I follow Conservation Bytes (https://conservationbytes.com), a blog maintained by Dr. Corey Bradshaw, an Australian scientist. He specializes in mathematical modeling of ecological processes, in particular, that of population genetics. In a recent post (https://conservationbytes.com/2018/04/03/why-populations-cant-be-saved-by-a-single-breeding-pair/ and https://theconversation.com/au), he warns that animal (or plant) populations should not become too low, otherwise the species will be lost, even though some individuals remain alive.
The basis of calculation is as follows: a population of 250 to 500 is needed to obtain fifty effective individuals, i.e., those who can breed. However, to retain evolutionary potential – to remain genetically flexible and diverse – the IUCN criteria suggest that at least 500 effective individuals are needed, which requires a population of 2,500 to 5,000.
Variation exist from species to species. But, is seems a “huge” number of individuals are needed for species to survive forever (say thousands of years). As you may know, most tiger and Asian elephant populations are lower than 2,500. If the Asian elephant population in and around the Nilgiris Biosphere Reserve (the largest in the world) seems to be of an adequate size, the tiger population of this region (also the largest in the word for this species) is only about 600 individuals.
The sad reality is that most tiger and Asian elephant populations are too low and the tiger is on the verge of extinction. The 2,000 tigers or so remaining in India is just a number of little importance. The number we should remember is that the largest population approximates 600 tigers, which means that the species is hanging by a thread. Other large populations in India and abroad are badly needed if we want to keep these species.