View of Upper and Lower High Falls from Above

Photograph by Tom Whiteley





Biostratigraphy is a sub-discipline of sedimentary geology that relies on the physical zonation of biota, both in time and space, in order to establish the relative stratigraphic position (i.e. older, younger, same age) of sedimentary rocks between different geographic localities. Although the basic rules of biostratigraphic zonation were established in the late 18th to early 19th centuries in Europe (ultimately resulting in the development of the Relative Geologic Time Scale), the implementation of biostratigraphic techniques was in use in the United States during the early to mid-1800's.

Some of the first geological surveys to be completed in the United States included those of the New York State Geological Survey. These surveys focused not only on New York's geological resources, but also emphasized the establishment of spatial and temporal relationships of stratigraphic units based on both lithologic and paleontologic composition. By the mid-1800's the New York Surveys had resulted in the development of a relative stratigraphic zonation based primarily on fossil distribution. New York localities are world famous for Cambrian through Devonian strata and fossils, but of particular importance to this website discussion is the contribution of the Ordovician rocks of central New York State to the establishment of a North American focused biochronology. The rocks found in the central New York Mohawk River Region, by definition of their fossil content, are now established as belonging to the Mohawkian Series of the Upper Ordovician Period.

The following material focuses on key fossil taxa present in the Trenton Limestone, their distribution or occurrence within the overall succession of Upper Ordovician strata, and their role in the establishment of the Upper Ordovician time scale. >>Back to Top


The goal of biostratigraphy is to use fossil occurrences within the rock record to establish correlations between time-equivalent rock strata as determined by the presence of a particular fossil species. Although the concept is generally straightforward, i.e. the presence of a specific fossil species in two geographic localities indicates the rocks containing the fossil specimens were deposited at about the same time, in practice biostratigraphic studies tend to be complex. The complexities of biostratigraphy result from aspects of the biology of the organisms including their environmental range, their evolutionary rates, as well as their tendency for preservation and probability of observation by the biostratigrapher.

Ultimately, the most rapidly evolving or short-lived, yet wide-ranging fossil taxa make the best biostratigraphic markers for correlation. If a given taxa is both wide-ranging and evolutionarily short-lived, and if it is robust enough to be preserved in the fossil record, then the taxa is often referred to as an index fossil. An index fossil identified in the rock record would constrain the age of the rock within which it is contained to a very specific interval of time when the organism lived.

Fossil taxa used in biochronologic investigations rarely satisfy all aspects of the ideal index fossil. That is, they often violate one or more of the following rules: 1), must have a widespread distribution (fossils tend to be limited to a small region or are found only in a particular depositional environment as opposed to globally); 2), must show rapid evolution (fossils change rapidly in preservable morphology so that distinctive identifiable species are easily recognized); 3), must be present in substantial numbers (so that fossils can be observed by the biostratigrapher); and 4), fossils should be robust mineralogically (so that depositional and diagenetic processes do not remove the fossils from the rock record).

Most often the best biostratigraphic markers or index fossils are taxa that live in the open water column either as free-floating plankton or as actively swimming nekton. Such organisms tend to be rapidly evolving, widely distributed and widely deposited. In contrast, benthic organisms which live on or very close to the seabottom, tend to be less widespread, fewer in numbers, and are typically found only in particular environments. Nonetheless, nektonic, planktonic, and benthic forms can be used to establish relative biostratigraphic age zonations. >>Back to Top


Throughout the majority of the Paleozoic Era, including the Ordovician Period, the most important fossil index taxa used by biostratigraphers belong to two main groups: graptolites and conodonts. Both of these, which are now extinct, are believed to have lived in open ocean settings as plankton (graptolites) and nekton (conodonts). These two taxa generally satisfy the qualifications for chronostratigraphic indexing. That is, their occurrence in the geologic record generally show evidence for: widespread distribution, high rates of evolutionary change, abundance in the fossil record, and fairly robust mineralogies. Both graptolites and conodonts are very useful for establishing the relative age of many Paleozoic rocks including those of the Trenton Group. In fact, some of the first biostratigraphic studies of the Trenton Limestones and equivalent Utica Group black shales led to the recognition and establishment of the North American Time Scale for Upper Ordovician time, based partly on graptolite biozonation.

