PCA, correlation analysis, constancy analysis, and MANOVA of my taxonomic data revealed that there local paleocommunities and paleocommunity types in my Yorktown samples, but no paleocommunities in the sense of Bennington and Bambach (1996). However, the occurrence of individual species can be affected by a plethora of paleoenvironmental (e.g., salinity, water depth, light penetration, turbidity, sediment conditions, etc.) and/or ecologic factors (e.g., predation, vegetation, bioturbation, food supply, etc.). The conditions are also frequently dependent (e.g., light penetration and sediment conditions can effect vegetation, which in turn can effect sediment conditions and bioturbation, etc.). Unfortunately, many of these factor are difficult to nail down with geological data, except for evidence from tracking the distributions of individual species. Since the distribution of species is what is being examined in this study, this evidence is of limited used.
The distributions of closely related species indicates that the species level may contain too much noise to clearly define paleocommunities. For instance, Astarte undulata and Astarte concentrica are both found in greatest abundance in the RBPT. However, A. concentrica is common at the Lieutenant's Run locality, and rare at the other localities, while A. undulata is common at all three localities sampled for the RBPT. Whatever conditions allowed for the abundance of A. concentric in LTR samples was apparently not present at the other localities, but what those exact conditions were can not really be determined with absolute certainty. Both A. undulata and A. concentrica undoubtedly occupied similar niches in the RBPT, and so classifying these two species into a single group based on their life history seems justified. Using this same logic, all species present in my data set could be classified into what might be termed "functional ecologic units" or guilds.
Species were assigned to guilds using several lines of evidence. The shell of bivalves contains a wealth of information about the life habit of the original occupant. Using guidelines originally set out in Stanley (1970), the probable food source (suspended organic material or detrital organic matter) and space utilization (e.g., epibyssally attached, shallow infaunal burrower, cementing) were determined. The life habit of other species within the same taxonomic family was then examined to see if they had similar life habits to that determined by the analysis of the shell morphology. Finally, each species was assigned a modern analog, and the life habit of that modern species compared to the life habit derived from morphology and comparison to other members of the family for the ancient species. Many of the species in the Yorktown fauna are still living, and when this was the case, the modern version of the species was assigned to its ancient counterpart. Almost all of the genera of the Yorktown fauna are extant, and extinct species were assigned a modern analog species with a similar morphology in the same genus whenever possible. Some of the genera in the Yorktown fauna are extinct (e.g. Chesapecten, Ecphora). In this case, the modern analog was assigned from a genus with a similar morphology in the same family. The procedure for the gastropods was similar, although the shell of gastropods does not contain the same wealth of life habit information as the bivalve shell, so the first step was not always very useful in determining the guild.
The guild structures of the MBPT and transition zone varied widely from locality to locality. As with the analysis of the taxonomic data, local paleoenvironmental conditions were the strongest controlling factor on the guild structure. There was no consistent guild structure even though members of the same guilds were present at almost all localities. The strong ecological interactions predicted by the ecological locking model were not recognizable in the guild structure of the MBPT and transition zone, and therefore ecological locking can probably be rejected for these Yorktown faunas.
The guild structure of the RBPT recurs with enough similarity to consider it a paleocommunity in the sense of Bennington and Bambach (1996). The species presence and abundance in each guild varied from locality to locality enough that the analysis on the taxonomic data indicated very large differences in the RBPT at different localities. However, the persistence in guild structure indicates that the niche structure was very similar. The availability of many of the niches in this paleocommunity type was controlled by the rubbliness of the substrata, which in turn was controlled by the high abundance of the bivalve Chama congregata. Thus, this appears to have been a biologically mediated system - when environmental conditions changed, and C. congregata was no longer present in high abundance, the community type ceased to exist, and thus the paleocommunity type is not found in strata higher in the section. While at first glance this might appear to be evidence for strong species interactions, and thus ecological locking, this is probably not the case. The importance of C. congregata to the maintenance of the RBPT was that it supplied rubble to the substrate, and since dead C. congregata were just as rubbly as living C. congregata, taphonomic feedback (Kidwell and Jablonski 1983, Kidwell 1986) might have been as important as actual interactions between living species. As long as the shells of dead C. congregata continued to be added to the sediment, this community type continued to exist. Once dead C. congregata shells stopped being supplied to the sediment, and the shells already present were buried, the community type guild structure was no longer viable, and the local communities changed according to the local environmental conditions. Thus, the local paleocommunities of the transition zone differed both from the RBPT and from each other.
The lack of consistent guild structures within the MBPT and transitions zone, plus the lack of a coordinated response in the local paleocommunities to the paleoenvironmental shift from RBPT to transition zone are not the expected observations for a system in which ecological locking was a major force. Therefore, there is no evidence for ecological locking in the guild structure of the Yorktown fauna. As with the analysis of the taxonomic data, local paleoenvironmental conditions appear to be more important than species interactions for determining local paleocommunities with paleocommunity types.
Back