If we believe that we should protect animals from unnecessary suffering and death than it seems that we should be focusing much more on reducing the non-human causes of animal suffering and death that occur almost continuously in the wild. - Amanda and William MacAskillI can't tell whether this is just a frivolous hot take (as Elizabeth Bruenig suggests) or if these people are in earnest, but this is not in any case a particularly compelling argument.
Its conclusion, after all, is that there are cases in which it is acceptable to kill. Equal consideration demands that we offer predators this defense as well, and that we specifically consider whether a predator killing its prey qualifies as "unnecessary suffering and death". Clearly zebras and the MacAskills both have reasons to consider predation unnecessary. But does Cecil the lion?
From here, you can take the argument in two directions.
On one hand, you can argue that the imperative of basic survival is always an absolute justification for killing. This is an ambitious position, but it appeals to moral intuitions that the MacAskills can't simply ignore - as they do when they compare predation to the scenario of "an infant with a handgun". That hypothetical acknowledges the actor's lack of moral agency, but it completely ignores the predator's interest in killing - one that the infant simply does not have. From behind the veil of ignorance, it seems defensible to insist on the right to do whatever it takes to survive. You would only reject this if you knew beforehand that you were prey - precisely the sort of incidental consideration the veil of ignorance tries to disregard.
This logic also gives us reason to reject their justification for non-judgmental intervention. We allow predators to survive not just because we think that they're innocent, but also because we would want them to let us survive if we were in their position. If we accept an absolute justification for survival-killing from behind the veil of ignorance, it follows trivially that we have to extend this courtesy to others.
Instead of insisting on an absolute license for survival-killing, we can, on the other hand, simply insist on a conditional license. This right is grounded in the same utilitarian logic that the MacAskills rely on in their justification for killing lions: survival-killing should be permissible if the suffering and death it prevents outweighs the suffering and death it causes.
So the right to kill becomes a kind of large-scale optimization problem. Since the authors suggest that this could be undertaken through "a rigorous risk analysis," it's worth insisting that the sort of analysis and control this would require is almost certainly mathematically impossible. Have these people not seen Jurassic Park? When they admit that "ecosystems are complex things," do they get that this is, mathematically, an admission that they are extraordinarily unpredictable and unknowable? Insofar as this argument depends on any degree of certainty about what will and will not happen if we wipe out a particular species, it is hilariously implausible right out of the gate, and certainly demands some minimal exposition before we accept it as a justification for killing.
The authors try to finesse this problem by insisting that they are only considering "the killing of individual predators," which is "unlikely to have knock-on effects on the ecosystem of the region." But this sets the bar for their argument far too low. They need to do more than rule out "major impacts" on an ecosystem; the utilitarian argument requires a counterfactual demonstration that the lives and suffering spared will outweigh the predator's survival.
This generally cannot be done. Consider, for example, the simple possibility that your target dies before it has the opportunity to kill again. This can obviously happen, particularly with omnivores and scavengers, with animals that go extended periods of time without eating, with animals the have short lifespans or that live near the bottom of the food chain, and so on. In that case, you will have done nothing to relieve anyone from suffering or death, though you will have certainly robbed that animal of the precious time it would have otherwise had.
Or consider the possibility that this animal would have done exactly what you are trying to do: it would have saved some animals by killing others. This obviously happens in cases of predator-on-predator predation, territorial struggles, herd hierarchy struggles, and so on, when a voracious predator is killed by one that is slightly less voracious. This, again, need not entail some kind of major impact on the ecosystem - it need only lead to more death and suffering than you tried to prevent by killing in the first place.
So even a conditional license to kill predators based on utilitarian considerations seems ultimately to demand a far greater degree of certainty than is generally possible, even on a case-by-case basis. There is simply no reason to assume that the first order effect of killing a predator will necessarily be justified by the impossible complex and unknowable cascade of second and third order effects that would follow. And even if we think it probable, it seems unlikely that we could establish it with the moral certainty that an affirmative justification for killing requires.
All of this seems fairly obvious to me. It may be a fun "thought experiment" to imagine who (and what) we would be justified in killing given certain knowledge of the future, but that kind of hubris is the province of Raskolnikov and Rumsfeld. Most ecosystems are far too complex to facilitate even case-by-case justifications for killing, much less the large-scale intervention that the MacAskills - even as they distance themselves from it - seem to think is possible.