The diagram to the below, modified from Holland (2003), represents a compendium of chronostratigraphic data for the Middle to Upper Ordovician of the eastern United States. In this diagram, the distribution of key index taxa including the graptolites and conodonts are shown relative to sequence stratigraphic interpretations, absolute age, and time-rock nomenclature (series and stages). Note that in most cases, the boundaries of graptolite and conodont biozones do not correspond directly with time-rock series and stage boundaries. This is a reflection of the practical difference between using the biostratigraphic distribution of fossils for chronology versus time-rock classification schemes which are based often on the lithologic expression of a rock succession at a given set of geographically constrained outcrop exposures. In addition to biostratigraphic and sequence stratigraphic zonation, also diagramed is the relative scale provided by the composite standard section (CSS). This represents a composite stratigraphic section for all of the Middle to Upper Ordovician rocks of central and eastern North America based on graphic correlations techniques (see Sweet, 1984 for more information). The diagram clearly shows the relative position of key graptolite and conodont species within each representative time sub-unit of the Middle to Upper Ordovician. In the rollover image, the key graptolite and conodont taxa present in the Trenton Limestone are in white. >>Back to Top


Although the diagram above shows all the chronostratigraphic information for the Middle to Upper Ordovician of central to eastern North America, the key interval of importance to the discussion of the Trenton Group here is that of the upper Mohawkian Series (roughly equivalent to the middle Caradocian Series of the European System). The Trenton Limestone represents a single lithostratigraphic group, but it is clearly composed of a variety of carbonate rock types, and lithostratigraphic sub-units deposited during a relatively long period of time (estimated at approximately 4-5 million years). Moreover as clearly shown above, the upper Mohawkian Series is subdivided into three stages (Ro.=Rocklandian, Ki.= Kirkfieldian, and Shermanian) based on biostratigraphic zonation. >>Back to Top

Graptolite and Conodont Biostratigraphy

The roll-over diagram above highlights key taxa useful for biostratigraphic studies of the Upper Mohawkian Series. Within the Trenton Group, the key graptolite taxa used for dating are Diplograptus multidens, Corynoides americanus, Orthograptus ruedemanni, and Climacograptus spiniferus. The key conodont taxa used are Phragmodus undatus, Phragmodus tenuis, Belodina confluens, Amorphognathus tvaerensis, and Amorphognathus superbus (Sweet, 1984; Mitchell and Bergström, 1991; Brett and Baird, 2002). The following two diagrams illustrate the use of graptolite and conodont biostratigraphic zonations from two different studies.

In the case of the first study, shown in the diagram below, Mitchell and Bergström (1991) establish correlations between distant geographic localities using principles of biostratigraphic zonation. Mitchell and Bergström used the position of graptolite zonal boundaries in the type sections of the Trenton Limestone (upper Mohawkian) and the type sections of the Kope Formation (lower Cincinnatian) to establish direct correlations between these localities. In this figure the position of key


































Upper Ordovician Chronostratigraphic Time Scale

Modified after Holland, (2003)

Original data compiled from:

Sweet, 1971; Bergström 1971; Sweet 1984; Holland & Patzkowsky 1998


Biostratigraphic Relationships Between the Cincinnatian and Underlying Mohawkian Series Between New York State and Ohio

Modified after Mitchell and Bergström, 1991


graptolite zonal boundaries are shown (i.e. O. ruedemanni, C. spiniferus, and G. pygmaeus) in relation to major lithostratigraphic divisions. In the case of the Trenton, only the middle and upper Trenton are shown and are referred to as the Denmark Ls. and the Cobourg Ls. (Kay, 1943). Using today's lithostratigraphic nomenclature, the Denmark Ls. would refer primarily to the Denley and Lower Rust Formations and the Cobourg would refer to the Upper Rust and Steuben Formations.

