Science: When it's time to call a tiger a tiger: Colin Tudge looks at the complexities of survival for the world's endangered species

Many animal species will become extinct in the next 100 years unless conservation in the wild is backed by breeding in captivity, which can have a huge impact. There are about 20,000 species of land vertebrate, and of these an estimated 2,000 may already be doomed in the wild and need breeding in safety.

Conditions in the wild can improve - given time and cash, poor countries can create the necessary reserves - and most captive-bred species can be returned successfully to the wild. The 2,000 candidates for captive breeding include some of the world's most spectacular creatures, such as tigers, rhinos, a host of primates, parrots and pheasants. And if all the world's zoos worked together, with government and public backing, they could support all of them through the next few difficult decades.

But many of those 2,000 are subdivided into 'races' or 'subspecies' - and this is where the arithmetic breaks down. For example, is it worthwhile, or even possible, to try to maintain Sumatran tigers separately from Siberian tigers, or to keep eastern black rhinos apart from western and southern? Or should we now say: 'A tiger is a tiger, and a black rhino is a black rhino', and just mix the subdivisions to produce generalised 'zoo hybrids'?

Chimpanzees are far more variable genetically than human beings and might be subdivided into half a dozen subspecies or more. Some parrot species break down into a dozen or so 'races'. Thus does 2,000 species - a manageable figure - grow into perhaps 10,000.

The dilemmas are huge and various. If zoos try to keep more subspecies than they can manage, then entire species may become extinct because there is a limited number of spaces per species in all the zoos combined, but any one population normally needs to contain several hundred individuals if it is to be viable. Thus, there are two subspecies of red panda, the 'common' and the larger Styan's. The present world total of red pandas is big enough to be viable, but the zoo population of each red panda subspecies, taken separately, probably is not.

The World Conservation Union has a 'Policy of Minimum Regret', which means, in effect, that subspecies should be kept apart until it becomes hopeless to do so. After all, once they are mixed, they cannot be unmixed. Thus most zoos aim at present to keep the different black rhino subspecies apart; but if numbers drop much more, it will be perverse not to combine them. There are dangers though: the three types are physically slightly different, and perhaps - no one knows - there may be significant ecological differences.

The two main subspecies of orang-utan, the Sumatran and Bornean, look different and have different chromosomes: so the case for keeping them separate is clear. But one-fifth of all the orangs in zoos are hybrids - bred before it was known that the two types had different chromosomes - and what is to be done with them? I know a serious zoo that wants to breed orangs, but has young hybrids. To breed from hybrids is irresponsible. To 'euthanase' such creatures looks very like murder. But orangs are long-lived: if the zoo keeps the hybrids, occupying space, it will have no spare facilities for breeding orangs until around 2030.

Yet with some creatures there is already a strong case for hybridisation. Just perhaps it is already time to mix the tigers. Sumatrans, adapted for tropical forests, are slimmer and smoother than the shaggy Siberians, adapted for snowy steppes. Clearly, Siberians contain genes that Sumatrans do not, and vice versa. But it now seems that Siberians and Sumatrans are merely subsets of Bengal tigers. So if we bred only Bengals, we should not lose any existing tiger genes; and a zoo hybrid, containing genes from all three subspecies, could in principle adapt quickly to any conceivable tiger habitat. Furthermore, by eliminating (or failing to breed from) existing zoo hybrids, as is already common practice, we are probably sacrificing at least some genes that have already disappeared from the wild, for the ancestors of some of those hybrids may have been caught decades ago. It seems too early, in the case of tigers, to abandon the 'Policy of Minimum Regret'. But deliberate hybridisation might be advocated in the foreseeable future.

The humble Partula tree-snails of the volcanic Pacific islands show how the dilemmas can be solved - at least in principle - by a protocol devised by Georgina Mace of London's Institute of Zoology and Professors Bryan Clarke of Nottingham University and James Murray of the University of Virginia, and applied by Paul Pearce Kelly of the London Zoo. The zoo keeps populations of Partula from Moorea, one of the Society Islands of the South Pacific. Originally, there were seven species of Partula on Moorea; all are now extinct in the wild, but several are being raised in captivity. Each is represented by several populations, which originally came from different valleys; and each such population is equivalent to a subspecies.

Under the Mace, Clarke and Murray protocol, each individual valley population is split into two. One portion is raised in pure form, to maintain its distinctness, and the other portion is mixed with populations of the same species which came from other valleys to create hybrids that are as genetically mixed as possible. Thus we have the best of all worlds: in the 'pure' types, any putative ecological specialisations are preserved; but the rough-

and-ready hybrids maintain the greatest genetic diversity, and should be most resilient and versatile. Eventually - soon, with luck - it should be possible to begin reintroducing the Partulas. Each pure kind should go back to its own valley and the hybrids can be put in valleys from which there are no survivors in captivity.

But of course, Partulas are small; Paul Pearce Kelly keeps entire populations, off display, in plastic lunch-boxes. It would be wonderful to adopt the same protocol for tigers and black rhinos; but there simply isn't the space. The ideal is clear, and tantalising. But for the world's endangered vertebrates, the dilemmas continue.

Colin Tudge's book 'The Engineer in the Garden' is published by Jonathan Cape, pounds 17.99. His book 'Last Animals at the Zoo', available from Oxford University Press, deals in more detail with captive breeding.

(Photograph omitted)