is no suggestion of a white gene anywhere in the pedigree of these tigers).
In other instances a zoo gets a pair of tigers and then breeds their descendants together, “making do” with the tigers they already have rather than obtaining new animals with the inherent risks involved in transport and introducing new tigers to each other. (There have been several instances of apparently healthy tigers dying soon after transfer to another zoo, and of tigers killing their newly-introduced mates).
These choices and events have nothing to do with breeding for white colour, although the same can occur with white tigers, for the same reasons.
2.2 Inbreeding of White-Gene-Carrier TigersIC’s for White-gene-carriers range from 0.0 to 0.55, with a cluster of 33% in the range of 0.20 to 0.29.
There is no genetic necessity to inbreed white tigers (see 3. Potential Pairings), so why does it happen ?
In addition to those outlined in 2.1, one more reason may be that white tigers are highly treasured and zoos are reluctant to part with them. When available, they are often youngsters from the same litter, so a sibling pair is obtained. The Central Zoo Authority directive to keep animals only in breeding pairs may have unintentionally increased this trend (if a zoo does not find a mate for an unpaired animal within 6 months they must transfer it to another zoo, so safer to obtain two together). 
Zoos also seem to do trades within the same group of zoos, causing tigers to be bred within the same family groups over and over. While some groups of zoos clearly keep good records and together have bred low-IC white tigers, I suspect that lack of complete and accurate pedigree information has contributed to some poor breeding choices elsewhere.
2.3 Disproportionate effect of the breeding choices of just 2 zoos
A closer look at the individual white tigers shows an interesting point – the disproportionate effect the choices of just one or two zoos can have on the mean IC.
Two zoos bred a large number of white-colour tigers from the same family line, all highly inbred. The ironic thing here is that the parents have been prolific breeders, with very high survivability of cubs, resulting in 18 white tigers with IC from 0.46 to 0.55
If these 18 tigers are removed from the calculations, the new mean ICs are:
- Overall mean IC = 0.1890
- Non-whites = 0.1934
- White-gene-carriers = 0.1859
Now the White-gene-carriers have a lower IC than the Non-whites!
Of course, these tigers should not be removed from the calculation, but doing so shows how the choices of just a few zoos can skew the averages.
2.4 IC is not a measure of fitness
Those 18 tigers also illustrate another important point – that the Inbreeding Coefficient is not a measure of health or survivability. It is a measure of risk when planning mating pairs: the IC of the planned cubs is the probability that gene alleles will be identical by descent; ie the risk that if there are deleterious recessive alleles present, that they may be doubled up and show their effects in the cubs. (This can also happen without any inbreeding). Perhaps this family line of white tigers does not have any deleterious alleles, or perhaps it is simply luck.
3. Potential pairings for Selective Breeding of White coat colour
Chart 3.1 below shows 698 possible pairings between family groups of White-gene-carrying tigers currently alive, and the resulting IC’s for prospective cubs. Each family group contains from one to several individual tigers with the same parents. Full details of the individuals are held in a separate table (not published).
All these pairings have been selected to produce white-coloured tigers (ww). Pairings between two (ww) parents will produce all white tigers; pairings between one (ww) and one (Ow) we would expect half white and half (Ow); etc.
3.2 The Results
A whopping 78.2% of these selective pairings for white-colour tigers would result in IC’s less than the current mean IC of the overall population, ie < 0.2077. This means that they would either decrease the mean IC, or have no effect on it.
16.6% of pairings would result in tiny IC’s less than 0.0550 (IC 5%), while a further 22.1% lie between 0.0550 and 0.1049 (IC 6-10%).
Clearly there is a lot of scope for breeding white tigers with very low IC in the current Bengal tiger population !
Of course other factors must be considered apart from IC, such as the logistics of pairing two tigers from distant zoos, and the temperament and health of each individual tiger.
An example on the chart that has already been attempted, is the pairing of a white tigress at Delhi (Family #14) with a heterozygous orange tiger from Mysore (Family #25). The IC of the cubs was a tiny 0.0396 (3.96%). Unfortunately the cubs did not survive, as often happens with a mother tiger’s first litter, but I hope the zoo will try again with this pair.
