FAUNAL GRADIENT ANALYSIS OF THIRD-ORDER CYCLES IN THE MIDDLE DEVONIAN HAMILTON GROUP IN NEW YORK STATE: A TEST OF FAUNAL RECURRENCE

Fossil assemblages from the Middle Devonian (Givetian) Hamilton Group of the Appalachian Basin provide numerous examples of faunal replacement within depositional cycles. The inference of faunal tracking relied largely on semi-quantitative documentation of recurrent associations (or “communities”), but this approach is not fully satisfactory, as many fossil species occur in several associations and biofacies appear to form transitional gradients rather than discrete groupings. In the present study, the ordination technique of Detrended Correspondence Analysis (DCA) was used to quantify and compare gradational biotic transitions in two successive third-order (~1.5-million year) sequences in the Hamilton Group, the bases of the Skaneateles and Ludlowville formations. Data consisted of bed-by-bed records of relative abundance of invertebrate fossil taxa across vertical facies transitions in both regressive-transgressive cycles. Samples from dysoxic, dark shale facies in the early highstand portions of the sequences, that uniformly yielded high scores on DCA axis 1 (DC1), were dominated by a few species of thin-shelled brachiopods and small bivalves. Fossiliferous samples from argillaceous limestones and calcareous siltstone/sandstone samples from the shallow, early transgressive portions of the sequences scored low on axis 1; these assemblages consisted mainly of large, thick-shelled brachiopods and corals. Samples of diverse brachiopod- and bivalve-dominated assemblages from mudstones, transitional between these end members, yielded intermediate DC1 scores. Hence, as in previous studies, DCA axis 1 is considered to represent a faunal gradient related to bathymetric factors. Plots of DC1 vs. stratigraphic position for each cycle display sub-symmetrical patterns, suggesting spatial shifting of faunal gradients in response to relative sea level change. Assemblages from analogous portions of the two cycles yielded similar DC1 scores indicating relative stability of depth-related gradients at the scales of 10⁶ years. Moreover, similarities between the two profiles suggest similar responses of many organisms to sea level change through time, an inference consistent with faunal tracking