PREDATORY TRILOBITES: COMBINING MORPHOLOGICAL, ICHNOLOGICAL, AND STRATIGRAPHIC DATA By Jacob Beasecker A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Geological Sciences—Master of Science 2022 ABSTRACT Interpretations of trilobite feeding habit are typically made using either morphological or ichnological data. Morphological data focus on functional interpretations of the hypostome and proximal limb segments. Ichnological interpretations for trilobite predator behavior are based on the co-occurrence of “worm” burrows with the trilobite trace fossil Rusophycus. In this study morphological, ichnological, and stratigraphic data is combined to test whether interpretations of trilobite predator behavior based on morphologic characters is congruent with the interpretations based on ichnofossils. A small number of Rusophycus ichnospecies is associated with “worm” (prey) burrows; in this study these were binned as two form ichnospecies: Rusophycus carleyi- type and Rusophycus pudicum-type. The probable trilobite trace-makers for R. carleyi-type traces associated with putative prey burrows are members of the Asaphida. Asaphid trilobites possessed spinose gnathobases and conterminant hypostomes so the morphological and ichnological data are consistent and support the interpretation that these trilobites were predatory. The stratigraphic range of R. carleyi-type traces associated with worm burrows coincides with the range of asaphid trilobites, supporting this association between specific trilobite taxa and their prey. R. pudicum-type burrows are attributed to trace makers from at least four different trilobite families. Most R. pudicum-type traces associated with a prey burrow are attributable to trilobites that possessed spinose gnathobases and conterminant hypostomes, compatible with interpretations that these trilobites were predatory. R. pudicum-type specimens associated with prey burrows are known only through the Early Devonian. The absence of R. pudicum-type trilobite hunting burrows after the Early Devonian may reflect the extinction of trilobites that pursued infaunal prey. Most R. pudicum-type traces are not associated with prey burrows, suggesting that these trilobites may have been opportunistic hunters. ACKNOWLEDGEMENTS I would like to thank my advisor Dr. Danita Brandt for all the help and knowledge she has imparted on me during my graduate studies. My committee members Dr. Michael Gottfried and Dr. Julie Libarkin. For financial support I would like to thank the Dry Dredgers of Cincinnati and the Department of Earth & Environmental Sciences of Michigan State University for their financial support. For allowing access to their collections, I would like to thank Dr. Glenn Storrs and Cameron Schwalbach from the Geier Collections and Research Center for allowing access to their collections. iii TABLE OF CONTENTS INTRODUCTION ...........................................................................................................................1 METHODS ......................................................................................................................................7 RESULTS ......................................................................................................................................14 DISCUSSION ................................................................................................................................16 CONCLUSION ..............................................................................................................................23 BIBLIOGRAPHY ..........................................................................................................................24 iv INTRODUCTION How do we determine the behavior of extinct animals, especially those that have no closely related living descendants? This study addresses this question with reference to trilobites, an extinct class of arthropods that appeared during the Cambrian Period and went extinct by the end of the Permian Period. Trilobite morphology showed great diversity, presumably reflecting different behavioral or ecological niches. However, we know little about how these animals lived because they have no closely related living relatives. Interpretations of trilobite feeding behavior have been based on aspects of trilobite morphology and, separately, on trace fossils interpreted as reflecting feeding behavior. This study addresses the following questions: Do interpretations of trilobite feeding made using trilobite morphology support interpretations of trilobite feeding based on trilobite trace fossils? Can combining these two data sets (trilobite morphology and trilobite trace fossils) lead to a better understanding of trilobite feeding habits? How did trilobites eat? Insects and crustaceans possess mandibles to assist in chewing whereas chelicerates possess a pair of chelicerae that are used by some to grasp prey or inject toxins by others. Trilobites possessed no movable jaw (Harrington 1959) and presumably used their appendages to gather food ventrally and pass it forward towards their backward-facing mouth (Fortey 1990) located towards the head region (cephalon). Two aspects of trilobite anatomy likely played a role in food ingestion: the hypostome and the gnathobases. The role of the hypostome in trilobite feeding.