The mature view is that it boils down to the “Chesterton’s Fence” concept. Rather than “humans bad, nature good”, we just don’t know if the result of intervention might be an unfit population / ecosystem.

The result of this, of course, is that we tend to intervene a lot when humans are affected (massive industrial footprints, screw-worms, etc) and a lot less when it’s irrelevant to human welfare. We are, by nature, biased towards the anthropic.