A crinoid fauna and a new species of Pycnocrinus from the Martinsburg Formation (Upper Ordovician), lower Hudson Valley, New York

Abstract

A new crinoid fauna has been discovered in the Upper Ordovician (Katian) Martinsburg Formation at a small shale quarry, locally known as the ‘Shale Bank,’ on the Shawangunk Ridge in Ulster County, NY. The assemblage, which is from a relatively low energy, offshore mud-bottom environment, includes four identified species, including a new species of glyptocrinid camerate, Pycnocrinus mohonkensis n. sp., described herein. Crinoid taxa in order of increasing branch density in the assemblage include (1) the dicyclic inadunate Merocrinus curtus with irregularly isotomous and heterotomous, non-pinnulate arms and a stout cylindrical column exceeding 700 mm; (2) the disparids Cincinnaticrinus varibrachialus, with heterotomous non pinnulate arms, and Ectenocrinus simplex, with extensively branched ramulate arms and meric columns of 460–500 mm; and (3) the camerate Pycnocrinus mohonkensis n. sp., with uniserial pinnulate arms and a somewhat shorter column. Some cylindrical stems with nodose and holomeric columnals are thought to belong to unknown camerate crinoids with pinnulate arms. Filtration theory is used to model food capture in the Martinsburg crinoids. Surprisingly, even densely pinnulate camerates were able to survive in this setting, suggesting that ambient currents attained velocities exceeding 25 cm/sec even in this offshore setting. Similar assemblages were widespread in eastern Laurentia during the Late Ordovician.

UUID: http://zoobank.org/23ca31e8-f572-4520-ba1d-891e3abb950d

Non-technical Summary

The fauna described in this paper was collected from a shale quarry (locally known as the ‘Shale Bank’) on the grounds of Mohonk Mountain House in the lower Hudson Valley, New Paltz, New York. Here we describe the crinoids of the Upper Ordovician (approximately 450 million years old) Martinsburg Formation. The Martinsburg consists of a medium dark-gray shale interbedded with sandstone beds that show graded bedding and cross laminae. The fauna includes a new species of Pycnocrinus, as well as the long-stemmed, small-crowned inadunate crinoids Ectenocrinus, Cincinnaticrinus, and Merocrinus, which are also present in several other Late Ordovician offshore faunas. The modes of life of the crinoids are reconstructed, based particularly on stem lengths and aspects of feeding related to the density of branching in the crinoids’ arms. This fossil occurrence probably represents a relatively low energy, offshore mud bottom environment that was episodically stirred by storm waves and currents.

Introduction

A common observation of Paleozoic crinoid paleoecology is that shallow water faunas are dominated by crinoids with dense pinnulate arms whereas deeper water faunas are mainly populated by animals with relatively open arms (e.g., Holterhoff, 1997, for the entire Paleozoic; Meyer et al., 2002; Brett et al., 2008, for Late Ordovician taxa from the Cincinnati, Ohio, area). The underlying variable is agitation because faster currents are required to force water through more dense filters, whereas slower currents are adequate for crinoids with more open filtration nets (e.g., Baumiller, 1993; Brower, 2007, 2011). However, it has long been recognized that certain pinnulate crinoids are more nearly ubiquitous and may occur abundantly in onshore as well as offshore, low energy settings (see Brett, 1984; Frest et al., 1999). Such is evidently the case with the new camerate species described herein, which occurs as part of an assemblage more dominated by open filtration fan crinoids, including Merocrinus and Cincinnaticrinus. Evidently, many “quiet water” offshore settings still had ambient currents strong enough to allow some dense filtered crinoids to survive. These transitional settings, that were common in the Paleozoic, fostered distinctive mixed assemblages dominated by suspension feeders, such as those described in this paper.

A new crinoid fauna from Martinsburg (Upper Ordovician, Katian, Edenian), from a small quarry on Mohonk Mountain House property on the Shawangunk Ridge (Figs. 1, 2), near New Paltz, Ulster County, eastern New York, includes four identified species of crinoids. In order of increasing filter density, these four crinoid taxa are the cladid Merocrinus curtus (Ulrich, 1879), with irregularly isotomous and heterotomous, non-pinnulate arms; the disparids Cincinnaticrinus varibrachialus Warn and Strimple, 1977 (heterotomous non pinnulate arms), and Ectenocrinus simplex (Hall, 1847), with extensively branched ramulate arms; and the glyptocrinid camerate Pycnocrinus mohonkensis n. sp. (uniserial pinnulate arms). Some cylindrical stems with nodose and holomeric columnals are also present, and these specimens are thought to belong to unknown camerate crinoids with pinnulate arms. These crinoids raise interesting questions as to the paleobiology of crinoids in offshore, mud-bottom environments. In particular, the theory of aerosol filtration provides some important insights into the niche differentiation of these crinoids.

In the present paper we describe the crinoids of the Martinsburg occurrence including a new species of Pycnocrinus. We discuss the paleoautecology of the crinoids and the community paleoecology of the occurrence and finally we compare the Martinsburg occurrence with other similar Late Ordovician crinoid assemblages and make some inferences about the type of environment that promoted this type of assemblage.

Geological setting

The crinoids were collected from the Upper Ordovician (Katian) Martinsburg Formation in the lower mid-Hudson Valley, New York. The precise age of this portion of the Martinsburg is not known; however, based on the macrofauna, it appears to be Edenian (middle Katian) age.

The fossiliferous Martinsburg consists of a medium dark-gray shale interbedded with fine-grained graywacke (litharenites); the sandstone beds exhibit graded bedding, small-scale hummocky cross lamination, and rare oscillation (symmetrical) ripple marks (Epstein and Lyttle, 1987). Gutter casts are common along with linear to sinusoidal horizontal burrow structures found on the silty beds. Together, these sedimentological features suggest deposition under a combination of distal storm waves and currents. Occasional prominent secondary pyrite layers are present and disseminated sphalerite, chalcopyrite, and galena can be found as patches throughout the matrix within the beds with some carbonaceous material (Feldman et al., 2012).

Ordovician plate convergence, which involved Laurentia (proto-North America) and the Taconic island arc, formed a deep basin into which thick sequences of mud and poorly sorted, muddy sands accumulated and eventually were consolidated into the Martinsburg Formation (Feldman et al., 2012). Epstein and Lyttle (2012) consider all the parautochthonous Ordovician siliciclastic sediments in the Wallkill Valley as the Martinsburg Formation of Middle to Upper Ordovician series. However, the parautochthonous sequence west and southwest of Albany, New York, is not contiguous with the rocks of the Wallkill Valley, nor has correlation of the rocks near Albany been done in sufficient detail to warrant using the names established for that area in the Wallkill Valley. The Penn Argyl Member of the Martinsburg is correlative with the shale and graywacke that Epstein and Lyttle (1987) mapped in the western part of the Wallkill Valley that lies unconformably below the Shawangunk Formation. They assigned these rocks to the Mamakating Formation. They report that the Mamakating Formation east of Wurtsboro, New York, consists of thick sequences of medium dark-gray, thin-bedded shale interbedded with very thin- to medium-bedded, very fine- to fine-grained greywacke sandstone (Epstein and Lyttle, 1987).

Paleoenvironments

The Martinsburg Formation (or Mamakating Formation) in eastern New York and Pennsylvania is part of a thick clastic wedge recording the early filling of the Taconic foreland basin. Sources of sediment included uplifted terrane to the east/southeast of New York (Rowley and Kidd, 1981; Kidd et al., 1995; Ettensohn, 2008). The youngest thrusts of the Taconic allochthon (Livingston thrust slice), including mainly shales and graywackes of the Sandbian Normanskill Formation, were emplaced onto parautochthonous Martinsburg/Snake Hill Shale. The adjacent peripheral foreland basin received flysch and wildflysch sediments derived from the incoming thrust slice (Vollmer and Bosworth, 1984; Kidd et al., 1995). Muds and rather immature siliciclastic sands, including recycled older Ordovician sandstones plus some reworked epidote and gneissic fragments, indicate erosion of uplifted basement and ophiolite from within the allochthonous accretionary wedge (Rowley and Kidd, 1981).

