Bulletiw OF THE | PEABODY MUSEUM OF NATURAL History YALE UNIVERSITY Systématic Revision of the Olenelloidea (Trilobita, Cambrian) Bruce S. Lieberman BULLETIN No. 45 OCTOBER 1999 “New HAVEN, CONNECTICUT Systematic Revision of the Olenelloidea (Trilobita, Cambrian) holotype. ennsylvania, U.S.A. ar [YPM 6657], 1 Dunb S getz arly Cambrian ellu E n Copyright © 1997 Peabod > Ole / P um of Natural History, / / Yale University. / y Muse / Photograph by William K. Sacco. Bulletin’ OF THE eas PEABODY MUSEUM OF Nardi Hispory YALE UNIVERSITY 18 OCTOBER 1999 NEw Haven, CONNECTICUT Systematic Revision of the Olenelloidea (Trilobita, Cambrian) Bruce S. Lieberman, Ph.D. Department of Geology University of Kansas Curatorial Affiliate Division of Invertebrate Paleontology Peabody Museum of Natural History Yale University BULLETIN OF THE PEABODY MUSEUM OF NATURAL HISTORY Curatorial Editor-in-Chief Jacques A. Gauthier Executive Editor Lawrence F. Gall Publications Editor Rosemary Volpe The Bulletin of the Peabody Museum of Natural History, Yale University, publishes independent monographs in the natural sciences based on research in the collections of the Yale Peabody Museum. Volumes are numbered consecutively and appear at irregular intervals. The Bulletin incorporates the Bulletin of the Bingham Oceanographic Collections, which ceased independent publication after Volume 19, Article 2 (1967). The series is now out of print. For submission guidelines and information on ordering Yale Peabody Museum publications, see our website or contact: Publications Office Peabody Museum of Natural History Yale University 170 Whitney Avenue P.O. Box 208118 New Haven, CT 06520-8118 U.S.A. Phone: (203) 432-3786 Fax: (203) 432-9816 E-mail: peabody.publications@yale.edu URL: http://www.peabody.yale.edu/publications ON THE Cover: Homotelus bromidensis [YPM 10186]. Early Ordovician, Oklahoma, U.S.A. Copyright © 1997 Peabody Museum of Natural History, Yale University. Photograph by B.S. Lieberman. Copyright © 1999 Peabody Museum of Natural History, Yale University. All rights reserved. This book may not be reproduced, in whole or in part, including illustrations, in any form (except by reviewers for the public press) without the written permission of the publisher. ISBN: 0-912532-47-5 Printed in the U.S.A. €) This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). CONTENTS vul__List of Figures vu List of Tables 1x Abstract 1x Keywords x Acknowledgments 3 Introduction 7 ONE * Phylogenetic Analysis of the Olenellinae 8 Systematic Paleontology 13. Included Taxa 133. Olenellus 27. Two Phylogenetic Analysis of the Mesonacinae 28 Systematic Paleontology 32 Included Taxa 32 Mesolenellus 38 Mesonacis 47. THREE * Phylogenetic Analysis of Elliptocephala 49 Systematic Paleontology 49 Included Taxa 49 Wanneria 53 Elliptocephala 65 FOUR * Phylogenetic Analysis of the Holmiidae 67 Systematic Paleontology 73 Included Taxa 73 Holmia 76 Baltobergstroemia new genus 80 Kjerulfia 81 Schmidtiellus 83 Palmettaspis 85 Esmeraldina 86 Holmiella 89 FIVE * Phylogenetic Analysis of Fritzolenellus 89 Systematic Paleontology 90 Included Taxa 90 Fritzolenellus 99 six * Phylogenetic Analysis of Mummaspis 99 Systematic Paleontology 99 Included Taxa 99 Mummaspis aT PEABODY MUSEUM BULLETIN 45 107. SEVEN * Phylogenetic Analysis of the Bristoliinae 114 Systematic Paleontology 114 Included Taxa u4 Laudonia 16 ~=Lochmanolenellus 8 Fremontella 120 Bristolia 129 EIGHT * Phylogenetic Analysis of Bolbolenellus 129 Systematic Paleontology 131 Included Taxa 131 Nephrolenellus 134 Bolbolenellus 141 NINE * Integrating Phylogenetic and Stratigraphic Data 145 References SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) FIGURES 79 90 94 100 108 FIGURE 1. FIGURE 2. FIGURE 3. FIGURE 4. FIGURE 5. FIGURE 6. FIGURE 7. FIGURE 8. FIGURE 9. FIGURE 10. FIGURE 11. FIGURE 12. FIGURE 13. FIGURE 14. FIGURE 15. FIGURE 16. FIGURE 17. FIGURE 18. FIGURE 19. FIGURE 20. FIGURE 21. FIGURE 22. FIGURE 23. A strict consensus of 40 most parsimonious trees of length 36 steps produced from analysis of character data in Table 2. Olenellus spp. Olenellus spp. Olenellus robsonensis (Burling). Olenellus getzi Dunbar. Olenellus getzi Dunbar. The most parsimonious tree of length 60 steps produced from analysis of character data in Table 4. Mesonacis fremonti (Walcott). Mesonacis spp. A strict consensus of two most parsimonious trees of length 53 steps produced from analysis of character data in Table 6. Wanneria walcottana (Wanner); Holmia kjerulfi (Linnarsson). Elliptocephala spp. A strict consensus of the six most parsimonious trees of length 90 steps produced from analysis of character data in Table 8. The weighted tree of length 32,490 steps, with the same topology as one of the most parsimonious trees used to derive the consensus tree shown in Figure 13. Esmeraldina rowei (Walcott); Palmettaspis consorta Fritz; Holmiella falcuta Pritz: The most parsimonious cladogram of length 5 steps produced from analysis of character data in Table 10. Fritzolenellus truemani (Walcott); Mummaspis occidens (Walcott); Mummaspis truncatooculatus (Fritz); Mummaspis muralensis (Fritz); Mummaspis macer (Walcott). The most parsimonious cladogram of length 9 steps produced from analysis of character data in Table 12. A strict consensus of the four most parsimonious cladograms of length 35 steps produced from analysis of character data in Table 14. Laudonia bispinata Harrington; Bristolia insolens (Resser); Bristolia harringtoni, new species; Bristolia mohavensis (Hazzard and Crickmay); Nephrolenellus jasperensis, new species; Bolbolenellus groenlandicus (Poulsen). The most parsimonious tree of length 14 steps produced from analysis of character data in Table 19. Bolbolenellus groenlandicus (Poulsen); Bolbolenellus altifrontatus (Fritz). A phylogenetic tree depicting topology and timing of patterns of speciation in the Olenelloidea. Vill PEABODY MuSsEUM BULLETIN 45 TABLES 10 Tasie1. Description of characters and character states used in phylogenetic analysis of Olenellus. 16 TABLE 2. Character state distributions for Olenellus and outgroup used in phylogenetic parsimony analysis. 30 TasLe3. Description of characters and character states used in phylogenetic analysis of Mesonacinae. 34 TasLe4. Character state distributions for Mesonacinae and outgroups used in phylogenetic parsimony analysis. 50 TABLE 5. Description of characters and character states used in phylogenetic analysis of Elliptocephala. TABLE 6. Character state distributions for Elliptocephala and outgroup used in phylogenetic analysis. WN tO 70 TaBLe7. Description of characters and character states used in phylogenetic analysis of Holmiinae. 76 Tasie 8s. Character state distributions for Holmiinae and outgroup used in phylogenetic analysis. g2 Tasieg. Description of characters and character states used in phylogenetic analysis of Fritzolenellus. 93 TABLE 10. Character state distributions for Fritzolenellus and outgroups used in phylogenetic analysis. 102 Tasie 1. Description of characters and character states used in phylogenetic analysis of Mummaspis. 104 TABLE 12. Character state distributions for Mummaspis and outgroup used in phylogenetic parsimony analysis. 10 ~= TaBLE 13. Description of characters and character states used in phylogenetic analysis of Bristolinae. us Tasre 14. Character state distributions for Bristoliinae and outgroups used in phylogenetic parsimony analysis. 1222 TasLe15. The first two principal components generated from analysis of morphometric data from species of the genus Bristolia. 124 TABLE 16. A summary of classification produced from a linear discriminant analysis. 126 TasLe 17. The results of a linear discriminant analysis. 133. TABLE 18. Description of characters and character states used in phylogenetic analysis of Bolbolenellus. 135 TABLE 19. Character state distributions for Bolbolenellus and outgroups used in phylogenetic parsimony analysis. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) Ibe ABSTRACT Phylogenetic analyses for all the available species comprising eight clades of trilobites within the monophyletic superfamily Olenelloidea (Trilobita, Cambrian) are presented. These trilobites are among the dominant elements of Early Cambrian faunas in Laurentia, and are the direct by-products of the Cambrian radiation. These phylogenetic analyses, along with a higher level phylogenetic analysis of the Olenelloidea, are used to obtain a perspective on patterns of speciation in this exclusively Early Cambrian trilobite group. A total of 78 species were considered. Based on phylogenetic patterns and the distribution of olenelloid species in the stratigraphic record, there appears to be an evolutionary gap of indeterminate duration in the early history of the Olenelloidea. This gap is in some way associated with the Cambrian radiation, and implies that part of the record of that radiation is effaced. Phylogenetic analysis also indicates substantial revision may be needed for Early Cambrian biostratigraphy. In particular, one of the important biostratigraphic zones in the Early Cambrian is the Olenellus zone. However, species that have been assigned to the genus Olenellus in past studies are in fact distantly related, and include representatives from several distinct clades within the Olenelloidea. Therefore, the genus Olenellus, as defined in past studies, is polyphyletic. The polyphyletic nature of this assemblage implies that past correlations that recognized an Olenellus zone may have little biostratigraphic meaning. In addition to phylogenetic analyses, three new species, Elliptocephala walcotti, Bristolia harringtoni and Nephrolenellus jasperensis, are described, and one new genus, Baltobergstroemia, is diagnosed. KEYWORDS Cladistics, Laurentia, stratigraphy, invertebrate paleontology. xX PEABODY MUSEUM BULLETIN 45 ACKNOWLEDGMENTS I thank N. Eldredge, L. Gall, A. Knoll, S. Westrop and T. White for their comments on earlier versions of this paper. I also thank L. Gall and R. Volpe for their assistance with editorial matters. In the course of this study, collections from the following institutions were used: American Museum of Natural History (AMNH); Geological Survey of Canada (GSC); University of Kansas Museum of Invertebrate Paleontology (KUMIP); Los Angeles County Museum (LACMIP); Museum of Comparative Zoology (MCZ); Geological Museum, University of Copenhagen (MGUH)); Palaeontological Institute Lund, Sweden (PIL); San Diego Natural History Museum (SDSNH); University of California at Riverside (UCR); Smithsonian Institution (USNM); Yale Peabody Museum (YPM). The assistance of the following in arranging the loan of specimens is greatly appreciated: T. White (YPM), J. McKnight (AMNH), J. Dougherty (GSC), A. Hart and R. Kaesler (KUMIP), L. Groves (LACMIP), F. Collier (MCZ), M. Bukh and W. Christensen (MGUH), M. Droser and M. Kooser (UCR), Per Ahlberg (PIL), T. Demere (SDSNH), and J. Thompson (USNM). This research was supported by National Science Foundation grant EAR-9505216. Systematic Revision of the Olenelloidea (Trilobita, Cambrian) Ve _ @pbiollonslO sii to: ’ (asiridinsD .& fidolin?) SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) Introduction Trilobites dominate Early Cambrian metazoan diversity, and the suborder Olenellina Wal- cott 1890 is the primary element in the Early Cambrian trilobite biota (Palmer and Repina 1993). Recently, the Olenellina were subjected to character analysis by Lieberman (1998), who concluded that the Olenellina as traditionally defined (e.g., Palmer and Repina 1993) actually represented a paraphyletic assemblage containing some superfamilies more closely related to the Redlichiina Richter 1933 than to other superfamilies within the “Olenellina.” However, within the Olenellina there is at least one monophyletic superfamily, the Olenel- loidea Walcott 1890. Lieberman (1998) also presented a cladogram displaying the evolu- tionary relationships of almost all the major generic clades within the Olenelloidea. That analysis considered a large number of genera, but only a small number of exponents (species) from each of the generic clades; however, this made it possible to recognize char- acters diagnostic for each of the clades. Moreover, it facilitated recognition of appropriate outgroup taxa that could be used in subsequent studies to elucidate patterns of specific dif- ferentiation within each of those clades. The present paper builds on Lieberman (1998) and pursues in greater detail phyloge- netic studies within the Olenelloidea. Here I present the results of eight phylogenetic analy- ses of all available species within almost all of these generic clades. These analyses include the genera: Olenellus Hall 1862; Mesolenellus Palmer and Repina 1993; Mesonacis Walcott 1885; Elliptocephala Emmons 1844; Holmia Matthew 1890; Kjerulfia Kiaer 1917; Schmidtiel- lus Moberg in Moberg and Segerberg 1906; Esmeraldina Resser and Howell 1938; Palmet- taspis Fritz 1995; Holmiella Fritz 1972; Fritzolenellus Lieberman 1998; Mummaspis Fritz 1992; Laudonia Harrington 1956; Lochmanolenellus Lieberman 1998; Fremontella Harring- ton 1956; Bristolia Harrington 1956; Nephrolenellus Palmer and Repina 1993; Bolbolenellus Palmer and Repina 1993; and one new genus, Baltobergstroemia. Three new species are de- scribed: Elliptocephala walcotti, Bristolia harringtoni and Nephrolenellus jasperensis. A mor- phometric analysis of relationships among some species of the genus Bristolia is also in- cluded. Along with each of these phylogenetic analyses, emended generic diagnoses are provided for all of the aforementioned genera in order to aid in future generic assignments. In addition, a revised subfamilial and familial classification and diagnosis are presented for the Olenelloidea. Type specimens were designated for the following species: neotypes for Olenellus thompsoni (Hall) and Mesonacis vermontana (Hall); and lectotypes for Olenellus clarki (Resser), O. nevadensis (Walcott), Elliptocephala logani (Walcott), Mummaspis trun- catooculatus (Fritz), M. oblisooculatus Fritz, M. muralensis (Fritz), Laudonia amputata Fritz and Bristolia insolens (Resser). Although it would be redundant to consider the results of Lieberman (1998) in great detail, a brief summation of the conclusions of that paper is worthwhile. Several taxa traditionally assigned to the Olenelloidea, such as Callavia Matthew 1897, Andalusiana Sdzuy 1961, Gabriellus Fritz 1992 and Cambropallas Geyer 1993, were excluded from that superfamily by Lieberman (1998) because they lacked some of its diagnostic characters. Within the Olenelloidea, Lieberman (1998) recognized two major clades. One of these, the family Olenellidae, contains the genera Olenellus, Mesonacis and Mesolenellus. This clade is 4 PeaBopy MuseuM BULLETIN No. 45 sister to a clade containing the rest of the taxa referable to the Olenelloidea. Within this sec- ond clade Lieberman (1998) recognized additional phylogenetic structure. For example, the genera Wanneria and Elliptocephala are successively distant outgroups to a large clade that includes the Holmiidae (treated as containing minimally the genera Holmiella, Esmeraldina, Kjerulfia and Holmia) and the remaining olenelloid genera. The remaining clade of olenel- loids can be further subdivided into a set of three genera, Laudonia, Mummaspis and Frit- zolenellus, that form successively distant outgroups to another clade that Lieberman (1998) referred to the family Biceratopsidae. The three genera Laudonia, Mummaspis and Frit- zolenellus were assigned to the paraphyletic family “Laudoniidae.” Within the Biceratopsi- dae Lieberman (1998) identified two subfamilies, the Bristoliinae, which contained the gen- era Bristolia, Fremontella and Lochmanolenellus, and the Biceratopsinae, which contained the genera Nephrolenellus, Bolbolenellus, Olenelloides, Biceratops and Peachella Walcott. The present study was motivated by the need to revise relationships within the Olenel- loidea, and place them within a rigorous framework. Past studies of the olenellids, such as Palmer and Repina (1993), have relied on circular and/or subjective criteria, placing them on quite dubious phylogenetic grounds. This has also had a negative impact on strati- graphic studies. For example, in some papers stratigraphic units were defined by the in- ferred phylogenetic position of the contained olenellid taxa, and this stratigraphic informa- tion was further used to determine evolutionary relationships. Moreover, what had originally been treated as the genus Olenellus (e.g., see Fritz 1992; Palmer and Repina 1993) actually represents a polyphyletic assemblage of taxa. Therefore, the eponymous Olenellus stratigraphic zone, an important division within the Early Cambrian, may be correlated on the basis of spurious criteria. From an evolutionary perspective, the Olenelloidea persisted during an important episode in the history of life, the Cambrian Radiation. Recently our understanding of the timing of this radiation, and the environmental changes that may have been associated with it, have greatly improved thanks to the studies of Knoll (1991, 1992), Bowring et al. (1993), Grotzinger et al. (1995) and Knoll et al. (1995), all detailed in Knoll (1996). It is now possi- ble to place evolutionary patterns during the Cambrian radiation into a detailed geologic context. This in turn makes it possible to make inferences about, evolutionary processes. With this in mind, information about evolutionary patterns from the phylogenetic analyses presented herein are combined with stratigraphic information and the results from Lieber- man (1998) to give an overall picture of patterns of evolution within the Olenelloidea. These patterns can then be used to adduce biogeographic patterns and also levels of speci- ation during the Early Cambrian. One pattern evident in this study, based on information from phylogeny and stratigraphy, is that much of the actual history of the diversification of the Cambrian Radiation may be missing, as Fortey and Owens (1990) suggested. The size of this gap is indeterminate. Although it is probably not too appreciable, it may be equivalent to the length of the Atdabanian stage. This pattern of limited gaps in the record is analogous to what has been found during other events of explosive radiation. For instance, Lieberman et al. (1991) and Lieberman (1993) recognized that much of the explosive adaptive radia- tion of the trilobite family Calmoniidae Delo 1935 was probably missing from the fossil record, with only the subsequent products of that radiation preserved. This paper is organized into several parts, each with a systematic paleontology section, and presenting the results of a phylogenetic analysis. Each phylogenetic analysis uses either the heuristic search option or, when possible, the exhaustive search option of PAUP vy. 3.1.1 (Swofford 1993) to uncover the most parsimonious tree(s), presented as a hypothesis of re- lationship, based on the character data analyzed. For each phylogenetic analysis several tests SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 5 were undertaken to determine the resiliency of the hypothesis of phylogenetic relationship presented. When possible, the protocol of Hillis (1991) was used to assess the amount of phylogenetic signal in the data. The distribution of all tree lengths was determined using PAUP (Swofford 1993). Bootstrap analyses were also conducted to assess the relative sup- port for each of the nodes in the most parsimonious tree(s). Whether or not all require- ments for rigorous statistical application of bootstrap analysis are met, it can serve as a qual- itative demonstration of the relative degree of support for specific cladogram topologies. Unless otherwise indicated, 100 bootstrap replications were employed in a heuristic, step- wise search that used five random replications per bootstrap replication, and groups were retained that were compatible with the 50% majority rule consensus tree. The confidence values for the nodes of the tree duplicated in the bootstrap analysis are presented. Finally, to further assess the support for various nodes of cladograms, the analysis described in Bre- mer (1994) was conducted. This method calculates consensus trees using all cladograms first one, then two, then three, etc., steps longer than the most parsimonious tree. The amount of branch support a node has is equal to the difference between the largest tree that node is still maintained in, minus the length of the most parsimonious tree. The total sup- port index of Bremer (1994) was also calculated. All morphological terminology follows Palmer and Repina (1993) and Lieberman (1998). quieted ts pre ul Soran ss vindles Seeley bet Se Gib od er Os rr +) rite 1 bY me my Ayn tyne comsteenenn ANY Alles a ace 0 ar echt cis eatin 01 90> 1 sda tegen Rc te tt Br tee icin ia ey vestige} Call breRianihs damit Paibabesa Leber un 2) wagon dembets Secparr ery, ents) ae) ee Se ieeat bettas, thee, ety: Gees mse ase res eahisoneel ee ee hte ee Oe es te SEs 088 cnt amternntng eel) bh ita. vi! © sii ay qui anagem en sets Sita -grentie: «+l, 0 pRieeumeial ar & _ We! were a Me eee be > oe wont ah ee Mae om ee rant & ee Sent Oe a - tens 2p chy »~ - a pias lta opcdrretntap Cia So - anay oh “can. ely: rie et) Foes @ | vee P . i ied'urnss ah wragtit a i ode arndiigya eng 04 i _ SG i A) eee hd wat % =ting eet PPA etd 104 4%, \ A) Wns bp ice Bip Op ote | chine Ap Sr shod desis op : See Bed ’ Pe \aanaey joe <- ‘or vow reer ashy - ee _ a ¥ = s 7 a - nly »@ n iV 1 ee | ok aie eek bes . " om nit ‘10 (pets Lae 4 . 7 } ® wd Vi D Ti a 7 = = rie : : ee elow ae Alew 6. Dee wend r i i= Siereaties - es . iL ieétianas Shad dont aati i fp vhs _ a ree a maggtg > i> Oa 7 ste» Pay i Vee J eee 19 Gentonhl oe 2 7 oi iru <«— @@ te Scotian a » ie i) - 068 i; (ye ean il tag Dae 66 Cis WA Ai j x, wea Ui 4 YS) wed Genta aa i if > *S a a jomets of shes: aletiie Ree a gall tent af viewed frarta (tad 1) 6 5yyleTmaeae a =i “Oe ets a1 & PRE Ny, Wide So Wes OE) on Gia dc, pen jieibts. Cetaeeaba andl wasted * ~ Seeemabnidelinarihasty treet Sea a pli Gig Oo’ . Anterior cephalic border . Plectrum . Length (sag.) of preglabellar field . Lateral margins of L4 2 . Lateral margins of L2 2S . Ocular lobe Primitive state (0) elevates anteriorly from anterior border furrow at approximately 45° angle relative to a dorsal, flat plane, then dorsally flattens, such that border is broadly convex (0) absent (0) preglabellar field absent (0) abaxial to lateral margins of LO (0) conjoined medially (0) diverging anteriorly, such that they bulge laterally relative to L1 (0) conjoined medially (0) with prominent furrow across entire margin Derived states (1) elevates anteriorly from anterior border furrow at approximately 70° to 90° angle relative to dorsal plane, then flattens, such that border is developed as a raised ridge (1) present (1) approximately 15% to 25% length (sag.) L4 (2) approximately 35% to 50% length (sag.) L4* (1) directly anterior of lateral margins of LO (2) adaxial to lateral margins of LO (1) not conjoined medially> (1) subparallel, such that they .do not bulge laterally relative to LI* (1) not conjoined medially4 (1) with furrow only present at anterior margin of ocular lobe Continued 4 The polarity of this character was determined using the outgroup criterion. However, polarity might be re- versed if ontogenetic criteria are used (Westrop, pers. comm. 1998). Specifically, some olenelloids during the course of their ontogeny show the progressive shortening and loss of the preglabellar field. b There is no evidence that this character is taphonomic because its states are independent of preservational regime. © As with character 3, the polarity of this character was determined using outgroup comparison. Ontogenetic criteria might potentially reverse the polarity. d There is no evidence that this character is taphonomic because its states are independent of preservational regime. