THE USE OF MODERN DEATH ASSEMBLAGES TO TEST THE GEOGRAPHIC MOSAIC THEORY OF COEVOLUTION

The Geographic Mosaic Theory of Coevolution (GMTC) suggests that coevolution is an inherently geographic process resulting from the genetic and ecological structure of species (Thompson, 2005). Testing the specific predictions of the theory in the fossil record is promising, but may not be possible if resultant patterns are erased in time-averaged fossil deposits. We suggest that modern death assemblages provide a good test of this bias, because of their minimal extent of time-averaging. We tested two of Thompson’s predictions:

1. Populations differ in the traits shaped by an interaction.

2. Coevolved traits will be well matched between species in some communities but sometimes mismatched in others.

We studied Busycon carica, a morphologically variable gastropod whose range encompasses two biogeographic provinces along the east coast of the US. Prediction 1 was tested by examining the distribution of anti-predator traits between the northern and southern provinces. Presumed anti-predatory adaptations (shell thickness, spinosity, and presence of a tumid ridge--a pronounced swelling on the shell’s canal) were characterized for 700 specimens from 67 localities. The tumid ridge, spine length, and mean thickness all increase in the southern province in concert with the appearance of powerful durophagous predators, and also vary among populations within provinces. These data support the first prediction.

To test prediction 2, correlations between the frequency of shell repair (an index of predation intensity) and the proportion of individuals possessing a tumid ridge were examined in local populations within the southern province. The abundance of these features varies within localities, and if Thompson’s prediction is met, there should be no correlation between repair frequency and trait abundance. Preliminary data suggest that the proportion of individuals possessing anti-predatory adaptations (i.e., tumid ridge) increases with higher predation intensity, suggesting that traits may be well-matched in several populations. This result does not support the second prediction.

Our results suggest that not all of the predictions of the GMTC may scale up into the fossil record. More studies are needed to address the role of time-averaging (and other taphonomic biases) in tests of the GMTC in the fossil record.