The portion of Martinsburg Formation exposed at Mohonk Mountain appears to record a strong pulse of sediment that resulted in encroachment of siliciclastic muds and silts into the interior of Laurentia during late Chatfieldian to middle Katian (Edenian) time. This tectonically related influx resulted in deposition of the relatively deep-water muds of the Utica black shale, Indian Castle Formation in east central New York (Baird and Brett, 2002), and portions of the Pennsylvania Embayment, Point Pleasant Basin, and Sebree trough in western Ohio, Indiana, and Kentucky subsurface. Thicker flysch-type sediments of eastern Pennsylvania include the Penn Argyl Shale. To the west, the Reedsville Formation in Pennsylvania and West Virginia represents shallower water muds and sands that accumulated along the western side of the foredeep. Dark- to medium-gray muds mixed with shelly limestones characterize the Edenian Kope/Clays Ferry Formation of Ohio, SE Indiana, and central Kentucky (Ettensohn, 2008; Brett et al., 2020).

The Mamakating facies of the Martinsburg represents an environment intermediate between deeper flysch-like graptolitic sediments and offshore muddy to sandy shelf sediments of more Reedsville-like facies. In this regard, the sampled beds bear close comparison with the lower portion of a prograded succession of the type Snake Hill Formation (Upper Ordovician, Chatfieldian; O. ruedemanni Zone) at Saratoga Lake documented by English et al. (2006). Those authors divide the 83-meter measured Snake Hill section into three ascending intervals of increasing grain size. (1) Lithofacies association A (0–14 m) consists of dark-gray to black shale with lenticular sublitharenite beds, interpreted as turbidites; the sandstones are slightly bioturbated and yield a moderate fauna of small brachiopods and bivalves. (2) Lithofacies association B (present at14–46 m and 52–83 m) consists of an equal mixture of dark-gray shale and irregularly bedded 10–80 cm sandstones and mudstone pebble debrites (regarded as massive turbidites, or grain flows). This facies association is moderately bioturbated with Teichichnus and Palaeophycus as hypichnia on sandstone bases and a rather diverse shelly fauna of brachiopods, bivalves, gastropods, cephalopods, and echinoderms (both crinoids and edrioasteroids). (3) Lithofacies association C (46–52 m) consists of quartz arenites with trough and herringbone cross-stratification; very few fossils other than fragments of brachiopod shells are present.

Although English et al. (2006) regarded the type Snake Hill succession as a giant transported block within the mélange near the frontal Taconic thrusts (Landing et al., 2003), they also argued that these rocks are parautochthonous, having been transported at most a few kilometers in comparison to the true Taconic allochthon (see also Landing, 1988). These facies are thought to have graded westward into dark, graptolitic shales of the Taconic trough, which, in turn, passed farther westward into calciturbidites and shales of the Dolgeville Formation. These facies graded westward into typical upper middle Trenton Group deeper shelf carbonates (Baird and Brett, 2002). English et al. (2006) further interpreted the Snake Hill facies as a succession of offshore slope muds with turbidites, storm dominated, sandy shelf, and tidally influenced sand bars that record a rare window into a prograding eastern shoreline area immediately adjacent to the Taconic allochthon during Chatfieldian time. The facies observed at the Mohonk Shale Bank are closely similar to lithofacies A, with perhaps a hint of transition in lithofacies B. We suggest that the Mamakating facies represent a synorogenic eastern clastic storm-influenced shelf–slope association analogous to and perhaps contiguous with those facies in the Snake Hill succession.

Much of the shale at Mohonk Mountain “Shale Bank” quarry is barren or contains only sparse graptolites, and the interbedded, graded graywackes have been interpreted as distal turbidites. However, certain interludes of shallowing occur, possibly of eustatic origin, such as those represented by the sampled beds at 12, 24, 28, and 35 feet in the Shale Bank. Portions of the seafloor locally must have been somewhat more elevated to depths where the bottom was intermittently affected by storm-generated waves and currents. The presence of gutter casts and small-scale wave-formed ripples/hummocks in sandstones associated with the fossil assemblages described herein indicate the presence of combined flows associated with tempestites. Distinctive trace fossils (Palaeophycus, Teichichnus) in some layers indicate at least temporary presence of benthic oxygen and life activity in the sediments. The presence of very thin hash beds of crinoid debris also indicates possible pauses in sedimentation that allowed slightly time-averaged skeletal remains to accumulate.

Taphonomy of the best-preserved fossils indicates that they are preserved in situ or with only minor disturbance. The occurrence of completely articulated crinoids at more than one level indicates very rapid burial (Ausich et al., 1999). Long stretches of crinoid columns are curved and bent over one another (see below), suggesting disruption by turbulent events. There is very slight indication of incipient disarticulation and displacement of pluricolumnals in columns. Arms of the crinoids are splayed rather than tightly in-folded, suggesting that the animals did not have sufficient time to react before being uprooted, killed, and buried. Partially articulated trilobite material is suggestive of semi-intact molts.

The seafloor must have been a relatively low energy setting much of the time. The moderate-low diversity assemblages of the Martinsburg (Table 1) suggest somewhat stressed low energy lower oxic conditions. However, the intermittent existence of crinoids, including rare densely pinnulate camerates, indicates that at times ambient current velocities exceeded ~25 cm/sec to permit these organisms to survive as passive suspension feeders (see below). Rare presence of eyeless Cryptolithus trilobites suggests turbid conditions perhaps near the dysphotic–aphotic boundary. This inference accords well with interpretations of Holland et al. (2001) for similar associations in the approximately coeval Kope Formation of the Cincinnati Arch region. Community paleoecology is discussed separately below.

Table 1. Faunal list of species from the Martinsburg “Shale Bank.”

Paleoautecology of Martinsburg crinoids

The niche structure of crinoids is basically two-dimensional (e.g., Ausich, 1980). Elevation relative to the seafloor forms the primary dimension, which is dictated by the length of the stem and the nature of the attachment device in Ordovician crinoids (e.g., Brower, 2011; Table 2, herein).

Table 2. Paleoecological data for suspension feeding crinoids of the Martinsburg Quarry. Detailed discussion is in the text and Brower (2005, 2007, 2008, 2010, 2011). The elevations and attachment types for Merocrinus curtus, Cincinnaticrinus varibrachialus, and Ectenocrinus simplex are taken from Brett et al. (2008) and Brower (2011). The elevations and attachment devices of Pycnocrinus mohonkensis n. sp. are not known. Most species of Pycnocrinus seem to have a distal coiled holdfast, located on or close to the substrate, and stems from about 10 to 400 mm long (Brett et al., 2008; Brower, 2010). The estimates for the food grooves of Merocrinus curtus, Cincinnaticrinus varibrachialus, and Ectenocrinus simplex are derived from Brower (2005, 2007, 2008, 2010, 2011). In camerate crinoids most food particles are caught by the tube feet along the pinnules. The average food groove width: pinnule width for another glyptocrinid is 0.632 (Brower, 1994). This figure is multiplied by the average pinnule width to determine the average food groove width for Pycnocrinus mohonkensis n. sp.

Elevation and attachment of the crinoids

The crinoids at the Martinsburg locality include the single crown of Pycnocrinus mohonkensis n. sp. and several stems and partial crowns of Merocrinus curtus, Ectenocrinus simplex (Hall) and rare Cincinnaticrinus (Table 2). Merocrinus curtus is an unusual cladid crinoid with a small crown and disproportionately long, robust, cylindrical columns up to 800 mm in length. The holdfast is either a cemented lump or possibly, in some, a distal coil (Brett et al., 2008). Ectenocrinus simplex is a well-known and common crinoid in the Walcott–Rust Quarry of New York, the Edenian of the Cincinnati Arch region, and elsewhere (see Brower, 2008, 2011, and references listed there). The holdfast is a lichenocrinid or modified lichenocrinid structure, which is cemented to other crinoid stems in the Walcott–Rust Quarry, but specimens from the midcontinent are attached to other types of shelly material such as brachiopods and echinoderm plates (Brett et al., 2008). Incomplete adult stem segments up to 460 mm in length are known (Brower, 2008, 2011). This length obviously represents a minimum elevation for E. simplex with respect to the seafloor. Likewise, Cincinnaticrinus varibrachialus had a lichenocrinus type holdfast and a relatively long column up to 500 mm long.