-As the food was passed forward it would move past the hypostome before entering the mouth opening for ingestion. Hypostomes are a calcified plate on the ventral surface (stomach side) located near the mouth and interpreted as having 1 function in feeding. Trilobite hypostomes come in a range of shapes and sizes (Harrington 1959), but are grouped into two main types, based on their relationship to the anterior border of the cephalon, as conterminant (attached to the anterior edge of the cephalon, Fig. 1b) and natant (detached, presumably supported by non-mineralized cuticle, Fig. 1a). A third hypostome type, impendent, is also attached to the anterior border (Fig. 1c) and will be regarded herein as a subtype of conterminant hypostome. Conterminant hypostomes have a calcified connection to the cephalon and a wider range of shapes than natant hypostomes. Natant hypostomes were not attached directly with the cephalon and are comparatively smaller than conterminant hypostome. This connection to the cephalon seen in conterminant hypostomes, compared to the unattached natant hypostome, is a key factor in interpretations of trilobite feeding behavior based on this morphologic feature. Conterminant hypostomes were much more rigid due to the calcified connection to the cephalon that the natant hypostome lacked (Fortey & Owens 1999). Fortey and Owens (1999) suggested that trilobites possessing conterminant hypostomes displayed predatory feeding behavior, the more rigid conterminant hypostomes allowed for the mastication and consumption of soft-bodied prey. In comparison trilobites with natant hypostomes were deposit feeders consuming organic matter that had settled onto the ocean floor using the appendages to shovel sediment into the oral cavity. 2 Figure 1. The three types of hypostomes (hypostomes indicated by arrows) illustrated. a). Natant hypostome. b) Conterminant hypostome. c) Impendent hypostome (from Fortey 1990, © The Palaeontological Association). The role of trilobite appendages in feeding.-Coxae are the proximal (toward the body) segment of the trilobite’s jointed appendages. Coxae may or may not possess spines; spinose coxae are known as gnathobases (Fig. 2). These spines are presumed to have functioned in masticating food and indicate a predatory feeding habit (Whittington 1977; Hughes 2001; Hegna 2009). As prey was being passed forward toward the mouth, these spines would have helped tear soft- bodied organisms up for easier consumption. (Whittington 1977, 1980; Hughes 2001). Figure 2. Illustration of trilobite appendage showing the segments and presence of spines. The coxae (most proximal segment) bears spinose gnathobases (gn) (from Ramsköld & Edgecombe 1996). 3 Evidence of trilobite feeding habit from trace fossils Trilobite trace fossils provide a second line of evidence for interpreting trilobite feeding behavior. Rusophycus is an ichnogenus attributed to bilateral arthropods, most commonly trilobites, that has been interpreted as reflecting resting or hunting behavior (Bergström 1973). Specimens also exist giving a direct line of evidence linking Rusophycus to trilobite trace makers where the trilobite is superimposed on top of the Rusophycus (Fig. 3). Figure 3. Direct evidence showing trilobites as a trace maker of Rusophycus. Specimen of a Flexicalymene superimposed on top of a Rusophycus pudicum (Osgood 1970). Left: Ventral side of the specimen giving us a look at the Rusophycus trace. Right: Dorsal surface of the trilobite sitting on top of the Rusophycus. Rusophycus specimens have been found intersecting presumed “worm” burrows at which point the “worm” burrow terminates (Fig. 4). These specimens have been described as evidence of interaction between the trilobite and “worm”, specifically, predation on the “worm” prey by the trilobite (Jensen 1990; Brandt et al. 1995; Selly et al. 2016). 4 Two ichnospecies of Rusophycus are known to co-occur with “worm” traces: R. carleyi and R. pudicum (Figure. 4). R. carleyi is associated with trilobites possessing predatory features (conterminant hypostome, spinose gnathobases, Table 1), preserves casts of the limb bases and in some cases the hypostome shape as well (Fig. 4a; Fortey & Owens 1999; Brandt et al. 1995; English & Babcock 2007). R. pudicum is a bean-shaped Rusophycus characterized by transverse “scratch marks” (Fig. 4b) also associated with trilobites that had spinose gnathobases (Table 1). Most R. pudicum-makers also had conterminant hypostomes. English & Babcock 2007 suggested that R. pudicum represented a more generalized feeding strategy. 