ECOLOGICAL ORDINATION PATTERNS AND ECOLOGICAL TURNOVER GRADIENTS IN MOLLUSK ASSOCIATIONS FROM THE QUATERNARY SEDIMENTARY SUCCESSIONS OF THE PO PLAIN, ITALY

Extensive core sampling of the Quaternary sedimentary successions of Po Plain produced a large quantitative dataset suitable for integrated analyses of paleobiological and sequence stratigraphic patterns. Building on previous research, we explore here changes in ecological (rather than taxonomic) composition of mollusk associations across the late Pleistocene and Holocene transgressive-regressive cycles. The Po Plain dataset includes 611 samples extracted from a network of 16 cores. The data include 131,780 specimens (234 species) of bivalves, gastropods, and scaphopods. Species abundance was converted to ecotype abundance. All species were classified into four types of ecological categories: Tiering, Mobility, Fixation, and Feeding. Within each of these four categories, multiple subcategories were used to subdivide species into finer ecotypes. At the finest ecological scale, the species present in the data encompassed 67 unique eco-categorical combinations (but over half of these combinations were represented by only one species). When grouped into coarser ecological categories (infaunal vs. epifaunal; mobile vs. non-mobile; etc.), the data was reduced to 11 unique combinations. The two resulting datasets were explored using Non-Metric Multidimensional Scaling (NMDS) and beta turnover gradient analysis using Bray-Curtis Similarity measure.

When the finely resolved ecotypes are used (67 variables), the resulting NMDS ordination and beta turnover gradients are nearly identical to the previous results obtained for genus and species-level taxonomic datasets indicating a tight correlation between water depth and faunal composition/turnover. This is most likely the case because most of the 67 combinations are represented by one or two abundant taxa, and thus the ecotype dataset is largely redundant with the taxonomic dataset. When these ecological categories are consolidated into coarser groupings, there is little relation between bathymetry and faunal composition and no pronounced gradient in beta turnover. The analyses based on eco-categorical specimen counts appear to be highly sensitive to the choice of categorization, with fine-scale eco-categories being largely repetitious with taxonomic data and coarse-scale categories offering a limited resolution in ordination analyses.