This is an excellent example showing the use of graptolite biozonation for biostratigraphic correlation, but unfortunately the distribution of graptolites within the carbonate-dominated Trenton Limestone and its equivalents is rather poorly constrained for lower stratigraphic intervals. The lack of preservation of graptolites in the shallow, well-oxygenated depositional settings of much of the Trenton and Lexington limestones (shown above) prevents more accurate biostratigraphic assessments within these strata. Therefore, further discussion of graptolite biozonation is deferred herein.

In addition to the graptolite biozones illustrated in the diagram above, Mitchell and Bergström (1991) indicate the approximate position of the Amorphognathus tvaerensis / Amorphognathus superbus conodont biozone boundary. This boundary was later found to correlate with the base of the base of the Poland Formation at Rathbun Brook, with a level in the middle of the C. americanus Zone, and well below the base of the Dolgeville (Mitchell et. al, 1994). The above diagram by Holland (2003) shows that the position of this conodont biozone boundary occurs within the Shermanian Stage and within the M6 sequence. Besides

    the study by Mitchell and Bergström (1991), earlier studies on conodont biostratigraphy by Sweet (1984) helped to elucidate the relative distribution of conodont forms for rocks across the United States. Sweet plotted the stratigraphic distribution of conodont forms in 61 individual geographic localities from the United States (similar in style to the figure to the right for the Black River to Trenton Groups of New York and Ontario), and created a 477 meter-thick composite stratigraphic section (CSS) using graphic correlation techniques. By plotting the distribution or range of conodont species against a vertical stratigraphic section  

for each geographic locality, Sweet generated a composite stratigraphic section for all localities combined. He thereby provided the total stratigraphic range of all species found in the Middle to Late Ordovician. Sweet then divided the CSS into 80 individual 6 meter-thick units, and into a series of chronozones based on the first appearance of certain conodont species. As a result of his approach, Sweet (1984) was able to construct a fairly high-resolution chronostratigraphic framework for the Mohawkian and Cincinnatian Series.

Sweet in 1971 recognized more than 13 individual "faunal units" but only 3 of these (faunal units 8, 9, and 10) apply to the Trenton limestone, and correspond to the P. undatus, P. tenuis, and B. confluens conodont-based chronozones established in the graphic correlation study. The following diagram modified from Sweet's graphic correlation study, highlights four of the conodont-based chronozones and their correlation across the United States from Nevada through eastern New York. Shaded in pink, green, gellow and purple are the chronozones that bracket the Trenton Limestone interval. Although more chronostratigraphic details and correlations are now established that modify Sweet's biostratigraphy slightly, his technique allowed the development of stage level chronostratigraphic correlations for these regions.


Range and Relative Abundance of Conodonts from the Black River and Trenton Groups of New York State and Ontario, Canada

Modified after Sweet, 1984



A major difference between the Mitchell and Bergström (1991), and Sweet (1984) studies is the use of two different conodont biochronologies. In the case of Mitchell and Bergström (1991), the A. tvaerensis and A. superbus conodont biozonation is established from studies of faunas found in the north Atlantic, while that of Sweet (1984) is focused on faunas from the midcontinent (see figure by Holland above). The different conodont biostratigraphic zonations result from the lack of north Atlantic conodonts in the majority of the midcontinental to eastern United States and vice versa. Moreover, both conodont biostratigraphic zonation schemes are required in order to establish concise inter-regional correlations. Unfortunately, in most cases the two faunas do not overlap in their geographic ranges, resulting in rather poor biostratigraphic constraints between north Atlantic and central and eastern United States conodont faunal provinces. In the case of the type Mohawkian interval in central New York, the conodont faunas are generally allied with those of the north Atlantic as is reflected in the figure by Mitchell and Bergström (1991). However, Sweet's (1984) graphic correlation integrated from midcontinent faunas found in sections from central New York and southern Ontario has helped to establish an overlap between the two conodont provinces (as is shown in the Holland figure above).