5 Figure 4. Comparison of R. carleyi and R. pudicum. Both fossils show intersection of a Paleophycus-like “worm” burrow at the approximate position of the trilobite mouth (indicated by circle), suggesting predation of the worm by the trilobite. Left: R. carleyi (from Brandt et al. 1995); anterior of the trace is toward the top; cast of genal spines, hypostome, and limbs are visible; Trace is about 16 centimeters long and 8 centimeters wide. Right: R. pudicum (from Hall, 1852); anterior of the trace is toward the top, showing typical transverse ridges interpreted as scratch marks. Rusophycus trace is approximately 2 centimeters wide by 3 centimeters long. 6 METHODS A systematic literature review was conducted using Google Scholar searching for trilobite morphology and ichnological data. The search parameters for morphological data were “trilobite hypostome”, and “trilobite gnathobases”. Ichnological data parameters were “Rusophycus”, and “trilobite predatory trace”. If a paper was not available through Google Scholar, it was obtained by Interlibrary Loan through the Michigan State University Library. Online museum collections from PRI (Paleontological Research Institution), Cincinnati Museum of Natural History Geir Center, Yale Peabody Museum, Field Museum of Natural History, and the National Museum of Natural History (Smithsonian) sites were searched for Rusophycus specimens associated with “worm” burrows. Photos, descriptions, and observations from trilobites for which the hypostome type is known were compiled in an annotated bibliography. An excel spreadsheet was used to organize data for the analyses using the statistical program R (version 4.0.2). R was chosen for the analyses because it is an open-source programming language with a wide range of tools and it has the capacity to create innovative, publication-quality plots (www.r-project.org/about.html). The focus of the literature search was on descriptions of trilobite traces with associated “worm” burrows that may indicate a predator/prey interaction. Morphological variables collected for the analyses included hypostome type (natant or conterminant) and presence or absence of spinose gnathobases of the presumed trace maker if the trace maker was indicated. If the trace-maker was not indicated, additional references were sought to identify the trilobite taxa present at that locality. 7 Ichnospecies of Rusophycus were placed into two categories: a distinctive type of Rusophycus in which the imprint of the trilobite’s ventral anatomy is preserved as a cast, which is referred to here as the R. carleyi (James 1885) type for the ichnogenus closely associated with this morphology, and a second category for bean-shaped Rusophycus that may or may not bear scratch marks but show no other trace of ventral anatomy, referred to here as the R. pudicum type, for R. pudicum (Hall 1852). If a study reported multiple samples of predatory Rusophycus-“worm” burrow interaction it was counted as one item in the data set unless the specimens were from different localities. If the trace-maker was not specified, geographic and stratigraphic information was used to canvass the literature to identify likely trace-makers. Faunal lists were examined, and the most likely trilobite(s) indicated as the probable trace-maker. Trace-makers interpreted from the supplemental literature (that is, not indicated in the original report of trilobite-worm interaction) are indicated by bold text in Table 1. The data were imported into R from excel as a data frame. A Cramer’s V analysis was used to test the association between the two morphological features (hypostome and gnathobases) and the hunting trace. Values of zero and one were assigned to the type of hypostome and the presence of spinose gnathobases, then individually tested for the association between that morphological feature and a hunting burrow trace (Table 1). Identifying possible trace-makers--Of the 18 samples of Rusophycus “worm” interactions, six of the samples originally had no designated trace maker. For reports that described trilobite predation traces but did not identify a trace-maker, candidate taxa were identified from other 8 papers describing the stratigraphy and paleontology of the area. Each of the decisions on a proxy trace-maker are described here: 1. Kurtz (1975) listed 53 taxa from the Davis Formation of Missouri, one agnostid trilobite and the rest assigned to the Order Ptychopariida. Trilobites of the Order Agnostida were not considered trace-makers due to their very small size and likely pelagic life habit (Esteve & Zamora 2014). The ptychopariids were the likely trace maker, given the size of Rusophycus traces recorded by Selley, et al. (2016). 2. Johnston et al. (2017) described the stratigraphy of the Cambrian of the Canadian Rockies, including the Lake Louise Formation from which Fenton and Fenton’s (1937) specimen was retrieved. Johnston et al. (2017) showed that the only group of trilobites found in the lower Cambrian at this location were of the Order Redlichiida. Fenton and Fenton (1937) predicted a possible Olenellus trace maker for the other trace fossils they described but did not assign a genus for the hunting burrow trace. Species of Olenellus have a conterminant hypostome and the large spinose gnathobases characteristic of predatory trilobites (Whittington 1980). Further, Olenellus belongs to the Order Redlichiida, all trilobites belonging to this order possess conterminant hypostome and have been linked to another hunting burrow trace (Seilacher 1955). 