The temporally consistent morphometric distinction between Pliocene and post-Pliocene specimens suggests that post-Pliocene specimens are assignable to a different species, or represent significant intraspecific geographic variation (Roopnarine 1995). Recognition of pseudosurvivor taxa such as C. cancellata is an important dimension in the study of extinction in the fossil record. Extinction intensities could be underestimated if pseudosurvivor taxa are not recognized. Such taxa may originate as cladogenetic descendants of the species that suffered extinction, or by the immigration of morphologically similar taxa into the region affected by the extinction.
Vermeij suggested that the process of escalation may be more evident in cases of the geographic replacement of a lineage that suffers extinction or habitat restriction, compared to cases of within-lineage evolution (Vermeij 1987). This study demonstrates both situations. After the extinctions, at the end of the Caloosahatchee Fm., there is a significant and dramatic decrease in the stereotypy of prey size selection by naticids of any given size (Fig. 9) in the Lower Pleistocene Bermont Fm. This could indicate that the Chione cancellata that were replacing the now extinct C. erosa were more escalated than their C. erosa predecessors, thereby forcing larger predators to select smaller, less escalated individuals. However, this interpretation can be refuted by three lines of evidence: (1) The most obvious anti-predatory trait on the shells of the two species is the elaborate concentric sculpture. While the sculpture probably functions to stabilize shells in sediment (Stanley 1970), it may also serve as an impediment to successful drilling (Vermeij 1980; Kitchell et al. 1981). Chione erosa however possessed much more elaborate and extensive sculpture than C. cancellata. (2) While body size does not vary significantly among our samples, C. erosa attained larger body sizes in general than did post-Pliocene C. cancellata (Roopnarine 1996). (3) Examination of the ontogenetic development of valve thickness in both species reveals that there was no significant difference between thickness of Pliocene Chione valves and those found in the Bermont Fm. (Fig. 10). Naticid predators therefore, were definitely not presented with a more escalated prey species when C. cancellata replaced C. erosa in the Early Pleistocene. Interestingly, absolute valve thickness of C. cancellata increased significantly through the Pleistocene and into the Recent, as shown by the relationship between valve thickness and size (Fig. 10). The reason for the increase of valve thickness is unknown, but could be the result of environmental factors, or escalation in response to predation. A number of environmental factors are known to influence shell size in bivalves, including water temperature. Warmer temperatures promote shell deposition, perhaps making it easier to mineralize a thicker shell, and coastal waters off Florida may have been slightly cooler during the Pliocene than they are today because of seasonal upwelling (Allmon et al. 1996). The subsequent post-Pliocene warming might therefore have made it possible for C. cancellata to increase valve thickness, but drilling mediated selection would ultimately reinforce the trend by favouring thicker shells.
There is a distinct possibility that some of these pattern can be explained by turnover in the naticid fauna itself. However, of the eight species of naticids known from the Floridian Pliocene, five are extant today. The Recent naticid fauna comprises approximately 14 species. In their compilation of post-Pliocene gastropod immigration to the tropical western Atlantic, Vermeij and Rosenberg found no evidence of naticid participation (Vermeij and Rosenberg 1994).
We hypothesize that the reduction in the stereotypy of prey size selection after the Plio-Pleistocene transition is a result of the different morphologies of C. erosa and C. cancellata interacting with naticid prey-handling behaviours. A stereotyped, unchanging handling behaviour may present novel challenges to the predator when faced with new prey, or the naticids may have altered their behaviours to cope with the new prey. The reduction of stereotypy was not the result of a new, more escalated prey taxon. Nevertheless, by the middle of the Pleistocene (Ft. Thompson Fm.), predation recovered somewhat, and selected prey size increased. But the mean location of drill holes also changed, suggesting an adjustment of predatory behaviour, or the appearance of new naticid species. These changes in size selection persist to the Recent. From the Middle Pleistocene on, specimens of C. cancellata have significantly thicker valves than either their conspecific ancestors or C. erosa, apparently escalating their defenses as a result of naticid drilling predation and the biogeographic effects of a regional mass extinction.
The reduction in the variation of drill hole location in the Pleistocene compared to the pre-extinction Pliocene does not seem to support Kelly and Hansen's prediction of a reduction in stereotypy, while prey size selection indicates the converse. These observations suggest that the reaction of predator-prey systems to extinction and replacement of prey species is more likely to be a mosaic of predatory responses, rather than a single and overall reduction of stereotypy. Moreover, because of the morphological similarity between the prey taxa, we may not expect the Kelley-Hansen hypothesis to necessarily be relevant in situations of geographic replacement. This further supports the need for careful morphological studies when assessing the ecological and evolutionary consequences of extinctions.