Can the Overkill Hypothesis be falsified?
For a hypothesis to have validity, scientists also have to define facts or events that could potentially prove it false. Martin identifies two kinds of facts which--if discovered--would imply that overkill is a false explanation of the Late Pleistocene extinctions.
(1) Extinction of animals too small or ecologically resistant to be vulnerable to human impact; and
(2) Extinction of Late Pleistocene faunas under circumstances that preclude an anthropogenic effect, i.e., the extinctions occurred before prehistoric human arrival.
In fact, we already know of a large number of small mammal extinctions. The overwhelming majority of these, however, occurred on islands, and causes other than simple human impacts may have played a leading role in their disappearance (for example, the introduction of exotic, aggressive species like Old World rats on Caribbean and Pacific islands).
Norman Owen-Smith and others have also developed an explanation for this consistent with the overkill hypothesis. This view, called the Keystone Herbivore Hypothesis, is based on the observation that very large mammals, like elephants, rhinos and hippos are prodigious environmental engineers. The overkill extinction of similarly-sized mammals in the Late Pleistocene would have transformed the environment for some smaller mammals and can account for their extinction as well. Testing this kind of dependency for non-extinct communities has proven difficult, however.
The second issue is more amenable to testing, because it fundamentally depends on dating only two points in time: when did people first get to a specific area, and when did the last populations of now-extinct species die out? In the case of North America, there is still a weak consensus that people first entered the continent in the very Late Pleistocene. However, the date keeps creeping backward. Recently, archaeologists working at the site of Monte Verde in southern Chile established beyond reasonable doubt that humans were present in the cone of South America, with a substantial material culture, as early as 13,500 calendar years ago. If it is true that their ancestors came over the Bering land bridge, it could be that human colonization of the New World began at least somewhat earlier, perhaps 20,000 (or even more) years ago.
For Martin's hypothesis, the problem with an earlier arrival schedule for humans is that the behavioral factor--naive megafaunal species lacking appropriate escape responses when faced by big-game hunters--becomes more problematic as an explanation of how these species might have been hunted to extinction. Also, dates of "last occurrence" of now-extinct species ought to show some temporal gradient if humans got into the continent early (and if overkill is indeed the leading cause of Late Pleistocene extinction). Yet there does not appear to be such a gradient in the few cases which have been explored.