3.3 Outcrossing – reducing the mean IC even furtherChart 3.1 does not include the outcrossings (with unrelated orange tigers) that are used to produce heterozygous (Ow) tigers, but these should, and do, occur. The outcrossings will produce White-gene-carriers with IC = 0.0000. Five of the family groups in the table are the direct result of previous outcrossings with wild-born tigers.
(Refer to “Inbreeding and Outcrossing” for explanation and examples of outcrossing).
4. Need for complete and accurate data
Availability of complete and accurate pedigree of tigers bred by all zoos is necessary to enable the best breeding choices to be made, and the information needs to be in a format that is usable by a zoo seeking a partner for their tiger.
Chart 3.1 (along with it’s table of individual tigers) is one method to show where to look for partners for each tiger. It also shows what would happen to the IC if a non-recommended pairing takes place – and thus perhaps discourage such pairings.
An incomplete or inaccurate dataset can give misleading results. Missing records can mean that good pairing opportunities are not identified. An error of just one digit in a studbook number can inadvertently change a tiger’s sire or dam to a completely different animal, and cause a major change in the IC calculation for all of its descendants, and their recommended pairings.
The Indian National Studbook for Bengal Tiger contains records of tigers going back to the 1950s. However, the studbook authors rely on the various zoos to supply the information, and often this information is simply not forthcoming. For example, for the latest national studbook (2018), only about 30% of zoos were noted as having provided updated information.
It would be advantageous for each zoo to be more proactive in ensuring that “their” pedigrees in the studbook are correct and complete – both for current and historic records.
5. Notes and Sources
* This article may be freely quoted but the source must be credited and cited as: “White Tigers: Current IC and Future Breeding Prospects – whitetigertruths.wordpress.com” OR a link given directly to this article.
* Inbreeding Coefficients were calculated using Wright’s formula to 15 generations, using Breed Mate Pedigree Software.
 Wildlife Institute of India (2018) National Studbook of Bengal Tiger (Panthera tigris tigris): IV Edition, Wildlife Institute of India, Dehradun and Central Zoo Authority, New Delhi.TR.No.-2018/34 Pages: 668.
 That seems to be the aim, but it is not entirely consistent. In the studbook, some known heterozygous and white tigers are labelled “Normal”, while some not carrying the white gene are labelled “White”.
 The use of OO, O?. Ow. and ww is my own convention which I have not seen used elsewhere. The standard method is to use a big W to signify the dominant orange gene with a small w for the recessive white gene. However I find that some online texts make it difficult to distinguish between WW and ww so I use the big O with little w instead to avoid misunderstandings.
Ow = orange-coated tiger carrying the white gene (ie heterozygous)
ww = white-coated tiger
OO = orange-coated tiger known to not carry the white gene
O? = orange-coated tiger not known to carry the white gene.
These O? (O/maybe) tigers have an ancestor that is either Ow or O? so there is a slight chance they do carry the white gene…. see 
 I read somewhere that one of the regional Species Survival Plans tiger population had a mean IC of 0.01, but that when the wild-born tigers were excluded, this jumped up to 0.10. Sorry, I cannot find the reference, so take this with a grain of salt if you will. It may mean that the Bengal tiger IC is not as high relative to other captive populations as it seems. Wild-born tigers do not belong in the mean IC calculation as they are not a product of captive breeding.
 There may be a case for including the O? tigers in with the White-gene-carriers instead, but if done, this does not change the results significantly. It would slightly decrease the mean IC for the White-gene-carriers from 0.2168 to 0.2150; and for the Non-whites from 0.1934 to 0.1903. There would still be a main cluster of 32% in the .20 to .29 group for White-gene-carriers; and still a main cluster of 27% in the .30 to .39 group for the Non-whites.
As these O? tigers probably do not carry the white gene, and are not usually used for white tiger breeding, AND it does not greatly affect the results, I have grouped them with the OO’s.
 Zoos in India 2014 – Legislation, Policy, Guidelines and Strategy
CZA, Edited & Revised by: B.S. Bonal, Inder Dhamija, B.R. Sharma, S.C. Sharma, Brij Kishor Gupta
 These errors can and have happened. The IC’s I have calculated for some of the White tiger pairings are higher than those given in the Studbook’s recommended pairings, because of two differences in Sires recorded in the Studbook. Ref “Proposed Corrections to the 2018 Indian National Studbook of Bengal Tigers”; 14pp; Dec 2018.