Pycnocrinus mohonkensis n. sp. is based on a single complete adult crown with a short stem segment. Several types of holdfasts are known within the genus. Pycnocrinus argutus (Walcott, 1883) (Walcott, 1883, 1884; see Brower, 2010, 2011) bears an open distal stem coil that was probably located on the seafloor or coiled around some soft object, such as a plant or sponge. The stem of P. argutus was probably relatively short. A complete juvenile was only 10–12 mm above the seafloor, but adult stems are longer with incomplete stems ranging 18–24mm long (Brower, 2010, 2011). A related holdfast is present in Pycnocrinus dyeri (Meek, 1872) (see Meek, 1873, p. 32, pl. 2, figs. 2a, b; Wachsmuth and Springer, 1897, p. 271, pl. 20, figs. 1a–c, pl. 21, figs. 3a–f, 6; Brower, 1974, p. 12, fig. 3) from the Maysvillian strata of the Cincinnati, Ohio, region. Stem lengths of P. dyeri from the Cincinnati, Ohio, area are 50–400 mm (Brett et al. (2008). At least one specimen of P. dyeri has a series of coils that are tightly wrapped around the column of another crinoid (Miller, 1880, p. 233, pl. 7, figs. 3b, c, listed as Glyptocrinus shafferi Miller, 1875, which is conspecific with Pycnocrinus dyeri; see Brower, 1974). The Cincinnatian species is commonly restored with its distal stem coiled around a branching bryozoan (Ausich, 1999; Brett et al., 2008). Brett et al. (2008) reported specimens of Pycnocrinus with open distal coils like those of P. argutus. Poorly preserved distal coils of a heteromorphic column were found in the Martinsburg fossil assemblages. These remains suggest that P. mohonkensis n. sp., like most other Pycnocrinus species, had a distally coiled columnar holdfast.

In contrast, the digitate holdfast of P. gerki Kolata, 1986, is attached to a strophomenid brachiopod shell (Kolata, 1986, fig. 3.1; Brower and Kile, 1994, pl. 5, fig. G). Incomplete stem segments in juvenile and adult specimens are 17–100 mm long.

If these data can be extrapolated to the Martinsburg crinoids, they suggest that both species were elevated well above the seafloor with adults of P. mohonkensis n. sp. being somewhat below those of Ectenocrinus simplex. Clearly all crinoids towered above the other suspension and filter feeders in the area, namely the abundant brachiopods, Sowerbyella and orthids, especially Cincinnetina, relatively rare bivalves, and perhaps the conulariids. Scavengers/detritus feeders (“collectors”), including the trilobite Cryptolithus and the ostracodes, presumably lived on the surface and perhaps swam slightly above the seafloor. The infaunal organisms that excavated the burrows probably ate organic detritus and micro-organisms within the sediment.

Feeding

All living crinoids are passive suspension feeders that catch microscopic food particles consisting of plant and animal plankton and organic detritus (see reviews in Meyer, 1982; Messing, 1997; Brower, 2006, 2007, 2011, and references cited therein). The animals are rheophillic and feeding takes place in the presence of gentle to moderate currents. Crinoid filtration fans are generally planar, parabolic, or conical, but may be irregular, and they are held at roughly right angles to the ambient current flow with the food grooves and their tube feet facing down current. The tube feet are in groups of three and most food items are caught by the longest of the three tube feet (here termed food-catching tube feet) by direct interception. The food items for modern crinoids generally range from about 20–150 microns in diameter and consist of a wide variety of food types, including phyto- and zooplankton and organic detritus. When not feeding, the arms of most crinoids are typically closed, and the tube feet are located under the covering plates of the food grooves.

Filtration theory has been used to model food capture in the Martinsburg crinoids (methods, computations, and underlying assumptions are described in detail by Baumiller, 1993, and Brower, 2007, 2011). In a detailed multivariate study of the diverse crinoids of the “Brechin Lagerstatte,” Cole et al. (2018) identified filtration mesh density and mesh area as the two most critical factors in separating different ecological groupings.

The structure of the arms and their filtration net dictate the nature of the food supply and the current velocities for successful feeding. Here, two characters are critical. (1) Tube-foot spacing equals the number of food-catching tube feet per mm along one side of an arm, ramule, or pinnule. This can be calculated for fossil crinoids from the size of the covering plates by drawing analogies with modern crinoids. (2) Width of the food grooves imposes an upper limit to the size of the food particles that can be collected. Two ambient current velocities can be derived from the nature of the filtration net and its food-catching tube feet. The minimum ambient current velocity (MinV) is the lowest velocity at which the crinoid can catch enough food particles to balance its energy needs. The maximum ambient current (MaxV) is the threshold flow velocity at which feeding must cease, because the velocity of water flowing though the filter becomes high enough to generate sufficient hydrodynamic drag on the food-catching tube feet, so they begin to buckle (Woodley, 1980; Brower, 2011).

Feeding habits of the Martinsburg crinoids

The results from the filtration models are summarized in Table 2. The four identified crinoids from the Martinsburg are morphologically and ecologically diverse, and they exhibit a wide range of arm types. From most open to most dense filtration nets, these are open filters with isotomous or irregularly heterotomous arms as in Merocrinus curtus, heterotomous filters with extensively branched or ramulate arms, like Cincinnaticrinus varibrachialus and Ectenocrinus simplex, and the uniserial pinnulate filter of Pycnocrinus mohonkensis n. sp. (Table 2). In general, the four species are separated both by elevations above the seafloor and the sizes of the average and largest food particles taken. The camerate crinoid Pycnocrinus mohonkensis n. sp. was closest to the seafloor, Cincinnaticrinus varibrachialus and Ectenocrinus simplex stood at about 460–500 mm, followed by Merocrinus curtus at 600–800 mm. In general, the largest food items were eaten by the crinoids at the highest level. Of the two species in the middle tier, Cincinnaticrinus varibrachialus probably consumed somewhat larger food material than Ectenocrinus simplex. In addition, the different crinoid species probably fed at somewhat different ranges of current velocities (Baumiller, 1993; Holterhoff, 1997; Brower, 2007, 2011, 2013).

The food grooves and covering plates of Ectenocrinus simplex are preserved in the specimens from the Walcott–Rust Quarry (Brower, 2008, 2011). Those of Pycnocrinus mohonkensis n. sp. are unknown and must be inferred. As in all pinnulate crinoids, most of the food particles must have been caught by the food-catching tube feet on the pinnules whereas those on the arms mainly served to convey food to the mouth (e.g., Brower, 2006, 2007). The average pinnule width is 0.20 mm. The ratio of food-groove width to pinnule width equals 0.632 for a well-preserved glyptocrinid (Brower, 1994), which produces a pinnular food-groove width of about 0.125 mm. The tube-foot spacing and covering plates for P. mohonkensis n. sp. are not known. The average value for four camerates with uniserial arms consists of about 11.5 food-catching tube feet per mm along one side of the pinnule, which corresponds to a food-catching tube foot with length and width of 185 and 29.4 μm, respectively (see Brower, 2007, for method of calculation).

The maximum food particle sizes that could have been collected are 159 μm and 125 μm for E. simplex and P. mohonkensis n. sp., respectively (Table 2). Within the available food in the plankton population, the size frequency distributions of the food particles for both crinoids are displaced toward larger food particles, because the animals are more efficient at catching larger food items than smaller ones. The combination of these two factors dictates that the food size frequency distribution of P. mohonkensis n. sp. is narrower and specialized toward smaller food particles than in E. simplex (see Baumiller, 1993, Brower, 2007, 2011). This pattern is clearly shown by the following means and standard deviations in microns for the plankton population (based on modern plankton measurements; J.C. Brower, unpublished data), and the food particles taken by E. simplex and P. mohonkensis n. sp., respectively: means, plankton size (estimated): 32.4 μm, mean particle size for E. simplex: 75.0 μm, and mean particle size for P. mohonkensis n. sp.: 51.1 μm; standard deviations 33.5, 41.0, and 21.2, respectively. On average, E. simplex consumed larger food items than P. mohonkensis n. sp., but the two crinoids overlapped widely with respect to the smaller food items.

The minimum and maximum current velocities for feeding of P. mohonkensis n. sp. exceed those of E. simplex. Both species could successfully feed at ambient current velocities of 4.2–25 cm/sec. The total range of ambient current velocities for P. mohonkensis n. sp. exceeded that of E. simplex (Table 2; AMNH 162964).