3. Bergström (1973) and Jensen (1990, 1997) both reported the occurrence of R. pudicum hunting burrows from Early Cambrian outcrops of Vastergotland, Sweden. Jensen attributed the traces to a species of Olenellus. This taxonomic assignment is extended to the traces recorded by Bergström. 4. The only body fossils listed in Walcott’s (1918) biostratigraphy of the Flathead Sandstone were brachiopods. Cloud & Bever (1973) compared trace fossils from the Flathead Sandstone to coeval strata found in Canada, and Australia and noted the presence of 9 Glossopleura, a trilobite marking the middle Cambrian, in the Flathead Sandstone, so this genus could be used as a proxy for Walcott’s (1918) trilobite trace-makers. However, Walcott (1918) originally regarded the overlapping “worm” and Rusophycus burrows he photographed as incidental superimposition rather than reflecting a predator/prey relationship, as the “worm” burrows are half the width of the trilobite burrows (large prey relative to the putative predator). Walcott’s (1918) paper is included here because subsequent studies of trilobite predation have included it as an example of trilobite predatory behavior (Tarhan et al. 2012). 5. Landing (1980) detailed the trilobite fauna from the Early Cambrian St. Johns Formation in Nova Scotia, which served here as the stratigraphic proxy for the Kings Formation described Pickerell & Blisset (1999). Among the non-agnostid trilobites listed by Landing (1980), the most likely candidate for a possible trace-maker is Callavia broeggeri. All possible candidates from the Cambrian of Nova Scotia belong to the Order Redlichiida, consistent with the other occurrences listed in Table 1. 6. The photos provided by Phelps (2015) of R. pudicum are from the Silurian. In the report of the Silurian outcrops in Kentucky Ettensohn et al. (2013) listed Calymene at several different outcrops. Calymene is a widespread and common genus in Paleozoic strata of eastern North America (Whitely et al. 2002). Calymene is associated with R. pudicum (Osgood 1970) and is the right size and shape to serve as the trace-maker for this specimen. Identifying hypostome type--Once potential trace-makers were identified, published descriptions of the morphology of the probable trace-maker taxa were used to determine hypostome type of the trace-maker. Because not all potential trace-makers were identified to genus or species, there is still the possibility of misattribution of hypostome type. For example, Kurtz (1975) did not include hypostome type in his systematic descriptions of ptychopairiid taxa. Fortey’s (1990) 10 conclusion that most ptychopairids possessed natant hypostomes was used as justification for hypostome type for these trilobites in Table 1. However, Fortey (1990) also noted that there were ptychopariids that had conterminant hypostomes (what Forety 1990 referred to the “Ptychopariid problem”. 11 Reference Trace Trace Maker* Hypostome Gnathobases Age Locality Fenton & R. Fenton pudicum olenellacean Conterminant present E. Cambrian Alberta, Canada Weber & Zhu R. Yunnan, China, Zhujiaqing Formation 2003 pudicum nd nd nd E. Cambrian R. seilacher 1955 pudicum redlichiida Conterminant present E. Cambrian Salt Range, Pakistan Jensen 1990, R. Vastergotland, Sweden, Mickwitzia Sandstone 1997 pudicum olenellacean Conterminant present E. Cambrian R. Vastergotland, Sweden Bergstrom 1973 pudicum olenellacean Conterminant present E. Cambrian Martinsson R. Paradoxides probably 1965 pudicum paradoxissimus* Cont in some present M. Cambrian Oland, Sweden R. Montana, USA, Flathead Sandstone Walcott 1918 pudicum glosopleura Conterminant present M. Cambrian Selley et al. R. 2016 pudicum ptychopariids natant present Cambrian Missouri, USA, L-M Davis formation Pickerill & R. M.-L. Blissett 1999 pudicum olenellacean Conterminant present Cambrian New Brunswick, Canada, King Square Formation Neto de R. E.-M. Portugal, Penha Garcia, Armorican Quartzite Carvalho 2006 pudicum asaphid Conterminant present Ordovician Formation English & R. Ohio, Oxford, USA Babcock 2007 pudicum Flexicalymene Conterminant present L. Ordovician Osgood & R. New York, USA, Clinton Group Drennen 1975 pudicum calymenid Conterminant present L. Silurian R. Phelps 2015 pudicum calymenid Conterminant present Silurian Kentucky R. Hall 1852 pudicum calymenid Conterminant present E. Silurian New York, USA R. Dalmanites New York, USA, Herkimer Formation Tarhan 2012 pudicum limulurus Conterminant nd Silurian R. Rhineland, Germany Wendorf 1990 pudicum Homalonotid Conterminant nd E. Devonian Table 1. Reports of Rusophycus-“worm” burrow interactions. Listed here are the references reporting the trace fossil, trace fossil taxon, trace maker, hypostome type, presence of gnathobases, age, and locality. Reports are listed in stratigraphic order for R. pudicum and R. carleyi-type traces. Trace makers in bold indicate instances trace-maker identification based on secondary literature (e.g., faunal lists). 