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 11 Table 1 continued. Primitive state Derived states 9. Posterior margin of ocular lobe (0) medial part of distal (1) medial part of distal extends back to margin of L1 margin of LO (2) distal tips of SO 10. Extraocular area (0) gently convex (1) prominently flattened 11. Anterior margins of T3, (0) deflected anteriorly (1) deflected anteriorly excluding spine relative to transverse line relative to transverse line at 5° to 10° angle at 0° to 5° angle 12. T3 extends back (0) length of entire prothorax (1) 6 to 8 thoracic segments 13. Ratio of width (tr.) of T7 (0) 0.50 to 0.64 (1) 0.70 to 0.75 to width (tr.) of T3 (excluding pleural spines) equal to the present analysis, which seeks to construct an informative (sensu Eldredge and Cracraft 1980) classification and takes a cladistic approach to supra-generic classification, where possible, with the caveat that to preserve taxonomic stability establishing a large number of monotypic families or subfamilies should be avoided. Herein, the Olenellinae is restricted to the genus Olenellus. The clade containing Mesonacis and Mesolenellus therefore must also be elevated to subfamilial status and is discussed in greater detail in the next section, which deals explicitly with those genera. The Olenellinae is characterized by the following shared characters when present in combination: 1. posterolateral margins of L4 divergent anteriorly; 2. transverse profile of ocular lobes flattened dorsally; 3. posterior tips of ocular lobes developed opposite medial part of distal margin of LO or distal tip of SO; 4. $3 not conjoined medially; 5. line between ends of $2 directed inward and posteriorly at 45 degree angle relative to transverse line; 6. S0 not conjoined medially; 7. lateral lobes on LO absent; 8. glabellar furrows moderately to strongly incised; 9. genal ridge prominently developed; 10. extraocular region opposite L1 broad, width (tr.) greater than or equal to 75% of width of glabella at L1; 11. genal spine near where it hits cephalic border directed posterolaterally at roughly 10 to 20 degree angle relative to sagittal line; 12. length (exsag.) of genal spine roughly equal to length (sag.) of first four to six thoracic segments; 13. genal spine angle developed opposite medial part of distal margin of LO; 14. intergenal angle developed adjacent to or directly behind genal spine angle; 15. intergenal angle with small, pointed spine; 16. intergenal angle deflected at roughly 10 degree angle relative to transverse line; 17. medial part of cephalic posterior border between intergenal angle and LO weakly flexed anteriorly; 18. anterior margin of 12 PEABODY MusEUM BULLETIN 45 Mesonacis fremonti Olenellus romensis Olenellus agellus Olenellus transitans Olenellus roddyi Olenellus nevadensis Olenellus clarki Olenellus parvofrontatus Olenellus thompsoni Olenellus robsonensis Olenellus geizi Olenellus crassimarginatus Figure 1. A strict consensus of 40 most parsimonious trees of length 36 steps produced from analysis of character data in Table 2 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The re- tention index is 0.71, and the consistency index is 0.72. The value of the g, statistic is —0.32. The following nodes in the text were supported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node 2 = 0.72; Node 3 = 0.53; Node 4 = 0.44; Node 5 = 0.63. Character states are placed at nodes, using MacClade v. 3.04 (Maddison and Maddison 1992), with the characters given in Table 1. The apomorphic state is given in parentheses. Square parentheses indicate equivocal character states that are ambiguous because of missing data, polymorphisms or multiple equally parsimonious resolutions. Equivocal characters are placed only at their basal phylogenetic position, and only unambiguous reversals are shown. Node 1, 5[0, 1], 7[0, 1], 9[0, 1, 2]; Node 2, 1(1), 2(1), 3(1); 4[1, 2], 5(1), 11[0, 1], 13(1); Node 3, 5(1), 7(1), 9(1); 11G), 12); Node, 4(1); Node 5, 6(1), 11[0, 1]. third thoracic pleural segment, before flexing strongly posteriorly, parallel to transverse line or weakly deflected posteriorly; 19. posterior margin of thoracic pleural furrow on third thoracic segment directed evenly posterolaterally; 20. thoracic pleural spines on seg- ments 15 to T8 developed as broad sweeping projections extending roughly four segments back; 21. length (exsag.) of thoracic pleural furrows (excluding those of T3) at medial part of segment equal to roughly 1.3 times length (exsag.) of posterior band of pleural segment; 22. width of thoracic pleural spines T5 to T8 at spine midlength more than two-thirds length (exsag.) of medial part of inner pleural region; 23. length (sag.) of spine on 15th thoracic segment roughly equal to length of whole prothorax; 24. base of spine on 15th thoracic segment nearly as wide (tr.) as axis of segment; 25. segments of opisthothorax SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 13 with reduced pleurae; 26. pygidium length (sag.) 1.5 times width (tr.); 27. posterior mar- gin of pygidium weakly convex. Of these characters, only character 25 is unique to the Olenellidae. Since the subfamily Olenellinae contains only a single genus, Olenellus, the generic and subfamilial diagnoses are identical. A synonymy and discussion for the outgroup used in phylogenetic analysis of the Olenellinae, Mesonacis fremonti, is given below under the heading of the genus Mesonacis. Included Taxa GENUS OLENELLUS HALL 1862 Type species: Olenus thompsoni Hall 1859. Assigned taxa: Olenellus agellus Resser and Howell 1938; O. getzi Dunbar 1925; O. par- vofrontatus Fritz 1991; O. roddyi Resser and Howell 1938; O. romensis Resser and Howell 1938; O. thompsoni crassimarginatus Walcott 1910; O. gilberti Meek 1874; O. sp. 5 Fritz 1972; Paedeumias transitans Walcott 1910; P. clarki Resser 1928; P. granulatus Palmer 1964; P. puertoblancoensis Lochman in Cooper et al. 1952; P. robsonensis Burling 1916; P. groen- landicus Poulsen 1958; P. turmalis Cowie 1968; P. sp. undet. Rasetti 1948; Callavia? nevaden- sis Walcott 1910. Diagnosis: Given above for Olenellinae. Discussion: Some of the taxa assigned to the genus and listed above were poorly preserved and therefore were not subjected to phylogenetic analysis. However, based on their posses- sion of the diagnostic characters of the Olenellinae and Olenellus, these taxa were tentatively assigned to the genus Olenellus, although not all of the diagnostic characters of the genus were always preserved. In the case of O. puertoblancoensis, this species is based on somewhat problematic material so its precise status as one or more species could not be considered by this analysis. Several other species traditionally assigned to the genus Olenellus belong instead to other genera within the Olenelloidea. These are discussed at length under the several generic clades listed below, including Fremontella, Bristolia, Mummaspis, Mesonacis, Bolbolenellus and Elliptocephala, along with reasons why these taxa must be excluded from Olenellus. Phylogenetic topology within the Olenellinae (Figure 1) suggests that the distinction between Olenellus (Olenellus) and O. (Paedeumias) may not be meaningful. Although phy- logenetic analysis predicts that there is a clade of species, some of which have traditionally been assigned to the genus Paedeumias (e.g., O. transitans, O. clarki, O. nevadensis, O. rod- dyi and O. parvofrontatus), this clade is nested within a group of other species traditionally assigned to the genus Olenellus. Retaining the genus or subgenus name Paedeumias would necessitate either making Olenellus paraphyletic, establishing a new paraphyletic genus, or creating two new monotypic genera. Instead, it makes more sense to drop Paedeumias, a name devoid of significance as a phylogenetic unit, and assign all taxa within the Olenelli- nae to the genus Olenellus. Rasetti (1966, 1967) assigned several specimens from the Lower Cambrian of New York State to the genera Olenellus and Paedeumias. Because of the small size of these specimens and their fragmentary state of preservation, their precise affinities could not be determined. However, they more closely resemble specimens of Elliptocephala, espe- cially in the condition of the glabellar furrows, and it appears likely they should be re- ferred to that genus instead of Olenellus. 14 PEABODY MusEUM BULLETIN 45 Figure 2. Olenellus spp. 1, 2. Olenellus romensis Resser and Howell, Rome Fm., Mason Creek, Salem, Virginia. 1 MCZ 2516, X1.5, 2. MCZ 2518°1.-1. 3"5"'Olenelins agellus Resser and Howell, Parker Slate, Parker Quarry, Georgia, Vermont. 3. MCZ 1793, x0.8. 4. AMNH 243, X1.1.5. YPM 72905, x1.1. 6. Olenellus roddyi Resser and Howell, Kinzers Shale, York, Pennsylvania, YPM 72904, x1.8. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 15 OLENELLUS THOMPSONI (HALL 1859) Olenellus thompsoni (Hall); Resser and Howell 1938:219, pl. 3, figs. 17-19 (see for more complete synonymy); Shaw 1955:792; Harrington 1956:59, text fig. 1a; Harrington et al. 1959:0192, fig. 132; Whittington 1989:114, figs. 3-5, 7 (only). Olenellus (Olenellus) thompsoni (Hall); Palmer and Repina 1993:22, fig. 3.1; Palmer and Re- pina 1997:404, fig. 255.1; Lieberman 1998:68, fig. 3.1. Olenellus agellus; Resser and Howell 1938:223, pl. 9, figs. 12 and 13 (only). Types: This species is in need of a neotype to stabilize its taxonomy. I designate YPM 8189, Lieberman (1998, fig. 3.1), a counterpart of an external mold of a nearly complete individual, the neotype of the species because of its good state of preservation and because records indicate it hails from the original type locality, Georgia Township, Vermont, probably Parker’s Quarry. Discussion: This species is treated in detail in Lieberman (1998). Material examined: AMNH 225, 231; MCZ 2451, 2460, 2467, 2469 and possibly 2482; YPM 8189. Because of its state of preservation, it could not be determined whether MCZ 2482 is a specimen of O. thompsoni or O. agellus. Occurrence: Vermont: the Parker Slate, Georgia, treated as in the middle upper Olenellus zone, following Palmer and Repina (1993). OLENELLUS TRANSITANS (WALCOTT 1910) Paedeumias transitans; Walcott 1910:305, pl. 34, fig. 1; Resser and Howell 1938:226, pl. 8, fig. 13; Harrington et al. 1959:0192, fig. 135.5; Olenellus (Paedeumias) transitans; Walcott. Palmer and Repina 1993:22, fig. 3.3; Palmer and Repina 1997:404, fig. 255.3; Lieberman 1998:70, figs. 3.2, 3.3. Paedeumias perkinsi; Resser and Howell 1938:226, pl. 8, fig. 1. Olenellus agellus; Resser and Howell 1938:223, pl. 9, fig. 11, y 12, 13. Olenellus brachycephalus (Emmons); Shaw 1955:791, pl. 75, figs. 1-3. Olenellus thompsoni (Hall); Whitfield 1884:151, pl. 15, figs. 1-4. Types: Lectotype, USNM 56808b, see Resser and Howell (1938). Discussion: This species is treated in detail in Lieberman (1998). Material examined: AMNH 223, 227, 243; MCZ 2427-2429, 2431, 2433, 2435, 2437, 2438, 2444, 2447, 2449-2451, 2453, 2455, 2461, 2465, 2467, 2470, 2474, 2476, 2478, 2480, 2522, 7368, 108661; and possibly YPM 72900. Because of its state of preservation, it could not be determined whether YPM 72900 is a specimen of O. transitans or O. agellus. Occurrence: Vermont: the Parker Slate, Georgia, treated as in the middle upper Olenellus zone, following Palmer and Repina (1993). OLENELLUS ROMENSIS RESSER AND HOWELL 1938 See Figure 2.15222. Olenellus romensis; Resser and Howell 1938:221, pl. 7, figs. 4-6, pl. 12, figs. 2, 3; Resser 1938:52, pl. 5, fig. 12. Olenellus buttsi; Resser 1938:52, pl. 5, figs. 9, 19. Types: Resser and Howell (1938) treated USNM 92718-92721 as the cotypes of this species, 16 PEABODY MUSEUM BULLETIN 45 Table 2. Character state distributions for Olenellus and outgroup used in phylogenetic parsimony analysis. Characters and alternative states are as listed in Table 1. Missing data are indicated by ?. Character states listed as X and Y are polymorphic, where X=(0&1) and Y=(1&2). ] es eS At 67 8a 90 | ree) Mesonacis fremont O00 0F 0; 00, 01070 OF0°40 Olenellus romensts POO OF XO OF 2 0 0) 01 Olenellus agellus OF Or Aces Oe 15070 Olenellus transitans LY aos Ore OT aaa O70r 1 Olenellus roddyi IDR DO 0) OnOn Olenellus clarki Peete? INP X: (OG tOr 21 LEO a Olenellus nevadenstis Hill GYG2 eX 10nlOniOenerl epee We Olenellus parvofrontatus le 2 LOM a) teed Olenellus thompsont B10) LO) (0) Od eal 08 ld 0) hh eG) Olenellus robsonensis Once O Ler OD Ibe) Olenellus getzi OP On OMe ay | 10-240 01 0 Olenellus crassimarginatus OPO Om 2 ratirel TO: ee Gace but they identify the specimen in their pl. 12, fig. 2 as the holotype of the species. This holo- type designation is valid. Discussion: Fritz (1991, pl. 17, figs. 11-14) figured a set of partial cephala from the Illtyd Formation, Yukon, Canada that he assigned questionably to O. romensis. These specimens clearly bear the characters diagnostic of Olenellus, and do match O. romensis in the con- dition of several characters. In particular, they have a similar anterior cephalic border, they both lack a plectrum, and they have the lateral margins of L4 distal of the lateral margins of L4. However, these cephala appear to differ from O. romensis in the condition of at least one character: the posterior margins of the ocular lobes extend back to the dis- tal tips of SO in O. romensis whereas in Fritz’s (1991) material they extend back to the me- dial part of the distal margin of LO. On the basis of this character difference they are ten- tatively treated as distinct from, though closely related to, O. romensis and are retained within the genus Olenellus. Material examined: MCZ 2516 (lot with 5 specimens), 2517, 2518 (lot with 11 speci- mens), 2520. Occurrence: In what Resser and Howell (1938) and Resser (1938) referred to as the Rome Formation, treated as late Early Cambrian, middle upper Olenellus zone, based on Rankin et al. (1989) and Barnaby and Read (1990) from the following localities: Virginia: Mason Creek, Salem; near Webster, NE of Roanoke; 2 mi SW of Blue Ridge Springs, 2 mi S of Max Meadows; Mason Creek, | mi E of Salem; 0.5 mi SE of Indian Rock; 1 mi E of Cleveland. Alabama: 1.5 mi N of Montevallo; 1.5 mi W of Montevallo. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 17 OLENELLUS AGELLUS RESSER AND HOWELL 1938 Figure 2.3 to 2:5 Olenellus agellus; Resser and Howell 1938:223, pl. 9, figs. 11-13. Types: Holotype, USNM 90806, see Resser and Howell (1938). Oddly, Resser and Howell (1938) state that AMNH 244 is the paratype of this species. This specimen is known to be the original type of O. thompsoni, which is missing, so there must be some typographical error in Resser and Howell (1938), particularly as they state on the previous page in the same publication that AMNH 244 is the missing holotype of O. thompsoni. Perhaps they meant AMNH 243, figured herein, or possibly AMNH 244 referred to several specimens, all of which are now lost. Discussion: Resser and Howell (1938) placed several specimens in synonymy with this species, but these could not be verified. This species is similar to O. thompsoni, differing chiefly in the condition of a preglabellar area between L4 and the anterior cephalic border furrow. The preglabellar area is present in O. agellus and absent in O. thompson1. At this time these taxa are treated as distinct species, as several specimens for each species had distinct and non-overlapping character states. Resser and Howell (1938) also asserted that these two taxa may have hailed from rocks with different lithologies. Material examined: AMNH 243; MCZ 1793, 2441, 2433, 2443, 2471, 3684, 7369, and possi- bly 2482; YPM 72905 and possibly 72900. Because of their state of preservation, it could not be determined whether YPM 72900 is a specimen of O. transitans or O. agellus, or whether MCZ 2482 is a specimen of O. thompsoni or O. agellus. Occurrence: Vermont: Parker Quarry, Parker Slate, Georgia, middle upper Olenellus zone, based on Palmer and Repina (1993). OLENELLUS PARVOFRONTATUS FRITZ 1991 Olenellus parvofrontatus; Fritz 1991:13, pl. 11, figs. 1-11, pl. 12, figs. 1, 2. Types: Holotype, GSC 91842, see Fritz (1991). Occurrence: Canada: Yukon territory, Unit 6, upper Illtyd Formation, upper Lower Cam- brian, Olenellus zone, Wernecke Mountains (see Fritz 1991). OLENELLUS RODDYI RESSER AND HOWELL 1938 Figures 2.6, 3.1 Olenellus roddyi; Resser and Howell 1938:221, pl. 6, figs. 6, 7. Olenellus similaris; Resser and Howell 1938:221, pl. 4, figs. 11-14. Olenellus jonasae; Resser and Howell 1938:222, pl. 8, figs. 4-6. Olenellus nodosus; Resser and Howell 1938:224, pl. 6, figs. 1-3. Olenellus nitidus; Resser and Howell 1938:222, pl. 8, figs. 7—9. Paedeumias glabrum; Resser and Howell 1938:226, pl. 8, figs. 2, 3. Paedeumias yorkense; Resser and Howell 1938:227, pl. 10, figs. 1, 2. Paedeumias eboracense; Resser and Howell 1938:227, pl. 8, figs. 14, 15. Paedeumias transitans; Walcott 1910:305, pl. 32, figs. 10, 12, 13, pl. 34, figs. 2-7. Olenellus thompsoni (Hall); Whittington 1989:114, figs. 1, 2, 6, 8-10, 12, 13, 15, 17, 19, D329 SP S4. 18 PEABODY MUSEUM BULLETIN 45 Figure 3. Olenellus spp. 1. Olenellus roddyi Resser and Howell, Kinzers Shale, York, Pennsylvania, YPM 72902, x1.4. 2-5. Olenellus clarki (Resser), Latham Shale, San Bernardino Co., California. 2. at upper end of Summit Wash on op- posite side of ridge, at elevation of 1551 m, 930 m SW of summit spring, 120 m W and 150 m N of SE corner sec 17, W side from old Providence Mine, Providence Mtns., SE 1/4 sec 17, T 11 N, R 14 E, UCR 7002/4, x1.0. 3. E slope of hill 1440 in S end of Marble Mtns., 834 ft E and 1999 ft N of the SW corner sec 12, T5 N,R14E, UCR 7271/2, x0.9. 4, 5. At the end of the dirt road on the opposite side of the hill from the quarry about 1.25 mi N 28 E of Cadiz station. Locality is in the same horizon as Hazzard and Crickmay’s (1933) M-5. Just above a prominently crossbedded quartzite and at the end of the dirt road, S end of Marble Mtns. 4. UCR 10/1135, x1.1. 5. UCR 10/1144, x1.1. 6, 7. Olenellus nevadensis (Walcott). 6. UCR 7271, locality is the same as 3, x0.9. 7. UCR 10/2005, locality is the same as 4, 5, X1.0. 8. Olenellus getzi Resser and Howell, Kinzers Fm., road metal quarry and walls of demolished tenement house on Noah L. Getz Farm, on Harrisburg Pike, about 1 mi N of Rohrerstown, and 2.5 mi W of Lancaster, Lancaster Co., Pennsylvania, YPM 94088, x1.1. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 19 ?Paedeumias yorkense Resser and Howell 1938:227, pl. 6, figs. 4, 5, pl. 9, figs. 5-7. ?Paedeumias transitans Walcott 1910:305, pl. 32, figs. 1-9, 11, pl. 33, figs. 2-5. Types: Holotype, USNM 90803, see Resser and Howell (1938). Diagnosis: Plectrum present; length (sag.) of preglabellar field approximately 15% to 25% length (sag.) of L4; lateral margins of L4 directly anterior of LO or medial of LO; $2 not con- joined medially; lateral margins of L2 bulging laterally relative to L1; $1 conjoined medially; ocular lobe with prominent furrow across entire margin; posterior margin of ocular lobe extends back to medial part of distal margin of LO; extraocular area gently convex; anterior margins of T3 deflected anteriorly relative to transverse line at 5 to 10 degree angle; T3 ex- tends back length of entire thorax; ratio of width (tr.) T7, at articulating tips, relative to width of T3 at same position, 0.70 to 0.75. Discussion: A new diagnosis is provided because historically there has been some confusion surrounding this species and other olenelloids from the Kinzers Shale. On the basis of this analysis, there are definitive records for three species of olenelloids in the Kinzers Shale: Wanneria walcottana (Wanner 1901), Olenellus getzi and O. roddyi. All other previously de- fined species are either synonyms or too poorly preserved to ascertain their taxonomic sta- tus. A large number of the olenelloid species that Resser and Howell (1938) described, in- cluding O. roddyi, O. similaris, O. nodosus, O. nitidus, O. jonasae, Paedeumias glabrum, P. yorkense and P. eboracense, appear to be identical, differing in no morphological characters, except perhaps those relating to structural deformation of the rocks where they occur, and are treated herein as comprising a single species. As all were established in the same publi- cation, and O. roddyi was arguably originally defined based on the most complete material, its species name is preserved. Some specimens assigned to P. yorkense by Resser and Howell (1938) and P. transitans by Walcott (1910) bear strong similarity to material typical of O. roddyi and O. getzi, but are known only from larval stages or thoracic material. This mate- rial could be conspecific with either O. getzi or O. roddyi, as these taxa do not differ mean- ingfully in the condition of their thoraxes, and the extent of ontogenetic variation at this time is not known. Therefore, these specimens are only questionably assigned to O. roddyi. Whittington (1989) assigned several beautifully preserved, complete individuals from the Kinzers Shale to O. thompsoni. However, this material differs from that typical of O. thompsoni in the condition of several characters, including: the anterior cephalic border rises from the extraocular area at a roughly 70 to 90 degree angle relative to a dorso-ventral plane and is then flattened such that it is developed as a raised ridge, rather than rising at a roughly 45 degree angle and then being broadly convex; a plectrum and relatively long (sag.) preglabellar field are present; the lateral margins of L4 are directly anterior of LO or medial of LO, rather than being distal of LO; $2 is not conjoined medially; $1 is conjoined medially; the anterior margins of T3 are deflected anteriorly at a roughly 5 to 10 degree angle relative to a transverse line, rather than roughly paralleling a transverse line; T3 ex- tends back the length of the entire thorax, rather than 6 to 8 thoracic segments; and the ratio of the width (tr.) of T7 (excluding the articulating tips) to T3 is 0.70 to 0.75 rather than 0.50 to 0.64. In the condition of these and other characters it is identical to O. roddyi, and there- fore his material is henceforth assigned to that species. Olenellus roddyi from the Kinzers Shale of Pennsylvania is similar to O. transitans from the Parker Slate of Vermont. However, they differ in the condition of a few qualitative char- acters. In particular, in O. roddyi the lateral margins of L4 are either directly anterior of LO or medial of LO, whereas in O. transitans they are only medial of LO; in O. roddyi the ocular lobe has a prominent furrow circumscribing its entire margin, whereas in O. transitans the 20 PEABODY MUSEUM BULLETIN 45 aw maine, ‘4 ea —— “aay “= 2 ont? amy Figure 4. 1. Olenellus robsonensis (Burling), ?Upper Mahto Fm., drift block on the slope of the Mural glacier below Mumm Peak, near Mount Robson, British Columbia, Canada, GSC 5272, holotype, x1.1. furrow is not visible at the anterior margin of the ocular lobe; and in O. roddyi the extraoc- ular area is gently convex, whereas in O. transitans it is prominently flattened. In addition to these qualitative characters, there exist quantitative character differ- ences between these taxa that allow them to be discriminated at a high level of statistical sig- nificance. One such quantitative character difference is the ratio of the length (sag.) of the preglabellar area relative to the length (sag.) of L4. The following ratios for these variables were calculated for all adequately preserved and available specimens of O. transitans (0.225, 0.375, 0.303, 0.304, 0.543, 0.391) and O. roddyi (0.144, 0.278, 0.179, 0.180, 0.207, 0.184, 0.141, 0.245, 0.209, 0.095, 0.082, 0.098, 0.176, 0.200, 0.205). The mean value of this ratio in O. transitans is 0.357+0.109, whereas the mean value of this ratio in O. roddyi is 0.18+0.06, and these values differ significantly (p<0.01 by the two sample t-test, Minitab 1995 v. 10Xtra [1995]). The median values of this ratio in O. transitans (0.34) and O. roddyi (0.18) differ (p<0.001 by Mann Whitney u-test, Minitab 1995 v. 10Xtra [1995]). Thus, on the basis of these character differences, these two species are treated as distinct, although they are as- sumed to be very closely related. Material examined: AMNH 24214 (3 specimens), 24215 (2 specimens), 24256, 24263 (4 SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 9 specimens), 24297 (3 specimens), 24300, 24871, 24872, 41556 (4 specimens); MCZ 2491, 2496, 2497, 2502, 2503, 2506, 2508, 2513, 2545, 5574 (3 specimens); YPM 8199, 9663, 35859, 35860, 35868, 35893, 35896, 72902—72904, 94037, 94167. Occurrence: Pennsylvania: the Kinzers Shale, middle upper Olenellus zone following Palmer and Repina (1993), 2 mi N of York, Fruitville, 3 mi N of Lancaster, Getz Quarry, 1 mi N of Rohrerstown, and 0.5 mi S of East Petersburg. OLENELLUS CLARKI (RESSER 1928) Figures 3.2 to 3.5 Paedeumias clarki; Resser 1928:9, pl. 3, figs. 1, 2; Riccio 1952:33, pl. 9, figs. 1-4. Olenellus clarki (Resser); Mount 1976:175, fig. 6; Palmer in Palmer and Halley 1979:68, pl. 3, figs. 1-5; Mount 1980:27, fig. 6. Types: Resser (1928) designated USNM 78393, a complete individual, and USNM 78394, a cephalon, as cotypes. In addition, Resser (1928) noted that two unnumbered cephala appear on the same slab as USNM 78393. Because of its excellent state of preservation, USNM 78393 is designated herein the lectotype, and the two unnumbered cephala and 78394 be- come paralectotypes. The type locality is the soft shale (probably the Latham Shale) at Bris- tol Mountain, near Cadiz, on the Santa Fe Railroad, about 100 mi E of Barstow, in the Mo- have Desert, California. Discussion: This species is closely related to O. gilberti, which unfortunately could not be considered in this phylogenetic analysis because adequate material was not available. How- ever, Palmer (Palmer and Halley 1979) provided characters that discriminate between these two taxa. Olenellus clarki and O. nevadensis cannot be consistently discriminated by the qualitative characters given in Table 1. However, these taxa are polymorphic for some of these characters, and not all of their polymorphic character states overlap. In particular, in O. clarki the length (sag.) of the preglabellar field is 15% to 25% of the length (sag.) of L4, whereas in O. nevadensis it is from 15% to 50% of the length (sag.) of L4; in O. clarki the lateral margins of L4 are either medial of the lateral margins of LO or directly anterior of them, whereas in O. nevadensis they are medial of LO; in O. clarki S2 is variably con- joined or not conjoined medially, whereas in O. nevadensis it is conjoined medially; and in O. clarki the posterior margins of the ocular lobes extend back to the distal tips of SO, whereas in O. nevadensis they extend back to SO or to the medial part of the distal margin of LO. Because these taxa have identical character states for many of the characters in Table 1 and overlap in the condition of some of the other characters, it is important to demon- strate that these taxa are indeed distinct and do not represent just a single, highly variable taxon. Bivariate character data show that O. clarki can be consistently discriminated from O. nevadensis, and therefore these are two distinct species. One quantitative difference be- tween these taxa is in the ratio of the maximum width of L4 relative to the maximum width of LO: for O. clarki (0.90, 1.0, 0.80, 0.86, 0.94, 0.90, 0.96, 0.96, 0.93, 0.88, 1.0, 0.89, 0.97, 1.0, 0.83, 0.88, 1.0, 0.83, 0.95, 1.01, 0.88, 0.97, 0.98, 0.94, 0.95, 1.0, 0.97, 1.01); for O. nevadensis (0.63, 0.72, 0.74, 0.77, 0.71, 0.65, 0.78). These ratios, 0.94+0.06 for O. clarki, 0.72+0.06 for O. nevadensis, are significantly different (p<0.00001 by the two sample t-test, Minitab 1995 v. 10Xtra [1995]). The median value of the ratio for O. clarki (0.95) differs from that of O. roddyi (0.72) (p<0.0001 by Mann-Whitney u-test, Minitab 1995 v. 10Xtra [1995]). Thus, on the basis of these character differences, these two species are treated as distinct although assumed to be very closely related. psp PEABODY MUSEUM BULLETIN 45 Figure 5. Olenellus getzi Dunbar, locality is the same as Figure 3.8. 1. YPM 94073, x0.9. 