The overall results are consistent with the expectations of filtration theory. The skeleton of the ramulate arms of E. simplex is more open and less extensively branched than the uniserial pinnulate arms of P. mohonkensis n. sp. (compare illustrations of E. simplex with the uniserial pinnulate camerate in Brower, 2006). The morphology of the food-catching tube feet is partially determined by the skeletal elements in the arms of the two taxa based on modern crinoids. The food-catching tube feet of P. mohonkensis n. sp. were shorter, stouter, and more closely spaced than those of E. simplex. The dense filtration fan (including both the skeleton and tube feet) of P. mohonkensis n. sp. was more resistant to fluid flow than the relatively open one of E. simplex; hence, P. mohonkensis n. sp. required a higher ambient current velocity to initiate successful feeding. The presumed short and stout food-catching tube feet of P. mohonkensis n. sp. had a lower aspect ratio than those of E. simplex; the aspect ratios (length:width) of these tube feet equal 6.3 and 10.0, respectively. Consequently, the pycnocrinid tube feet probably would have withstood higher ambient current velocities and greater drag forces before they began to buckle than the more gracile tube feet of E. simplex (Woodley, 1980; Brower, 2011).

Other factors: respiration

It is interesting that contrary to expectations of the aerosol filtration theory the longest-stemmed crinoids are those with very small crowns and non-pinnulate arms. Camerates require higher current velocities to flush seawater efficiently through their dense pinnulate arms. Thus, it might be predicted that those adapted to lower energy, offshore environments, would have had long columns to elevate the crowns into the strongest currents. In contrast, non-pinnulate crinoids with open filtration nets like most disparids would not require higher current velocities and might be expected in the short column group. However, the distribution patterns of the crinoids do not bear out these predictions. In fact, in offshore facies like the Kope and Martinsburg, those camerates that do occur actually have relatively short columns, whereas the small-crowned disparids, which dominated these deeper water settings, have much longer columns that elevated the crowns 0.5–1 m above the substrate.

Observations of crinoids in deeper water settings in the Ordovician–Devonian show that dominant disparids and cladids (e.g., Ectenocrinus, Cincinnaticrinus, homocrinids, anemesocrinids, dendrocrinids, merocrinids) had relatively long columns, even at very small (juvenile?) stages, indicating that long slender columnals were added rapidly early in ontogeny (Brett, 1984). We suspect that the long columns were, in part, an adaptation for elevating the crown rapidly above seafloor. With relatively fewer tube feet, these crinoids would have had limited capacity for oxygen uptake so that elevation above dysoxic substrates would have been a selective advantage. In a similar vein, Gorzelak et al. (2020) noted that an extremely long stem in scyphocrinitid camerate crinoids, which they interpreted as being benthic, was functional not only for more efficient feeding but also for elevating the crown above unfavorable low-oxygen conditions on the bottom.

In contrast, in some distal settings, camerates, with greater surface area of tube feet, could survive under lower oxygen levels closer to the substrate. Evidently, these crinoids still experienced strong enough currents at least some of the time to filter plankton. Alternatively, given the ability to coil distal columns around substrates in many of these small camerates, these may have survived as secondary tier dwellers on taller objects, including on long-stemmed crinoids. Small camerates are known to have attached relatively high up on columns of Eulcalyptocrinites in offshore mudstone facies of the Rochester Shale (Brett and Eckert, 1982).

Community paleoecology

Crinoid assemblages described herein form a part of relatively low diversity assemblages with about 21 species dominated by small, generalized brachiopods, small protobranch bivalves, orthocone nautiloids, conulariids, and rare trilobites, as well as the crinoids described herein (Table 1). This assemblage closely resembles those ascribed to the offshore, deeper, dysoxic setting and yielding a low axis-1 score in gradient analysis (detrended correspondence analysis or DCA) of Holland et al. (2001). In the following sections, we discuss the trophic structure of this assemblage and the possibilities of niche partitioning.

Trophic structure

Turpaeva (1957) listed four common relationships among Arctic and Boreal communities. (1) A community is typically dominated by one trophic group; (2) if the most dominant species in the community belongs to one trophic group, the next most dominant species belongs to a different trophic group (commonly, the third most dominant species belongs to still a third trophic group); (3) among the various species of a community that belong to a given trophic group, a single species commonly dominates the group in terms of biomass; and (4) thus, the several most dominant species in the benthic community use the available food resources more fully than if they fed at a single resource, and feeding competition is minimized. This type of community structuring has been identified in many paleocommunities. For example, Feldman (1980) documented this type of trophic structure in the marine communities of the Middle Devonian Onondaga Limestone.

Feldman (1980) reported on the relative abundance of major invertebrate taxa within brachiopod-dominated communities of the Onondaga Limestone (Devonian, Eifelian) in New York. His paleoecologic conclusions regarding the trophic structure, based on Turpaeva's (1957, as cited in Walker, 1972) analysis of Recent Arctic and Boreal communities, can be applied to the Sowerbyella–Cincinnetina Community from the Martinsburg Formation on the Shawangunk Ridge in the lower mid-Hudson Valley.

The assemblage of the Martinsburg Shale, including the crinoids described herein, is dominated by brachiopods, particularly Sowerbyella sericea (Sowerby, 1839) and Cincinnetina multisecta (Meek, 1873b), which formerly was assigned to Dalmanella or Onniella. Based on these dominant taxa, we term this assemblage the Sowerbyella–Cincinnetina Community. In the Sowerbyella–Cincinnetina Community, one trophic group, brachiopods, dominates at least in terms of biovolume. The next most dominant group, crinoids, is dominated by a different trophic group: high-level passive suspension feeders (Table 3). The next most abundant groups (in order of dominance) are nuculid bivalves (infaunal deposit feeders), trilobites (probably scavenger or soft-prey predators), conulariids (mid-level suspension feeders), cnidarians (high-level micropredators), and cephalopods (predators).

Table 3. Trophic levels of taxa within the Cincinnetina–Sowerbyella Community listed according to relative abundance.

Bretsky (1970) reported a Sowerbyella–Onniella Community from the Upper Reedsville Formation and portions of equivalent Martinsburg strata (Upper Ordovician) in the central Appalachians. The typical locality of the Sowerbyella–Onniella Community is approximately 200 miles southwest of the Shawangunk Martinsburg community. Interestingly, both communities appear to occur in shales in similar offshore positions (see Bretsky, 1970, figs. 24, 25). Bretsky's (1970) faunal constituents, which also occur in the Martinsburg in the lower mid-Hudson Valley, contain the brachiopods Cincinnetina multisecta, which formerly was assigned to Onniella, Sowerbyella sericea, and Rafinesquina “alternata” (Conrad, 1838), the protobranch bivalve Lyrodesma postriatum (Emmons, 1842), trilobites (Isotelus, Cryptolithus), and crinoids (columns). Without abundance data, we can only make a rough comparison with the trophic structure of the Shawangunk Sowerbyella–Cincinnetina Community, but a strong similarity exists.

Lehman and Pope (1989) reported on “Lithofacies 3,” comprising mudstones and shales in the Upper Ordovician Reedsville Formation at Swatara Gap, Pennsylvania, with a fauna that is dominated by infaunal feeders such as Pseudolingula rectilateralis (Emmons, 1842), Colpomya faba intermedia (Ruedemann, 1926), and Cryptolithus bellulus. These taxa represent three different trophic groups. Lehman and Pope's (1989) Lithofacies 1 and 2 (in lenses) contain 12 taxa that represent several trophic groups. Because the authors did not give any relative abundance data, it is difficult to compare their fauna with that of the Martinsburg in terms of dominance.

Comparison of the Martinsburg crinoid fauna with other Ordovician faunas

There are several other Late Ordovician fossil assemblages that show some similarities with that of the Martinsburg in terms of crinoids as well as other taxa. We briefly surveyed these to provide comparative data that may yield insights into paleoecology.