12 Table 1 (cont’d) Brandt et al. Southwestern Ohio, Corryville, USA, Great Lake 1995 R. carleyi Isotelus Conterminant present L. Ordovician Formation English & Ohio, Oxford, USA Babcock 2007 R. carleyi Isotelus Conterminant present L. Ordovician 13 RESULTS Frequency data of reported trilobite hunting burrows are shown Fig. 5. Of the 16 reported R. pudicum traces, 14 specimens for which a trace-maker could be assigned possessed a conterminant hypostome. Eight of these 14 specimens are associated with trilobites that had both a conterminant hypostome and spinose gnathobases. The remaining six R. pudicum specimens were associated with probable trace-makers through secondary literature (e.g., locality faunal lists). Only one presumed predatory trace of R. pudicum type remains unassociated with a probable trace-maker. One specimen of R. pudicum is attributed to a group of trilobites that possessed a natant hypostome. This datum could potentially be relocated, as Fortey (1990) noted that taxonomic assignment of ptychopairiid trilobites is problamatic. It is reasonable to question whether this trace maker possessed a conterminant hypostome. All reports of R. carleyi tentatively interpreted as predatory traces (left column in Fig. 5) are associated with the Asaphida trilobite genus Isotelus, which had a conterminant hypostome and spinose gnathobases. No trilobites with natant hypostomes are associated with presumed predatory R. carleyi-type traces. Sample size is very small; only two specimens are reported here from two different authors. Additional specimens of R. carleyi-type predation traces are known, but all came from the same locality, so are counted as one occurrence. Results of Cramer’s V analysis returned a significant value at the 95% interval showing a strong association between the morphological features and hunting burrow traces. This only shows that both morphologic features share strong association with these hunting burrow traces and does not determine that one is more important than the other. This does further reinforce past interpretations (Fortey & Owens 1999) regarding these specific morphological features and their link to feeding behavior. 14 Figure 5. Balloon graph showing frequency data of reported Rusophycus interactions with “worm” burrows. R. carleyi shown on the left, and R. pudicum on the right. The horizontal axis indicates hypostome type; the vertical axis indicates the presence or absence of spinose gnathobases. 15 DISCUSSION The relationship between hypostome type and presumed trilobite predation behavior The data in Table 1 show that trilobites with a conterminant hypostome are dominantly associated with Rusophycus/ “worm” burrow interactions. Of 18 entries in Table 1, 16 are attributed to trilobites associated with a conterminant hypostome, the identity of 1 trace-maker- specimens is unknown, and a single specimen is attributed to a trilobite with a natant hypostome. These data are consistent with the interpretations made by Fortey & Owens (1999), in which they identified the possession of a conterminant hypostome as a key indicator of possible predatory habit. The relationship between spinose gnathobases and predatory behavior All the trilobite trace-makers that are identified in Table 1 possessed spinose gnathobases. If spinose gnathobases were predominantly adaptations for predatory behavior, the stratigraphic record of trilobites with these structures might be expected to overlap the record of trilobite hunting burrows. Trilobites with spinose gnathobases are known from the Cambrian to the Permian (Harrington 1959; Stürmer & Bergström 1973; Whittington 1975, 1980, 1992). However, hunting burrows are known only from the Cambrian (Weber & Zhu 2003) to the Early Devonian (Wendorf 1990). The fact that the trace fossil record does not track the entire stratigraphic range of trilobites with spines gnathobases could be a consequence of the vagaries of the trace fossil record (Seitz & Brandt 2019), in other words, there may not have been conditions conducive to preserving trilobite/worm trace fossils after the Devonian. However, well-preserved Rusophycus are known from post-Devonian strata, suggesting an alternative 16 hypothesis, that this pattern represents a shift in trilobite behavior from pursuit of infaunal prey to other, epifaunal or nektic prey. The relationship between presumed trilobite predation traces and specific ichnotaxa Every report of R. carleyi associated with presumed predatory trace is attributed to the trilobite genus Isotelus. This genus has a large, serrated—forked conterminant hypostome and spinose gnathobases. Conterminant hypostomes and spinose gnathobases have been interpreted as morphological features representing a predatory feeding habit within trilobites (Fortey & Owens 1999). Trilobites belonging to the Family Asaphida, which includes the genus Isotelus, have been interpreted as predators (English & Babcock 2007). To summarize, the association of asaphid trilobites with R. carleyi-type traces supports the interpretation that R. carleyi-type traces were produced predominantly, if not exclusively, by trilobites that possessed conterminant hypostomes (Fortey & Owens 1999). R. carleyi type traces associated with “worm” burrows are shown to be produced by the same genus of trilobite, Isotelus. Isotelus possessed a large hypostome with serrations and a fork-like shape. In combination with a large