Time-averaging mixes species of non-contemporaneous living assemblages, so paleocommunities do not directly record life assemblages. However, if the rate of stratigraphic accumulation is comparable to or greater than the rate of (environmentally mediated) biological change, the gradation from one environment to another becomes geologically sensible. To test for the presence of gradational change therefore requires sampling at a high level of stratigraphic resolution. The well-exposed and geographically extensive conformable transition from silty sands to muddy silts in the middle sequence of the Yorktown Formation (Pliocene) is ideal for such testing.

Replicate samples taken at 30 cm intervals at various localities yielded 140 samples containing 145 species of mollusks. The samples were analyzed using cluster analyses, factor analyses, ANOVA, and MANOVA. The factor analysis revealed two well-formed groupings (dominated by Chama congregata in sand, and Mulinia congesta in mud), with a well-defined transitional gradient between them. This gradient is not seen in sections from the landward edge of the deposits, where a lower rate of stratigraphic accumulation may have eliminated the evidence of a gradient.

While each stratigraphic section records the same basic directional changes in paleocommunities through time, a test of the species rank-abundance between collections within the paleocommunity groupings shows very little correlation. Therefore, the paleocommunity gradient produced by environmental change does not reveal evidence that biological interactions control the changes in paleocommunity structure through time.