Among the most similar assemblages in terms of crinoids, brachiopods, and trilobites are the uppermost Mohawkian to Edenian assemblages of the Cincinnati Arch region (see Ausich, 1999; Meyer et al., 2002; Brett et al., 2008). Merocrinus curtus, Cincinnaticrinus varibrachialus, and Ectenocrinus simplex occur in the lower part of the Kope Formation (Edenian) in the Cincinnati Ohio area (Meyer et al., 2002), although Merocrinus is largely restricted to the lowermost Fulton Submember (Brett et al., 2008). The pinnulate glyptocrinid Glyptocrinus decadactylus Hall, 1847, is recorded, albeit rarely from the lower part of the Kope Formation by Holland et al. (2001) and scattered ossicles have been reported from the upper Kope sequence by Meyer et al. (2002). Similar trilobites (Cryptolithus) and brachiopods (Sowerbyella, Cincinnetina) are present in the Martinsburg Quarry and the lower Kope (compare Table 1 with Holland et al., 2001). The lower Kope mainly consists of interbedded terrigenous mudstone, calcisiltite, and skeletal packstone (~20% carbonate) that was deposited offshore on a gently inclined ramp below wave base of all but the most severe storms (see Holland et al., 2001; Meyer et al., 2002). Up to 50% of skeletal packstones obtained from the upper Kope Formation and the overlying Fairview Formation are crinoidal packstones, which probably formed in slightly more agitated waters between storm and normal wave base (Holland et al., 2001; Meyer et al., 2002).

Kallmeyer and Ausich (2016) described a relatively deep-water assemblage from the Kope Formation in northern Kentucky. Crinoids present are Glyptocrinus nodosus Kallmeyer and Ausich, 2016 (dense uniserial pinnulate filter), Plicodendrocrinus casei (Meek, 1871) (irregularly isotomous to hetrotomous, non-pinnulate open filter), and Ectenocrinus simplex (Hall, 1847). Although Plicodendrocrinus casei is not known from any of the other faunas, other dendrocrinids with similar arms occur in the Walcott–Rust Quarry with Dendrocrinus gregarius Billings, 1857, and Dendrocrinus sp. from Swatara Gap. However, the other taxa at this locality have little in common with the Martinsburg quarry, the Walcott–Rust Quarry, the other Kope Formation faunas, or the Swatara Gap fossils (compare Table 1 herein with table 1 in Kallmeyer and Ausich, 2016). The Kope occurrences may record a slightly variant local community type. However, associated beds above and below the crinoid level have very similar faunas, including abundant Ectenocrinus and brachiopod assemblages rich in Sowerbyella, Cincinnetina multisecta, and Dalmanella emacerata (Hall, 1860), suggesting that the overall environmental setting was rather similar. Indeed, it is important to note that these communities were not discrete entities, but no doubt formed parts of a gradient. Quite possibly those with glyptocrinids may have been slightly shallower but intergradational with disparid-dominated assemblages.

Another similar fauna is the Cincinnaticrinus varibrachialus–Ectenocrinus simplex assemblage of the Walcott–Rust Quarry in the Spillway Member of the Rust Formation of the Trenton Group (upper Shermanian or Trentonian near the base of the Orthograptus ruedemanni Graptolite Zone) near Trenton Falls, New York (Brower, 2011). Both units contain numerous specimens of Merocrinus curtus, Ectenocrinus simplex, and Cincinnaticrinus varibrachialus. Another species of the pinnulate glyptocrinid Pycnocrinus, P. argutus (Walcott, 1883) (see also Walcott, 1884), occurs in moderate abundance in the Walcott–Rust Quarry. The conulariids, brachiopods, bivalves, and trilobites of the two localities are also similar (compare Table 1 with Brower, 2011). The background sediments in the Rust Quarry are comprised of siliciclastic and carbonate muds, which were buried under carbonate mudflows and/or distal turbidites. The habitat lay in relatively deep water at the base of a carbonate ramp that was below the reach of most storms but still within the photic zone (Brett, 1999; Brett et al., 1999; Brower, 2005).

There are also less-significant similarities with the Swatara Gap fauna of the Reedsville Shale (at least partly equivalent to the Martinsburg) of Pennsylvania (see Lehman and Pope, 1989, for paleoecology and fauna). Based on personal observations, Ectenocrinus simplex, an undescribed species of the pinnulate glyptocrinid Abludoglyptocrinus Kolata, 1982, and dendrocrinids (irregularly isotomous to heterotomous, non-pinnulate open filters) dominate the 10 species of crinoids listed by Lehman and Pope (1989). However, the Swatara Gap crinoids are more diverse (ten species) than those in the Martinsburg quarry with only three or four taxa. Several brachiopods and trilobites are obtained from both localities (compare Table 1 with Lehman and Pope, 1989, table 1). These authors postulated that the Reedsville beds formed rapidly as storm deposited tempestites in moderately distal marine waters that were less than 30 meters deep, within the photic zone.

Finally, the Martinsburg, Kope, and Wolcott–Rust crinoid faunas contrast strongly with the slightly older Katian (Mohawkian, Chatfieldian) “Brechin Lagerstätte” from the Upper Ordovician Simcoe Group of southern Ontario, Canada, that recently was analyzed in detail (Ausich et al., 2018; Cole et al., 2018, 2020; Wright et al., 2019), which has been characterized as the most diverse Ordovician crinoid fauna. The latter shares a few similarities (e.g., association with Sowerbyella and dalmanellid brachiopods). However, it has a much greater diversity of crinoids with approximately 27 genera versus six and four genera for the Kope and Martinsburg faunas, respectively. Ausich et al. (1994) characterized the Early Paleozoic Crinoid Macroevolutionary Fauna as being dominated by disparids and diplobathrid camerates. The Martinsburg, Kope, and Wolcott–Rust faunas are reflective of this grouping, as disparids are the most common forms, in both the Martinsburg and the better-known Cincinnatian (Edenian) crinoid assemblages. Small camerates are present in low abundance in all assemblages, although these are mainly monobathrid glyptocrinids rather than diplobathrans. Glyptocrinus nodosus Kallmeyer and Ausich, 2016, is found very rarely in the Kope Formation, G. decadactylus Hall occurs in large numbers in the Fairview Formation (see Ausich, 1999; Brett et al., 2008). Pycnocrinus dyeri itself is typical of assemblages in the Maysvillian, especially upper Fairview, which record somewhat shallower more diverse communities in the Cincinnati region, while P. mohonkensis n. sp. in the Martinsburg derives from offshore mud-bottom settings and is more analogous in its environment to G. nodosus.

In terms of composition, the diverse Brechin assemblage is an outlier relative to the other, more offshore crinoid associations, being numerically dominated by cladid inadunates, especially the early flexible, Cupulocrinus, and secondarily by camerates, followed by disparids, based upon rank-order abundance. This contrasts with the Martinsburg–Kope–Wolcott–Rust assemblages, which are dominated by the disparids Cincinnaticrinus and Ectenocrinus, neither of which is present in the diverse Brechin crinoid assemblages. The faunas share Merocrinus (present in the Martinsburg and abundant in the lowermost Kope) and Iocrinus in the Kope.

Very different overall faunal diversities are observed in the various temporally related faunas, with at least 18 species in the Martinsburg quarry, 27 at the Kallmeyer and Ausich (2016) locality, 48 in the overall Kope sequence (Holland et al., 2001), 73 species at Swatara Gap (Lehman and Pope, 1989), 75 in the Rust Quarry (Brett, 1999), and more than 130 species (97 genera, including 27 crinoids) in the Kirkfield or Brechin Lagerstätte (faunal list of Liberty, 1969, and Cole et al., 2020). The underlying reasons for the differential diversities remain uncertain, but the sample sizes at the various localities are strongly different and clearly affect the species diversity in as much as far fewer fossils have been found at the Martinsburg quarry than at any of the other localities, whereas the Swatara Gap and Rust assemblages have been studied in detail from many years. The diversity of the Kope recorded in Kallmeyer and Ausich (2016) reflects a partial diversity of the lower Kope; a larger sample taking in a broader spectrum of facies would yield diversities comparable to the Rust Quarry or Martinsburg. However, there are probably also real differences that relate to ecology. One of the most distinctive differences of the Martinsburg occurrence and all others is the lack of bryozoans; in contrast, several species of ramose, bifoliate, and encrusting bryozoans are recorded for the Kope, Swatara Gap, Rust quarry, and Brechin assemblages. The cause of the absent bryozoans is unclear but might be related to the higher sedimentation rates of the more proximal Martinsburg. Other differences are probably related to the varying nature of energy/water-flow conditions that most likely dictate which taxa are dominant at a given locality.

The differences between the high- and low-diversity associations reflect both evolutionary and paleoecological factors. Some of the diverse camerates and inadunates of the early Katian assemblages apparently are at least regionally extirpated in the later Ordovician (Paton et al., 2019). However, differences also reflect disparity of environments. The Brechin Lagerstatte and the Kope probably represent opposite ends of a faunal gradient from high-diversity shallow-subtidal assemblages to deep-subtidal and offshore mud-bottomed assemblages.

