The problem is speed. Human presence often leads to much quicker change than other causes – for example, it's much easier for us to reach Antarctica, and hence deliver whatever is hitching a ride with us.
Colonizing life forms tend to dislodge indigenous species, often driving them into extinction. This almost always results in a net loss of diversity in the short term, because the colonizers are very much like their cousins back home, while the natives may be something else entirely (think Australia's marsupial mammals being driven off by Old World placentals). Of course the colonizers will stray away from their roots eventually, but it will take many thousands of years before the original level of diversity is restored.
You could say that's "business as usual", and might as well be – but there's no telling what secrets those gone species take with them. And for scientists that's a big problem, because we revel in discovery, but what if a newly discovered species is driven extinct before its habits can be documented? What if we can never get to grasp its biochemistry? What new medicines and chemicals could be lost forever, or postponed for many decades?
Zoologists and other natural world researchers are always struggling to hold back time, trying to keep their subjects as immutable as possible, so they can learn the most about them before they slip away, changed or killed off. And that's the extent of their powers, really: hard as they try, they can only slow change down, not hold it back forever.
So it's not that human-facilitated change is "bad" or "unnatural"; it's just that it is so fast, that it could disrupt research that could cope with other, slower change factors. And that is bad, because there's no telling what knowledge is lost in the process.