conterminant hypostome Isotelus also possessed large spines on its appendages and preserved digestive tracts have been found to be sediment- free (English and Babcock 2007). These features have been interpreted as representing a predatory feeding habit. The association between R. carleyi hunting traces with exclusively Isotelus supports a predatory habit for these trilobites as well as other trilobites possessing similar morphological features. R. carleyi trace fossils with no “worm” burrow interaction are largely associated with the trilobite group asaphids, which Isotelus belongs to. Including the other trilobites that have been determined to be trace makers of R. carleyi, the hypostome association with these trace fossils is exclusively conterminant. Feeding habits are complex, and 17 it should be noted that Isotelus and other trilobites possessing similar features may not have been solely predatory but may have been opportunistic hunters. In contrast to the clear association of R. carleyi with asaphid trilobites, R. pudicum traces do not show a clear association with one specific trace maker. R. pudicum trace-makers include trilobites from the families Calymenidae, Dalmanitidae, Homalonotidae, and Ollenellidae and represent taxa with both conterminant and natant-type hypostomes. However, all reports apart from one for R. pudicum traces associated with worm burrows (Table 1) are attributed to taxa that possessed conterminant hypostomes strengthening the relationship between hypostome type and feeding behavior put forward by Fortey & Owens (1999). The relationship between stratigraphic range and trilobite-prey infaunal interactions R. carleyi that are associated with worm burrows are known only from the Ordovician Period, whereas the total stratigraphic range of R. carleyi extends from the Cambrian to the end of the Ordovician. R. carleyi hunting burrows are known from two occurrences (Brandt et al. 1995, English & Babcock 2007), which suggests that the absence of R. carleyi-type hunting burrows is a sampling artifact. However, the absence of R.carleyi after the Ordovician tracks the extinction of the olenid and many asaphid trilobites( Harrington 1959), which are the only trilobites known to be associated with the R. carleyi-type hunting burrows, supporting the hypothesis that the asaphids were the primary producers of these hunting burrows. R. pudicum-type Rusophycus are known from the Cambrian to the Triassic, but R. pudicum-worm burrow associations are known only through the Early Devonian (Wenndorf 1990; Table 3). Post Permian R. pudicum specimens were not produced by trilobites, and as 18 trilobites were extinct by the end of the Permian, and many post-Permian Rusophycus are known from non-marine strata (Würzburg et al. 2001). R. pudicum hunting burrows are more widely reported, occur over a longer stratigraphic range, and are associated with a greater diversity of trace makers than R. carleyi-type hunting burrows. The difference in abundance, diversity of trace makers, and stratigraphic range may indicate a larger diversity of feeding behaviors of R. pudicum-type trace makers compared to the primarily predatory behavior of the R. carleyi trace-makers. R. pudicum hunting burrows may be indicative of trilobites that displayed deposit-feeding as their primary behavior but were facultative predators. The absence of R. pudicum-type trilobite hunting burrows after the Early Devonian may reflect the extinction of trilobites that pursued infaunal prey. A variety of post- Devonian trilobites excavated R. pudicum traces (Table 3), but these are not associated with “worm” burrows. 19 Reference Trace Maker Hypostome Age Locality Type Crimes Olenid Conterminant L. Cambrian Wales 1975 Orłowski Olenid Conterminant L. Cambrian Poland et al. 1971 Gibb et al. Asaphellus Conterminant E. Ordovician Morocco 2010 Draper Asaphid Conterminant E. Ordovician Australia 1980 Baldwin Pleisomeaglaspis Conterminant E. Ordovician France 1977 Fillion & Asaphid Conterminant E. Ordovician Newfoundland Pickerill 1990 Gutiérrez- Asaphellus Conterminant E. Ordovician Morocco Marco et al. 2017 Neto de asaphid Conterminant E-M. Portugal Carvalho Ordovician 2006 Neto de Nd Nd E-M. Portugal Carvalho et Ordovician al. 2014 Brandt et Isotelus Conterminant L. Ordovician Ohio al. 1995 English & Isotelus Conterminant Ordovician Ohio Babcock 2007 Babcock Nd Nd 2003 Table 2. Reports of R. carleyi-type traces with no interaction with a “worm” burrow. Reports are listed in stratigraphic order. nd = not determined. Listed in the table are the reference reporting the trace, reported trace maker, hypostome type, age and locality of the trace fossil. 