Much of the Middle to Upper Ordovician series level chronology is based on graptolite and conodont biostratigraphy, but at higher resolutions (i.e. stage level resolution), the recognition of graptolite and conodont biozonation becomes problematic. Often other tools, such as graphic correlation, must be used to provide better control. In practice, most index fossils have some limitations in their application. In the case of graptolites and conodonts, they also have limitations based on their preservation potential (poor graptolite preservation in well-oxygenated, high-energy facies), on their abundance (conodont elements are very small and tend to be very sparse except where concentrated due to sedimentary condensation processes), and their geographic range. Moreover, individual species from both taxonomic groups show relatively low turn-over rates, and have fairly long durations. While graptolites and conodonts are excellent for establishing series level correlations, their use at higher-resolutions including sub-stage level analysis may require the use of additional tools, such as graphic correlation. >>Back to Top

Macrofaunal Biostratigraphy: Concepts of "BioZones"

A variety of benthic taxa (brachiopods, bryozoans, molluscs, trilobites, and echinoderms) have been used historically to define individual Trenton sub-units throughout much of New York State and Ontario. As mentioned in discussions


Conodont Biostratigraphic Correlations for the Upper Ordovician of the U.S.

Modified after Sweet, 1984


in the "Cast of Geologists" section of this website, early emphasis was placed on the description and classification of the numerous fossil taxa from Trenton Falls and equivalent localities. Many species were described, but not until the end of the 19th century had much progress been made toward elucidating a biologic zonation within the Trenton Limestone. From the turn of the century and continuing until the mid 1900's, researchers such as White (1896); Prosser and Cummings (1896); Raymond (1903); and Kay (1937; 1943; 1968) began to investigate and report on the biostratigraphic zonation of the type Trenton Limestone at Trenton Falls. Based on these studies, research primarily by Marshall Kay, helped to develop a much higher series of both litho, and biostratigraphic terms for use in the sub-division and correlation of the Trenton Limestone.

There has been much confusion in the literature as to the proper stratigraphic terminology to use (rock terms vs. time-rock terms) especially outside of the New York/Ontario/Quebec region. The modern lithostratigraphic terminology (from Brett and Baird, 2002) is shown in the figure to the right. This figure indicates the relative position of key faunal "zones" relative to lithostratigraphic intervals.

In the diagram, it is first apparent that the majority of well-characterized faunal zones are at the base of the Trenton Limestone and include those of the Rocklandian, Kirkfieldian, and Shorehamian (early Shermanian). In contrast, much of the middle Trenton, Denley and Rust Formations, is typified by a variety of taxa that are "typical Trenton" species and are more or less common throughout this part of the Trenton Limestone. In this sense, Kay was unable to further divide this interval. Kay (1968) referred to the entire Middle Trenton using the biostratigraphic-based time-rock term: Shermanian. The figure modified from Kay's


1968 paper also illustrates the relative position of time-rock and rock terminologies.

Today, the remainder of the Upper Trenton is considered Shermanian in age. Originally, Kay recognized the development of the Rafinesquina deltoidea zone within the uppermost Trenton Steuben Formation as a unique faunal zone and correlated it with outcrops in Ontario. This uppermost Trenton interval containing Rafinesquina deltoidea and an abundance of gastropod species was delineated by the time-rock term Cobourgian. The Cobourgian aged rocks in Ontario were correlated with the upper Rust and Steuben Limestones at Trenton Falls (as shown in Kay's figure below), as Rafinesquina deltoidea first appears in the upper Rust Formation and expands dramatically in the Steuben.

Based on the occurrence of these faunal zones, Kay (1937; 1943; 1968) established a number of time-rock terms to emphasize the distinction between biostratigraphic and lithostratigraphic terminologies. In the figure below, Kay illustrates the hierarchical system of series, stage, and sub-stage level classifications and the relative position of lithostratigraphic boundaries. Kay (1968) considered the entire Trenton Limestone as belonging to one series which he termed Trentonian (although it was also used as a rock-term), and further recognized the Wildernessian, and Barneveldian intervals as sub-series time-rock terms. Moreover, Kay identified four time-rock stages (Rocklandian, Kirkfieldian, Shermanian, and Cobourgian), and an additional series of sub-stages within the Shermanian (Shorehamian and Denmarkian).