2. YPM 35873, x0.9. Material examined: LACMIP 200-A (2 specimens), 200-B, 3376, 4908-21, 5872 (3 speci- mens), 9690; MCZ 7372; SDSNH 17001, 20707 (4 specimens), 20708 (10 specimens), 20802 (2 specimens); UCR 10 (5 specimens), 10/197, 10/586, 10/985, 10/1135, 10/1144, 10/1160, 10/1421, 10/2009, 10/2022, 7002/4, 7271/2, 7899. Occurrence: Carrara Formation, upper Olenellus zone, Funeral Mountains, Resting Springs Range, Eagle Mountain, Grapevine Mountains, Salt Spring Hills, all in California; Nevada Test Site and Desert Range, Nevada; and from the Latham Shale, treated as in the Bristolia zonule, upper Olenellus zone, at the southern end of the Marble Mountains, near Chamb- less in the Mojave Desert portion of San Bernardino County, California, at the end of the dirt road on the opposite side of the hill from the limestone quarry (Chambless Limestone), about 1.25 mi N, 28 degrees E of Cadiz railroad station, Hazzard and Crickmay’s (1933) lo- cality M-5. Locality lies just above the crossbedded Zabriskie Quartzite at the end of the dirt road. Fossils in gray-brown shale, equal to the upper Latham Shale, 6 to 10 m thick. It is 416 ft W, 216 ft S of NE corner sec 11, T 5 N, R 14 E, San Bernardino Baseline and Meridian, on the USGS 15 ft Danby Quadrangle. Also from Latham Shale, near Summit Springs, W side of Providence Mountains, San Bernardino County, California. Riccio (1952) also reports the species from the Latham Shale, Marble Mountains, 190 m W of the limestone quarry, SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 23 0.5 mi E of Cadiz, in the Mojave Desert portion of San Bernardino County, California, pos- sibly equivalent to Hazzard and Crickmay’s (1933) locality M-5. OLENELLUS NEVADENSIS (WALCOTT 1910) Figures 3.6, 3:7 Callavia? nevadensis; Walcott 1910:285, pl. 38, fig. 12. Paedeumias nevadensis (Walcott); Resser 1928:9, pl. 3, figs. 3-7; Riccio 1952:33, pl. 9, figs. 5, 6. Olenellus nevadensis (Walcott); Mount 1976:175, fig. 10; Palmer in Palmer and Halley 1979:73, pl. 4, figs. 10, 13, 17; Mount 1980:27, fig. 10; Fritz 1991:13, non pl. 12, figs. 3-5. Types: The type locality of this species, according to Walcott (1910), is from the Pioche For- mation, at the summit of Prospect Mountain, the Eureka District, Nevada. Walcott (1910) figured three specimens that he assigned to this species; however, they represent three sep- arate taxa. One of these (Walcott 1910, pl. 38, fig. 13) appears to be a specimen of Bristolia anteros Palmer (Palmer and Halley 1979) or belongs to a species closely related to that taxon. The other (Walcott 1910, pl. 38, fig. 14) is another olenelloid species of indetermi- nate taxonomy, though it clearly is not a specimen of O. nevadensis. Resser (1928) restricted nevadensis to Walcott’s specimen (1910, pl. 38, fig. 12), and this decision is supported here. Resser (1928) illustrated several specimens that he referred to as plesiotypes, but these are not from the type locality and therefore cannot be considered valid types. These are con- specific with O. nevadensis, but hail from Resser’s (1928) Bristol Mountain section in Cali- fornia. The only possible valid type of this species would be Walcott’s specimen (1910, pl. 38, fig. 12), USNM 56799a, and is designated as a lectotype here. Discussion: Fritz (1991) illustrated two specimens from the Lower Cambrian Illtyd Forma- tion of the Yukon that he questionably assigned to O. nevadensis. These specimens are frag- mentary, but similar to O. nevadensis in that they bear a long (sag.) preglabellar area, a plec- trum and a narrow (tr.) L4. However, they do differ from typical O. nevadensis in the condition of a few characters, including: the ocular lobes extend further posteriorly than those typical of O. nevadensis; and the distal tips of S2 extend further laterally than those typ- ical of O. nevadensis. Based on these character differences, the Yukon material is provision- ally excluded from O. nevadensis, though it is held to be closely related to that species. Be- cause Fritz’s (1991) material is so fragmentary, it is not treated in any greater detail herein. Quantitative means for distinguishing this species from O. clarki are presented above. Material examined: LACMIP 200A, 3376; SDSNH 16783-16785, 16812, 20710 (10 speci- mens), 20804 (2 specimens); UCR 10, 10/175, 10/2005, 4079, 7271. Occurrence: California: Carrara Formation, Bristolia zonule, upper Olenellus zone, Funeral Mountains and Grapevine Mountains; Latham Shale, treated as in the Bristolia zonule, upper Olenellus zone, at the southern end of the Marble Mountains, near Chambless in the Mojave Desert portion of San Bernardino County, at the end of the dirt road on the opposite side of the hill from the limestone quarry (Chambless Limestone), about 1.25 mi N, 28 degrees E of Cadiz railroad station, Hazzard and Crickmay’s (1933) locality M-5. Locality lies just above the crossbedded Zabriskie Quartzite at the end of the dirt road. Fossils in gray-brown shale, equal to the upper Latham Shale, which is 6 to 10 m thick. It is 416 ft W, 216 ft S of NE cor- ner sec 11, T 5 N, R 14 E, San Bernardino Baseline and Meridian, on the USGS 15 ft Danby Quadrangle. Riccio (1952) also reports the species from the Latham Shale, Marble Moun- tains, 190 m W of the limestone quarry, 0.5 mi E of Cadiz, in the Mojave Desert portion of 24 PEABODY MUSEUM BULLETIN 45 Figure 6. Olenellus getzi Dunbar, locality is the same as Figure 3.8. 1. YPM 94061, x0.8. San Bernardino County, possibly equivalent to Hazzard and Crickmay’s (1933) locality M-5. Nevada: Carrara Formation, Bristolia zonule, upper Olenellus zone, Desert Range, Nevada. OLENELLUS ROBSONENSIS (BURLING 1916) Figure 4.1 Paedeumias robsonensis; Burling 1916:53. Olenellus robsonensis (Burling); Fritz 1972:15; Whittington 1989:121, figs. 20, 25; Fritz 1997-9: Types: Holotype, GSC 5272, see Fritz (1972). Discussion: This species is based on a single spectacular specirnen that possesses more than 42 thoracic segments, including the usual complement of 13 prothoracic segments. Such a large number of opisthothoracic segments is unknown for any other olenelloid (as defined in Lieberman 1998), and therefore represents a distinct autapomorphy of the species; how- ever, Kleptothule rasmusseni Budd 1995, a trilobed arthropod from the Lower Cambrian of Greenland, although presumably distantly related to O. robsonensis, also bears a many seg- mented thorax. In all other features of its morphology, O. robsonensis matches those char- acters typical of the genus Olenellus. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) i) WN Material examined: GSC 5272, the holotype. Occurrence: Canada, British Columbia: ‘Upper Mahto Formation, ¢middle Olenellus zone (according to Fritz 1992), drift block on the slope of the Mural Glacier below Mumm Peak, near Mount Robson. OLENELLUS GETZI DUNBAR 1925 Figures 3.8, 5.1, 5.2, 6.1 Olenellus getzi; Dunbar 1925:307, figs. 1, 2; Resser and Howell 1938:221, pl. 5, figs. 1-5; Fritz 1972:11; Whittington 1989:120. Olenellus alius; Resser and Howell 1938:224, pl. 7, figs. 7, 8. ?Olenellus peculiaris; Resser and Howell 1938:223, pl. 6, fig. 10. ?Paedeumias yorkense; Resser and Howell 1938:227, pl. 6, figs. 4, 5, pl. 9, figs. 5-7. Types: Holotype, YPM 14365, see Dunbar (1925). Discussion: As mentioned above, some of the specimens assigned to P. yorkense by Resser and Howell (1938) and P. transitans by Walcott (1910) bear a strong similarity to material typical of O. roddyi and O. getzi but are known only from larval stages or thoracic mater- ial. This material could be conspecific with either O. getzi or O. roddyi, as these taxa do not meaningfully differ in the condition of their thoraxes and the extent of ontogenetic varia- tion is not known. Therefore, these specimens are only questionably assigned to O. getzi. Consistent differences between O. getzi and O. roddyi include a more convex anterior cephalic border in the former; a plectrum missing in O. getzi but present in O. roddyi; the lateral margins of L2 in O. getzi not deflected laterally relative to L1, as in O. roddyi; the posterior margins of the ocular lobes extend back to SO in O. getzi, but back to the medial part of the distal margin of LO in O. roddyi; and T3 extends further posteriorly in O. rod- dyi than in O. getzi. Fritz (1972) and Whittington (1989) suggested that O. getzi was conspecific with O. thompsoni. Although these taxa are closely related, there are consistent differences between them and therefore cannot be considered conspecific (they also are known from disjunct lo- calities). Such character differences include: in O. thompsoni the lateral margins of L4 are distal of the lateral margins of LO, while in O. getzi they are directly anterior of LO; in O. getzi the lateral margins of L2 are not deflected laterally relative to L1, as in O. thompsont; the posterior margins of the ocular lobes extend back to SO in O. getzi, but back to the medial part of the distal margin of LO in O. thompsoni; and in O. getzi the anterior margins of T3 are deflected anteriorly relative to a transverse line at a 5 to 10 degree angle, while in O. thompsoni they roughly parallel a transverse line. Material examined: MCZ 2488; YPM 9294, 9297, 9298, 14365, the holotype, 31962, 35867, 35869-35871, 35873, 35874, 35887-35891, 35894, 63304, 94060-94062, 94073, 94088, 94090, 94160, 94168, 94306-94308, 94310, 94312. Occurrence: Pennsylvania: Kinzers Shale, middle upper Olenellus zone, 2 mi N of York, and Noah Getz Farm, | mi N of Rohrerstown. OLENELLUS CRASSIMARGINATUS WALCOTT 1910 Olenellus thompsoni crassimarginatus; Walcott 1910:340, pl. 35, figs. 8, 10. Olenellus crassimarginatus Walcott; Resser and Howell 1938:224, pl. 7, fig. 2. Olenellus wanneri; Resser and Howell 1938:224, pl. 7, fig. 2. 26 PEABODY MUSEUM BULLETIN 45 Olenellus latilimbatus; Resser and Howell 1938:222, pl. 6, fig. 9. Olenellus thompsoni Hall; Walcott 1886:167, pl. 17, fig. 1. Olenellus thompsoni Hall; Walcott 1891, pl. 83, fig. 1b. Types: Walcott (1910) figured three specimens that he assigned to his new variety, O. thomp- soni crassimarginatus. One of these specimens (Walcott 1910, pl. 35, fig. 9) shows a poorly preserved indeterminate species of olenelloid. It clearly is not O. crassimarginatus because it lacks the prominent iutergenal spines, has a relatively broad L4 and has the glabellar fur- rows conjoined. Walcott (1910, pl. 35, fig. 8) shows the only specimen that hails from the type locality, Parker’s Quarry, Georgia, Vermont, and appears to be Walcott’s (1910) in- tended type of the species. This specimen is therefore the holotype, USNM 56836a. Resser and Howell (1938) inexplicably designated USNM 90810 the holotype of the species. Discussion: One of the specimens Walcott (1910, pl. 35, fig. 9) figured and assigned to what is here treated as O. crassimarginatus is poorly preserved. However, based on the illustrated material, it appears that this specimen belongs to some other indeterminate olenellid species (see above). Olenellus wanneri and O. latilimbatus are identical to O. crassimargina- tus in the characters considered in this analysis, and therefore they are treated as conspecific, with O. wanneri and O. latilimbatus representing junior subjective synonyms. Olenellus crassimarginatus is the only species of olenellid known from both the Kinzers Shale and the Parker’s Slate. Occurrence: Vermont: Parker Slate, middle upper Olenellus zone, Parker’s Quarry, Georgia. Pennsylvania: Kinzers Shale, middle upper Olenellus zone, 0.5 mi S of East Petersburg, 2 m1 N of York, and Fruitville, 3 mi N of Lancaster. Phylogenetic Analysis of the Mesonacinae The Mesonacinae clade includes the genera Mesolenellus and Mesonacis, which are in turn sisters to the species of Olenellus. Evidence for this sister group relationship is discussed in greater detail below, and justification for reviving the subfamilial term Mesonacinae is also presented. A total of 10 taxa was subjected to phylogenetic analysis, including all available and ad- equately preserved species of the genera Mesonacis and Mesolenellus, and two outgroup taxa: Olenellus thompsoni and O. transitans. These taxa are appropriate outgroups to the genus Mesonacis based on the higher level phylogenetic analysis presented in Lieberman (1998). Phylogenetic patterns were determined by parsimony analysis of 25 holaspid exoskeletal characters (Table 3). The codings for the taxa analyzed are given in Table 4. All characters were treated as unordered (nonadditive). These data were subjected to an exhaustive search on PAUP v. 3.1.1 (Swofford 1993). One most parsimonious tree (Figure 7) was recovered of length 60 steps, consistency index = 0.57, and retention index = 0.57. The distribution of all tree lengths (Hillis 1991) was significantly left-skewed relative to the distribution of tree lengths produced by random data. The g, statistic was —0.59, which is statistically significant (p<0.01), indicating these data have a strong phylogenetic signal. The confidence values for nodes of the consensus tree duplicated in bootstrap analy- sis are given in Figure 7. Using the method of Bremer (1994), five trees of length less than or equal to 61 steps, 44 trees of length less than or equal to 62 steps, 138 trees of length less than or equal to 63 steps, and 451 trees of length less than or equal to 64 steps were recov- ered before the analysis was terminated because the consensus cladogram was a complete polytomy. The amount of branch support for the various nodes is given in Figure 7. The total support index (Bremer 1994) for the tree is 0.10, a moderately low value according to the examples given in Bremer (1994). Any potential incongruence between stratigraphy and phylogeny, or any potential gap in the history of the Mesonacinae, centers around the first occurrences of species within the two major clades within that subfamily. These two clades, Mesolenellus and Mes- onacis, are sisters to one another, but there is some controversy about the stratigraphic first occurrence of species within the genus Mesolenellus. Species of that genus first appear ei- ther in the “Nevadella” zone or in the Olenellus zone, whereas species of Mesonacis are dis- tributed throughout the Olenellus zone. If the former stratigraphic assignment is accepted, this would imply a potentially large stratigraphic gap in the history of this group. The stratigraphic first appearance of Mesolenellus hyperborea, the type of the genus Mesolenel- lus, was treated as in the “Nevadella” zone by Palmer and Peel (1979), but this first ap- pearance was later adjusted upward into the Olenellus zone by Palmer and Repina (1993). It is not clear, based on the biostratigraphic evidence, why the stratigraphic first appear- ance of this species was subsequently changed. There appears to be more congruence be- tween stratigraphic first appearance and phylogenetic position within the genus Mesonacis. The basal member of that clade, according to the phylogeny shown in Figure 7, M. fremonti 28 PEABODY MUSEUM BULLETIN 45 (Walcott 1910), appears early in the Olenellus zone according to Nelson (1976) and Palmer and Halley (1979), and subsequent species appear later in the Olenellus zone, for example, M. vermontanus (Hall 1859). Systematic Paleontology SUPERFAMILY Olenelloidea Walcott 1890 FAMILY Olenellidae Walcott 1890 SUBFAMILY Mesonacinae Walcott 1890 As defined in Palmer and Repina (1993), the Olenellinae is identical to the Olenellidae, and contains the same genera. Here the Olenellinae is restricted to the genus Olenellus, and discussed above. The clade containing Mesonacis and Mesolenellus therefore must also be elevated to subfamilial status. Walcott (1890) originally proposed the family group name Olenellidae, but rejected it because it was a homophone of Olenidae Burmeister 1843. In its stead he established the Mesonacidae (as Mesonacidiae), and continued its use in several important papers (e.g., Walcott 1910). As discussed in Whittington (1989), ho- mophony is not grounds to reject a family group name, and Whittington (1989) rightly stated that a family group name must be based on a genus contained within the taxon. Whittington (1989) further stated that Resser (1928) showed the difficulties of distin- guishing between species referred to under Mesonacis or Olenellus, and Whittington used this to argue that Mesonacis is a synonym of the former (because the establishment of Olenellus predates the establishment of Mesonacis). In reality, Resser (1928) did not show the difficulties of distinguishing between Mesonacis and Olenellus, but gave clear criteria for differentiating between them (Resser 1928:3—4)! Resser (1928) stated that both Mes- onacis and Olenellus are good genera and can be consistently distinguished by several characters, including the condition of the pleurae on the opisthothoracic segments, the relative position of the ocular lobes and the position of the genal spine angle. However, Resser and Howell (1938) retreated somewhat from this position, and claimed that generic differences do not exist between Olenellus thompsoni and M. vermontanus. They further claimed that some of the differences elucidated in Resser (1928) were still valid, but “not believed to be of generic importance” (Resser and Howell 1938:217). Subsequent commentaries on the taxonomy of the olenellids (e.g., Harrington et al. 1959) treated Mesonacis as a synonym of Olenellus. Palmer and Repina (1993) treated Mesonacis (and also Mesolenellus) as a subgenus of Olenellus on the grounds that species within these gen- era could not always be easily distinguished (Palmer, pers. comm. 1996). However, once clear diagnostic characters for these genera were formulated, it became possible to elevate them to full generic status. Diagnostic characters for both Mesolenellus and Mesonacis are given below under their respective generic headings and clear means of distinguishing be- tween these taxa and Olenellus are provided. Because it is clear that Mesonacis and Mesolenellus represent a good clade, and because this clade is of equal rank to the Olenellinae, an appropriate subfamilial name for them is needed. Hence, the taxon Mesonacinae of Walcott (1890) is re-established, as the subfamily is based around M. vermontanus, the type of the genus Mesonacis. The Mesonacinae can be defined by the possession of the following diagnostic char- acters in combination (thoracic and pygidial characters are not preserved in all species as- signed to the Mesonacinae): 1. anterior part of ocular lobe close to L4; 2. ocular lobe with prominent furrow; 3. width (tr.) of interocular area approximately equal to width of ocular SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 29 Olenellus thompsoni Olenellus transitans Mesolenellus svalbardensis Mesolenellus hyperborea Mesonacis fremonti Mesonacis bonnensis Mesonacis vermontanus Mesonacis eagerensis Mesonacis hamoculus Mesonacis cylindricus Figure 7. The most parsimonious tree of length 60 steps produced from analysis of character data in Table 4 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The retention index is 0.57, and the consistency index is 0.57. The value of the g1 statistic is 0.59. The following nodes in the text were sup- ported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node 1 = 0.49; Node 2 = 0.92; Node 3 = 0.40; Node 4 = 0.45; Node 5 = 0.28; Node 6 = 0.55; Node 7 = 0.29. The following branch support values (Bremer 1994) were recovered for the following nodes: Node 1 = 1; Node 2 = 3; Node 6 = 1. Total tree support (Bremer 1994) is 0.10. Character states are placed at nodes, using MacClade v. 3.04 (Maddison and Maddison 1992), with the characters given in Table 3. The apomorphic state is given in parentheses. Square parentheses indicate equivocal character states that are ambiguous because of missing data or multiple equally parsimonious resolutions. Equivocal characters are placed only at their basal phylogenetic position, and only unambiguous reversals are shown. Node 1, 3[0, 1], 11(1), 12(1), 14[0, 1], 16(1), 17(1), 19(1); Node 2, 3(1), 5[0, 1,2], 6(1), 8(1), 14(0), 21(1), 24(1); Node 3, 2(1), 7(1), 14(1); Node 4, 1[0, 1], 11[1, 2], 15(1), 18(1), 20(1), 25(1); Node 5, 5[0, 2], 10[0, 1], 14[1, 2], 22(2); Node 6, 5(2), 6(2), 10(1), 12(2), 14(2); Node 7, 3(1), 13(1), 21(1). lobe; 4. axial part of LO with node; 5. extraocular area flattened, lacking prominent anasto- mosing ridges (except in M. svalbardensis); 6. extraocular region opposite L1 40% to 50% width (tr.) of glabella at L1 (except in M. fremonti); 7. intergenal angle directed anteriorly at angle greater than or equal to 30 degrees relative to transverse line; 8. anterior margin of thoracic pleural furrow on third segment parallels a transverse line before flexing strongly posteriorly when proceeding from proximal to distal edge; 9. width (tr.) of thoracic pleural spines T5 to T8 at spine midlength less than or equal to half length (exsag.) of correspond- ing segment between spine and axis; 10. segments of opisthothorax with prominent pleu- rae (except in M. fremont); 11. pygidium length (sag.) 1.5 times width (tr.). None of these 30 PEABODY MUSEUM BULLETIN 45 Table 3. Description of characters and character states used in phylogenetic analysis of Mesonacinae. (0) represents the primitive state and (1) and (2) are derived states. _ . Anterior margin of third thoracic segment (T3) relative to transverse line bo . Posterior margins of ocular lobes opposite Ww . Anterior part of frontal lobe — . Length (exsag.) of medial part of thoracic pleural furrows On . Medialmost part of intergenal angle arches anteriorly 6. Length (exsag.) of genal spine co . Lateral margins of glabella opposite L2, when proceeding anteriorly \O . Plectrum 10. Thoracic pleural spines behind T3—T4 extend back roughly — transverse line approximately Primitive state (0) parallel or flexing no more than 5° anteriorly (0) medial part of distal margin of LO (0) intersects anterior border furrow (0) long, 1.5 times as long as posterior pleural band (0) three-fourths of way between point behind distal tip of ocular lobe and genal spine angle (0) approximately equal to 4—5 times the length (sag.) of LO (0) not conjoined medially (0) diverging such that glabella expands laterally (0) absent (0) 4-7 segments . Angle intergenal angle forms with (0) 5°-15° Derived states (1) flexing at least 20° anteriorly (1) medial part of distal margin of L1 (1) does not intersect anterior border furrow (1) short, equal to length of posterior pleural band (1) behind distal tip of ocular lobe (2) at a point midway between point behind distal tip of ocular lobe and genal spine angle (1) approximately equal to 8—10 times the length (sag.) of LO (2) approximately equal to - 2 times the length (sag.) of LO (1) conjoined medially (2) very faintly incised medially (1) not diverging such that glabella does not expand laterally (1) present (1) 1-2 segments (1) 30°-35° (2) 45°-50° Continued SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) Table 3 continued. 12. Length (exsag.) of anterior cephalic border between frontal lobe and genal spine angle 14. Third thoracic pleural spine extends back 15. Medial part of cephalic posterior border between lateral margins of LO and intergenal angle 16. Lateral margins of frontal lobe 17. Lateral margins of glabella between posterior margin of LO and point midway forward on distal tip of L1 18. Line from posterior tip of ocular lobe to junction of posterior margin of ocular lobe with glabella forms roughly 19. Width (tr.) of axis relative to width of pleural segment for T3 excluding spine 20. Anterior cephalic border Primitive state (0) 1.3—1.5 times length (exsag.) of L2 at distal tip (0) medial tip behind distal tip such that inner edge is inclined posteriorly (0) 6-8 thoracic segments (0) posteriorly directed (0) distal to lateral margins of LO (0) converging, such that glabella contracts (0) 0°—5° angle with sagittal line (0) 60%—70% (0) elevates anteriorly from anterior border furrow at approximately 45° angle relative to a dorsal, flat plane, then dorsally flattened, such that border is broadly convex Derived states (1) roughly 0.75 times length (exsag.) of L2 at distal tip (2) roughly 0.40 length (exsag.) of L2 at distal tip (1) medial and distal tip as far forward, transverse (2) very faintly incised medially (1) length of entire thorax (2) 3-4 thoracic segments (1) parallels transverse line (1) directly anterior to lateral margins of LO (1) subparallel, such that glabella is of constant width (1) 15°-20° angle with sagittal line (1) 100%-105% (1) elevates anteriorly from anterior border furrow at approximately 90° angle relative to dorsal plane, then flattened, such that border is developed as raised ridge Continued 32 PEABODY MUSEUM BULLETIN 45 Table 3 continued. Primitive state Derived states 2SO0 (0) not conjoined medially —_ (1) conjoined medially 22. Genal spine angle opposite (0) medial part of margin (1) distal edge of SO of LO (2) medial part of margin of L1 (3) medial part of margin of L2 3 Rs? (0) does not contact axial (1) contacts axial furrows furrows 24. Glabellar furrows (0) prominently incised (1) faintly incised 25. Thoracic pleural spines on all (0) width (tr.) equal to half (1) width (tr.) roughly equal segments but T3, at medial part length (exsag.) of to one quarter length of spine corresponding pleural (exag.) of corresponding segment at medial part pleural segment at between spine and axis medial part between spine and axis characters are unique to the Mesonacinae, meaning they all show some homoplasy or are symplesiomorphic within the Olenelloidea. Included Taxa GENUS MESOLENELLUS PALMER AND REPINA 1993 Type species: Holmia hyperborea Poulsen 1974. Assigned taxa: Olenellus svalbardensis Kielan 1960. Diagnosis: Anterior cephalic border anterolateral of frontal lobe length (exsag.) 