all a bit grey..
i definately believe in preserving distinct populations (and potential subspecies) wherever possible, and i generally beleive its okay to be a little more liberal when classifying subspecies that it is when defining a seperate species.
when it comes to zoos, there are many circumstances where i agree that zoos made a decent decision to manage a population of hybridised decent rather than attempt to juggle a number of smaller genetically distinct populations or capture more animals from the wild to focus on one. however, what worries me is that in other situations, zoos might not do the right thing and preserve a distinct race when they have a clear opportunity. elephants are a good example. i'll bet the vast majority of breeding-age elephants in captivity outside asia are of the mainland race. their decendents were most likely from the wilds of burma, india, thailand or malaysia etc.. however, elephants are relatively difficult to breed in captivity and so of the zoos that do try (which is admittedly very few in comparison) any opportunity seems to be taken up. so you have zoo like hannover as you mentioned deliberately creating hybrid calves. no doubt the oregon zoo will breed their bornean elephant when they get the chance also. unfortunately zoos seem to be intent on managing elephants at the species level, ignoring the fact that most of their animals can be traced to a pure mainland ancestory. this means these hybrids will no douubt eventually find themselves polluting the main zoo breeding pool with their hybridised genes and you'll end up with a hybrid elephant population in zoos. i have little doubt that eventually (be it 20-50 years away), elephant will become so rare in the wild that zoo-bred animals will be mated with wild animals in an attempt to increase genetic diversity. and thus, there is the begining of the end for our elephant subspecies. like everything else, even our animals will become a monoculture. which, in the case of elephants, probably wouldn't make much difference to the ecology of the area, is nonetheless a pretty boring concept in my mind.
another interesting concept is the problem of species, that rather than having geographical boundaries separating subspecies like a river or mountain range, have one continuous didtribution over a large range. we talked of the seabirds (can't remember the species) in scotland that spread their range in one direction around the globe and by the time they reached scotland again had become a seperate species. however following their range, to this day you can still trace a flow of genes from one "species" to the other as the populations still interbreed with their neighbours to this day. maybe a more practical (in zoo terms) example is plains zebras. historically at teh soutehr end of africa you had a distinctly stripeless and brown coloured zebra at the norther end you had a two-toned heavily striped race. in the middle are intermediate forms, like the most common subspecies we keep in our australian zoos. but if hypothetically you crossbred the far northern race with the now extinct quagga (the far southern one) would the ruslting progeny be legitimately considered a pure bred chapman's zebra? (the intermediate race)....
none of it seems (pardon the pun) that black and white to me....