The Brechin or Kirkfield fauna records relatively shallow, shallow- to mid-subtidal conditions with overall higher current velocities and low-sedimentation/clean-water conditions with heterogeneous substrates, including local hardgrounds with patches of unconsolidated skeletal sands and muds (Cole et al., 2019, 2020; Paton et al., 2019). Conversely, the Kope–Martinsburg type assemblages probably inhabited deeper, lower energy environments with more monotonous mud to sand substrates. The Wolcott–Rust assemblage probably records an intermediate position with lower energy and somewhat less substrate heterogeneity, but fully oxic bottom conditions and relatively low background sedimentation.

The type Snake Hill Formation (Upper Ordovician, Chatfieldian; O. ruedemanni Zone) exposed at Saratoga Lake also provides an important faunal analog to the Martinsburg. The lower portions of this formation have yielded a diverse fauna (nearly 90 species), including most of the species recorded here from the Martinsburg at Mohonk Mountain. The most comparable strata are their lowest Lithofacies A, which comprises dark-gray to nearly black shales with sublitharenite sandstones regarded by English et al. (2006) as turbidites. The fauna of these beds is of moderate diversity with dominance of Dalmanella and Sowerbyella and abundant bivalves and rare Cryptolithus trilobites. These beds yielded a few specimens of articulated Cincinnaticrinus.

These comparisons and the lithologic and stratigraphic relations of the Ordovician Martinsburg quarry with the crinoids suggest that it was deposited in a similar outer shelf setting below the wave base of most storms. As noted subsequently, the morphologies of the food-gathering filtration fans of the Martinsburg crinoids are consistent with this supposition. At least some other common constituents of the Martinsburg fauna accord with this scenario, namely the presence of the articulate brachiopods Sowerbyella, Cincinnetina, and perhaps Dalmanella, and the trilobites Cryptolithus and Isotelus (see Holland et al., 2001). It seems that densely pinnulate crinoids formed minor constituents of moderately deep to deep water Ordovician crinoid faunas. It is perhaps noteworthy that all of the crinoids with these dense pinnulate arms are glyptocrinids. Typical shallow water agitated Ordovician crinoid habitats are mostly populated by crinoids with pinnulate arms and dense filtration fans (e.g., Meyer et al., 2002; Brett et al., 2008). As noted by Holterhoff (1997), this pattern persists throughout the Paleozoic to some degree.

Terminology

Most terminology is from Ubaghs (1978). The AB, BC, DE, and AE interrays are lateral interrays. The plates incorporating fixed brachials into the calyx between two rays are called interradials. Similar plates within a ray are termed intraradials. The tube-foot spacing for the Martinsburg crinoids equals the number of covering plates per mm on one side of an arm or pinnule; the relationship is valid because one food-catching tube foot is grouped with a single covering plate (e.g., Brower, 2006, 2007). The height of a covering plate (measured parallel to the long axis of the arm or pinnule) is calculated by the reciprocal of the tube-foot spacing and vice versa. The food-groove width comprises the width across a set of two covering plates on opposite sides of an arm segment or pinnule. Shape ratios are listed in the conventional style, for example width:height, where the width is divided by the height. The measurements in this paper were made with a binocular microscope at magnifications of 10–50×.

Materials and methods

Specimens were collected over several years from the Upper Ordovician (Katian) Martinsburg Formation in the lower mid-Hudson Valley, New York. The Martinsburg crops out in a shale pit, locally referred to as the “Shale Bank,” at the junction of Garden Road and Terrace Road, about 1.6 km from the gate house on Mountain Rest Road on the grounds of Mohonk Mountain House, near New Paltz, Ulster, County, New York (Figs. 1, 2). The shale is tectonically strained and is locally referred to as a ‘pencil shale.’ The rock is quarried and crushed for use on the carriage roads to slow erosion and provide firm footing for hikers. Crinoids, at least as disarticulated columnals, occur at four distinct levels. The fossiliferous Martinsburg in the “Shale Bank” consists of a medium dark-gray shale interbedded with fine-grained greywacke (litharenites); the graywacke beds show graded bedding and cross laminae, some of which are parallel, and others are gently hummocky (Epstein and Lyttle, 1987).

Repository and institutional abbreviation

Specimens are reposited in the Division of Paleontology (Invertebrates), American Museum of Natural History (AMNH), New York, NY, USA.

Systematic paleontology

This research utilizes a revised classification for crinoids based on Ausich (1998), Cole et al. (2017), and Wright et al. (2017),

Class Crinoidea Miller, 1821
Subclass Camerata Wachsmuth and Springer, 1885
Infraclass Eucamerata Cole, 2017
Order Monobathrida Moore and Laudon, 1943
Suborder Glyptocrinina Moore, 1952
Superfamily Glyptocrinoidea Zittel, 1879
Family Glyptocrinidae Zittel, 1879
Genus Pycnocrinus Miller, 1883

Type species

Glyptocrinus shafferi Miller, 1875, which is conspecific with Glyptocrinus dyeri Meek, 1872 (see Brower, 1974) from the Upper Ordovician Maysvillian of Ohio.

Pycnocrinus mohonkensis new species
 Figures 3, 4

Figure 1. Sketch map of locality in Ulster County, southeastern New York where fossiliferous Martinsburg Shale crops out in a location on Mohonk Mountain House property (M) called the “Shale Bank.”

Figure 2. Schematic stratigraphic column of the upper Martinsburg Formation at the “Shale Bank,” Mohonk Mountain House.

Figure 3. Pycnocrinus mohonkensis n. sp.; holotype specimen AMNH FI 162955. (1) Overview of crown and partial column; (2) enlargement of calyx (aboral cup) and proximal arms; (3) enlargement of arms; note long pinnules and bifurcation on secundibrachial 16 or 17. Martinsburg fossil bed 12’ above the base of the section; “Shale Bank,” Mohonk Mountain House. Note 10 mm bar scale.

Figure 4. Sketch of aboral cup of Pycnocrinus mohonkensis n. sp. holotype, AMNH FI 162955; R = radial plate; P = primanal.

Holotype

AMNH FI 162955, Upper Ordovician (Katian), Martinsburg Formation, from a shale quarry on the grounds of Mohonk Mountain House, New Paltz, New York.

Diagnosis

A species of Pycnocrinus with slender and relatively conical calyx; ornamentation of calyx dominated by elevated median-ray ridges on rays and strong ridge on anal series; plates with vague nodes and short stellate ridges. Arms branching once, on secundibrachial 16 or 17; arms and pinnules long and slender compared to calyx.

Description

Crown with long arms; ratio of crown height: “calyx size” (height from calyx base to distal margin of axillary primibrachial) 8.9 mm.

Calyx conical, with straight sides; the ratio of calyx width at the level of primibrachial 2 to “calyx size” = 1.33; ornamentation a large and elevated median-ray ridge on rays and a prominent ridge on the anal series plates; sides of ray plates and anal series plates and interradial plates depressed, with vague nodes or short stellate ridges. Distal fixed brachials range from distal margin of secundibrachial 2 in interrays to distal margin of secundibrachial 3 or 4 within the rays. D ray and parts of C and E rays observed. Basals pentagonal, with marginal rim; width:height ratio = 1.2. Radials in lateral contact, septagonal, width:height = 1.1. Primibrachial 1 hexagonal; width:height = 1.1. Primibrachial 2, axillary, with five or six sides, width:height = 1.1. Secundibrachial 1 with seven sides; width:height = 0.85. Secundibrachial 2 with six or seven sides; lacking obvious fixed pinnule; width:height = 1.1.

One interradial in proximal range, hexagonal, ending at primibrachial 1 level. Second range of interradials with two hexagonal plates, ending at level of primibrachial 2 or secundibrachial 1. Higher ranges with uncertain structure. Intraradials present within the D ray; one plate in two proximal ranges, ending at level of secundibrachial 2; two plates present in next two or three ranges, ending at level of secundibrachials 3 or 4.

Primanal, above C and D ray radials, septagonal; width:height = 0.58; primanal followed by three plates, of which the middle one belongs to the anal series. Anal series plates elongate; width:height ratios of proximal to distal four plates range from 0.65 to 0.39. Proximal two ranges of CD interradial plates with one pentagonal or hexagonal plate, these two ranges end at levels of primibrachial 2 and secundibrachial 1; three higher ranges with two or three irregular plates, ending at secundibrachial 2 level.