0.75 times length (exsag.) of L2 at distal tip; anterior cephalic border elevates from anterior border furrow at approximately 45 degree angle relative to dorsal flat plane, then intersects raised lineament, anterior of this point, border lies in flattened dorsal plane before deflecting ventrally; anterior part of frontal lobe does not intersect anterior border furrow; lateral margins of frontal lobe deflected as far laterally as lateral margins of LO; distal margins of glabella opposite L2 subparallel; S2 straight, not conjoined medially; lateral margins of glabella between posterior margin of LO and point midway forward on distal tip of L1 subparallel; line from posterior tip of ocular lobe to junction of ocular lobes with glabella forms roughly 5 degree angle with sagittal line; SO conjoined medially; lateral lobes pre- sent on LO; posterior margin of ocular lobes opposite medial part of distal margin of LO; SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 33 glabellar furrows faintly incised; intergenal angle forms 30 to 35 degree angle with trans- verse line; genal spine length (exsag.) approximately length (sag.) of 8 to 10 thoracic seg- ments; pleural spine of T3 extends back 6 to 8 thoracic segments; anterior margin of T3 roughly parallels a transverse line medially; posterior margin of pleural furrow of T3 par- allels transverse line before flexing posterolaterally; thoracic pleural spines behind T3 ex- tend back 4 to 7 segments posteriorly; spines wide (tr.), one-half length (exsag.) of corre- sponding pleural segment at its medial part; width (tr.) of thoracic axis 100% of width of pleural field excluding spines; base of large axial spine on T15 less than half width (tr.) of axis of segment. Discussion: Originally Palmer and Repina (1993) treated Mesolenellus as a subgenus of Olenellus. However, based on the phylogeny of the Olenelloidea in Lieberman (1998) this would necessitate lumping both Mesonacis and Mesolenellus within the genus Olenellus. Palmer (pers. comm. 1996) now believes that the differences among the taxa Mesonacis, Mesolenellus and Olenellus are sufficient to warrant generic status, and rec- ommends treatment as separate genera. His recommendation is followed here, as clear diagnostic characters exist for Mesolenellus and indeed all of the Mesonacinae, such that it can be discriminated from Olenellus. In particular, Mesolenellus differs from Olenel- lus (which includes what was formerly referred to as O. [Paedeumias]) in the condition of the following characters: the anterolateral parts of the glabella are more prominently separated from the extraocular area in Mesolenellus than in Olenellus (not considering the anterior border furrow); in Mesolenellus the posterior margins of L4 are subparal- lel, in Olenellus they are divergent anteriorly; in Mesolenellus the lateral margins of the glabella opposite L2 are subparallel, in Olenellus they are divergent; in Mesolenellus the lateral margins of the glabella between the posterior margin of LO and a point midway forward on the distal tip of L1 are subparallel, in Olenellus they are convergent; in Mesolenellus SO is conjoined medially, in Olenellus it is not; in Mesolenellus the lateral lobes on LO are present, in Olenellus they are absent; in Mesolenellus the glabellar fur- rows are faintly incised, in Olenellus they are prominently incised; in Mesolenellus the extraocular area is relatively narrower than it is in Olenellus; in Mesolenellus the length (exsag.) of the genal spine is equal to the length (sag.) of the first 8 to 10 thoracic seg- ments, in Olenellus it is equal to the length of the first 4 to 5 segments; in Mesolenellus the intergenal angle is developed posterior of the lateral margins of the ocular lobes, or at a point halfway between the ocular lobes and the genal spine angle, whereas in Olenellus it is developed adjacent to the genal spine angle; in Mesolenellus the integenal angle has a distinct dorsal swelling, whereas in Olenellus it has a spine; in Mesolenellus the intergenal angle is directed anteriorly at a roughly 30 to 35 degree angle relative to a transverse line, whereas in Olenellus it is directed anteriorly at a roughly 5 to 10 de- gree angle; in Mesolenellus the posterior margin of the thoracic pleural furrow of T3 has the medial part parallel to a transverse line with the lateral part deflected weakly poste- riorly, whereas in Olenellus it is directed evenly posterolaterally; in Mesolenellus the width (tr.) of the thoracic axis on segments 5 to 8 is equal to the width of the pleural segment excluding the spine, in Olenellus the axis is 60% to 70% of the width of the pleurae excluding the spine; in Mesolenellus the width (tr.) of the thoracic pleural spines on segments 5 to 8 at the spine midlength is less than half the length (exsag.) of the me- dial part of the inner pleural region, while in Olenellus they are relatively broader, more than two thirds the length of the medial part of the inner pleural region; in Mesolenel- lus the base of the spine on T15 is less than half the width (tr.) of the axis of the seg- ment, in Olenellus it is nearly as wide as the segment; and in Mesolenellus the 34 PEABODY MUSEUM BULLETIN 45 Table 4. Character state distributions for Mesonacinae and outgroups used in phylogenetic parsimony analysis. Characters and alternative states are as listed in Table 3. Missing data are indicated by ?. ] Par Fae ope De Pg 1S) Meese 5 16 7464900 LEZ yaar ae G7182910 LASS 5 Olenellus thompsoni OFOEO; Or 0r0"0r0 7070" 20" 0"0"070 0" 0207070 00000 Olenellus transitans OOS OFOr0 OL 10)” Ol Oo Oa OTOrOm 00000 Mesolenellus svalbardensis 00102101410 LOL O50 els TOwIeO 10010 Mesolenellus hyperborea OFO Mee LOO. 0s IlOLO. Ll OnE 13,4400 Mesonacis vermontanus Hed) .O.- 0522 3000,0.1 2, DAQEDE Tigslizelily Ipmk 02001 Mesonacis bonnensis COM SOLORZAOFLNOs VEZ ADs Tee Hea 00011 Mesonacis fremonti Oar 070"0"041"0"0"0 OVO OOM TOF 0 00000 Mesonacis eagerensts O00 2721 00" 12.0) 2; OnOR iit 01001 Mesonacis hamoculus So leletee 0 ele OnO) 2 Ee as aR oa Le aT hae LED AGONe Mesonacis cylindricus Col esOe 1 OL Ge 2s Ouk cc (ON Oii mile eee 20 0K2 opisthothorax has prominent pleurae, in Olenellus it does not (not all thoracic charac- ters can be discerned in both species of Mesolenellus). This genus occurs in strata either assigned to the “Nevadella” zone or the lower middle Olenellus zone. MESOLENELLUS HYPERBOREA (POULSEN 1974) Holmia hyperborea; Poulsen 1974:84, pl. 1, figs. 4-7, pl. 2, figs. 1-6, pl. 3, figs. 1-8; Palmer and Peel 1979:33, fig. 3. Olenellus (Mesolenellus) hyperborea (Poulsen); Palmer and Repina 1993:22, fig. 3.7; Palmer and Repina 1997:408, fig. 255.4. Mesolenellus hyperborea (Poulsen); Lieberman 1998:70. Types: Holotype, MGUH 13008, see Poulsen (1974). Discussion: All specimens of this species, except MGUH 13945, have the genal spine angle opposite the medial part of the distal margin of L2. In this one specimen the genal spine angle is opposite the medial part of the distal margin of L1. However, in MGUH 13945 the posterior part of the cephalon is also superimposed over the last two thoracic segments, and the intergenal angle appears to be deformed, with a concomitant displacement of the genal spine posteriorly. Because the condition of the genal spine angle in this specimen appears to be affected by deformation, this character (character 22), was coded with the state it has in all the other specimens of the species. Poulsen (1974) suggested that this species was closely related to Holmia mirabilis Poulsen 1958, here assigned to the genus Elliptocephala, and presented some statistical evi- dence as support. He calculated regression lines for measurements of log glabellar length versus several variables including log intergenal width and log border width. In some cases SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 35 Poulsen (1974) recovered regression lines with similar slopes across what he regarded as highly divergent taxa (e.g., Olenellus gilberti and Holmia kjerulfi Linnarsson 1871); in other cases he found divergent slopes across what he believed to be closely related taxa, and finally he found similar slopes across what he believed to be closely related taxa (e.g., Elliptocephala mirabilis and Mesolenellus hyperborea). First, there is insufficient taxonomic sampling to determine whether this statistical cri- teria for recognizing taxonomic entities has any validity. There is also a strong circular ele- ment to his reasoning that E. mirabilis and M. hyperborea are closely related. Thus, although the results of his analysis are interesting, it is hard to determine their broader significance for classification within the Olenelloidea. Moreover, based on phylogenetic analysis of the Olenelloidea in Lieberman (1998), M. hyperborea is distantly related to Holmia and Ellipto- cephala, and the close relationship of M. hyperborea to H. kjerulfi or E. mirabilis suggested by Poulsen (1974) cannot be accepted at this time. Differences between M. hyperborea and H. kjerulfi include (the asterisk denotes genus- level apomorphy of Mesolenellus): in M. hyperborea the anterior cephalic border near but not directly anterior of L4 has a length (exsag.) less than half the length of LO (sag.), in H. kjerulfi it is equal to the length of LO; in M. hyperborea the anterior cephalic border is promi- nently separated from the extraocular area by a furrow, and in H. Ajerulfi it is not; in M. hy- perborea* L4 does not contact the anterior border furrow, in H. kjerulfi it does; in M. hy- perborea L4 does not expand prominently dorsally, in H. kjerulfi it does; in M. hyperborea the lateral margins of L4 are proximal to the lateral margins of LO, in H. kjerulfi they are dis- tal; in M. hyperborea the posterior margins of L4 are subparallel, in H. kjerulfi they are di- vergent anteriorly; in M. hyperborea the ocular lobes are flattened in transverse profile, in H. kjerulfi they are convex; in M. hyperborea S3 is not conjoined medially, in H. kjerulfi it is; in M. hyperborea a line between the ends of 82 is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line, in H. kjerulfi such a line parallels a transverse line; in M. hyperborea L2 and L3 merge distally, in H. kjerulfi they do not; in M. hyperborea S2 is straight, in H. kjerulfi it is convex anteriorly; in M. hyperborea the distal sector of SO is straight, in H. kjerulfi it is convex anteriorly; in M. hyperborea SO is conjoined medially, in H. kjerulfi it is not; in M. hyperborea the posterior margin of LO is roughly transverse, in H. kjerulfi it is convex posteriorly; in M. hyperborea the extraocular area is flattened, in H. kjerulfi it is prominently vaulted; in M. hyperborea the genal spine near where it hits the cephalic border is directed posterolaterally at a roughly 30 to 35 degree angle relative to a sagittal line, in H. kjerulfi it is directed posterolaterally at a roughly 10 to 20 degree angle; in M. hyperborea* the genal spine is relatively much longer (exsag.) than it is in H. kjerulft; in M. hyperborea the genal spine angle is further anterior than it is in H. kjerulfi; in M. hyper- borea the intergenal spine is represented by a distinct dorsal swelling or node, in H. kjerulfi it is a small pointed spine; in M. hyperborea the posterior border between the intergenal angle and LO is transverse, in H. kjerulfi it flexes posteriorly; in M. hyperborea T3 (third tho- racic segment) is macropleural, in H. kjerulfi it is not; in M. hyperborea* the anterior and posterior margins of the thoracic pleural furrow on the third segment proceeding from the proximal to distal edge parallels a transverse line before flexing strongly posteriorly, while in H. kjerulfi they are directed weakly posteriorly before flexing more strongly posteriorly; in M. hyperborea* the pleural spines on segments 5 to 8 are relatively much longer (exsag. ) and broader (tr.) than they are in H. kjerulfi; in M. hyperborea the prothoracic axial rings lack lateral furrows, whereas they have two prominent anterolateral lobes in H. kjerulfi; in M. hyperborea the thoracic pleural furrows extend onto the spines, in H. kjerulfi they do not; in M. hyperborea the boundary between the thoracic pleural furrow and the anterior band 36 PEABODY MUSEUM BULLETIN 45 is sharp, in H. kjerulfi it is gradational; and the axial spine on T15 is relatively much longer (sag.) in M. hyperborea than it is in H. kjerulfi. Many of these character states are present in other olenelloids, but the extensive differences between the two species suggests that as- signment of M. hyperborea to Holmia is untenable. Material examined: MGUH 13008, the holotype, 13011, 13017, 13020, 13021, 13945. Occurrence: Denmark, Greenland: the upper part of the Buen Formation, above Brillesg, 1.5 km E of Borglum Elv, Peary Land, northeastern end of the island, treated as in the “Nevadella” zone by Palmer and Peel (1979), and in the lower middle Olenellus zone by Palmer and Repina (1993). MESOLENELLUS SVALBARDENSIS (KIELAN 1960) Olenellus svalbardensis; Kielan 1960:84, pl. 1, figs. 1-3, pl. 2, figs. 2, 4, 5, non 1, 3, pl. 3, figs. 2; 3,5, non’1, 4) pl. 45 figs: Vy 2. Olenellus cf. svalbardensis; Kielan 1960; Poulsen 1974:82, pl. 1, figs. 1-3. Types: Holotype, Laboratory of Geology, Polish Academy of Sciences, Cracow, Sv-I/14, see Kielan (1960). Diagnosis: Plectrum present; medial part of intergenal angle arches anteriorly at point mid- way between distal tip of ocular lobe and genal spine angle; S2 does not contact axial fur- rows; medial part of cephalic posterior border flexes posteriorly; genal spine angle devel- oped opposite medial part of distal margin of LO; length (exsag.) of medial part of thoracic pleural furrows approximately 1.5 times length (exsag.) posterior pleural band. Description: Length (exsag.) of anterior cephalic border length at point between lateral margin of L4 and genal spine angle equal to 75% length (exsag.) L2 at distal tip; anterior cephalic border declined at 45 degree angle relative to dorsal plane, before becoming dor- sally flattened in region of anterior border furrow; plectrum present; frontal lobe does not contact anterior border furrow; anterior margin of frontal lobe roughly transverse; lateral margins of frontal lobe deflected as far laterally as lateral margins of LO; lateral margins of glabella opposite margin of L2 are subparallel; glabellar furrows faintly incised; $2 trans- verse, medial edge declined posteriorly, not conjoined medially; L2 and L3 typically merge distally; line from posterior edge of ocular lobe to junction of posterior margin of lobe with glabella roughly forms 10 degree angle relative to sagittal line; SO transverse, S1 convex an- teriorly, both contact axial furrows, with medial edges declined posteriorly; lateral margins of glabella subparallel between LO and point midway forward on distal tip of L1; posterior edge of ocular lobe opposite medial part of distal tip of S0; $1 conjoined medially; lateral lobes present on LO; posterior margin of LO weakly convex posteriorly, with faint axial node; genal spines developed as long projections, length (exsag.) of 8 to 10 thoracic segments (sag.), sweeping backward at roughly 20 degree angle relative to sagittal line; genal spine angle opposite medial part of distal margin of LO; intergenal angle prominently developed, without spine, inclined at 30 to 35 degree angle relative to transverse line; extraocular re- gion width (tr.) approximately 40% to 50% width between distal-most tips of ocular lobes; prominent anastomosing ridges visible on extraocular region; faint intergenal and genal ridges visible; posterior margin of cephalic posterior border between LO and intergenal angle flexing posterolaterally. Thorax divided into pro- and opisthothorax; faint axial nodes present on medial part of thoracic axial rings; axial rings at fourth segment 30% width (tr.) of pleural field, ex- cluding spines; spines of T3 macropleural, extend as far posteriorly as T6 to T8; medially, SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 37 anterior margin of T3 parallels transverse line; spines of T3 deflected posteriorly at roughly 30 degree angle relative to sagittal line; anterior margin of pleural furrow of T3 parallels transverse line before flexing posterolaterally; posterior margin of pleural furrow of T3 par- allels transverse line before flexing posterolaterally; anterior margins of thoracic pleural fur- rows prominently separated from anterior band by distinct change in slope; thoracic pleural furrows extend onto spines; thoracic pleural spines behind T4 extend three to four thoracic segments back, lie in roughly same dorso-ventral plane as pleural segments; prominent spine on axial ring of T15, length (sag.) approximately equal length of entire thorax. Pygidium not sufficiently preserved. Discussion: Kielan (1960) figured a series of specimens from the Slakli Series of West Spits- bergen that she assigned to Olenellus svalbardensis. Her sample appears to contain two dis- tinct morphotypes. One of these, which includes the specimen she designated the holotype of the species, is characterized by the following features: the anterior cephalic border is de- clined at a 45 degree angle relative to a dorsal plane before becoming dorsally flattened in the region of the anterior border furrow; anterolateral margins of the frontal lobe are di- rected posteriorly at a roughly 10 degree angle relative to a transverse line; the lateral mar- gins of L4 are deflected as far laterally as the lateral margins of LO; the sides of the glabella opposite L2 and L3 are subparallel; lateral lobes are present on LO; the posterior margins of the ocular lobes extend back to the medial part of LO; and the posterior border of the cephalon parallels a transverse line medially and then distal of a point midway between the distal tip of the ocular lobe and the genal spine it flexes anteriorly such that the intergenal angle forms a roughly 30 to 35 degree angle with a transverse line. The second morphotype is characterized by the following morphology: the anterior cephalic border, when proceeding anteriorly, elevates from the anterior border furrow at a 90 degree angle relative to a dorsal plane, then is dorsally flattened; the anterolateral mar- gins of the frontal lobe are directed posterolaterally at a roughly 40 degree angle relative to a transverse line; the lateral margins of L4 are distal to the lateral margins of LO; the lateral margins of the glabella opposite L2 are deflected laterally such that the width (tr.) of the glabella expands, opposite L3 they are deflected medially such that the width (tr.) of the glabella contracts; lateral lobes are absent from LO; the posterior margins of the ocular lobes extend back to SO; and the posterior border of the cephalon parallels a transverse line be- fore weakly curving anteriorly distally. The second morphotype, which does not include the holotype of Kielan’s (1960) species, may represent a new species of Olenellus, based on the state of some of its charac- ters, including: the anterior margins of the frontal lobe are directed posteriorly at a roughly 40 degree angle relative to a transverse line; the posterior edge of the ocular lobe extends back to the medial part of the distal margin of LO; the lateral margins of L4 are distal to the lateral margins of LO; the ocular lobes smoothly merge with the extraocular area; the an- terolateral margins of the frontal lobe of the glabella are not prominently separated from the extraocular area; the surface of the interocular area is developed as a flattened shelf; the cephalic posterior border flexes weakly posterolaterally between LO and the intergenal angle; L2 and L3 merge laterally; the anterior cephalic border is developed as a narrow flattened ridge; the lateral margins of the glabella bulge laterally at L2; the genal spine angle is oppo- site LO; there is an axial node on LO; the posterior margin of LO is relatively transverse. This taxonomic assignment will remain tentative until all the material for this species can be considered in greater detail. The other morphotype within Kielan’s (1960) O. svalbardensis, which includes the holotype, represents a species of Mesolenellus, based on its possession of the characters di- 38 PEABODY MUSEUM BULLETIN 45 agnostic for the Mesonacinae and Mesolenellus elucidated in this analysis and recovered from Lieberman (1998). In addition, the phylogenetic analysis of the Mesonacinae con- ducted here suggests that this species is sister to Mesolenellus hyperborea, so it is henceforth treated as a species of Mesolenellus. Kielan (1960) originally suggested that this species was closely related to Fritzolenellus truemani (Walcott 1913). As mentioned above, Kielan’s (1960) species consisted of two dis- tinct taxa. The one treated as Mesolenellus svalbardensis is distantly related to Fritzolenellus, based on phylogenetic topology within the Olenelloidea from Lieberman (1998), and any similarities between the two taxa must be viewed as primitive retentions or convergences (see Lieberman [1998] for character evidence separating Fritzolenellus from Mesolenellus). One of the few characters that both M. hyperborea and F. truemani share in common is the presence of lateral lobes on LO. The other specimens of Kielan’s (1960) that are referable to Olenellus do appear to match the diagnostic characters of the genus Olenellus rather than Fritzolenellus, as mentioned above. Most of the specimens that belong to M. svalbardensis have faintly incised glabellar furrows, but the holotype and another specimen illustrated in Kielan (1960) do show faint traces of the glabellar furrows. Thus, these specimens were used for the purpose of coding the characters that describe the glabellar furrows. Poulsen (1974:82, pl. 1, figs. 1-3) figured specimens from the Schley Fjord Formation, Peary Land, in eastern North Greenland, which he referred to as O. cf. svalbardensis. The specimens illustrated appear similar if not conspecific with Kielan’s (1960) O. svalbardensis, and are provisionally assigned here to that species. One possible difference between the Greenland and Spitsbergen material is that, in the specimens Poulsen (1974) illustrated, the lateral margins of the glabella appear to be diverging somewhat more strongly anteriorly than they do in the Spitsbergen material. Knoll and Swett (1987) and Knoll et al. (1989) commented on the close relationship between faunas in Spitsbergen and Greenland, and thus it is not surprising that some species might be distributed in both regions. Occurrence: Norway: Spitsbergen, the lower part of the Slakli Series, Sofiekammen Forma- tion, on the northern side of Hornsund, West Spitsbergen, co-occurring with a probable specimen of Nevadella Raw 1936, though this requires further analysis. This species either occurs in the Olenellus zone or the “Nevadella” zone, congruent with the first appearance of M. hyperborea. GENUS MESONACIS WALCOTT 1885 Type species: Olenus vermontanus Hall 1859. Assigned taxa: Olenellus fremonti Walcott 1910; O. bonnensis Resser and Howell 1938; O. ea- gerensis Best 1952; O. hamoculus Cowie and McNamara 1978; O. cylindricus Palmer in Palmer and Halley 1979. Diagnosis: Anterolateral margins of frontal lobe of glabella not prominently separated from extraocular area by furrow; posterior margins of frontal lobe expanding where ocular lobes intersect with it; anterior portions of posterior margins of ocular lobes opposite medial part of distal margins of L1; posterodistal margins of L3 formed by ocular lobes; length (exsag.) of genal spine less than or equal to length (sag.) of first four thoracic segments; lateral lobes absent from LO; intergenal angle forms 30 to 50 degree angle with transverse line; lateral margins of each prosomal thoracic axial ring converging when proceeding from anterior to posterior edges; field of thoracic pleural furrows at medial portion long (exsag.), equal to approximately 1.3 times length of posterior pleural band; anterior margin of T3 medially SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 39 flexes anteriorly; posterior margin of thoracic pleural furrow on T3 directed evenly poste- riorly laterally; width (tr.) of thoracic axis 100% width of pleural field excluding spines. Discussion: This genus is distributed throughout Laurentia, from the Great Basin to Ver- mont, Newfoundland and Scotland, in strata of the Olenellus zone. As discussed above under the heading Mesonacinae, there is strong character evidence to support a separation of the genus Mesonacis (as well as Mesolenellus) from the genus Olenellus, including what was formerly referred to as O. (Paedeumias), and here it is treated as distinct. Character dif- ferences between the genus Mesonacis and Olenellus include: 1. in Mesonacis the transverse profile of the ocular lobes is convex dorsally, whereas it is flattened dorsally in Olenellus; 2. in Mesonacis the posterior tips of the ocular lobes are developed opposite the medial part of the distal margin of L1, in Olenellus they are developed opposite the medial part of the dis- tal margin of LO; 3. in Mesonacis a line from the posterior tip of the ocular lobe to the junc- tion of the posterior margin of the lobe with the glabella forms a 15 to 20 degree angle with a sagittal line (except in M. fremonti), whereas in Olenellus it forms a 0 to 5 degree angle; 4. in Mesonacis the intergenal angle is directed anteriorly at least 30 degrees relative to a trans- verse line, whereas in Olenellus it is directed anteriorly about 10 degrees; 5. in Mesonacis the thoracic pleural spines on all segments but the third are narrow (tr.), roughly one quarter of the length (exsag.) of the corresponding pleural segment at its medial part between the spine and axis (except in M. fremonti), whereas in Olenellus they are relatively broader, about half the width of the corresponding pleural segment at its medial part between the spine and axis; 6. in Mesonacis the width (tr.) of the axis is 100% to 105% the width of T3, excluding the spine, while in Olenellus it is 60% to 70%; 7. and in Mesonacis the base of the spine on T15 is narrow (tr.), less than half the width of the axis (except in M. fremonti), whereas in Olenellus it is equal to the width of the axis. It is clear from this character list that species of Mesonacis can now be clearly distin- guished from specimens of Olenellus. The only species that shows any variability in the de- velopment of characters typical of Mesonacis is M. fremonti. This is not surprising, as this species is a near basal member of the Mesonacis clade (obviously the causal relationship is uncertain here because the relationships among species of Mesonacis were at least partly de- termined by reference to the characters given above). Thus, the contention of Resser and Howell (1938), Harrington et al. (1959), Whittington (1989) and Palmer and Repina (1993) that species of Mesonacis cannot clearly be distinguished from species of Olenellus no longer appears valid. Resser (1928) described Mesonacis bristolensis. Harrington (1956) made this species the type of his genus Bristolia. There are many character differences between Mesonacis and Bristolia, elucidated below under the latter genus, and it is clear that B. bristolensis can no longer be assigned to Mesonacis. Best (1952) illustrated the species O. schofieldi Best 1952, which bears some similar- ity to species of Mesonacis. However, it lacks the diagnostic characteristics of that genus that differentiate it from members of the genus Olenellus, and at this time it is excluded from Mesonacis. In particular, in O. schofieldi the intergenal angle deflects anteriorly at a 10 to 15 degree angle relative to a transverse line; a line from the posterior tip of the oc- ular lobe to the junction of the posterior margin of the lobe with the glabella forms a 0 to 5 degree angle with a sagittal line; the ocular lobe is dorsally flattened; and the width (tr.) of the thoracic axis on the third segment is 60% to 70% of the width of the third pleural segment, excluding the spine. The position of the posterior tips of the ocular lobes in O. schofieldi is variably developed. In some specimens, including the holotype, they ex- tend back to SO, is more typical of Olenellus. However, in one small, incomplete specimen 40 PEABODY MUSEUM BULLETIN 45 Figure 8. 1, 2. Mesonacis fremonti (Walcott), Latham Shale, San Bernardino Co., California. 