20 Hypostome Reference Trace Maker Age Locality Type Bergström & Olenellus Conterminant E. Cambrian Greenland Peel 1988 Jensen 1990 olenellacean Conterminant E. Cambrian Sweden Selley et al. Missouri, Nd Nd L. Cambrian 2016 USA Gutschick & Rodriguez Proteus Natant L. Ordovician Western U.S 1977 Osgood 1970 Flexicalymene Conterminant L. Ordovician Ohio Osgood 1970 Cryptolithus Natant L. Ordovician Ohio English & Kentucky, Flexicalymene Conterminant Ordovician Babcock 2007 USA Hall 1852 calymenid Conterminant E. Silurian New York Osgood & calymenid Conterminant M. Silurian New York Drennen 1975 Kentucky, Phelps 2015 calymenid Conterminant Silurian USA Dawson 1864 Phillipsia Conterminant Carboniferous Nova Scotia Bergström Kierulfia Conterminant Carboniferous Ohio 1970 Table 3. Reports of R. pudicum traces with no “worm” burrow interaction, associated hypostome type, stratigraphic and geographic occurrence. Reports are listed in stratigraphic order. nd = not determined. Listed in the table are the reference reporting the trace, reported trace maker, hypostome type, age, and locality of the trace fossil. 21 Figure 6. Stratigraphic range of R. carleyi and R. pudicum. The red represents the range of co- occurrence with worm burrows for that ichnospecies. 22 CONCLUSION Hypotheses of trilobite feeding habits that are constructed as either/or scenarios (predators/not predators) obscure the probable complexity of trilobite paleobiology and paleoecology. The overall rarity of trilobite hunting burrows likely reflects intraspecific variability in trilobite feeding habit. Given the morphologic diversity of trilobites, it is reasonable to conclude that some trilobites were trophic generalists, perhaps primarily detritivores, but also opportunistic predators and/or scavengers. The disappearance of R. carleyi- type hunting burrows after the Ordovician can likely be attributed to the extinction of the trilobite taxa that made them (members of the Order Asaphida). The absence of R. pudicum-type hunting burrows after the Early Devonian may reflect the extinction of the trilobite infaunal predator niche. 23 BIBLIOGRAPHY Babcock L.E. 2003: Trilobites in Paleozoic predator-prey systems, and their role in reorganization of Early Paleozoic ecosystems. In Kelley P.H., Kowalewski M., Hansen T.A. (eds): Predator—Prey Interactions in the Fossil Record, 55-92, Topics in Geobiology 20. Baldwin, C.T. 1977: Rusophycus morgati: An asaphid produced trace fossil from the Cambro- Ordovician of Brittany and Northwest Spain. Journal of Paleontology 51, 411-425. Bergström, J. 1970: Rusophycus as an indication of early Cambrian age. Geological Journal Special Issue 3, 35-42. Bergström, J. 1973: Organization, life and systematics of trilobites. Fossils and Strata 2, 1-69 +5 plates. Bergström, J. & Peel, J.S. 1988: Lower Cambrian trace fossils from northern Greenland. Grønlands Geologiske Undersøgelse Rapport 137, 43-53. Brandt, D.S., Meyer, D.L. & Lask, P.B. 1995: Isotelus (Trilobita) "hunting burrow" from Upper Ordovician Strata, Ohio. Journal of Paleontology 69, 1079-1083. Cloud, P.E., and Bever, J.E. 1973: Trace fossils from the Flathead Sandstone, Fremont County, Wyoming, compared with early Cambrian forms from California and Australia, Journal of Paleontology 47, 883–885. Crimes, T.P. 1975: The production and preservation of trilobite resting and furrowing traces. Lethaia 8, 35–48. Dawson, J.W. 1864: On the fossils of the genus Rusophycus. Canadian Naturalist and Geologist 1, 363-367. Draper, J.J. 1980: Rusophycus (Early Ordovician ichnofossil) from the Mithaka Formation, Georgiana Basin. BMR Journal of Australian Geology & Geophysics 5, 57-61. English, A. & Babcock, L.E. 2007: Feeding behaviour of two Ordovician trilobites inferred from trace fossils and non-biomineralised anatomy, Ohio and Kentucky, USA. Memoirs of the Association of Australasian Palaeontologists 34, 537-544. Esteve, J., & Zamora, S. 2014: Enrolled agnostids from Cambrian of Spain provide new insights about the mode of life for these forms. Bulletin of Geosciences 89, 283-291. Ettensohn, Fr., Lierman, R.T., Mason, C.E., Andrews W.M., Hendricks, R.T., Phelps, D.J. & Gordon, L.A. 2013 The Silurian of central Kentucky, U.S.A.: Stratigraphy, palaeoenvironments and palaeoecology 44, 159-189. Fenton, C.L. & Fenton, M.A. 1937: Trilobite nests and feeding burrows. American Midland Naturalist 18, 446-451. 24 Fillion, D. & Pickerill, R.K. 1990: Ichnology of the Upper Cambrian? to Lower Ordovician Bell Island and Wabana groups of easrtern Newfoundland, Canada. Palaeontographica Canadiana 7. Fortey, R.A. 1990: Ontogeny, hypostome attachment and trilobite classification. Palaeontology 33, 529-576. Fortey, R.A. & Owens, R.M. 1999: Feeding habits in trilobites. Palaeontology 42, 429-465. Gibb, S., Chatteton, B.D.E. & Gingras, M.K. 2010: Rusophycus carleyi (James, 1995), Trace fossils from the Lower Ordovician of southern Morocco, and the trilobites that made them. Ichnos 17, 271-283. Gutiérrez-Marco, J.C., García-Bellido, D.C., Rábino, I., & Sá, A.A. 2017: Digestive and appendicular soft-parts, with behavioural implications in a large Ordovician trilobite from the Fezouata Lagerstatte, Morocco. Scientific Reports 7. Gutschick, R.C. & Rodriguez, J. 1977: Late Devonian-Early Mississippian trace fossil and environments along the Cordilleran Miogeocline, western United States. Geological Journal Special Issue 9, 195-208. Hall, J. 1852: Natural History of New York, Palaeontology volume 2. Containing Descriptions of the Organic Remains of the Lower Middle Divisionof the New York System (Equivalent of the Lower Silurian Rocks of Europe), 1-362. C. Van Benthuysen, Albany. Harrington, H.J. 