Although the biostratigraphic based stage-level classifications established by Kay work fairly well for correlation in the New York, Ontario, Quebec regions, historically, the application of this complex of time-rock terms outside of the region has been difficult. The development of Kay's high-resolution stratigraphic framework was the result of many years of research and integration of multiple correlation methods including lithostratigraphy, event stratigraphy (K-bentonite correlation), as well as biostratigraphy. Unfortunately, due to facies change, biogeographic provincialism of key taxa, and the variable preservation of correlated




Litho- and Biostratigraphic Terminology for the Trenton Limestone


    ------------------Canada-------------------|--------------------New York---------


Time-Stratigraphic Series, Stages, and Sub-Stages of the Trenton Group

Image modified from Kay (1968)




K-bentonite horizons, the application of the Upper Mohawkian Time Scale of Kay (1968), has been problematic and is as yet unresolved. >>Back to Top

Macrofaunal Biostratigraphy: Coenocorrelation and Gradient Analysis

Concepts of single taxon "biozone" correlation proved to be difficult, especially in the Trenton outside of New York. Some of the most important contributions made by Marshall Kay were toward the delineation of individual lithostratigraphic units in the Trenton and the use of K-bentonites to establish direct time-line correlations between outcrop sections. Moreover in a biostratigraphic sense, the "typical Trenton" faunas although of poor use in biozonation, showed distinctive patterns in abundance. The figure below compiled from data available online at by Professor Ray Gildner (Union College, Schenectady, New York) illustrates the vertical ranges and relative abundance of many taxa of brachiopods, bryozoans, trilobites and molluscs. Although the distribution and relative abundances (indicated by the width of the colored boxes) of individual taxa appear to be random, it is seen that individual

    taxa can be grouped in assemblages of fossils that often occur together and tend to remain in association. Kay often made the observation that "typical" Trenton taxa were rarely found individually or admixed with taxa from another assemblages. These observations were never fully investigated, but Kay attributed these patterns to minor changes in depositional environment and water depth. Based on modern lithologic descriptions, and the development of small-scale cycles within larger-scale depositional sequences, it is quite reasonable to attribute Kay's small-scale changes in fossil assemblages to a variety of environmental changes impacting the ecology of fossil communities. In the late 1970's, John Cisne and colleagues from Cornell University began to quantitatively document the small-scale variations in fossil  

    assmemblages both in vertical dimension (time) and in the lateral dimension (space). By using changes in fossil assemblages in time and space, Cisne and Rabe (1978) applied concepts of coenocorrelation (community correlation) and gradient analysis to generate correlations, and to explain trends in the vertical and lateral changes in community composition as observed by Kay. Cisne et al. (1982) used the position of K-bentonite (volcanic ash) horizons in localities from Trenton Falls eastward into the the Mohawk Valley to constrain individual stratigraphic intervals. Then by collecting fossil presence/absence and abundance data from individual stratigraphic sampling stations, they applied a  

Relative Abundance Chart for Key Taxa from

the Middle Trenton Interval (Shermanian Age)

Image compiled from data by Gildner (2003).

    variety of statistical techniques to compare their data both vertically between sampling stations and along transect downslope into the Taconic Foreland Basin. The diagram to the right, from Cisne et al. (1982), illustrates the use of relative abundance of individual taxa from a given stratigraphic interval (which in this case is located within the lower Denley, their 15 m marker) and the calculated reciprocal averaging score statistic. Along this single time-space transect from Trenton Falls (0 km) to Sprakers, New York (~60 km), the relative abundance values for each taxon observed change sequentially. These authors attribute the gradient from west to east to water depth increase  

from shallow to deep from the Trenton Shelf at Trenton Falls into the Taconic Foreland Basin near Canajoharie and Sprakers.

Using this method, Cisne and Rabe (1978) demonstrated the patterns of community change across the transect during several time intervals as shown in the following figure. In this diagram, three time-specific comparisons were made between four main localities including: Trenton Falls, Rathbun Brook, North Creek, and Dolgeville.