1. S end of Marble Mtns., near Cadiz (loc. 3376, LACMIP), LACMIP 34973, x0.8. locality is the same as Figure 3.4, UCR 10, x0.9. (Best 1952, pl. 1, fig. 4) they may extend back only to the medial part of the distal margin of L1, which is typical of Mesonacis. Because this specimen is poorly preserved, and rep- resents a small individual, the ocular lobes are treated as extending back to SO, but it is important to mention that there is some potential variation here. The thoracic pleural spines are too poorly preserved in O. schofieldi to determine whether or not their condi- tion matches that typically found in species of Mesonacis. On the basis of these character differences O. schofieldi must be excluded from the genus Mesonacis, and instead proba- bly belongs within the genus Olenellus. It will be treated in greater detail in the analysis of the genus Olenellus being undertaken. Some specimens of Mesonacis (e.g., M. bonnensis and M. eagerensis) have a raised transverse line that extends from the interocular area to the posterior cephalic border. This character was assumed to have little phylogenetic significance, and was instead treated as reflecting some taphonomic factor. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 4l MESONACIS VERMONTANUS (HALL 1859) Olenellus vermontanus (Hall); Resser and Howell 1938:220, pl. 4, figs. 15, 16; Shaw 1955:793, pl. 75, fig. 10 (see for more complete synonymy); Whittington 1989:120, figs. 14, 16, 18. Olenellus (Mesonacis) vermontanus (Hall); Palmer and Repina 1993:22, fig. 3.2; Palmer and Repina 1997:408, fig. 255.2. Mesonacis vermontanus (Hall); Lieberman 1998:67. Olenellus georgiensis; Resser and Howell 1938:220, pl. 5, fig. 7, non fig. 6. Types: Hall’s (1859) original type of the species, AMNH 230, was reported lost by Resser and Howell (1938) and Whittington (1989). Thorough additional examination of the col- lections of the AMNH failed to produce the specimen. Therefore, a new type is needed. Wal- cott specimen from the type locality (1910, pl. 26, fig. 3), also illustrated by Resser and How- ell (1938, pl. 4, fig. 16), is complete and beautifully preserved. That specimen, USNM 15399a, is designated the neotype of the species. Resser and Howell (1938) referred to this specimen as a plesiotype. Material examined: MCZ 2434, and possibly 108662, which is incompletely preserved. Discussion: The holotype of Olenellus georgiensis Resser and Howell bears strong similarity to M. vermontanus and, although deformed, appears to be morphologically indistinguish- able from that species, as Shaw (1955) recognized. Henceforth it is treated as a junior sub- jective synonym of M. vermontanus. However, the specimen illustrated as O. georgiensis (Resser and Howell 1938, pl. 5, fig. 6) is not conspecific with M. vermontanus and instead appears to be a specimen of O. thompsoni (Hall). In particular, the pleural spines of T3, and the other thoracic pleural spines, are much longer than those in M. vermontanus and match the condition found in O. thompsoni. Further characters indicating that this specimen be- longs to O. thompsoni rather than M. vermontanus include: the ocular lobes extend back to the medial part of the distal margin of LO rather than the medial part of the distal margin of L1; the intergenal angle forms approximately a 0 to 10 degree angle with a transverse line, rather than a 30 to 35 degree angle; the extraocular area is relatively broad; and the lateral margins of the glabella between the posterior margin of LO and the distal tip of S1 converge. Occurrence: Vermont: Parker Slate, Parker’s Quarry, Georgia, middle upper Olenellus zone, according to Palmer and Repina (1993). MESONACIS FREMONTI (WALCOTT 1910) Figures 8:15 8.2594 Olenellus fremonti; Walcott 1910:320, pl. 37, figs. 1, 2; Riccio 1952:30, pl. 7, fig. 6; Mount 1976:176, fig. 7; Palmer in Palmer and Halley 1979:70, pl. 3, figs. 14-17; Mount 1980:22, fig. 7. Olenellus (Fremontia) fremonti Walcott; Lochman in Cooper et al. 1952:91, pl. 18, figs. 4, 5. Mesonacis fremonti (Walcott); Resser 1928:6, pl. 1, figs. 3-9, pl. 2, fig. 9, pl. 3, fig. 8. Fremontia fremonti (Walcott); Harrington 1956:57, text fig. 1b, non pl. 15, fig. 6. Fremontia sp.; Nelson 1976:31, pl. 8 (left side of plate). Types: Lectotype, USNM 56819a. Walcott (1910) figured and assigned several specimens to his new species O. fremonti and gave as the type locality for this species his locality 52, Prospect Peak, Eureka District, Nevada. Resser (1928) excluded all but one of the specimens from Walcott’s (1910) type locality from his M. fremonti (Walcott 1910, pl. 37, fig. 2). None 42 PEABODY MUSEUM BULLETIN 45 of the other specimens from Walcott’s (1910) locality 52 were assigned to any other species by Resser (1928), nor were grounds for removing these specimens from fremonti given. Sub- sequently Harrington (1956) designated USNM 56819a as lectotype for the species. (Lochman in Cooper et al. 1952 suggested that this specimen could potentially be the type for the species.) This was necessary to insure taxonomic stability. Conceivably, any one of the specimens from the type locality not treated as new taxa and described, or lumped with previously described taxa, could serve as the type specimen of M. fremonti. Harrington (1956) gave good grounds for the designation of his lectotype and provided justification for excluding most of Walcott’s (1910) other specimens of fremonti from that species. Unfortu- nately, the specimen he chose to illustrate as representative of Fremontia fremonti, U3SNM 56819n (Walcott 1910, pl. 37, fig. 18) from the type locality of fremonti, is actually a species of Bristolia, closely related or perhaps conspecific to B. fragilis Palmer (Palmer and Halley [1979]) (see the discussion of characters separating Bristolia from Mesonacis under the genus Bristolia, below). Palmer, in Palmer and Halley (1979), for some unspecified reason ignored Harrington’s (1956) lectotype designation and mistakenly asserted that Resser (1928), by removing all but one of the specimens from the type locality from his M. fre- monti, had automatically designated that specimen (Walcott 1910, pl. 37, fig. 2; USNM 56819a) the holotype of the species. This is incorrect, as there were several specimens from the type locality whose taxonomic affinities had not been established. In addition, Resser (1928) had designated no type specimen. Harrington’s (1956) definition serves as the es- tablishment of the lectotype of the species. If Resser (1928) had designated that specimen as a type, he would have validly designated a lectotype for O. fremonti. Discussion: Raw (1936) treated O. fremonti as the type of the genus Fremontia Raw 1936, and Harrington (1956) elaborated on this. It is conceivable that Fremontia could remain a valid genus; however, there are several arguments against this. First, based on phylogenetic topology within the genus Mesonacis (Figure 7), the type of the genus Mesonacis, M. ver- montanus, is nested within a clade that is sister to M. fremonti. If M. fremonti were to be treated as part of Fremontia, then either M. bonnensis would have to be assigned to a para- phyletic “Fremontia,” or it would need to be treated as a monotypic genus. This of course assumes that it is desirable for the genus Mesonacis to remain a yalid taxonomic concept, that is, monophyletic. Establishing evolutionary patterns within the genus Mesonacis was one of the goals of this analysis, and thus it is essential that Mesonacis be monophyletic. In order to avoid naming new monotypic genera, for purposes of taxonomic stability and to avoid establishing paraphyletic genera, M. fremonti is lumped within the rest of the genus Mesonacis. It is conceivable that the genus Mesonacis could be lumped within a Fremontia that included F. fremonti. However, since Mesonacis is a more established name, has prece- dence and is most frequently used in the literature, it seems preferable to retain that name rather than Fremontia. In the phylogeny of the genus Mesonacis, M. fremonti maps as a near basal species within that genus, and characters separating the other species of Mesonacis from this species can be seen in Figure 7, and in Tables 3 and 4. Not surprisingly, primitive retentions found in M. fremonti are also common to species of Olenellus. Some of these include the relatively broad (tr.) extraocular area; the absence of pleurae of the opisthothorax in M. fremonti (ad- mittedly indeterminate in some species of Mesonacis); and the relatively long (sag.) axial spine on the 15th segment, which is also quite broad (tr.) near the axis. Lochman in Cooper et al. (1952) figured a few specimens that she assigned to Olenel- lus (Fremontia) fremonti. One of these (Cooper et al. 1952, pl. 18, fig. 5,) is poorly preserved but appears to possess the diagnostic characters of the species, and is thus here provision- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 43 ally assigned to M. fremonti. Some of the specimens Palmer and Halley (1979) figured as M. fremonti are extremely deformed, but appear to bear the diagnostic characteristics of the species and are treated as members of M. fremonti. This species shows some variation in the condition of the posterior cephalic border between LO and the intergenal angle. In some specimens it is transverse, and in others it is weakly deflecting posteriorly. Material examined: LACMIP 4908-18, 5870 and 34973 (and large numbers of unnumbered specimens in the general stratigraphic collections); SDSNH 168383, 16874, 16954, 16955, 16974, 19099, 20709 (8 specimens), 20806 (2 specimens), 26220; UCR 10/2004 and a large number of unnumbered specimens in the general stratigraphic and taxonomic collections assigned to lots 10, 7270 and 7271; MCZ 2544, 2545, 7370; USNM 56819i. Occurrence: California: In the middle part of the Latham Shale, southern end of the Mar- ble Mountains, E slope of hill 1440, 834 ft E and 1999 ft N of the SW corner sec 12, T 5 N, R 14 E, and in the Latham Shale, 190 m W of the limestone quarry, 0.5 mi E of Cadiz, in the Mojave Desert portion of San Bernardino County, possibly equivalent to Hazzard’s (1933) locality M-5, treated as in the Bristolia zonule, upper Olenellus zone; from the upper Poleta and lower Harkless Formations, lower Olenellus zone, and the Mule Spring Limestone, upper Olenellus zone, White/Inyo Mountains region (Nelson 1976); the Pyramid Shale Member, Carrara Formation, Olenellus zone, Funeral Mountains, Resting Springs Range, and Salt Spring Hills, White/Inyo Mountains region (Palmer and Halley 1979). Nevada: In arenaceous shales at the summit of Prospect Mountain, Eureka County. MESONACIS BONNENSIS (RESSER AND HOWELL 1938) Olenellus bonnensis; Resser and Howell 1938:222, pl. 7, fig. 3. Olenellus terranovicus; Resser and Howell 1938:222, pl. 6, fig. 8. Olenellus brevoculus; Resser and Howell 1938:225, pl. 8, figs. 18, 19; Best 1952:19. Types: Holotype, USNM 90808, see Resser and Howell (1938). Discussion: The three separate species that Resser and Howell (1938) figured and described are identical and come from the same locality. Thus, they are treated as belonging to the same species. The name of the species, based on the best preserved material, is retained. Occurrence: Canada: Newfoundland, Forteau Formation, Olenellus zone, E shore of East Arm, Bonne Bay, west coast of Newfoundland. MESONACIS EAGERENSIS (BEST 1952) Olenellus eagerensis; Best 1952:19, pl. 1, figs. 5-9. ?Olenellus eagerensis; Best 1952:19, pl. 1, figs. 10-12. Types: Holotype, University of British Columbia GT 101, see Best (1952). Discussion: Some of the specimens illustrated by Best (1952, pl. 1, figs. 6, 8) appear to be somewhat deformed and crushed sagittally, causing L4 to appear relatively shorter (sag.) than it should. However, the holotype and the other figured specimens of this species have the characteristic long (sag.) L4, circa 1.5 times the length of LO and L1. Best (1952, pl. 1, figs. 11, 12) figured two specimens that he claimed were closely related varieties of O. ea- gerensis. These specimens have a more dorsally expanded L4, a narrower extraocular area and a genal spine angle further anterior than what is found in the holotype and the 44 PEABODY MUSEUM BULLETIN 45 Figure 9. Mesonacis spp. 1. Mesonacis fremonti (Walcott), Pioche Fm., Prospect Mt., Eureka District, Nevada, USNM 56819i, X1.5. 2. Mesonacis cylindricus (Palmer in Palmer and Halley 1979), 10 to 10.5 ft above base of Latham Shale, in small NE trending gully on E side of hill 1440 in S end of Marble Mtns., 834 ft. E and 1,999 ft N of SW corner sec 12, T 5 N, R 14 E, San Bernardino Co., California, UCR 7897, x1.4. paratypes of the adult specimens of M. eagerensis. These specimens appear to be highly de- formed and it is hard to determine if they are truly different from M. eagerensis. At this time they are questionably assigned to M. eagerensis. Similarly, Best (1952, pl. 1, fig. 9) illustrated a small cephalon that shows some morphological similarity to M. eagerensis, but may also be similar to the specimens in Best (1952, pl. 1, figs. 10-12) in having a narrow extraocular area and an expanded L4. Again, it is questionably assigned to M. eagerensis. Occurrence: Canada: British Columbia, the Eager Formation, the Olenellus zone, just S of the Fort Steele-St. Eugene Mission road, 6 mi NE of Cranbrook. MESONACIS HAMOCULUS (COWIE AND MCNAMARA 1978) Olenellus hamoculus; Cowie and McNamara 1978:627, pl. 70, figs. 3-9; McNamara 1978:641, text figs. 4e, f. ; Olenellus (Angustolenellus) hamoculus; Palmer and Repina 1993:22, fig. 3.6. Types: Holotype, BGS 13302, see Cowie and McNamara (1978). Discussion: Cowie and McNamara (1978) placed this species in the genus Olenellus and considered it closely related to O. lapworthi Peach and Horne 1892, O. reticulatus Peach 1894, O. intermedius Peach 1894 and Olenelloides armatus Peach 1894. What Cowie and McNamara (1978) and McNamara (1978) referred to as Olenellus reticulatus and O. lap- worthi, two species from the middle Olenellus zone of northwest Scotland, actually appear to belong to the genus Fritzolenellus and appear to bear all the diagnostic characters of that genus (see the discussion below under the genus Fritzolenellus). Fritzolenellus is distantly related to Mesonacis, based on the phylogenetic analysis conducted in Lieberman (1998), and thus the contention of Cowie and McNamara (1978) that M. hamoculus is closely re- lated to either F. reticulatus or F. lapworthi must be strongly questioned. Because these Scottish taxa comprise distantly related species that happen to co-occur in rocks of simi- lar age from the same region, the paedomorphic and evolutionary pathway that McNa- mara (1978) saw connecting these species, as well as Olenelloides armatus, cannot be ac- cepted. Similar ages and geographic distributions are not generally accepted as a means of establishing hypotheses of evolutionary relationship. Olenellus intermedius Peach was also SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 45 placed into the paedomorphic evolutionary pathway with F. reticulatus, F. lapworthi and M. hamoculus by McNamara (1978). Olenellus intermedius is known from two poorly pre- served specimens, both early ontogenetic stages; however, these appear to be conspecific with, or belong to, a species closely related to either F. reticulatus or F. lapworthi. This species is discussed in greater detail under the genus Fritzolenellus, below. Species of Mesonacis and Fritzolenellus can be distinguished by the following charac- ters: in Mesonacis L4 does not expand prominently dorsally, in Fritzolenellus it does; in Mes- onacis the anterolateral margins of the ocular lobes smoothly merge with the extraocular area, in Fritzolenellus the anterolateral margins of the ocular lobes are separated from the extraocular area by a furrow; in Mesonacis the interocular area is developed as a flattened shelf, in Fritzolenellus it is arched; in Mesonacis the lateral lobes on LO are absent, in Frit- zolenellus they are present (this character is not clearly visible in F. reticulatus or F. lapwor- thi as these species are poorly preserved); in Mesonacis the intergenal angle is directed ante- riorly at least 30 degrees relative to a transverse line, in Fritzolenellus it is directed anteriorly only 0 to 10 degrees; in Mesonacis the anterior margin of the third thoracic pleural segment is directed anteriorly before flexing posterolaterally, in Fritzolenellus it parallels a transverse line; in Mesonacis the anterior margin of the thoracic pleural furrow on T3, when proceed- ing from the proximal to the distal edge, parallels a transverse line, before flexing postero- laterally, while in Fritzolenellus it flexes weakly posterolaterally; in Mesonacis the length (exsag.) of the thoracic pleural furrows (excluding T3) at the medial part of the segment are equal to 1.5 times the length of the posterior band of the pleural segment, in and Frit- zolenellus their length is less than or equal to the length of the posterior pleural band; and in Mesonacis the base of the spine on the 15th thoracic segment is less than half the width (tr.) of the axis of the segment (except in M. fremonti), whereas in Fritzolenellus it is nearly as wide (tr.) as the axis of the segment. Palmer and Repina (1993) made M. hamoculus the type of their new subgenus Olenel- lus (Angustolenellus), but M. hamoculus is nested within a clade of several species assigned to the genus Mesonacis. The name Angustolenellus could be retained, but this would entail either erecting several new monotypic genera or establishing paraphyletic genera. This is not desirable, and hence Angustolenellus is relegated to a synonym of Mesonacis. Material examined: BGS 13302, the holotype, and BGS 13298, a paratype. Occurrence: Scotland: The “Fucoid” Beds, middle Olenellus zone according to Palmer and Repina (1993), from a roadside quarry 170 m E 10 degrees S of NE inlet of Loch Ae, Suther- land, NW Scotland. MESONACIS CYLINDRICUS (PALMER IN PALMER AND HALLEY 1979) Figure 9.2 Olenellus cylindricus; Palmer in Palmer and Halley 1979:69, pl. 2, figs. 9, 10, 13 non 14. Types: Holotype, USNM 177197, see Palmer and Halley (1979). Discussion: Palmer questionably assigned a specimen to this species (Palmer and Halley 1979, pl. 2, fig. 14) that differs in the condition of several character states from that found in M. cylindricus. In particular, this specimen lacks an advanced genal spine angle; has a transverse intergenal angle instead of forming an angle with a transverse line of roughly 50 degrees; has a faintly incised, transverse S2 that is not conjoined, does not contact the axial furrows and has the medial edges directed posteriorly, rather than a convex S2 that is con- joined medially, contacts the axial furrows and has the distal and medial edges as far for- 46 PEABODY MUSEUM BULLETIN 45 ward; the cephalic posterior border is transverse between LO and the intergenal angle rather than being directed posteriorly; and SO and S1 are not conjoined, rather than being con- joined. On the basis of these character differences this specimen is excluded from M. cylin- dricus. It may instead be referable to Wanneria humilis Robison and Hintze 1972. Material examined: UCR 7897. Occurrence: California: Eagle Mountain Shale, Carrara Formation, the Olenellus arcuatus zonule, upper Olenellus zone, Grapevine Mountains, White/Inyo Mountains region, ac- cording to Palmer and Halley (1979); and 10 ft to 10.5 ft above base of Latham Shale, upper Olenellus zone, in small NE trending gully on E side of hill 1440 in S end of Marble Moun- tains, 834 ft E and 1999 ft N of SW corner sec 12, T 5 N, R 14 E, San Bernardino County. Phylogenetic Analysis of Elliptocephala The genus Elliptocephala is a moderately diverse clade within the Olenelloidea (see Palmer and Repina 1993; Lieberman 1998). Using the diagnostic characters of Elliptocephala rec- ognized by a higher level phylogenetic analysis of the Olenelloidea given in Lieberman (1998), it was possible to recognize 11 previously described species and one new species from the Lower Cambrian Olenellus zone of Laurentia and the Schmidtiellus mickwitzi zone of Baltica that belong within that genus. Eleven of these species were subjected to a phylo- genetic analysis to determine their relationships. A total of 12 taxa was subjected to phylogenetic analysis, including 11 ingroup taxa. All available adequately preserved species within the genus were analyzed. Additional species possibly referable to Elliptocephala are discussed below. Wanneria walcottana was used as the outgroup taxon in phylogenetic analysis, based on phylogenetic topology within the Olenelloidea determined in Lieberman (1998). According to Lieberman (1998), Ellipto- cephala and Wanneria are very closely related, though not sister taxa, and fewer character differences separate Wanneria from Elliptocephala than separate Elliptocephala from any other taxon considered in Lieberman’s (1998) analysis. Evolutionary relationships within Elliptocephala were determined by parsimony analysis of 20 exoskeletal characters (Tables 5 and 6). Unfortunately, only cephalic charac- ters could be used because of the dearth of thoracopygidia known for species of Ellipto- cephala. These data were subjected to an exhaustive search using PAUP v. 3.1.1. All charac- ters were treated as unordered. Two most parsimonious trees of length 53 steps were recovered. These trees differed solely in the relative placement of Elliptocephala parvifrons (Fritz 1972) and E. logani, and a strict consensus of these two trees is shown in Figure 10. The retention index is 0.53 and the consistency index is 0.43. The g; statistic was —0.33. This value differs significantly (p<0.05) from values derived from distributions using random data (Hillis 1991) and suggests a strong and robust phylogenetic signal in the character data given in Table 6. Confidence values for the nodes of the consensus tree duplicated in the bootstrap analysis are given in Figure 10. Thirty-six trees of length less than or equal to 54 steps were recovered before the analysis of Bremer (1994) was terminated because the consensus cladogram was a complete polytomy. Thus, the total support index (Bremer 1994) for the tree is very low. Lieberman (1998) suggested that there might be some incongruence between phylo- genetic position and stratigraphic occurrence, particularly centered around the phyloge- netic placement of the genera Elliptocephala, Wanneria (assigned in Lieberman [1998] to the paraphyletic “Wanneriidae”) and the Holmiidae. The latter comprises several genera, in- cluding Holmia, Kjerulfia, Esmeraldina, Schmidtiellus and Holmiella, relative to the Olenel- lidae. The Holmiidae in Lieberman (1998) are derived within the Olenelloidea relative to several taxa, including the Olenellidae, Elliptocephala and Wanneria. Past studies (e.g., Palmer and Repina 1993) have suggested that the Olenellidae do not appear until the sub- sequent Olenellus zone, with Elliptocephala and Wanneria not appearing until the lower 48 PEABODY MUSEUM BULLETIN 45 7) = 7) = 3 2) — BY v 3 = 2 o S ¢ 7) iS E 7 AS) = S Ss a) © — = 2 © iS) c < 9 & 3 is) nn ON N (oe) \oO So i) Oo = on oO’ N oo \O S = i=) Wanneria walcottana 0 — _ Elliptocephala asaphoides | ee eS 6 OD — Elliptocephala sequomalus _ Elliptocephala mediocris Elliptocephala walcotti Elliptocephala bicensis Elliptocephala laxocules Elliptocephala paraoculus Elliptocephala parvifrons Elliptocephala logani Le ce cee ce ce a) ES | | SS CO OFF Ke OK CO SS &] 2 oS Se eS S&S ISIS Ss dis OS OOS lO OS) Elliptocephala mirabilis aS St CO SS ES OlUCOCCUCCOCUCO RS NO |= =| =| S| RB LO KH KH YH KH CO — So oO eo oe = So es oS ES —|& CO Fe Fe ee et KF CO CO COC SS SSS Sore oS oS oS > © > eee Saws i Ss s SO SS] OSS ia Ss) ©) S&S oS eC Oo iS oS Oo CS i SCOrFrOoOrrFoFrFrFOoOoOo°o SGoOorrFrrePrFoo9oecoeo°o |S —|-& CO CO eee Fe KF OK CO SSS a Ss ow S&S oS 2 iS Eo eo oS is © ~ ~ ~ Elliptocephala lundgreni additional species of Elliptocephala in this analysis reveals that characters 3 and 4 still reli- ably distinguish Wanneria from all members of Elliptocephala, though admittedly no py- gidial remains for this genus are known aside from the type of the genus. However, the two most derived species of Elliptocephala relative to Wanneria, E. mirabilis (Poulsen 1958) and E. lundgreni do possess a relatively broad anterior cephalic border (character 1). In addi- tion, three species of Elliptocephala considered here, E. paraoculus (Fritz 1972), E. laxocules (Fritz 1972) and E. mediocris (Poulsen 1958), do not have a prominently developed parafrontal band (character 2). This indicates that some of the characters separating Wan- neria from Elliptocephala do show homoplasy, and this underscores the contention of Lieberman (1998) that the branch separating Wanneria and Elliptocephala from one an- other in the phylogeny of the Olenelloidea is a short one. This matches the conclusion of Bergstr6m (1973) and Ahlberg et al. (1986), who suggested that Elliptocephala and Wan- neria were Closely related. However, one of these two characters reliably distinguishes Wanneria from E. logani: the relative position of the posterior margin of the ocular lobe may show ontogenetic variation (Westrop, pers. comm. 1998). Specifically, during the ontogeny of some olenel- loids the ocular lobes become progressively shortened (see Fritz 1972, pl. 9, figs. 1, 5, 6, 12). Specimens of W. walcottana are typically much larger than specimens of E. logani. This size difference, and thus the relative position of the posterior margin of the ocular lobes, might be related to taphonomic factors. For example, perhaps large specimens of W. walcottana are more likely to be found than large specimens of E. logani because the habitat that large specimens of W. walcottana occupied is easily sampled for fossils, and SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 53 in reality both taxa show similar size distributions (with size being a proxy for age). If this is in fact the case, then these taxa would be differentiated on spurious characteristics. On the other hand, the differences in size could reflect true, evolutionary differences between W. walcottana and E. logani since two closely related taxa are likely to differ in minor fea- tures, such as overall size, with only a simple heterochronic shift involved in producing the morphological differences. Because these two taxa come from geographically dis- parate regions that have been extensively sampled for fossils, and because the W. walcot- tana morphotype is not known from the localities where E. logani is found, and vice versa, the position put forward here is that the size related differences between the two taxa are evolutionary, and not taphonomic. Additional cephalic characters distinguish Wanneria from all members of the genus Elliptocephala. Specifically, the anterior margins of the frontal lobe at each side of the midline are deflected posteriorly relative to a transverse line at a larger angle in Ellipto- cephala than in Wanneria, and the posterior margin of LO is of greater convexity in Ellip- tocephala than in Wanneria. Thoracic characters separating Wanneria from Elliptocephala (only preserved in E. asaphoides and E. bicensis [Walcott 1910]) include: the distal part of the posterior margin of the pleural furrow of T3 is directed weakly anterolaterally in Wan- neria, whereas it is directed posterolaterally in Elliptocephala; pleural spines T5 to T8 sweep roughly four segments back in Wanneria as opposed to roughly one to two seg- ments back in Elliptocephala; the pleural furrows extend only half the width of the inner pleural region in Wanneria, whereas they extend onto the spine in Elliptocephala; the tho- racic pleural furrows (excluding T3) at the medial portion of the segment relative to the length (exsag.) of the posterior band of the pleural segment are long in Wanneria, 1.3 to 1.5 times the length of the posterior band, whereas they are short, equal to the length of the posterior band, in Elliptocephala; and the spine on the 15th thoracic segment is rela- tively much longer in Elliptocephala (only known in E. asaphoides) than it is in Wanneria. A pygidial character separating Wanneria from Elliptocephala (only preserved in E. as- aphoides) is the condition of the posterior margin, which is weakly convex in Ellipto- cephala and bifurcated with a median notch in Wanneria. Thus, based on all this charac- ter evidence, and following Lieberman (1998), the genus Wanneria is still considered to be closely related to Elliptocephala, but the two do not share a sister group relationship. As the branch between Elliptocephala and Wanneria is a short one, these two genera were placed within a single paraphyletic family, the “Wanneriidae,’ rather than erecting mono- typic families for each of them. Occurrence: Pennsylvania: the Kinzers Formation, Lancaster County, treated as in the mid- dle upper Olenellus zone, following Palmer and Repina (1993). GENUS ELLIPTOCEPHALA EMMONS 1844 Type species: Elliptocephala asaphoides Emmons 1844 Assigned taxa: Olenellus sequomalus Fritz 1972; Wanneria mediocris Poulsen 1958; Ellipto- cephala walcotti n. sp.; Callavia bicensis Walcott 1910; Olenellus laxocules Fritz 1972; Olenel- lus paraoculus Fritz 1972; Wanneria parvifrons Fritz 1972; Olenellus logani Walcott 1910; Holmia mirabilis Poulsen 1958; Olenellus praenuntius Cowie 1968; Olenellus lundgrent Moberg 1892. Diagnosis: Length (sag.) of L4 equal to roughly 1.5 times length of LO and L1 (sag.); ante- rior margins of frontal lobe at each side of midline deflected posteriorly at roughly 40 de- gree angle relative to transverse line; lateral margins of L4 distal to lateral margins of LO; 54 PEABODY MUSEUM BULLETIN 45 Figure 11. 