1959: General description of Trilobita. In Harrington, H.J. et al.: Treatise on Invertebrate Paleontology Part O, Trilobita, 038-0117. University of Kansas Press and the Geological Society of America, Lawrence, Kansas. Hegna, T. 2010: The function of forks: Isotelus-type hypostomes and trilobite feeding. Lethaia 43, 411-419. Hughes, N.C. 2001: Ecologic evolution of Cambrian trilobites. In Zhuravlev, A.Yu. & Riding, R. (eds): The ecology of the Cambrian radiation, 370-403. Columbia University Press, New York. James, J.F. 1885: The fucoids of the Cincinnati Group. Journal of the Cincinnati Society of Natural History 7, 151-166. Jensen, S. 1990: Predation by early Cambrian trilobites on infaunal worms - evidence from the Swedish Mickwitzia Sandstone. Lethaia 23, 29–42. Jensen, S. 1997: Trace fossils from the Lower Cambrian Mickwitzia sandstone, south-central Sweden. Fossils and Strata 42, 1–111. Johnston, P.A., Collom, C.J. & Desjardins, P. 2017: Lower to Middle Cambrian of the southern Canadian Rockies, The Geological Society of America Field Guide, 48. Kurtz, V.E. 1975: Franconian (Upper Cambrian) trilobite faunas from the Elvins Group of southeast Missouri. Journal of Paleontology 19, 1009-1043. 25 Landing, E., Nowlan, G. S., & Fletcher, T. P. 1980: A microfauna associated with Early Cambrian trilobites of the Callavia Zone, northern Antigonish Highlands, Nova Scotia. Canadian Journal of Earth Sciences, 17, 400-418. Martinsson, A. 1965: Aspects of a middle Cambrian thanatotope on Öland. Geologiska Föreningens i Stockholm Förhandlingar 87, 181-230. Neto de Carvalho, C. 2006: Roller coaster behavior in the Cruziana rugosa group from Penha Garcia (Portugal): implications for the feeding program of trilobites. Ichnos 13, 255-265. Neto de Carvalho, C., Rodrigues, J.C. & Baucon, A. 2014: “Fossil art”: the importance and value of the palaeobiodiversity in the Naturtejo Global Geopark, under UNESCO (Portugal). Laboratório Nacional de Geologia e Energia, Comunicações Geológicas 101 91-99. Orłowski, S., Radwański, A. & Roniewicz, P. 1971: Ichnospecific variability of the Upper Cambrian Rusophycus from the Holy Cross Mts. Acta Geologica Polonica 21, 341-348. Osgood, R.G. 1970: Trace fossils of the Cincinnati area. Palaeontographica Americana 41, 281- 444. Osgood, R.G. & Drennen, W.T. III 1975: Trilobite trace fossils from the Clinton Group (Silurian) of east-central New York state. Bulletins of American Paleontology 67, 299-349. Phelps, D. 2015: Downloaded from https://www.uky.edu/OtherOrgs/KPS/images/rusowrm1.jpg Phelps, D. 2015: Downloaded from http://www.uky.edu/OtherOrgs/KPS/images/rusowrm2.jpg Pickerill, R.K. & Blissett, D. 1999: A predatory Rusophycus burrow from the Cambrian of southern New Brunswick, eastern Canada. Atlantic Geology 35, 185-196. Ramskold, L. & Edgecombe G.D. 1996: Trilobite appendage structure Eoredlichia reconsidered. Alcheringa 20, 269-76. Seilacher, A. 1955: Spuren und Lebenweise der Trilobiten, Spuren und Fazies im Unterkambrium. In Schindewolf, O.H. & Seilacher, A. Beiträge zur Kenntnis des Kambriums in der Salt Range (Pakistan). Akademie der Wissenschaften und der Literatur zu Mainz, Abhandlungen der mathematisch-naturwissenschaftliche Klasse 1955, p. 86-143. Seitz, M., & Brandt, D. 2019: Defining the ichnogenus Arthrophycus using numerical taxonomic methods. Ichnos 26, 58-79. Selly, T., Huntley, J.W., Shelton, K.L. & Schiffbauer, J.D. 2016: Ichnofossil record of selective predation by Cambrian trilobites. Palaeogeography, Palaeoclimatology, Palaeoecology 444, 28- 38. 26 Stürmer, W. and Bergström, J. 1973: New discoveries on trilobites by x-ray. Paläontologische Zeitschrift 47, 104-141. Tarhan, L.G., Jensen, S. & Droser, M.L. 2012: Furrows and firmgrounds: evidence for predation and implications for Paleozoic substrate evolution in Rusophycus from the Silurian of east- central New York. Lethaia 45, 329-341. Walcott, C.D. 1918: Appendages of trilobites. Smithsonian Miscellaneous Collections 67, 115- 216. Weber, B. & Zhu, M. 2003: Arthropod trace fossils from the Zhujiaqing Formation (Meishucunian, Yunnan) and their palaeobiological implications. Progress in Natural Science 13, 795–800. Wenndorf, K.W. 1990: Homalonotinae (Trilobita) aus dem Rheinischen Unter-Devon: Homalonotinae (Trilobita) from the Lower Devonian of the Rheinisches Schiefergebirge. Palaeontographica Abteilung A 211, 1-184. Whitely, T. E., Kloc, G. J. & Brett, C.E. 2002: Trilobites of New York. Comstock Publishing Association, Ithaca, NY. Whittington, H.B. 1975: Trilobites with appendages from the Middle Cambrian Burgess Shale, British Columbia. Fossils and Strata 4, 97-136. Whittington, H.B 1977: The Middle Cambrian trilobite Naraoia, Burgess Shale, British Columbia. Philosophical Transactions of the Royal Society of London, Series B 280, 409–443. Whittington, H.B. 1980: Exoskeleton, moult stage, appendage morphology, and habits of Middle Cambrian trilobite Olenoides serratus. Palaeontology 23, 171-204. Würzburg, S.M., Kraków, U.A. & Würzburg, K.M. 2001: Upper Triassic (Keuper) non-marine trace fossils from the Haßberge area (Franconia, south-eastern Germany). Paläontologische Zeitschrift, 75, 71-96. 27