Tegmen unknown.

Arms two per ray; brachials uniserial and slightly cuneiform, pinnulate, branching once on secundibrachial 16 or 17. One arm with 123 tertibrachials, which would give a total of about 525 brachs per ray. Location of proximal pinnule uncertain, not obvious on secundibrachial 2 unlike many other glyptocrinids; proximal pinnule present on intraray side of secundibrachial 4; higher non-axillary brachials with pinnules alternating from side to side of arms. Average width:height ratio for secundibrachials 7 to 16 equals 2.64. Tertibrachials become narrower from proximal to distal; average width:height ratios are 2.2 for tertibrachials 1–30, 1.7 for tertibrachials 31–70; 0.91 for most distal plates. Pinnules long and slender, widely separated; maximum observed length 10 mm. Pinnulars long and slender; average width:height ratio 0.19; average width 0.20 mm. Food grooves, covering plates, and articular surfaces unknown. Food-catching tube feet located on pinnulars. Calculated width of pinnular food groove is 0.125 mm (based on an average food groove width:pinnule width ratio of 0.632 for another glyptocrinid; Brower, 1994). The calculated food groove width is like that of other glyptocrinids, as discussed by Brower (2006, 2007, 2010).

Only proximal part of column preserved, partly crushed, originally round. Two orders of columnals present immediately below calyx: nodals and first internodals with nodose sides, and average heights of 1.1 mm and 0.83 mm, respectively. Three orders of columnals probably present in distal part of stem, all with nodose sides and average heights of 0.38, 0.27, and 0.15 mm. Axial canal and articular surfaces unknown.

Etymology

Mohonkensis, referring to the shale quarry at Mohonk Mountain House on the Shawangunk Ridge, lower Hudson Valley, New York.

Remarks

The new species is only known from a single complete crown with a short and partially crushed stem segment. Pycnocrinus mohonkensis n. sp. is most closely related to P. argutus (Walcott, 1883) (see also Walcott, 1884; Brower, 2010, p. 633, figs. 4, 5) from the Spillway Member of the Rust Formation at the Rust Quarry in the vicinity of Trenton Falls, New York. Both species bear four arms per ray with arms that branch on secundibrachials 15 to 19 in P. argutus and 16 or 17 in P. mohonkensis n. sp. The new species differs from P. argutus in having a much narrower calyx with stronger median ray and anitaxial ridges and incipient stellate ridges or nodes, along with narrower and longer arms and pinnules in comparison to the calyx and “calyx size.”

Pycnocrinus mohonkensis n. sp. also resembles P. dyeri (Meek, 1872) (see Meek, 1873a, p. 32, pl. 2, figs. 2a, b; Wachsmuth and Springer, 1897, p. 271, pl. 20, figs. 1a–c, pl. 21, figs. 3a–f, 6; Brower, 1974, p. 12, fig. 3) from younger rocks in the Maysvillian of the Cincinnati, Ohio, region and its relative P. altilis Eckert (1987, p. 854, figs. 3, 4) from the upper Katian of Canada. The number of secundibrachials varies from 8 to 13 in P. dyeri and from 10 to 12 in P. altilis versus the 16 or 17 plates of the new species. The six to nine fixed secundibrachials of P. dyeri and P. altilis exceed those of the comparable sized holotype of P. mohonkensis n. sp. with two to four plates. The wider calyx outlines of P. dyeri and P. altilis are easily distinguished from the more slender profile of P. mohonkensis n. sp. In addition, the stellate ridges of P. dyeri and P. altilis are much more prominent than the weak and partially developed ridges or nodes of the new species.

As in P. mohonkensis n. sp., P. gerki Kolata (1986, p. 716, figs. 2.7, 3.1–3.6, 4; Brower and Kile, 1994, p. 34, pl. 5) from the Guttenberg Formation of Wisconsin and P. sardesoni Brower and Veinus, 1978 (see Brower and Veinus, 1978, p. 416, pl. 11, text-figs. 5A–C) from the Decorah Shale of the Twin Cities are characterized by four arms per ray but the 16 or 17 secundibrachials of the new species outnumber the six secundibrachials of the two Midcontinent forms. The calyx of P. mohonkensis n. sp. is narrower than those of P. gerki and P. sardesoni and more numerous fixed brachials are present in P. sardesoni. The short stellate ridges or nodes of the new species differ from the calyx ornament of P. gerki and P. sardesoni.

Specimens of Pycnocrinus multibrachialis Brower and Veinus, 1978 (see Brower and Veinus, 1978, p. 421, pl. 12, fig. 6, text-fig. 5D) from the Decorah Shale of the Twin Cities and P. ramulosus (Billings, 1857) (see Billings, 1859, p. 57, pl. 7, figs. 2a–f, pl. 8, figs. 1a–e; Wachsmuth and Springer, 1897, p. 273, pl. 20, figs. 5a, b; Wilson, 1946, p. 28) from various Upper Ordovician units in Quebec and Ontario bear from at least 8 to 16 arms per ray, which are thus much more numerous than the four arms per ray of P. mohonkensis n. sp. The arm structure suggests that P. multibrachialis and P. ramulosus are only distantly related to P. mohonkensis n. sp.

Glyptocrinus marginatus Billings, 1857 (see Billings, 1857, p. 260, 1859, p. 59, pl. 9, fig. 1a; Wachsmuth and Springer, 1897, p. 275, pl. 20, fig. 2; Wilson, 1946, p. 27) from the Hull Limestone of Quebec is a problematic species, which is based on the holotype. The crinoid is poorly known, but its arms only appear to branch once on secundibrachial 5. The calyx plates have marginal rims, relatively weak median ray, and anitaxial ridges. Fixed pinnules are absent and the CD interray is much wider and contains more plates that in typical pycnocrinids. These characters separate P. mohonkensis n. sp. from G. marginatus, which, at present, is tentatively regarded as an aberrant glyptocrinid.

Subclass Pentacrinoidea Jaekel, 1918
Infraclass Inadunata Wachsmuth and Springer, 1885
Parvclass Cladida Moore and Laudon, 1943
Family Merocrinidae Miller, 1890
Genus Merocrinus Walcott, 1883 [1884]

Type species

Dendrocrinus? curtus Ulrich, 1879, p. 18.

Merocrinus curtus (Ulrich, 1879)
 Figures 5, 6.1, 7.4, 8.1, 8.2

1879

Dendrocrinus? curtus Ulrich, p. 18, pl. 7, fig. 14.

1883

Merocrinus corroboratus Walcott, p. 4 (advanced publication).

1883

Merocrinus typus Walcott, p. 3 (advanced publication).

1884

Merocrinus typus Walcott, p. 209, pl. 17, fig. 5.

1884

Merocrinus corroboratus Walcott, p. 210, pl. 17, fig. 6.

1889

Merocrinus typus; Miller, p. 262.

Figure 5. Merocrinus curtus (Ulrich, 1879). (1, 2) Crown of AMNH FI 162956 (see also Figs. 7.4, 8.1) showing heterotomous arms and mold of calyx; (2) enlargement of mold of dorsal cup; (3) cylindrical column of Merocrinus showing minor deformation, possibly parasitic in nature; AMNH FI162958 “Shale Bank,” Mohonk Mountain House.

Figure 6. Small slab showing two distinctive types of crinoid columns. (1) AMNH FI 162962 cylindrical homeomorphic column of Merocrinus with even-height holomeric columnals; (2–4) AMNH FI 162961a–c uncompressed and compressed (collapsed) trimerous columns of Ectenocrinus simplex (Hall, 1847).

Figure 7. Overview of large slab of dark Martinsburg shale from ~12’ above base of section at “Shale Bank,” Mohonk Mountain House showing several significant specimens. (1) Uncompressed cylindrical homeomorphic columns of Ectenocrinus simplex (Hall, 1847), AMNH FI 162959a, compressed and collapsed trimerous columns of E. simplex; (2, 3) AMNH FI 162959b, c; (4) small pluricolumnal of Ectenocrinus showing trimere suture, AMNH FI 162963 (see drawing in Fig. 8.4); (5) crown, including external mold of dorsal cup of Merocrinus curtus, AMNH FI 162956 (detailed in Figs. 5, 8.1); (6) mold of partial dorsal cup of E. simplex with attached short pluricolumnal AMNH FI 162960, inset (7) showing enlargement of this specimen (see drawing in Fig. 8.3).