1, 2. Wanneria walcottana (Wanner), Kinzers Shale, Pennsylvania. 1. near Lancaster, YPM 9295, x0.9. 2. locality is the same as Figure 3.8, YPM 29269, x1.1. 3. Holmia kjerulfi (Linnarsson) Tomten, Ringsaker, Norway, PIL Lo 4456t, x1.4. preocular furrow on L4 directed inward and forward from glabellar margin; posterior margins of L4 divergent anteriorly; distal margins of L3 convex outward; middle sector of S3 convex anteriorly; L2 and L3 do not merge; distal margins of L2 diverging anteriorly; line between ends of S2 transverse; $2 convex anteriorly; ocular lobe with prominent ocu- lar furrow; line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms 10 to 20 degree angle with sagittal line; anterodistal margins of L3 formed by ocular lobes; intergenal angle developed posterior of point halfway between oc- ular lobes and genal spine angle; genal and intergenal ridges prominently developed; dis- tal sector of SO with proximal end well posterior of distal end; axial part of LO with node; SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 55 posterior margin of LO convex posteriorly; lateral lobes on LO absent; thoracic pleural fur- rows extend onto spines; length (exsag.) of furrows equal to length of posterior band; an- terior margin of thoracic pleural furrow on T3 parallels a transverse line before flexing strongly posteriorly; long spine on 15th thoracic segment, spine broad (tr.) at its base; py- gidium twice as broad as long, posterior margin weakly convex. Discussion: This genus is widespread, distributed throughout eastern (New York State, Que- bec, Labrador and Greenland) and western Laurentia (Northwest Territories), and into Baltica as well. Thus it occurs in the Acado-Baltic and Pacific Provinces of the Olenellid Realm (sensu Cowie and McNamara 1978), in strata assigned to the lower and middle Olenellus zone in Laurentia and in strata assigned to the Schmidtiellus mickwitzi zone, the basal trilobitic strata, in Baltica. As discussed above under Wanneria, Elliptocephala is closely related to Wanneria but does not share a sister taxon relationship with it. As mentioned above under the genus Olenellus, Rasetti (1966, 1967) assigned several specimens to the genera Olenellus and Paedeumias. However, these specimens instead ap- pear referable to the genus Elliptocephala since they have many of the characters diagnostic of that taxon. In particular, the anterior margins of frontal lobe at each side of midline are deflected posteriorly at roughly 40 degree angle relative to transverse line; the preocular fur- row on L4 is directed inward and forward from the glabellar margin; the distal margins of L3 are convex outward; the middle sector of S3 is convex anteriorly; L2 and L3 do not merge; the distal margins of L2 are diverging anteriorly; a line between the ends of S2 is transverse; S2 is convex anteriorly; the ocular lobe has a prominent ocular furrow; a line from the posterior tip of the ocular lobe to the junction of the posterior margin of the lobe with the glabella forms a 10 to 20 degree angle with a sagittal line; the anterodistal margins of L3 are formed by the ocular lobes; the distal sector of SO has the proximal end well pos- terior of the distal end; the axial part of LO has a node; the posterior margin of LO is convex posteriorly; and the lateral lobes on LO are absent. These specimens are probably closely re- lated to Elliptocephala asaphoides. Fritz (1991) figured a partial cephalon, which he referred to as Wanneria sp. 1, that ap- pears to satisfy all the diagnostic characteristics of Elliptocephala and belong to E. parvifrons, though it does show subtle differences from this taxon (discussed below under the species). However, since it is from an early ontogenetic stage and poorly preserved, it is not treated as distinct at this time. ELLIPTOCEPHALA ASAPHOIDES EMMONS 1844 Figure 12.1 Elliptocephala asaphoides Emmons; Walcott 1910:269, pl. 24, figs. 1-10, pl. 25, figs. 1-18, pl. 44, fig. 3 (see for more complete synonymy); Tasch 1952:486, figs. la—d; Lochman 1956:1376, pl. 6, figs. 2-21; Whittington 1957:935, pl. 115, figs. 1-6; Harrington et al. 1959:0194, fig. 135.1; Bergstrém 1973:307; Whittington 1989:128, figs. 45, 46, 48, 50, 51; Palmer and Repina 1993:26, fig. 6.7; Palmer and Repina 1997:414, fig. 262.4; Lieberman 1998:67. Types: Lectotype, USNM 18350a, designated by Palmer and Repina (1993). Material examined: AMNH 222; USNM 18350a, the lectotype; YPM 8190, 35863. Occurrence: New York: the Greenwich Formation, eastern New York State, within the mid- dle Olenellus zone, following Palmer and Repina (1993), detailed locality information given in Landing and Bartowski (1996) and references therein. 56 PEABODY MUSEUM BULLETIN 45 Figure 12. Elliptocephala spp. 1. Elliptocephala asaphoides Emmons, Greenwich Fm., Greenwich, Washington Co., New York, YPM 8190, x0.7. 2-6. Sekwi Fm. type section, 0.2 mi SE of June Lake, Mackenzie Mtns., District of Mackenzie, Canada. 2. Elliptocephala sequomalus (Fritz), GSC 27352, the holotype, x1.6. 3. Elliptocephala wal- cotti, new species, GSC 27291, the holotype, x1.6. 4. Elliptocephala laxocules (Fritz), GSC 27333, the paratype, x1.6. 5. Elliptocephala logani (Walcott), GSC 27376, x1.6. 6. Elliptocephala parvifrons (Fritz), GSC 27355, the holotype, x1.7. 7, 8. Elliptocephala lundgreni (Moberg), Norretorp Fm. at Tunbyholm, east Scania, Sweden, PIL Lo 1398t internal and external molds, x1.6. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 57 ELLIPTOCEPHALA SEQUOMALUS (FRITZ 1972) Figure 12.2 Olenellus sequomalus; Fritz 1972:15, pl. 12, figs. 14-25. Types: Holotype, GSC 27352, see Fritz (1972). Discussion: Fritz (1972) treated this species as belonging to the genus Olenellus. However, it lacks some of the diagnostic characteristics of that genus, and instead appears to bear the hallmark characteristics of the genus Elliptocephala. In particular, the lateral margins of L4 are distal to the lateral margins of LO; the preocular furrow on L4 is directed inward and for- ward from the glabellar margin; the middle sector of $3 is convex anteriorly; L2 and L3 do not merge; a line between the ends of S2 is transverse; $2 is convex anteriorly; the interge- nal angle is developed posterior of a point halfway between the ocular lobes and the genal spine angle; and the posterior margin of LO is strongly convex posteriorly. Thus, because of the possession of these characters, and the absence of other characters diagnostic of Olenel- lus, it must be excluded from that genus and instead assigned to Elliptocephala. Material examined: GSC 27352, the holotype. Occurrence: Canada, Northwest Territories: the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1668 ft above the base of the formation, middle Olenellus zone, following Fritz (1972), south end of the Sekwi Range, Mackenzie Moun- tains, District of Mackenzie. ELLIPTOCEPHALA MEDIOCRIS (POULSEN 1958) Wanneria mediocris; Poulsen 1958:18, pl. 5, figs. 1-5. Wanneria troelseni; Poulsen 1958:21, pl. 5, figs. 6-9. Wanneria ruginosa; Poulsen 1958:19, pl. 6, figs. 1-5. Wanneria subglabra; Poulsen 1958:20, pl. 6, figs. 6-8. Types: Holotype, MUGH 10696, see Poulsen (1958). Discussion: Poulsen (1958) was correct in recognizing the close relationship between this species and W. walcottana. However, E. mediocris is more closely related to, and better fits the diagnostic characters of, the genus Elliptocephala rather than those of the genus Wan- neria, and thus it is treated as belonging to Elliptocephala. Poulsen (1958) recognized several species of Wanneria from northwestern Green- land; specifically, his W. troelseni, W. ruginosa and W. subglabra in the collections of the MUGH. On re-examination these appear to be identical to E. mediocris. This material con- sists of variably preserved cephala, and most of the differences between these “species” seem to center around the fact that they represent substantially different ontogenetic stages. Even then, prominent similarities emerge in the condition of the posterior cephalic border, which lateral to the intergenal angle flexes posterolaterally; the jaggedly convex S2 and $3; the conjoined SO, $1, $2 and S3; the rounded and broad (tr.) L4; and in the relative position of the posterior margins of the ocular lobes. Some differences can be seen be- tween E. mediocris and W. ruginosa, particularly in the latter’s relatively slightly longer (exsag.) anterior cephalic border and more prominently incised S2 and S3. However, these characters vary even within the two illustrated specimens of W. ruginosa. The specimens of W. ruginosa are half the size of E. mediocris and from a presumably much earlier onto- genetic stage. Because these character differences seemed insignificant, all these species 58 PEABODY MUSEUM BULLETIN 45 were treated as conspecific and the specimens of E. mediocris were used for coding charac- ters, as these appear to come from the latest ontogenetic stages available for this taxon. Material examined: MUGH 10696, the holotype; MUGH 10698, holotype of W. troelseni; MUGH 10699, holotype of W. ruginosa. Occurrence: Denmark: Greenland, W of Blomsterbaekken, Kap Leiper, and Marshall Bugt, Inglefield Land, NW Greenland, Lower Cambrian Wulff River Formation, questionably placed somewhere within the Olenellus zone. ELLIPTOCEPHALA WALCOTTI LIEBERMAN NEW SPECIES Figure 12.3 Olenellus truemani Walcott; Fritz 1972:16, pl. 9, figs. 1-14. Types: Holotype, GSC 27291, designated here, the best specimen of Fritz’s (1972) material housed in the GSC. Diagnosis: Length (exsag.) of anterior cephalic border between lateral margins of L4 and genal spine angle equal to 50% to 60% length (sag.) of LO; L4 does not contact anterior border furrow and does not expand prominently dorsally; plectrum not visible in dorsal view; prominent parafrontal band visible, width (tr.) near ocular lobes equal to length (sag.) medially; ocular lobe only subtly separated from extraocular area by shelf; poste- rior end of ocular lobes extend back to S0; S3 jaggedly convex; $3, $2, S1 and SO conjoined medially; S2 and $3 prominently incised; S1 transverse; intergenal angle represented by distinct dorsal swelling; posterior cephalic border between intergenal angle and genal spine transverse. Description: Anterior cephalic border narrow, flattened ledge, length (exsag.) between lat- eral margins of L4 and genal spine angle equal to 50% to 60% length (sag.) of LO; frontal lobe does not contact anterior border furrow; prominent parafrontal band visible in dor- sal view; plectrum not visible; anterior margin of frontal lobe at each side of midline de- flected posteriorly at roughly 40 degree angle relative to a transverse line; lateral margins of frontal lobe distal to lateral margins of LO; lateral margins of L4.divergent anteriorly; an- terodistal margins of L3 formed by ocular lobes, distal margins of L3 convex outward; S3 jaggedly convex, conjoined medially; $2 convex anteriorly, conjoined medially; L2 and L3 do not merge distally; line from anterior to posterior edge of ocular lobe forms roughly 10 degree angle relative to sagittal line; SO, S1 transverse, contact axial furrows, medial edges declined posteriorly, conjoined medially; width (tr.) of glabella opposite margins of L1 constricted slightly relative to width at lateral margins of LO; posterior edge of ocular lobe opposite distal tip of SO; ocular lobes weakly elevated from extraocular area; posterior mar- gin of LO convex posteriorly, with axial node medially; genal spines of length (exsag.) ap- proximately 4 to 5 times length (sag.) of LO, sweeping posterolaterally at roughly 15 degree angle relative to sagittal line; genal spine angle opposite LO; intergenal angle developed as faint dorsal projection, inclined at 10 to 15 degree angle relative to transverse line; ex- traocular area opposite L1 broad, width (tr.) approximately 75% width of glabella at L1; faint anastomosing ridges visible on extraocular area; faint intergenal and genal ridges vis- ible; posterior margin of cephalic posterior border between LO and intergenal angle weakly flexing posterolaterally. Discussion: Originally Fritz (1972) treated this species as conspecific with O. truemani. However, Fritz (1992) recognized that this identification was no longer tenable, though he still treated this species as belonging to the genus Olenellus. First, Olenellus truemant SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 59 belongs to Fritzolenellus and differs in a large number of characters from members of the genus Olenellus (characters elucidated in Lieberman [1998]) so that generic assignment can no longer be upheld. In addition, what Fritz (1972) identified as O. truemani lacks some of the diagnostic characters of Fritzolenellus and instead matches those diagnostic of the genus Elliptocephala (given above). In particular, this new species, E. walcotti, dif- fers from the type of Fritzolenellus, F. truemani, in the condition of the following charac- ters: E. walcotti has the length (exsag.) of the anterior cephalic border near but not di- rectly in front of L4 equal to 50% to 60% length (sag.) of LO, while in FE truemani it is equal to the length of LO; E. walcotti has a prominent parafrontal band, F. truemani does not; the posterior tips of the ocular lobes extend back to SO in E. walcotti, in FE. truemani they extend back to the medial part of the distal margin of L1; E. walcotti lacks lateral lobes on LO, F. truemani has these lobes; E. walcotti has the interocular area sloping evenly from the tip of the ocular lobe to the glabella, in F. truemani the interocular area is some- what arched or swollen; E. walcotti has SO conjoined, in F. truemani SO is not conjoined; and E. walcotti has a more prominently developed genal ridge than does F. truemani. These and other characters mark E. walcotti as distinct from F. truemani and group it with species of the genus Elliptocephala rather than with Fritzolenellus or Olenellus. Material examined: GSC 27291, the holotype. Occurrence: Canada: Northwest territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1271 ft above the base of the formation, lower Olenellus zone, fol- lowing Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. ELLIPTOCEPHALA PRAENUNTIUS (COwlIE 1968) Olenellus praenuntius; Cowie 1968:9, pl. 1, figs. 1-10; Fritz 1972:17. Types: Holotype, GSC 18040, see Cowie (1968). Discussion: This species is nearly identical to E. walcotti, new species. It differs from that species solely in the relative placement of the posterior tip of the ocular lobes, which is op- posite the medial part of the distal margin of LO in Elliptocephala praenuntius, whereas it is opposite SO in E. walcotti. These character states do not seem to vary meaningfully within either species. Because E. praenuntius is poorly preserved and because the original material could not be obtained for study, it was not subjected to phylogenetic analysis, but instead simply placed within Elliptocephala and considered to be closely related to E. walcotti. Occurrence: Canada: 430 ft above the base of the Kane Basin Formation, possibly the lower Olenellus zone, 6 mi NW of the head of Scoresby Bay, eastern Ellesmere Island. ELLIPTOCEPHALA BICENSIS (WALCOTT 1910) Callavia bicensis; Walcott 1910:277, pl. 41, figs. 9, 9a; Rasetti 1948:10, pl. 2, figs. 1-3. Types: Holotype, USNM 56794, see Walcott (1910) and Rasetti (1948). Discussion: Originally, this species was assigned to the genus Callavia. However, the type species of Callavia lacks the diagnostic characteristics of the superfamily Olenelloidea and must henceforth be excluded from that superfamily (Lieberman 1998). In contrast, E. bi- censis clearly bears the hallmark characteristics of the Olenelloidea: the ocular lobes merge with the posterior margin of L4, and the length (sag.) of L4 is equal to the length of LO and L1. In addition, although poorly preserved, the species appears to have all the diagnostic 60 PEABODY MUSEUM BULLETIN 45 characters outlined above of the genus Elliptocephala, based on higher level phylogenetic analysis of the Olenelloidea, and thus can be treated as belonging to that genus. The age of this species is difficult to determine, as it comes from a limestone boulder within a con- glomerate, and the boulder can no longer be definitively ascertained (Rasetti 1948). How- ever, all limestone boulders subsequently collected from near this Bic, Quebec locality all contain ptychoparids that probably correlate with the Olenellus zone. Thus, E. bicensis is provisionally placed within that zone in the Early Cambrian. Occurrence: Canada: Quebec, limestone boulder believed to be from the Early Cambrian Olenellus zone, in ?Lower Ordovician conglomerate, from near Bic (Rasetti 1948). ELLIPTOCEPHALA LAXOCULES (FRITZ 1972) Figure 12.4 Olenellus laxocules; Fritz 1972:13, pl. 11, figs. 7-21. Types: Holotype, GSC 27335, see Fritz (1972). Discussion: Based on the character evidence discussed above under E. sequomalus, although originally assigned to Olenellus E. laxocules must henceforth be assigned to Elliptocephala. Material examined: GSC 27333, a paratype. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1668 ft above the base of the formation, middle Olenellus zone, following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. ELLIPTOCEPHALA PARAOCULUS (FRITZ 1972) Olenellus paraoculus; Fritz 1972:14, pl. 15, figs. 8-22. Types: Holotype, GSC 27394, see Fritz (1972). Discussion: Based on the character evidence discussed above under E. sequomalus, although originally assigned to Olenellus, E. paraoculus must henceforth be assigned to Elliptocephala. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1746 ft to 1945 ft above the base of the forma- tion, middle Olenellus zone, following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. ELLIPTOCEPHALA PARVIFRONS (FRITZ 1972) Figure 12.6 Wanneria parvifrons; Fritz 1972:30, pl. 13, figs. 1-5. ¢Wanneria sp. 1; Fritz 1991:18, pl. 3, fig. 2. Types: Holotype, GSC 27355, see Fritz (1972). Discussion: Fritz (1972) originally assigned this species to the genus Wanneria. Al- though Wanneria is closely related to Elliptocephala, Elliptocephala parvifrons actually appears to be more closely related to the genus Elliptocephala than to Wanneria since it bears the diagnostic characteristics of Elliptocephala and lacks some of the diagnostic characteristics of Wanneria. In particular, E. parvifrons does not have the length (exsag.) SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 61 of the anterior cephalic border between the lateral margins of the frontal lobe and the genal spine angle equal to the length (sag.) of LO (the condition in W. walcottana), but rather the length is equal to about half the length of LO; the parafrontal band is promi- nently visible in E. parvifrons, whereas it is not visible in W. walcottana; the ocular lobes in E. parvifrons extend back to SO rather than the medial part of the distal margin of L1; the anterior margins of the frontal lobe at each side of the midline in E. parvifrons are deflected posteriorly at a roughly 40 degree angle relative to transverse line, rather than at a roughly 10 degree angle as in W. walcottana; the ocular lobe is separated from the extraocular area by a shelf in E. parvifrons, whereas it smoothly merges with the ex- traocular area in W. walcottana; and the posterior margin of LO is convex posteriorly in E. parvifrons, whereas it is roughly transverse in W. walcottana. Thus, based on this character evidence, henceforth this species is assigned to the genus Elliptocephala and excluded from Wanneria. Fritz (1991) figured a partial cephalon, referred to as Wanneria sp. 1, that appears to satisfy all the diagnostic characteristics of Elliptocephala. Although from an early ontoge- netic stage, and poorly preserved, it appears to show no differences from E. parvifrons. Al- though Fritz (1991) claimed that there were differences between his Wanneria sp. 1 and E. parvifrons, when material from similar ontogenetic stages are compared these differences do not appear to be valid. Thus, at this time his Wanneria sp. | is treated as conspecific with E. parvifrons. Material examined: GSC 27355, the holotype. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1668 ft above the base of the formation, middle Olenellus zone, following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. Yukon Territory, the Illtyd Formation, Fritz’s (1991) Unit 3, Olenel- lus zone, Early Cambrian, Wernecke Mountains. ELLIPTOCEPHALA LOGANI (WALCOTT 1910) Figure 12.5 Olenellus logani; Walcott 1910:333, pl. 41, figs. 5, 6. Wanneria logani (Walcott); Poulsen 1958:16; Fritz 1972:29, pl. 14, figs. 1-14, pl. 16, figs. 1-7; Fritz 1991:17, pl. 3, figs. 16-17; Lieberman 1998:70. Types: Walcott (1910) figured two USNM specimens that were casts of GSC 414d and 414e. These specimens serve as Walcott’s (1910) original syntype material. The better of these specimens (Walcott 1910, pl. 41, fig. 5), GSC 414d, is here designated the lectotype. The other specimen, GSC 414e, is the paralectotype. Discussion: Fritz (1972) concluded that this species belonged to the genus Wanneria. How- ever, on the basis of character evidence identical to that given above under E. parvifrons, it is clear that this species must henceforth be assigned to the genus Elliptocephala. Ellipto- cephala logani has a larger geographic range than any other species of olenelloid considered in this study. Material examined: GSC 27376. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 1668 ft to 1937 ft above the base of the forma- tion, middle Olenellus zone following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie; Yukon Territory, the Illtyd Formation, Fritz’s (1991) Unit 62 PEABODY MUSEUM BULLETIN 45 3, Olenellus zone, Early Cambrian, Wernecke Mountains; Newfoundland, the Forteau For- mation, Olenellus zone, Early Cambrian, L’Anse au Loup, straits of Belle Isle, Labrador. ELLIPTOCEPHALA MIRABILIS (POULSEN 1958) Holmia mirabilis; Poulsen 1958:15, pl. 3, figs. 5-8; Bergstrom 1973:309. Wanneria? mirabilis (Poulsen); Fritz 1973:13. Types: Holotype, MUGH 10691, designated in Poulsen (1958). Discussion: Poulsen (1958) assigned this species to the genus Holmia. This species is only known from cephalic remains, but there are some cephalic characters that differentiate members of the genus Elliptocephala from the type species of the genus Holmia, H. kjerulfi. The condition of these characters indicates that Poulsen’s (1958) species should be as- signed to the genus Elliptocephala. In particular, in E. mirabilis the anterior cephalic bor- der is prominently separated from the extraocular area by a furrow and in H. kjerulfi such a furrow is not prominently developed; in E. mirabilis the parafrontal band is much more prominently developed in dorsal view than it is in H. kjerulfi; in E. mirabilis the pre-ocu- lar furrow on L4 is directed inward and forward from the glabellar margin, in H. kjerulfi it is not prominently developed; in E. mirabilis the ocular lobe is separated from the ex- traocular area by a prominent shelf, whereas the ocular lobe smoothly merges with the ex- traocular area in H. kjerulfi; in E. mirabilis SO is prominently conjoined medially, in H. kjerulfi it is not; in E. mirabilis, lateral lobes on LO are absent, in H. kjerulfi they are pre- sent; and in E. mirabilis the intergenal angle is developed posterior of a point halfway be- tween the ocular lobes and the genal spine, while in H. kjerulfi it is developed directly pos- terior of the lateral margins of the ocular lobe. These characters indicate that E. mirabilis should be assigned to the genus Elliptocephala. However, it is not that easy to distinguish every species of Elliptocephala from H. kjerulfi, particularly on the basis of cephalic mate- rial. Following the phylogenetic topology within the Olenelloidea from Lieberman (1998), the genus Elliptocephala is not that distantly related from the Holmiidae. This agrees with the conclusions of Ahlberg et al. (1986), though the two are not sister taxa. Therefore, based on the material that he had available to him, Poulsen’s (1958) assignment is not that far off. Bergstrom (1973) asserted that this species probably belonged with Wanneria, which he considered to be closely related to Elliptocephala. Again, this analysis suggests that his taxonomic assignment was fairly accurate. In addition, it so happens that all of the characters elucidated in Lieberman (1998) that distinguish the entire family Holmiidae from members of the genus Elliptocephala are characters of the thorax. These are of no assistance when trying to demonstrate that the cephalon Poulsen (1958) referred to as H. mirabilis should be excluded from the Holmiidae. However, the close match between the diagnostic characters of Elliptocephala and the char- acters of E. mirabilis, as well as the failure of that species to adequately match the diagnoses of any of the genera within the Holmiidae, suggest that this species belongs to the genus EI- liptocephala. Fritz (1973) suggested that this species belonged within the genus Wanneria. As dis- cussed above, the genus Elliptocephala is closely related to Wanneria, though they are not sis- ter taxa. Species of these genera can clearly be distinguished by the characters given above. Material examined: MUGH 10691, the holotype. Occurrence: Denmark: Greenland, Marshall Bugt, Inglefield Land, NW Greenland, Lower Cambrian Wulff River Formation, questionably placed somewhere within the Olenellus zone. