Can the Keystone Herbivore Hypothesis be falsified?

The information that would presumably supply the strongest test of the keystone herbivore hypothesis would involve (1) a demonstration that the largest herbivores in fact died out first; and (2) that microenvironmental change in areas once occupied by megaherbivores occurred hard on the heels of their disappearance.

The first point could be retrieved from a dense radiometric record. Although there are claims that some of the megaherbivores died out very early during the period of losses, on the time scale involved (perhaps only 400 years) it may be very difficult to show that all the variables can be accounted for. Among these is the nature of radiocarbon dates themselves, which are not precise dates in the calendrical sense, but are instead statistical statements of the probability that the real age of a dated specimen falls within an agreed-upon error of estimate.

There is an even greater problem with the second point. Although palynological investigations may supply a reasonably accurate picture of long-term changes in vegetation in an area, they are only indirect indicators of cause. It may rarely be possible to determine whether a given change--especially in a small area--is more likely due to loss of megaherbivores than to environmental change due to other factors.