The purpose of reintroductions is to restore native species to their historical habitats. Minimally, to do this we need to have a knowledge of the biology of the species to be introduced, its current status, and its historic status. I will give a few examples later, but first we need to consider whether reintroductions are ever justified. Some would say no, that if the habitat is good, species will find their way back. In my experience, these folks live in riparian areas where river and stream banks provide corridors for natural recolonization. Now imagine the prairie pothole region of the upper midwest, where wetlands are islands isolated in a sea of corn/bean/wheat fields. There is no way short of tornado dispersion (see It's Raining Frogs and Fishes) that amphibians are going to recolonize some of these areas. We have a simple choice, reintroduce native species or write off the habitat. Which is the greater evil?
How do we know that a species is extirpated? Good point -- absence, is difficult, if not impossible, to prove. We determine absence by lack of presence, in the context of a particular species' biology. For example, we can be more confident of absence where habitats are discrete (potholes) and animals short lived.
Imagine an isolated northern Iowa wetland where we know that Frank Blanchard collected cricket frogs (Acris crepitans) in 1920, during his first visit to the site. We've seined this same pothole each summer for the past eight summers and listened for calling frogs -- no cricket frogs. We know that cricket frogs prefer aquatic habitats (that they will not travel great distances over land), we know that they only live two or three years in nature. Are they gone? A better question is what is the probability that they are gone? Now consider these same questions for American toads. They can travel long distances over land, and we know that in captivity they can live 36+ years. In my view, reintroductions would be more justified in the case of the cricket frog than the toad. Another factor would be the patterns of decline in these species. Cricket frogs are disappearing hard and fast in the upper Midwest, whereas American toads appear to be doing fine. There is more urgency to conserve cricket frogs than American toads.
A few other issues have been raised. How do I define "successful reintroductions"? In my sense, successful reintroductions of cricket frogs means that no animals of any life history stage were collected for several years prior to the reintroduction, and that since the reintroduction animals have continuously been found. Of course we cannot predict the future, we can only say that reintroductions have been successful up to this point in time. Without genetic analysis of the original and the reintroduced animals (assuming there are differences and the reintroduced animals are not marked) we cannot scientifically prove that the animals that we now see are the reintroduced animals (or their offspring) or the newly-awakened original animals. Does this mean that we should not be reintroducing species unless we know the genetic makeup of the animals to be brought in? If so, and we get our results, how do we know whether the new animals carry the same genes as the original animals when the original population has been extirpated? In fact, museum collections may be able to help us out.
Whether the environmental role of the reintroduced species is the same as the original population is probably less dependent on intraspecific differences than it is on habitat/ecosystem alterations since the original extirpation. For example, if many of the invertebrates in the area have also been extirpated, a food habits study conducted now would yield different results than one done at the turn of the century. Image this. A food habits study was done at the turn of the century on a population of amphibian species X. The population gets wiped out. You reintroduce animals, they take, and after a few years repeat the food habits survey. You note differences in prey species taken, and their proportions. Do you then account for this by genetic differences in the sensory-motor apparatus used to detect and capture prey, or do you chalk it up to differences in prey availability? This seems like a ridiculous choice but there are cases where the first option may apply, for example in polymorphisms. If the eastern tiger salamanders of NW Iowa were extirpated and animals were reintroduced from other populations, there is some chance that the cannibal morph larval genes would not make it back. Again, the biology of each species must be considered.
Species reintroductions can be a useful tool to determine whether a habitat can still support the species (and therefore does not have animals because since the extirpation they have not found their way back). They can also be a useful tool to determine the cause of the original extirpation if the cause is still present (Gary Feller's work).
These are the guidelines I recommend for species reintroductions:
It would also be useful to collect genetic information from the introduced animals, but obviously this cannot always be done.
If guideline 2 is followed and animals are reintroduced to a viable but undiscovered population, the effect would be, presumably, to mimic natural immigration. This would, of course, depend on how far away one has to travel to find a viable host population.
Generally, I not a fan of moving animals around. Sticking to amphibian examples, bullfrog introductions continue to decimate native amphibians. Where habitats can support a species that was historically there, and where it is generally conceded that animals will not recolonize on their own, reintroductions are the only alternative to having the species absent from that site. The less habitat available to a species, the more seriously reintroductions should be considered (remember the heath hen example).
If general guidelines are proposed for when species should be reintroduced, I strongly suggest they consider the unique biology of particular species.
I also warn that there will always be disagreement when it comes to reintroductions. We range in our approaches from being completely hands off to completely proactive. Where controversy is particularly severe, I suggest that in making these decisions we use a form of peer review. If we can convincingly argue for reintroductions to a non-biased group of herpetologists, it is probably reasonable to perform the transplants.
While we discuss reintroductions as introductions, it is ironic to realize that in many cases reintroductions are necessary because introductions of non-native species extirpated native populations. In my view non-native introductions are always going to be a more severe disturbance than reintroductions; we should not confuse the two.
Two final queries.
First, if one opposes responsible reintroductions, I would hope that one also opposes introductions. If you then follow up on your convictions and eliminate the introduced species, have you fixed the environmental problem? I would argue no, not until you have made an attempt to restore the damage done by the introduction to the original community. Without reintroductions, how do you restore an original community in the face of our fragmented landscape? Is there any alternative to reintroductions in a fragmented landscape?
Second, if one opposes responsible reintroductions I suppose one must also oppose habitat restorations, for many of the same reasons -- we really do not know what was there originally, we cannot reproduce exactly the same species assemblages (or to be silly, the same plants in the same places), many of these species may have died out on their own, if they really belonged there they would still be there, etc. In short, we do not know any of these things (indeed in many cases we cannot know any of these things), so why bother? If anyone champions restorations but opposes reintroductions, I'd like to see their logic.