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 63 ELLIPTOCEPHALA LUNDGRENI (MOBERG 1892) Figures 12.7, 12.8 Wanneria? lundgreni; Bergstrom 1973:304, figs. 17, 18, 19a (see for more complete syn- onymy). Kjerulfia? lundgreni; Ahlberg, Bergstrom and Johansson 1986:52. Holmia lundgrent; Fritz 1973:13. Holmia? lundgrent; Fritz 1995:714. Types: Lectotype, PIL LO 1398T, designated in Bergstrom (1973). Discussion: This species has an autapomorphy that easily distinguishes it from other mem- bers of the genus Elliptocephala: near the ocular lobe, the width (tr.) of the parafrontal band on L4 is twice the length (sag.) of the parafrontal band at the anterior end of the glabella. In all other species of Elliptocephala the width of the parafrontal band is approximately equal to the length of the parafrontal band. Bergstrom (1973) questionably assigned this species to the genus Wanneria. However, he recognized (as did Ahlberg et al. 1986), the close relationship between Elliptocephala and Wanneria and concluded that either one of these generic assignments could hold for his W.? Iundgreni. In fact, he suggested that his W.? lund- greni might actually share more features with E. asaphoides than it does with W. walcottana. This analysis and the analysis of Lieberman (1998) are essentially concordant with his con- clusion. As discussed above, Elliptocephala and Wanneria are closely related, though not sis- ter taxa. On the basis of the character evidence presented under E. parvifrons, this species is assigned to Elliptocephala rather than to Wanneria. Bergstrom (1973) figured additional thoracic material, which he also assigned to this species. This material is not articulated with the cephalic material, and thus at this time it is not treated as conspecific with E. Iundgreni. In addition, it is rather poorly preserved, making any conclusions on this material and its affinities difficult. Ahlberg et al. (1986) questionably assigned this species to the genus Kjerulfia. As dis- cussed above under E. mirabilis, only thoracic characters unequivocally distinguish mem- bers of the genus Elliptocephala from the family Holmiidae, and E. Iundgreni is only defi- nitely known from cephalic material. However, several cephalic characters make it possible to distinguish E. lundgreni from the type of the genus Kjerulfia, K. lata Kiaer 1917. In par- ticular: 1. in E. Jundgreni the anterior cephalic border between the lateral margins of the frontal lobe and the genal spine angle is short, with the length (exsag.) equal to one-half the length (sag.) of LO, in K. lata the length is equal to the length of LO; 2. in E. lundgreni the anterior border is prominently separated from the extraocular area by a furrow, in K. lata such a furrow is not prominently developed; 3. in E. Jundgreni the parafrontal band is much more prominently developed in dorsal view than it is in K. lata; 4. in E. lundgreni the lateral margins of L4 are distal to the lateral margins of LO, in K. lata they are proximal to the lateral margins of LO; 5. in E. lundgreni the ocular lobe is separated from the extraoc- ular area by a prominent shelf, whereas the ocular lobe smoothly merges with the extraoc- ular area in K. lata; 6. in E. lundgreni the distal margins of L3 are convex outward, whereas in K. lata they are straight; 7. in E. lundgreni a line between the ends of S2 is transverse, whereas in K. lata a line between the ends of S2 is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line; 8. in E. Jundgreni a node is present on the axial part of LO, in K. lata a spine is present; 9. in E. Iundgreni, lateral lobes on LO are absent, in K. lata they are present; 10. in E. lundgreni the extraocular area is prominently vaulted, whereas in K. Jata it is gently convex. On the basis of these character differences E. 64 PEABODY MUSEUM BULLETIN 45 lundgreni must be excluded from the genus Kjerulfia. Similar character differences emerge between E. lundgreni and H. kjerulfi, except for 4, 6, 7 and 10. These taxa also differ in the condition of the following: in E. lundgreni SO is con- joined medially, in H. kjerulfi it is not; in E. lundgreni the intergenal angle forms a roughly 10 degree angle relative to a transverse line, in H. kjerulfi it forms a roughly 45 degree angle; and in E. lundgreni the posterior tips of the ocular lobes extend back to the medial part of the distal margin of LO, while in H. kjerulfi they extend back to the distal tips of SO. A sim- ilar set of characters discriminate this species from other holmiids. Based on the phylogeny in Figure 10 this species appears to be a derived member of the genus Elliptocephala, though it appears early in the fossil record. Indeed, it is one of the earliest trilobites in the fossil record of Baltica (Ahlberg et al. 1986; Ahlberg 1991). All other members of the genus Elliptocephala do not appear before strata equivalent with the lower middle Olenellus zone. In conjunction with phylogenetic topology within the genus, this suggests that there is a substantial gap in the history of Elliptocephala, with the implication that a large portion of the Cambrian Radiation may be missing from the fossil record. Material examined: PIL LO 1398T, the lectotype. Occurrence: Sweden: The upper part of the Norretorp Formation, Lower Cambrian, at Tun- byholm, Flagabro, Gladsax, and S of Brantevik in eastern Scania. Trollskogen at Hallestad in central Scania. All treated as in the Schmidtiellus mickwitzi zone, following Ahlberg et al. (1986) and Ahlberg (1991). Phylogenetic Analysis of the Holmiidae The Holmiidae is a diverse family of seven genera and at least 17 species, some of which are the earliest trilobite species to appear in the fossil record of Baltica. It includes taxa distrib- uted throughout Baltica and Western Laurentia (in the Great Basin and northwestern Canada), and also in Morocco. The family and the genera within it are revised, and one new genus is diagnosed, “Baltobergstroemia.” The bearing of taxonomy within the family on stratigraphic boundaries in Baltica is also discussed. A total of 18 taxa was subjected to phylogenetic analysis. This includes all species that in past studies have been assigned to the genera Esmeraldina, Kjerulfia, Holmia, Holmiella, Palmettaspis and Schmidtiellus. It also includes one outgroup taxon, Wanneria walcottana. This taxon is an appropriate outgroup to these genera, based on the higher level phyloge- netic analysis presented in Lieberman (1998). Phylogenetic patterns were determined by parsimony analysis of 35 holaspid exoskeletal characters. These are given in Table 7. The codings for the taxa analyzed are given in Table 8. All characters were treated as unordered (nonadditive). These data were subjected to a heuristic search on PAUP v. 3.1.1 (Swofford 1993). The six most parsimonious trees of length 90 steps were recovered, with a consistency index of 0.52 and a retention index of 0.69. A strict consensus of these trees is shown in Fig- ure 13. The bootstrap confidence values for the nodes of the consensus tree duplicated in the bootstrap analysis are given in Figure 13. Almost all of the nodes have bootstrap sup- port, but the value of this support varies depending on which node is being considered. Using the method of Bremer (1994), 156 trees of length less than or equal to 91 steps, 1,790 trees of length less than or equal to 92 steps and 8,600 trees of length less than or equal to 93 steps were recovered before the analysis was terminated because of insufficient computer memory. Only one of the nodes received any branch support (see Figure 13). It is conceiv- able that this node had even greater support, but because of insufficient computer memory this could not be determined. The total support index (Bremer 1994) for the tree is at least 0.03, a low value according to the examples given in Bremer (1994). Most of the uncertainty in the tree centers around the poorly known “Baltobergstroemia” sp. from Morocco. To de- termine which of the six most parsimonious trees is best supported by the data, a weight- ing procedure was used on PAUP vy. 3.1.1 (Swofford 1993), with characters reweighted ac- cording to the rescaled consistency index, with the maximum value (best fit) used and a base weight of 1,000 assigned (standard protocol in PAUP). A single most parsimonious tree of length 32,490 steps was recovered (shown in Figure 14). There is some incongruence between stratigraphic and phylogenetic patterns in the phylogeny presented in Figure 14, particularly in the timing of relative appearance of some of the Baltic species. The basal taxa within the Holmiidae, members of the genus “Holmua,” appear relatively late in the stratigraphic column, in the Holmia kjerulfi group zone, relative to their phylogenetic position. By contrast, species of the genera Schmidtiellus and “Balto- bergstroemia, which are derived relative to species of the genus “Holmia,” appear in the first trilobite bearing beds in Baltica, in the Schmidtiellus mickwitzi zone. The Laurentian taxa, 66 PEABODY MUSEUM BULLETIN 45 Wanneria walcottana “Holmia” palpebra “Holmia” lapponica Holmia kjerulfi “Baltobergstroemia” sp. Schmidtiellus mickwitzi torelli Schmidtiellus new species Schmidtiellus reetae Baltobergstroemia mobergi “Baltobergstroemia” sulcata “Palmettaspis” cometes Palmettaspis consorta Palmettaspis lidensis Palmettaspis parallela Esmeraldina rowei Holmiella falcuta Holmiella preancora Kjerulfia lata NO —_ © ak io) O1 at NO Figure 13. A strict consensus of the six most parsimonious trees of length 90 steps produced from analysis of charac- ter data in Table 8 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using a heuristic search with a stepwise addition sequence and 100 random replications. The retention index is 0.69, and the con- sistency index is 0.52. The following nodes in the text were supported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node 2 = 0.67; Node 3 = 0.52; Node 4 = 0.41; Node 5 = 0.49: Node 6 = 0.61; Node 7 = 0.58; Node 8 = 0.68, Node 9 = 0.49; Node 10 = 0.34; Node 11 = 0.46; Node 13 = 1.00. The following branch support value (Bremer 1994) was recovered for Node 13 = 3+. Total tree support (Bremer 1994) is 0.03. Character states are placed at nodes, using MacClade v. 3.04 (Maddison and Maddison 1992), with the characters given in Table 7. The apomorphic state is given in parentheses. Square parentheses indicate equivocal character states that are ambiguous because of missing data, polymorphisms or multiple equally parsimonious resolutions. Equivocal characters are placed only at a basal phyloge- netic ps and only See et reversals are shown. Node 1, 4[0, 1], 5[0, 1, 2], 9(1), 12(1), 19[0, . 10 I ee ), 25[0, 1], 29[0, 1], 33(1), 34[0, 2]; Node 2, 5(1), 18(1), 25(1) 26(1 ); oe 3, sa 1], 6( 22([0, ese a eee 4(1), se 12(2), 2 8(1), 22(2), 25(1); Node 5, an a Auer Node 6, 3(1), rage 29(0), 34(1), 35(1); Node hoe J, 1911.3), 23(0); Node 8,201) sis Z 13(1), L400) 15Cb), 20), 2202), 250), 27 (0); Lo Ly, 271 20(0, 1]; Node 10, Basan enone 12,5 (1) 3 LOCH); pA (1); Node oe 5(L)K16G), 2010s; ed 13530) F410 @)F (2) Fas (0) 2a; 20(2), 23(0), 24(0), 27(0), 32(1), 33(0). SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 67 sister to the clade containing the genera Schmidtiellus and “Baltobergstroemia,” appear in the basal “Nevadella” zone. The precise correlation between the North American and Baltic sec- tions is not clear, but these zones may both correlate with the upper part of the Atdabanian Stage according to Palmer and Repina (1993) and Geyer and Palmer (1995), implying rea- sonable concordance in this section of the tree. However, some caution must be exercised in extrapolating too much from these general stratigraphic patterns to the phylogenetic pat- terns, because stratigraphic correlations for the Early Cambrian, particularly those of Palmer and Repina (1993), are based on the presumed phylogenetic position of the taxa contained within those strata. This may create some circularity. Systematic Paleontology SUPERFAMILY Olenelloidea Walcott 1890 FamiLy Holmiidae Hupé 1953 The Holmiidae has been defined in several ways over the years, and it is worth briefly pre- senting this taxonomic history here to show the areas of congruence between this study and past studies, while also highlighting incongruence and its potential source. Hupé (1953) de- fined the Holmiidae as a subfamily (Holmiinae) within the Olenellidae and specified that it definitively contained the genera Holmia, Kjerulfia and Bondonella Hupé 1953. Lieberman (1998) excluded Bondonella from the Olenelloidea as it lacked some of the diagnostic char- acteristics of that superfamily. In particular, in Bondonella the posterior part of L3 does not bulge laterally relative to L1, and the ocular lobes merge with the entire lateral margin of the frontal lobe, rather than just with the posterolateral part of L4. Lieberman (1998) placed this taxon within the Judomioidea Repina, a superfamily sister to the Olenelloidea. On this basis, that taxon must be excluded from the Holmiidae, which is nested within the Olenel- loidea. The other genera listed appear correctly assigned to the Holmiidae. Harrington et al. (1959) partly retained Hupe’s (1953) taxonomy, but excluded Kjerulfia from the Holmiidae while assigning Schmidtiellus to it. Character evidence suggests that Kjerulfia should be in- cluded within the Holmiidae, but Schmidtiellus belongs there as well. Bergstr6m (1973) treated the Holmiidae as a family containing the genera Holmia, Elliptocephala, Esmerald- ina, Schmidtiellus and Wannerta. Elliptocephala and Wanneria are closely related but are not sister taxa, according to the phylogeny of the Olenelloidea in Lieberman (1998) (and also discussed in greater detail above). These two genera are in turn closely related to the Holmiidae, but do not share a sis- ter group relationship with that family. Conceivably they could be placed within the Holmi- idae, but this would entail either making the Holmiidae paraphyletic or placing all olenel- loid taxa outside the Olenellidae within that family. The former is not advocated because the Holmiidae as defined here is a good clade. The latter taxonomic decision was not fol- lowed because it is believed that additional families within the Olenelloidea can be recog- nized, and thus a more informative classification can potentially be made. This analysis and that of Lieberman (1998) are in accord with the other genera Bergstrém (1973) assigned to the Homiidae. Repina (1979) assigned the genera Holmia, Schmidtiellus, Kjerulfia, Elliptocephala, Bondonella, Andalusiana and, questionably, Holmiella to the Holmiinae, one of two sub- families she assigned to the Holmiidae. To the other subfamily, the Callaviinae (Poulsen in Harrington et al. 1959), she assigned the genus Callavia. Grounds for exclusion of some of these genera from the Holmiidae have already been presented above. This study and that of 68 PEABODY MUSEUM BULLETIN 45 o — % 5 m7) i) 3 Sy iho s , i oo VS i oe ~ o ° n ” ~ % wo AY = a yg Ss 2 Onl 6 sr c % o & & = o s 2g © x & Lis — — w — —— y FogetileS oe ae Se Se; WR)! Nah inaemg 8 Cae. Fs E.. se ee S 2 8 Be ge ee 2 Q Q = 8 ° 9° w = o Q ~ 3 oS 2 8 =. Gar ar a eS ae Ge a en pe ee ae er i a a oe Ae ee > >= 0 77) D D 77) Q Q Q & ~ Q 0 z z > & o oo = 2) o © & nm ~ a 3 o WY — ~~ ~ ~ _— 5 8S eS Ue SS See se ee woke) Sun SA Rie be alias a Ce te +2: Se Se se ws tc © © £2 2 2m ose SES 52 So eo go fs Sf a nh om & Be. 6. a oe Sb 5S Srey ee to ge om. = SS ap Ol Gee oT Wy oe = oOo? ak OO = oe 00 QO ak > —_ O1 5 12 N oO Figure 14. The weighted tree of length 32,490 steps, with the same topology as one of the most parsimonious trees used to derive the consensus tree shown in Figure 13. Characters were reweighted according to the rescaled consistency index with a base weight of 1,000 steps, and the maximum value was used (standard protocol of PAUP v. 3.1.1 [Swofford 1993]). Then a heuristic search with a stepwise addition sequence and 100 random replications was used to construct the tree. Character states were placed at nodes, using MacClade v. 3.04 (Maddison and Mad- dison 1992), with the characters given in Table 7. Only differences from the cladogram in Figure 13 are shown: Node 1, 4(1), 5(1), 9(1), 12(1), 18[0, 1], 19(1), 23(1), 24(1), 25(1), 29(1), 33(1), 34(2); Node 3, 22[0, 1, 2], 28(1); Node 4, 2[0, 1], 5[1, 2], 6(1), 18(1), 22(2); Node 5, 5(2), 12(2); Node 6, 8(1), 25(0); Node 7, 26(1); Node 8, 29[0, 31), 3415 2)s Neder, 20). 4/0: 1), 13(1); 1401), 151). 201); 271). Lieberman (1998) support the contention that Holmiella, as well as Schmidtiellus and Kjerulfia, belong with the Holmiuidae. Andalusiana is known from poorly preserved material; however, it appears to lack some of the diagnostic characters of the Olenelloidea. In particular, the ocular lobes contact but do not merge with the ocular lobe; the ocular lobe is in contact with the entire lateral margin of the frontal lobe; and the dorso-ventral elevation of the ocular lobes gradually de- creases between the distal most tip of the ocular lobe and the point where the ocular lobes contact the frontal lobe. However, one diagnostic character of the Olenelloidea this genus possesses is the condition of L3, which bulges laterally relative to L1. On the basis of these character differences, this genus is excluded from the Olenelloidea, and thus is also excluded from the Holmiidae. Its relationships to the other superfamilies within the Olenellina—the SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 69 Nevadioidea Hupé 1953, the Judomioidea, the Olenelloidea and the paraphyletic “Fal- lotaspidoidea” Hupé 1953—will be considered in a subsequent paper. The type of the genus Callavia, C. broeggeri (Walcott), also lacks some of the diagnos- tic characters of the Olenelloidea. In particular, the ocular lobes merge with the entire mar- gin of the frontal lobe, L3 does not bulge laterally relative to L1, and the length (sag.) of the frontal lobe is less than the total (sag.) length of LO and L1. Lieberman (1998) therefore ex- cluded it from the Olenelloidea and provisionally assigned it to the Judomioidea. Because its propinquity does not lie with the Olenelloidea, it will be discussed in a later paper, along with Andalusiana and several other taxa. Ahlberg et al. (1986) correctly recognized that the Callaviinae could not be related to the Holmiidae. They assigned the four genera Holmia, Andalusiana, Kjerulfia and Schmidtiellus to the Holmiidae. The status of these genera have been evaluated above. Palmer and Repina’s (1993) classification was discussed in detail in Lieberman (1998). They assigned two subfamilies, Holmiinae and Callaviinae, to the Holmiidae, following Re- pina (1979). To the Holmiinae they assigned Holmia, Andalusiana, Elliptocephala, Holmiella and Schmitdiellus. Palmer and Repina (1993) also assigned Kjerulfia to the Callaviinae. As mentioned above, Callavia needs to be excluded from the Olenelloidea. By contrast, Kjerul- fia is a bona fide member of the Holmiidae. Inexplicably, Palmer and Repina (1993) omit- ted Esmeraldina from their concept of the Holmiidae and, indeed, did not even discuss it in their paper on the Olenellina. However, this was rectified in Palmer and Repina (1997). Es- meraldina is a member of the Olenelloidea and the Holmiidae, and evidence for this is pre- sented below. Geyer and Palmer (1995) incorrectly assigned two taxa to the Holmiidae that appear not to belong within the Olenelloidea. Iyouella Geyer and Palmer 1995 lacks the defining characters of the Olenelloidea. In particular, L3 does not bulge laterally relative to L1, the ocular lobe contacts but does not merge with the lateral margins of the frontal lobe, and L4 is very short (sag.), roughly two-thirds the length (sag.) of LO and LI. It instead belongs within the paraphyletic “Fallotaspidoidea.” Geyer and Palmer (1995) also assigned the genus Cambropallas Geyer 1993 to the Holmiidae. As discussed in Lieberman (1998), species of this genus lack some of the diagnostic characters of the Olenelloidea, and Cambropallas must be excluded from the Olenelloidea because the ocular lobes fuse with the entire lateral margin of the frontal lobe. Lieberman (1998) placed this genus within the Judomioidea. In- terestingly, Geyer (1993) also had suggested that this genus sat outside the Olenelloidea, though his position changed in subsequent publications. Lieberman (1998) recently defined the Holmiidae as a clade including, but not re- stricted to, the genera Holmia, Kjerulfia, Esmeraldina and Holmiella. This analysis extends the character analysis of Lieberman (1998), presents characters diagnostic for the Holmi- idae and discusses all of the genera and species within the family. Characters diagnostic of the Holmiidae include: 1. anterior cephalic border between the genal spine angle and the anterolateral margins of the frontal lobe with length (exsag.) equal to length (sag.) of LO; 2. plectrum absent; 3. pre-ocular furrow on frontal lobe when visible directed inward and for- ward from glabellar margin; 4. line from posterior tip of ocular lobe to junction of poste- rior margin of lobe with glabella forms 0 to 25 degree angle with sagittal line; 5. posterior tips of ocular lobes developed opposite SO or medial part of distal margin of LO; 6. width (tr.) of interocular area greater than or equal to width of ocular lobe; 7. S3 gently or jaggedly convex; 8. L2 and L3 do not merge; 9. $2 convex; 10. distal margins of L2 when proceeding anteriorly either diverging or subparallel; 11. distal sector of SO convex or sinuous, with proximal end well posterior of distal end; 12. axial part of LO with node or spine present; 7O Table 7. PEABODY MUSEUM BULLETIN 45 Description of characters and character states used in phylogenetic analysis of Holmiinae; (0) represents the primitive state, (1), (2) and (3) represent derived states. — . Abrupt dorsal notch-like truncation of ocular lobes at margins of L4 2,99 SS) . Lateral margins of glabella opposite L2, when proceeding anteriorly 4. Lateral lobes on LO nn . Posterior margin of ocular lobe opposite 6. Ornamentation on LO . Frontal lobe “I co . Lateral margins of frontal lobe 9. Intergenal angle developed 10. Angle intergenal angle forms with transverse line 11. Genal spine angle opposite 12. Extraocular region Primitive state (0) absent (0) jaggedly convex or carat shaped (0) diverging (0) absent (0) medial part of margin of L1 (0) node (0) contacts anterior border furrow (0) distal to lateral margins of LO (0) midway between ocular lobe and genal spine (0) 0° to 15° (0) medial part of margin of LO (0) flattened Derived states (1) present (1) evenly convex (1) subparallel (1) present (1) distal tip of SO (2) medial part of margin of LO (1) spine (1) does not contact anterior border furrow (1) anterior of lateral margins of LO (1) directly behind distal tip of ocular lobe Gh) 35) tot>, (2) 80° to 90° (1) medial part of margin of L1 (2) medial part of margin of L3 or medial part of frontal lobe (1) prominently vaulted (2) gently sloping from ocular lobe to lateral border furrow Continued SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 71 Table 7 continued. 13. Primitive state Anterior cephalic border 14. Anterolateral margins of frontal lobe . Length (sag.) of frontal lobe 16. Anterodistal margins 20. 2 —_— of L3 formed by . Distal margins of L3 oil . Extraocular region opposite of L1 Orientation of lateral margin of genal spine near spine angle relative to a sagittal line . Genal spine extends back approximately (0) not prominently separated from extraocular area by furrow (0) not prominently separated from anterior extraocular area by furrow (0) equal to 1.5 times length (sag.) of LO and L1 (0) ocular lobes (0) convex outward (0) not conjoined medially (0) broad, width approximately 75% width of glabella at L1 (0) deflected at roughly 10° to 15°angle (0) first 4 to 5 thoracic segments Derived states (1) prominently separated by furrow (1) prominently separated from anterior extraocular area by furrow (1) equal to 1 to 1.1 times length (sag.) of LO and L1 (1) axial furrows (1) straight (1) conjoined medially (1) narrow, width approximately 45% to 55% width of glabella at L1 (2) very narrow, width approximately 10% to 15% width of glabella at L1 (3) very narrow, width approximately 25% width of glabella at L1 (1) deflected at roughly 30° to 40° angle (2) deflected at roughly 60° to 70° angle (1) at least 8 thoracic segments Continued 72. PEABODY MUSEUM BULLETIN 45 Table 7 continued. 2: pia 26. Thoracic pleural spines of Tl to T4 no 493 Frontal lobe Line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms 27. Anterior cephalic border 28. 29: 30. 