For a point of contrast, consider this creationist quote (via Pharyngula):
If man is an animal, the Constitution was written by animals and for animals. This preposterous conclusion destroys the Constitution. The Aguillard Humanists leave us with no Constitution and no constitutional rights of any kind if they allow us to teach only that man is an animal.
I've previously discussed the fallacy which underlies this sort of thinking. But I'd like to add: an ethical system which rests upon the foundation of a magical spark (or whatever they think it is that makes us not animals) is a vulnerable ethical system indeed. An atheistic ethics has no such vulnerability: it begins by recognizing the value of all human beings (and perhaps some non-human ones too), avoiding the folly of making this value rest upon contingencies that might well turn out to be false.
Humans have value, period. If God exists, that's fine, it doesn't change the fact that we have value in ourselves. And if he doesn't exist, then again, that's no great problem. Atheistic ethics thus has a stability and strength in its foundations that sets it apart from theistic approaches. By making morality depend on God, you make it weaker. Morality becomes contingent on the question whether God exists. If he doesn't -- as well might be the case -- then you're screwed.
Further, conditionalizing our value in such a way is inherently disrespectful to humanity, suggesting that people have no real value in themselves, and wouldn't deserve your compassion unless God added that extra magic spark to make it worth your while. Such a view strikes me as morally bankrupt and utterly repugnant, especially compared to the more "unconditional" value bestowed by a humanistic ethics.
Often theists say that without the threat of divine punishment, we'd have no reason to be moral. But that just shows that they were never really moral to begin with. (There's nothing particularly virtuous about refraining from murder solely because of your fear that you'll get caught. The truly virtuous person wouldn't even want to do evil in the first place.) A rationalistic ethics is not so vulnerable to contingencies, as its normative force persists whether one is likely to face divine judgment or not. This is clearly preferable, as we shouldn't want an ethical system which would lose all force the moment one starts to question the existence of God (as all inquisitive people inevitably do at some point). A well-functioning society requires the prevailing morality to have a stronger foundation than that, and this is precisely what an atheistic ethics can provide. (And avoiding arbitrariness is surely another point in its favour.)