In this preliminary paleoecological analysis, the abundance of one bivalve family, the arcids, exhibited little difference between two samples while another bivalve family, the pectinids, exhibited great variability in dominance. However, the presence and absence of species within the arcids and their individual abundances did vary greatly between samples. Among pectinids, the species presence/absence data alone would not have differentiated the two samples but the species abundances would have been informative. Thus, these different data sets may provide non-overlapping lines of evidence for subtle differences in paleocommunity structure. In this case, the differences in species compositions and abundances between samples in stratigraphic proximity may not have clear significance in terms of the reconstruction of food webs or paleoecological roles but the differences may suggest subtle paleoenvironmental or paleoecological conditions useful for reconstructing the context for evolutionary change. In addition, the paleocommunity analysis allows for the identification of species that may be sensitive to ecological changes over short time intervals and thus may warrant additional evolutionary analysis. Other proxies examined in Dominican Republic Rio Gurabo samples have suggested similar paleoecological differences between other closely spaced stratigraphic samples (Tang et al. 2003; Chan et al. 2003).

Our preliminary paleocommunity analysis suggested that "P." thompsoni may be a good candidate for looking at morphological differences, which may occur when paleoecological conditions change significantly. In the stratigraphically lower sample, "P." thomsponi comprised only 6.2% of the total bivalves present whereas in the upper sample, it comprised 50.7% of the total bivalves counted. This difference suggests a significant change in ecological conditions as well as a significant change in the ecological role for "P." thompsoni, making this species a good candidate for investigating whether there are measurable morphological responses correlated with these paleoecological changes.

The results indicate that there is a significant difference in morphological patterns between the two samples. Although there is no size difference between the two populations of "P." thompsoni, detailed landmark analysis indicates marked differences of shape between the samples and more surprisingly, a change in growth patterns as well. In the lower interval, we documented a clear difference between smaller and larger individuals whereas in the upper sample, there is no allometric growth signal.

Whereas it is impossible without further study to determine the selection pressures that may have led to the different morphologies of these two "P." thompsoni populations, the shape changes are consistent with biomechanical interpretations of the paleoenvironmental conditions of the two samples. The analysis suggests that both the shape difference and the allometry are related to morphological landmarks surrounding the byssal notch. Bivalves pass bundles of byssal threads through the byssal notch to attach to hard substrates, and these morphological structures are related to life habits (such as swimming in pectinids), water energy, and even substrate composition (Waite et al. 2002). Based on sedimentary information, it does appear that sample 052500-1C—which is an amalgamated shell bed—would have experienced higher energy levels than the other sample. This interpretation is consistent with the hypothesis that the byssal notch is wider in 052500-1C so that individuals have a larger bundle of byssal threads as holdfasts attached to the substrate.

This difference in hydraulic energy regimes between the two paleoenvironments may also explain the significant change in growth patterns exhibited by the two populations. While it is difficult to differentiate between genetically based differences and ecophenotypic variation in fossil populations, the difference in growth patterns strongly suggests a genetic or developmental change. In pectinids, byssal threads are usually lost during ontogeny so that adults are free-lying on the sea floor, although given "P." thompsoni's small size, the byssus might have been retained throughout life. The fact that there are differences in the byssal notch between small and large individuals in sample 052500-1C suggests that there are size- or age-dependent factors reflecting a greater need for physical attachment to a hard substrate at larger size. Perhaps larger individuals in this sample with higher water energy were subjected to disproportionately greater hydrodynamic forces than in the other population and hence required relatively denser byssal bundles, with greater surface areas of attachment (Vogel 1988). Also suggestive is the fact that this species was much less abundant in 052500-1C where hydraulic energy is inferred to be higher.

As this study has shown, morphological change need not be interpreted solely as a species-level phenomenon, but can and should be considered in a paleocommunity context. The geographic mosaic theory of co-evolution suggests that microevolutionary variation is influenced by local conditions and biotic interactions, and the Dominican Republic material represents a prime opportunity to test this assertion. A major direction of further research should therefore be the characterization of morphological evolution in multiple co-occurring lineages, for example among pectinids and arcids, coupled with increased study of the potential direct and indirect biotic interactions among those lineages. The use of both quantitative morphometric techniques to evaluate population-level differences and quantitative paleoecological techniques to evaluate community-level differences can provide a powerful approach for characterizing these linkages within the context of evolutionary paleoecology.