DISTINGUISHING BETWEEN CLOSELY RELATED SPECIES WITHIN A LINEAGE: GEOMETRIC MORPHOMETRICS OF MULINIA OF THE ATLANTIC COASTAL PLAIN, U.S.A

Taxonomic uncertainties and difficulties in interpreting the Mulinia lineage are well recognized in the literature despite the ubiquitous presence of this mactrid in the Neogene of the Atlantic Coastal Plain and elsewhere. Initial species identification, dating back to late 1800’s, was based on ambiguous qualitative shell characteristics and led to frequent misidentifications of fossil and recent material. M. congesta, though highly abundant, is restricted to Pliocene formations such as the Yorktown and its age equivalents. M. lateralis first appears in the lower Pleistocene James River City Formation and its equivalents and is proposed to be the descendant species of M. congesta based on similar geographic coverage and population dynamics. Current diagnostic characteristics used to distinguish between the two Mulinia species remain qualitative, mainly shell thickness and location in the stratigraphic section. There are few remarks within the literature that comment on a possible shape difference between the two species, though rigorous quantitative studies have been lacking.

Preliminary analysis of 388 valves reveals that differences between the two Mulinia species are significant enough to distinguish them from one another based on 15 geometric landmarks using the Procrustes method. Discriminant analysis yielded classificatory errors rates of 3% when individual specimens were reclassified into a priori groups based on average shape values of each species. Using M. congesta as a reference, thin-plate spline analysis suggests the umbo of M. lateralis moved towards the anterior of the valve, the pallial sinus is closer to the posterior of the valve and both muscle scars moved into the interior of the valve slightly. In addition to these allometry-free shape differences, the two species differ in terms of size (á=0.05, p <0.0001, Wilcoxon), with M. congesta having greater size variability and overall size than M. lateralis . These results imply that quantitative traits can be used successfully to distinguish Mulinia species, and support the taxonomic validity of the two species.