Figure 8. Diagrammatic sketches of selected Martinsburg crinoids preserved on large shale slab shown in Figure 7. (1, 2) Merocrinus curtus (Ulrich, 1879). (1) AMNH FI 162956, shown in Figure 5.1 and 5.2, 7.5 showing low dicyclic dorsal cup (based on mold) with multi-branched heterotomous arms; arrow points to arm fragment with cover plates. (2) Reconstruction of AMNH FI 162957 partial crown and articulated long column; sketch involves considerable inference as specimen is highly compacted and partially decalcified. (3, 4) Ectenocrinus simplex (Hall, 1847); (3) AMNH FI 162960, reconstruction of mold of partial dorsal cup with small attached pluricolumnal, shown in Figure 7.6, 7.7; (4) AMNH FI 162963, pluricolumnal showing collapsed trimeric constriction with longitudinal sutures between trimeres, shown on Figure 7.4.

Holotype

USNM 42103; partial crown and proximal column; Walcott–Rust quarry, Grant, New York; Ulrich, 1879, p. 18, pl. 7, fig. 1.

Diagnosis

A species of Merocrinus with smooth aboral cup and arm plates. Cup cylindrical with relatively wide base; radials fully in contact, with wide radial facets; infrabasals high and radials short compared to cup height. Arms with relatively wide brachials, especially the proximal ones, compared to most merocrinids; arms with bilateral heterotomy above the primibrachials; highest observed axillary located on quartibrachials; arms with at least eight branches per ray; roughly average numbers of brachials in various parts of arms for genus. Anal tube straight, at least up to proximal tertibrachials.

Material

AMNH FI 162956 crown including calyx mold; AMNH FI 16957, poorly preserved small crown and column; AMNH FI 162958 and AMNH FI 162959 columns.

Remarks

A complete treatment of M. curtus from the Upper Ordovician of the Cincinnati, Ohio, area and the Rust Quarry of New York is given by Brower (2010, p. 635–640, figs. 6.1–6.4, 7). The taxonomy of this species is complicated and two widely accepted described species, M. corroboratus Walcott and M. typus Walcott, were considered as junior synonyms of M. curtus (Ulrich) based on the statistical analysis of Brower (2010; partial synonymy listed above).

The Martinsburg crinoids include a reasonably well-preserved partial crown of a small specimen, a rather battered aboral cup with the proximal arms and a long stem segment, along with numerous well-preserved long stem segments. The two specimens with cups and parts of the arms and columns are complete enough so the identification is definite. The associated round, homeomorphic, non-tapering columns with no pentameres, and the crenulate articular surfaces with a round to roundly pentalobate axial canal clearly belong to M. curtus. The Martinsburg partial crown provides new information about the species because one of its distal arm segments has a series of open covering plates along its food groove. The tube foot spacing provides valuable information about the feeding habits of the species.

Subclass Pentacrinoidea Jaekel, 1918
Infraclass Inadunata Wachsmuth and Springer, 1885
Parvclass Disparida Moore and Laudon, 1943
Family Homocrinidae Kirk, 1914
Genus Ectenocrinus Miller, 1889

Type species

Heterocrinus simplex Hall, 1847, p. 280.

Ectenocrinus simplex (Hall, 1847)
 Figures 6.2–6.4, 7.1–7.4, 7.6, 7.7, 8.3, 8.4

1847

Heterocrinus simplex Hall, p. 280, pl. 76, figs. 2a–d.

1977

Ectenocrinus simplex (Hall); Warn and Strimple, p. 84.

1992

Ectenocrinus simplex; Brower, p. 979.

1997

Ectenocrinus simplex; Brower, p. 442.

2008

Ectenocrinus simplex; Brower, p. 64–67, figs. 9–11.

Holotype

NYSM 16070 (New York State Museum), a crown and short pluricolumnal, collected from Sugar River Formation, 10 m above the base of the outcrop on Moose River upstream from junction with Black River, Boonville, NY; Hall, 1847, p. 280, pl. 76, fig. 2.

Diagnosis

A species of Ectenocrinus characterized by a round column with pentalobate axial canal; columnal sutures not strongly crenulate. Aboral cup with angular outline and relatively wide base. Crown plates smooth.

Material

AMNH FI 162959a–c columns; AMNH FI 162960 mold of partial cup and short pluricolumnal; AMNH FI 162961a–c columns; AMNH FI 162963 small pluricolumnal.

Remarks

The most detailed descriptions of morphology and ontogeny and discussions of living habits of specimens throughout North America are in Warn and Strimple (1977) and Brower (1992, 1997, 2008). Numerous long trimeric column segments from the quarry can be assigned to E. simplex. Unfortunately, the articular surfaces are not observable in any of the stems. In addition, a small partial aboral cup has the lower part of the radial circlet, and the basal plates and an associated tapering trimeric column segment. The species identification is definite. The wide base and other proportions of the cup fit with E. simplex rather than the narrow base characteristic of E. triangulus Titus, 1989 (see Titus, 1989, p. 90, figs. 4.1–4.23, 5.2–5.8, 5.12) from the stratigraphically older Denley Limestone of New York. The cup of E. sp. (Kolata, 1976, p. 447, pl. 1, figs. 6, 7) from the Big Horn Limestone of Wyoming is much more rounded than that of E. simplex. Also, the sutures between the adjacent columnals of the Wyoming taxon are much more crenulate than those of E. simplex. Although most of the Martinsburg columns are partially flattened (e.g., Figs. 6.3, 6.4, 7.2), the original cross section was round, as shown in non-collapsed specimens (e.g., Figs. 6.2, 7.3), which is the case in normal specimens of E. simplex. The stem of E. triangulus varies from triangular to subround.

Family Cincinnaticrinidae Warn and Strimple, 1977
Subfamily Cincinnaticrininae Warn and Strimple, 1977
Genus Cincinnaticrinus Warn and Strimple, 1977

Type species

Cincinnaticrinus varibrachialus Warn and Strimple, 1977.

Cincinnaticrinus varibrachialus Warn and Strimple, 1977
 Figure 9.1

1873a

Heterocrinus heterodactylus? Meek, pl. 1, fig. 1a, b.

1977

Cincinnaticrinus varibrachialus Warn and Strimple, p. 41–42, pl. 1, figs. 1, 2, pls. 3–5, text-fig 8.

2005

Cincinnaticrinus varibrachialus; Brower, p. 171–173, figs. 1–6.

Figure 9. Pluricolumnals. (1) long pluricolumnal of Cincinnaticrinus, AMNH FI 162964; (2) columnals and large lumen; probable glyptocrinid column, AMNH FI 162965.

Holotype

UCGM 3871 (University of Cincinnati Geological Museum; now Cincinnati Museum Center) specimen of crown and stem originally listed as Heterocrinus heterodactylus Hall?, 1847, by Meek (1873a, p. 12) and illustrated as Heterocrinus heterodactylus? (Meek, 1873a, pl. 1, fig. 1a, b). Subsequently designated as the type of Cincinnaticrinus varibrachialus by Warn and Strimple (1977, p. 41, pls. 3–5, text-fig 8; Brower, 2005, p. 171–173, figs. 1–6).

Diagnosis

A species of Cincinnaticrinus characterized by a conical cup; distal cup width: proximal cup width 1.4 or greater.

Material

AMNH FI 162964 long portion of nodose column.

Remarks

Full descriptions, taxonomic notes and history, and an outline of paleoecology are available in the references listed in the partial synonymy. One of the Martinsburg specimens, an external mold of a column about 80 mm long, is placed in C. varibrachialus. The columnals are heteromorphic and consist of several orders of nodose plates along with some having smooth sides; pentameres can be identified on part of the column. Overall, the stem closely resembles specimens of C. varibrachialus from the Cincinnati, Ohio area and the Trenton Rust Quarry of New York. In addition to differences in aboral cup shapes, the nodose columnals of C. varibrachialus are much higher than those of C. pentagonus (Ulrich, 1882) (see Warn and Strimple, 1977, p. 55–57, pl. 6) from the Maysvillian and Richmondian of the Cincinnati, Ohio, area.