Relative Abundance Chart for Key Taxa from one transect interval (m15) from the Middle Trenton Denley Formation (Shermanian Age) between Trenton Falls and Sprakers, New York

Image modified from Cisne et al. (1982).


    As observed from the diagram here and shown schematically on Gildner's figure above, most species from the middle Trenton at Trenton Falls, although variable in their relative abundance, are generally not present at all sampling stations. Moreover, when considered laterally across the transects, relative abundances of individual taxa are equally variable and show some patchy distributions. Notice the occurrence of two assemblages that show alternation in their relative abundance: Paucicrura / Sowerbyella assemblage and the Sphenothallus / Isotelus assemblage. In the figure, note the two incursions of the Sphenothallus / Isotelus assemblage, once in the vicinity of standard section marker 8 and again around standard section 20. Although both assemblages are present, the incursion (via increase in relative abundances) of the Sphenothallus / Isotelus assemblage again suggests, as mentioned previously, an environmental/ecological change  

favoring the change to the later assemblage. It is clearly suggested by Cisne and Rabe's figure above (1978), that the shifting pattern in relative abundances of these two assemblages is the result of water depth changes. The authors demonstrate the decrease in the reciprocal averaging score statistics both along transects at any one time and between transects during different times. In these examples, Cisne and colleagues have shown that the development of faunal assemblages at any given stratigraphic interval can change in time and space. In their estimation, the change in relative abundance of individual taxa is the result of environmental change resulting from sea-level fluctuation and migration of the Taconic Foreland Basin into the Mohawk Valley region.

Similar to techniques used by Cisne and his colleagues, Ray Gildner has presented data on his website using the premise that if the presence of a fossil taxon in a given stratigraphic layer reflects the preferred living environment for that species, then the relative abundance of the species should be a measure of the environmental conditions optimum for the growth of the organism. Gildner (2003) used another statistical analysis similar to those used in gradient analysis and coenocorrelation. He has applied ordination methodology to try to establish environmental gradients using relative abundances of a variety of taxa. In his estimation, changes in relative abundance should reflect changes in any number of things: temperature, salinity, turbidity, depth, etc.


Relative Abundance Chart for Key Taxa from three transect intervals (m3, m15 & m40) from the Middle Trenton Denley Formation (Shermanian Age) between Trenton Falls and Dolgeville, New York

Image modified from Cisne and Rabe (1978).


Similar to the reciprocal averaging technique used by Cisne, Gildner used a statistical treatment or correspondence analysis called detrended correspondence analysis (DCA), to explain the variation in relative abundances between samples, between time-slices, and between localities. DCA places each sample along a series of constructed axes in order to establish a gradient explaining the largest variation in the data. In the case of the Mohawk Valley relative abundance data, the greatest axis of variation is illustrated here. The most prominent gradient shows values ranging from low (dominantly to the left and bottom) to high (dominantly the the right and top). Using the ordination score, Gildner has color-coded the values to indicate the range of scores, from low (yellow), to high (red). Based on the DCA ordination scores, the shift in faunal assemblages


and relative abundance of each taxa is clearly tied to changes in sea-level. Although Trenton Falls shows the shallowest scores overall, a number of patterns are developed and illustrated in the rollover image above. The most obvious pattern observed using the DCA data is the large-scale emplacement of deep water facies to the east of Trenton Falls. However, upon closer investigation, higher-order patterns are apparent. Even within the Trenton Falls locality, there appears to be a series of deepening and shallowing phases indicated by the faunal relative abundances.

Biostratigraphic studies of the Trenton Limestone vary widely and have been developed for a variety of different purposes. Nonetheless, all studies have helped to elucidate important stratigraphic details surrounding the nature of the Trenton Limestone and its depositional history. >>Back to Top


DCA Ordination Scores

Image modified from Gildner (2003). by Professor Ray Gildner

(Union College, Schenectady, New York)


© 2004 President and Fellows of Harvard College