3 —_ Se ao: 34. developed as Medial part of intergenal angle Lateral margins of prothoracic pleural spines inclined at angle relative to sagittal line Medial and distal tips of S3 . Medial and distal tips of S2 Pygidium Posterior margin of pygidium Posterior margins of first four thoracic segments near spine bases, relative to orientations of margins on spines 5. Ornament on occipital lobe Primitive state (0) sweep back 4 segments (0) not conjoined medially (0) not conjoined medially (0) does not expand prominently dorsally (0) 0° to 15° angle with sagittal line (0) flattened plane, angled relative to dorsal plane (0) with distinct node (0) 10° to 20° (0) as far forward (0) as far forward (0) length (sag.) equal to 0.8 to 1 times width (tr.) (0) with median notch (0) both flexing weakly posteriorly at roughly same angle Derived states (1) sweep back 2 segments (2) sweep back 1 or no segments (1) conjoined medially (1) conjoined medially (1) expands dorsally at least to level of dorsalmost surface of eyes (1) 20° to 25° angle with sagittal line (1) rounded ridge (1) with small spine (1) 40° to 50° (1) distal tip further forward than medial tip (1) distal tip further forward than medial tip (1) broad, twice as wide (tr.) as long (sag.) (1) weakly convex (1) roughly transverse (2) flexing slightly anteriorly before flexing posteriorly (0) posterior of midline of LO (1) anterior of midline of LO SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 73 13. posterior margin of LO convex; 14. glabellar furrows moderately incised; 15. extraocular area gently convex or prominently vaulted; 16. width (tr.) extraocular region opposite L1 equal to 15% to 55% width of glabella at L1; 17. genal spine extends back 4 to 8 thoracic segments; 18. intergenal angle developed posterior of lateral margins of ocular lobes or pos- terior of point halfway between ocular lobes and genal spine; 19. intergenal spine with dis- tinct node or spine; 20. posterior cephalic border between intergenal angle and LO flexes posteriorly; 21. thorax broken up into pro- and opisthothorax; 22. anterior margin of T3 medially parallels a transverse line or is very weakly directed posteriorly; 23. T3 normal; 24. anterior margin of thoracic pleural furrow on T3, when proceeding from proximal to dis- tal edge directed weakly posterolaterally before flexing strongly posterolaterally; 25. protho- racic axial rings with two prominent anterolateral lobes; 26. lateral margins of individual prothoracic axial rings prominently converging when proceeding from anterior to posterior edges; 27. single nodes present on medial part of thoracic axial rings; 28. thoracic pleural furrows extend width of inner pleural region; 29. width (tr.) of thoracic pleural spines T5 to T8 at spine midlength less than half length (exsag.) of medial part of inner pleural re- gion; 30. prominent spine at 15th thoracic axial ring or at pro-opisthothoracic boundary; 31. base of spine of width (tr.) roughly half width of axis; and 32. pygidium broad, width (tr.) one to two times length (sag.). On the basis of these characters, all members of the Holmiidae discussed below can be assigned to that family, whereas all other taxa can easily be excluded. Taxa in this family are found in the “Nevadella” zone of western Laurentia (the Great Basin), and in the Schmidtiellus mickwitzi and the various Holmia zones of Baltica. The subfamily Holmiinae is treated as containing the same taxa as does the Holmiidae. Several Polish taxa that clearly belong to the Holmiidae could not be subjected to phy- logenetic analysis as they are poorly preserved and illustrated and were not available for study. The taxonomic status of these species could not be investigated. Included in this group are: Holmia glabra Orlowski 1974; H. orienta Orlowski 1974; H. grandis Kiaer1916; Schmidtiellus panowi (Samsonowicz) 1959; S. nodosus Orlowski 1985; and Kjerulfia orcina Orlowski 1974. Included Taxa SUBFAMILY Holmiinae Hupé 1953 GENUS HoLMIA MATTHEW 1890 Type species: Paradoxides kjerulfi Linnarsson 1871. Assigned taxa: Holmia lapponica Ahlberg and Bergstrém 1983; Kjerulfia? palpebra Ahlberg 1984. Diagnosis: Anterior cephalic border not prominently separated from extraocular area by furrow; frontal lobe of glabella contacts anterior border furrow; anterolateral parts of glabella not prominently separated from extraocular area by furrow; prominent parafrontal band not visible in dorsal view; anterior margins of frontal lobe at each side of midline deflected posteriorly at roughly 40 degree angle relative to transverse line; length (sag.) of frontal lobe equal to 1.5 times length of LO and L1 medially; L4 expands dorsally; lateral margins of L4 distal to lateral margins of LO; abrupt dorsal notch-like truncation of ocular lobes at margin of L4 absent; ocular lobe smoothly merges into extraocular area; oc- ular lobe with prominent furrow; transverse profile of ocular lobes convex dorsally; sur- face of interocular area developed as flattened shelf; anterodistal margins of L3 formed by 74 PEABODY MUSEUM BULLETIN 45 ocular lobes; distal margins of L3 convex outward; $3 conjoined medially, jaggedly convex; medial and distal tips of $3 as far forward; lateral margins of glabella at L2 when proceed- ing anteriorly diverging; L1 to L3 strongly transversely convex; line between ends of S2 transverse; SO not conjoined medially; distal sector of SO convex anteriorly; posterior mar- gin of ocular lobe opposite distal tip of SO; LO with axial node posterior of midline; lateral lobes present on LO; extraocular region prominently vaulted; genal spine near genal spine angle directed posterolaterally at roughly 10 to 20 degree angle relative to sagittal line; genal spine extends back four to five thoracic segments; genal spine angle opposite medial part of distal margin of LO; intergenal angle posterior of lateral margins of ocular lobes; posterior margin of thoracic pleural furrow on T3 directed evenly posterolaterally; lateral margins of prothoracic pleural spines inclined at roughly 45 degree angle relative to sagit- tal line; posterior margins of first four thoracic segments near spine bases flex slightly an- teriorly before flexing posteriorly; boundary between thoracic pleural furrow and anterior band gradational; length (exsag.) of thoracic pleural furrows at medial part of thoracic segment | to 1.3 times length of pleural furrows on distal part of segment; length (exsag.) of thoracic pleural furrows at medial part of segment equal to length (exsag.) of posterior band; spine on 15th thoracic segment short, length (sag.) of two thoracic segments; opisthothorax with prominent pleurae which continue orientation of prothoracic seg- ments; pygidium with length (sag.) equal to width (tr.); posterior margin of pygidium weakly convex. Discussion: Several taxa that in the past have been assigned to this genus must on the basis of this analysis be excluded. In particular, Baltobergstroemia mobergi (Bergstrom 1973) and “B.” sulcata (Bergstrom 1973) must be excluded from the genus Holmia because of the con- dition of the following characters: the posterior margin of the ocular lobe is opposite the medial part of the distal margin of LO instead of opposite the distal tip of S0; the lateral margins of L4 are directly anterior to the lateral margins of LO instead of distal to them; the extraocular region is less prominently vaulted; and the frontal lobe does not expand promi- nently dorsally. An apparently new Schmitdiellus species (formerly H. cf. mobergi in Ahlberg et al. 1986) must be excluded from Holmia because of the condition of, the following characters, in addition to the ones listed above for the genus “Baltobergstroemia”: the anterior cephalic border is prominently separated from the extraocular area by a furrow; S3 is evenly, rather than jaggedly, convex; the distal tip of S2 is further forward than the medial tip, instead of being at the same position; the lateral margins of the glabella opposite L2 are subparallel rather than diverging anteriorly; and the axial node on LO is anterior of the midline of LO rather than posterior of the midline. All these species are discussed in greater detail below. What was formerly referred to as Holmia mirabilis by Poulsen (1958) also must be ex- cluded from the genus Holmia. This species is discussed in detail above under the genus EI- liptocephala, to which it is assigned, and character evidence for this exclusion is also pro- vided there. As construed here, Holmia is defined as a paraphyletic clade centered around the type of the genus, H. kjerulfi. This was done rather than erecting a new paraphyletic genus, or two monotypic genera, because the morphological differences separating these species from one another seemed slight. Moreover, the nodes separating these species are not extremely resilient, according to tests done on the cladistic analysis conducted here. Thus, it seemed prudent to lump these two species into the genus Holmia even though that makes the genus paraphyletic. The shuttermark convention of Wiley (1979) is used to designate the paraphyly of Holmia. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 75 HOLMIA KJERULFI (LINNARSSON 1871) Figure 11.3 Holmia kjerulfi (Linnarsson); Bergstr6m 1973:287, fig. 2 (see for more complete synonymy); Ahlberg 1984:256; Whittington 1988:579, figs. 2, 3; Whittington 1990:37, figs. 13-19; Palmer and Repina 1993:25, fig. 6.1; Palmer and Repina 1997:414, fig. 261.1; Lieberman 1998:67- Types: Lectotype, Swedish Geological Survey, Uppsala, Sweden, 5329a, b, see Bergstr6m (i973): Discussion: Holmia kjerulfi marginata Orlowski 1974, treated as a distinct species by Or- lowski (1985), is closely related to H. kjerulfi, as Ahlberg et al. (1986) argued. However, this material is so poorly preserved that at this time it is hard to ascertain whether or not it is conspecific with H. kjerulfi, represents a distinct subspecies, or rather deserves full specific status. One potential difference between this Polish material and the Norwegian material is that in the former the extraocular region is relatively narrower. Ahlberg and Bergstrém (1983) illustrate a small internal mold that they questionably referred to H. kjerulfi. It is dif- ficult to tell the veracity of this assignment because of the state of the material. Material examined: MCZ 2524; PIL Lo4456t. Occurrence: Norway: The Holmia shale, in the Holmia kjerulfi zone, Tomten Farm, Ringsaker (Ahlberg 1985). “HOLMIA” LAPPONICA AHLBERG AND BERGSTROM 1983 Holmia lapponica; Ahlberg and Bergstr6m 1983:242, fig. 2; Ahlberg 1985 fig. 3; Ahlberg et al. 1986:47, fig. 5. Types: Holotype, Swedish Geological Survey, Uppsala, Sweden, 1231, see Ahlberg and Bergstrom (1983). Discussion: This species is assigned to a paraphyletic “Holmia” because it differs from the type of that genus in the condition of characters 18, 26 and 28, listed in Tables 5 and 6. Occurrence: Sweden: Grammajukku Formation, in the Holmia kjerulfi zone, on the NE shore of Lake Langvattnet near Langsjoby, 10 km W of the village of Storuman, county of Vasterbotten, southern Swedish Lapland (Ahlberg and Bergstr6m 1983). “HOLMIA” PALPEBRA (AHLBERG 1984) Kjerulfia? palpebra; Ahlberg 1984:257, figs. 6, 7; Ahlberg et al. 1986:41, fig. 1. Holmia? sp.; Ahlberg 1984:256, figs. 4, 5. Types: Holotype, Swedish Geological Survey, Uppsala, Sweden, 3916, see Ahlberg (1984). Discussion: Ahlberg’s (1984) Holmia sp. appears to be identical to the type of his K.? palpe- bra, and the specimens hail from the same locality and horizon. Because of this, they are treated as conspecific. This species can no longer be assigned to the genus Kjerulfia because it differs from the type of that species in the condition of several characters, including: the anterior cephalic border is developed as a rounded ridge, rather than as a flattened, inclined plane; the lateral margins of L4 are deflected further laterally than the lateral margins of LO, rather than being directly anterior of the lateral margins of LO; the frontal lobe expands more 76 PEABODY MUSEUM BULLETIN 45 Table 8. Character state distributions for Holmiinae and outgroup used in phylogenetic analysis. Characters and alternative states are as listed in Table 7. Missing data are indicated by ?. Character states listed as X are polymorphic, where X=(0&1). L259 ¥4. 526-7. 8.90 | 232479. 6: 758, 910 Wanneria walcottana OF0"0"0"0' 00-0 00 "080" 0"0 000 O00 Holmia kjerulfi OOO 0 ONO sal 0,1 10207010) 050-150 “Holmia” lapponica 00.0) dell 0. Ov0.1.0.) .0).15 00,020 050 “Holmia” palpebra O20.0), % ISO ROKO 0 0.1050,020),0, 1.30 Baltobergstroemia mobergi OLO50) 20 lO Ty Ore OZ sONOs OFOVORT EO “Baltobergstroemia” sulcata OFOF0FD 271-08 LAL OTS OPDOROFORONO MINS Hz “Baltobergstroemia” sp. Om Or 2> <0" OM Os" Or270 "Or Ordnorle iG Kjerulfia lata OT Ort tO 0 0.0)" 02, 0 OnCz0 alee Schmidtiellus mickwitzi torelli Ol 2 OSs ee Call Ca Omi Ons aac Schmidtiellus sp. Oleted 2a, On Os. 40 22 1NOM Owl 350 Schmidtiellus reetae OO SAs 2 AOA 0) AO 25 OLO KORO RIE O Palmettaspis consorta Ea Osde Leal PCL a Or Ona Palmettaspis parallela LPS cece Amelie leit Oem Om eel olerte() wienleatia Palmettaspis lidensis | Ts ait i, td DN Hi Lge Lh Ln oe La Ul le “Palmettaspis” cometes Oak OR ORO TO SOME On Oeil, © Esmeraldina rowei 110 On 1050 Ls0e On tes teal Holmiella preancora desl O: OnOjel 2 tea Zales lee ty eles Op 2 Holmiella falcuta LES isl dc OnO0 ela2 Ae leona Continued prominently dorsally; the medial and distal tips of S2 are as far forward instead of having the distal tip further forward than the medial tip; the extraocular region is more prominently vaulted; the extraocular region is relatively narrower; and the intergenal angle lacks a small spine. This species is assigned to a paraphyletic “Holmia” because it differs from the type of that genus in the condition of characters 18, 26 and 28, listed in Tables 5 and 6. Occurrence: Sweden: Upper part of the Grammajukku Formation, in the Holmia kjerulfi zone, E bank of the Torbacken rivulet, about 600 m SSW of Delliknas, the Laisvall area, cen- tral Swedish Lapland (Ahlberg 1984). GENUS BALTOBERGSTROEMIA LIEBERMAN NEW GENUS Type species: Holmia mobergi Bergstr6m 1973. Assigned taxa: Holmia sulcata Bergstr6m 1973; Holmia inusitata Ahlberg and Bergstrém in Ahlberg et al. (1986); Holmia? sp. in Geyer and Palmer 1995. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) Ta Table 8 continued. i) i) i) i) Ne N NO i) i) Oo Oo Wo ies) eS) ie) — i) eS) cS nan oO’ N Co \O So _ i) Oo > nn Wanneria walcottana O70 070707050"0"0' 07) 0800 070 Holmia kjerulfi Oniexasie 0-01 1 Ome nOrO 1 2,0 “Holmia” lapponica OSO se a OFOL IO. O01 22.0 “Holmia” palpebra Ce caleal Vig ARON GAOln | 10s 29 610 Baltobergstroemia mobergi iO TOO Oe ¢ 2,0 “Baltobergstroemia” sulcata Ae OSO, Ll See QeesOre Feach() “Baltobergstroemia” sp. 2 Gale Ge I OO) Cia ithe Oars: 5° Kjerulfia lata OR aalmiOnlsOoteee I~ Nise Schmidtiellus mickwitzi torelhi SO IGO I OC20. 0; . SO eel ey Schmidtiellus sp. ee OnImOalelv cca: 1 Rae al Schmidtiellus reetae On lel) OsTOMIEORO BCs al 1 Palmettaspis consorta 2 Oe Ome 280 Palmettaspis parallela 2h Beer OM ONO Ok ca1k0 Palmettaspis lidensis Ce AOE ses On 1000) 50 “Palmettaspis” cometes 1 aad GOs 250) part Ogres 0 Esmeraldina rowet 12 Oar Os | OF On 2. 0 Holmiella preancora CoO) ASO OAT. Os, 1,100 Holmiella falcuta SePOLO SOMONE sO ON O70 Diagnosis: Anterior cephalic border not prominently separated from extraocular area by furrow; frontal lobe contacts anterior border furrow, not expanding prominently dorsally; distal margins of L4 deflected as far or farther laterally than distal margins of LO; length (sag.) of frontal lobe 1.5 times length of LO and L1; anterodistal margins of L3 formed by ocular lobes; lateral margins of L3 convex outward; $3 and S2 with medial and distal tips as far forward; lateral margins of glabella opposite L2 diverging when proceeding anteriorly; S1, $2 and $3 conjoined medially; posterior margin of ocular lobe opposite medial part of distal margin of LO; lateral lobes on LO present; LO with node or spine posterior of midline of LO; intergenal angle directly behind distal tip of ocular lobe; intergenal angle forms 0 to 15 degree angle with transverse line; intergenal angle with small spine; extraocular region gently sloping from eye to lateral border furrow; genal spine angle opposite medial part of distal margin of LO. Discussion: In the most parsimonious consensus cladogram (Figure 13) two of the species in this genus, B. mobergi and “B.” sulcata, sit in a polytomy that is sister to a clade con- 78 PEABODY MUSEUM BULLETIN 45 taining the genera Kjerulfia and Schmidtiellus. The poorly preserved “B.” sp. was variably placed in the six most parsimonious trees (consensus in Figure 13) either sister to the clade including the Great Basin taxa and the genera Kjerulfia, Schmitdiellus and the rest of “Baltobergstroemia’; sister solely to the clade of Great Basin taxa; or sister solely to the clade comprising the genera Kjerulfia, Schmitdiellus and the rest of “Baltobergstroemia.” However, in the weighted tree (Figure 14), this species is one node down the tree from “B.” sulcata, which is in turn sister to a clade consisting of B. mobergi and the genera Kjerulfia and Schmidtiellus. These species could have been treated as three monotypic genera, but because the differences separating them were slight, they are known from rel- atively limited material and the nodes separating them are not particularly resilient, ac- cording to tests of the cladogram in Figure 13, it seemed most prudent to lump these taxa into a single genus, even though that makes the genus paraphyletic (according to the weighted tree) or even polyphyletic (according to some of the unweighted equally most parsimonious trees). Etymology: The genus is named by combining “Balto” for Baltica, the region where two of these species are found, with the latinized “bergstroem,” for Jan Bergstr6m, whose research has significantly increased our knowledge of Cambrian trilobites. BALTOBERGSTROEMIA MOBERGI (BERGSTROM 1973) Holmia mobergi; Bergstr6m 1973:288, figs. 3-6 (see for more complete synonymy); Ahlberg et al. 1986:48, fig. 1. ¢Holmia inusitata; Ahlberg and Bergstrém in Ahlberg et al. 1986:43, fig. 3. Types: Holotype Palaeontological Institute, University of Lund (PIL), Lund, Sweden, LO 4457T, see Bergstrom (1973). Discussion: Baltobergstroemia inusitata (Ahlberg and Bergstrom) appears almost identical to B. mobergi, though both species, particularly the former, are known from extremely lim- ited material. The only difference between these taxa is the condition of the occipital spine, whose base appears somewhat broader in B. inusitata than in B. mobergi. Baltobergstroemia inusitata was not incorporated into phylogenetic analysis because its character codings were identical to those given for B. mobergi in Table 8 (the difference in the condition of the occipital spine was not coded because it would have been an autapomorphy). Addi- tional differences may emerge in the condition of the anterior cephalic border, not pre- served in B. inusitata, and also in the condition of the extraocular area. Therefore, at this time, these two species are not synonymized. If these species are identical, it would imply that the Holmia inusitata zone that Ahlberg et al. (1986) established would not be valid. In- stead, it would be indistinguishable from the underlying Schmidtiellus mickwitzi zone, from which B. mobergi is known. This species is excluded from the genus Holmia because of the condition of the fol- lowing characters: 1. the lateral margins of L4 are anterior of the lateral margins of LO, rather than being deflected distal to the lateral margins of LO; 2. L4 is less prominently ex- panded dorsally; 3. $1 is conjoined medially; 4. a line from the posterior margin of the oc- ular lobe to the junction of the posterior margin of the lobe with the glabella forms a larger angle with a sagittal line than it does in H. kjerulfi; 5. the posterior margin of the ocular lobe is opposite the medial part of the distal margin of LO rather than the distal tip of SO; 6. there is a spine rather than a node on LO; 7. the extraocular area is gently sloping from the eye to the lateral border furrow, rather than being prominently vaulted; 8. the thoracic SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 79 Figure 15. 1-3. Base of the Montenegro Mbr., Campito Fm., Barrel Springs, Silver Peak Range, Nevada. 1, 3. Esmeraldina rowei (Walcott). 1. YPM 72929, x1.6. 3. YPM 35865, 2.3. 2. Palmettaspis consorta Fritz, YPM 72908, X1.7. 4. Holmiella falcuta Fritz, locality is the same as Figure 12.2, GSC 27273, the paratype, x1.5. pleural spines are relatively slightly shorter than those found in H. kjerulfi. Occurrence: Sweden: Norretorp Formation, Schmidtiellus mickwitzi zone, Early Cambrian, on the shore between Brantevik and Gislovshammar, eastern Scania. “BALTOBERGSTROEMIA SULCATA (BERGSTROM 1973) Holmia sulcata; Bergstr6m 1973:292, figs. 7, 8; Ahlberg et al. 1986:41, fig. 41. Types: Holotype, Palaeontological Institute, University of Lund, Lund, Sweden (PIL), LO 4462T, see Bergstrém (1973). Discussion: The generic name of this species is placed within paraphyletic shuttermarks be- cause it differs from the type of the genus in the condition of only one character, part of the reason why the two species were placed within the same genus: the condition of the poste- rior margin of the ocular lobe relative to the junction of the posterior margin of the lobe with the glabella. In B. mobergi the posterior margin of the lobe is rotated outward about 10 degrees relative to its condition in “B.” sulcata. This species is excluded from the genus Holmia because of the condition of characters 1 to 3, 5 and 6, listed above under B. mobergi. Occurrence: Sweden: Believed to be in the Holmia kjerulfi zone, from Forsemolla in An- drarum and Brantevik-Gislovshammar, eastern Scania. 80 PEABODY MUSEUM BULLETIN 45 “BALTOBERGSTROEMIA SP. Holmia? sp.; Geyer and Palmer 1995:468, fig. 3.10. Discussion: This species is assigned to the paraphyletic genus “Baltobergstroemia” because it differs from B. mobergi in the condition of the following characters: the lateral margins of L4 are distal to the lateral margins of LO; L4 expands more prominently dorsally; and the pos- terior margin of the lobe is rotated inward about 10 degrees relative to its condition in B. mobergi. This species is excluded from the genus Holmia because of the condition of charac- ters 3 and 5, listed above under B. mobergi. It was not described as a new species because of its poor state of preservation and because it is known from limited material. However, it was included in this phylogenetic analysis due to its relevance to Cambrian biogeography. Occurrence: Morocco: Issafen Formation, Sectigena zone, western Anti-Atlas Area, Adai Sec- tion (see Geyer and Palmer 1995). GENUS KJERULFIA KIAER 1917 Type species: Kjerulfia lata Kiaer 1917. Assigned taxa: Monotypic. Diagnosis: Anterior cephalic border not prominently separated from extraocular area by deep furrow, developed as flattened, angled plane; frontal lobe contacts anterior border fur- row, does not expand prominently dorsally; lateral margins of L4 directly anterior to lateral margins of LO; anterolateral margins of glabella not prominently separated from extraocu- lar area by furrow; length (sag.) of frontal lobe equal to 1.5 times length (sag.) of LO and L1; pre-ocular furrow on frontal lobe directed inward and forward from glabellar margin; an- terodistal margins of L3 formed by ocular lobes; postero- and anterodistal margins of L3 with same orientation; lateral margins of glabella opposite L2 diverging when proceeding anteriorly; posterior margins of ocular lobes opposite distal tip of SO; line from posterior margin of ocular lobe to junction of posterior margin of lobe with glabella forms 20 to 25 degree angle relative to sagittal line; $3 evenly convex, distal tip further forward than medial tip; distal tip of $2 further forward than medial tip; SO, $1, S2 and $3 conjoined medially; SO sinuous; LO with spine posterior of midline of LO; lateral lobes on LO present; intergenal angle developed midway between ocular lobe and genal spine, with small spine; intergenal angle forms 10 to 15 degree angle with transverse line; extraocular area gently sloping from eye to lateral border furrow, width (tr.) opposite L1 equal to 45% to 55% width of glabella at L1; genal spine angle opposite medial part of distal margin of LO; genal spine close to genal spine angle forms roughly 10 to 15 degree angle relative to sagittal line; genal spine ex- tends back roughly 4 to 5 thoracic segments. Discussion: As several new characters were discovered in the course of this analysis, a new di- agnosis for the genus is presented. Several taxa that in the past have been assigned to this genus must be excluded. In particular, Kjerulfia? palpebra Ahlberg instead appears to belong in the genus “Holmia.” Character evidence for its exclusion from the genus Kjerulfia is given above under “H.” palpebra. Geyer and Palmer (1995:472, figs. 5.6, 5.9, 5.12) illustrated four extremely poorly preserved and deformed partial cephala from the Asrir Formation, Secti- gena zone, in the western Anti-Atlas area of Morocco. Two of these they referred to Kjerul- fia? sp. and two as Holmiidae? genus and species undetermined. As far as can be determined, these specimens belong to the same species: they are morphologically identical, but are so poorly preserved little more can be said about their relationships. The only conclusive fea- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 81 ture that would group these taxa with Kjerulfia lata is the axial spine, posterior of the mid- line on LO. It is interesting that in addition to “Baltobergstroemia” sp. (discussed above) an- other species belonging to the Holmiidae may be found in Early Cambrian Moroccan strata. Kjerulfia schwarzbachi Ahlberg and Bergstrém (Ahlberg et al. 1986) is known from ex- tremely poorly preserved and deformed material, and thus was not considered in phyloge- netic analysis. It is difficult to determine its precise generic relationships. One feature it shares with Kjerulfia lata is the axial spine, posterior of the midline on LO. KJERULFIA LATA KIAER 1917 Kjerulfia lata; Kiaer 1917:73, pl. 9-13, 14, figs. 1-2; Nikolaisen and Henningsmoen 1990:62, figs. 3, 7a—c (see for more complete synonymy); Palmer and Repina 1993:27, figs. 6.9, 6.10; Palmer and Repina 1997:416, figs. 2a, b; Lieberman 1998:67. Types: Lectotype, Paleontologisk Museum Oslo 61376, see Nikolaisen and Henningsmoen (1990). Discussion: Palmer and Repina (1993, fig. 6.10) figured a partial thorax that they assigned to this species (also illustrated in Palmer and Repina 1997, fig. 2b). Since it is completely dis- articulated from any cephalic material it is impossible to determine whether this assignment is correct, and therefore thoracic characters for this taxon were coded as “?” in Table 8. Occurrence: Norway: Middle part of the massive-bedded quartzite member of the Doul- basgaissa Formation, questionably in the Holmia kjerulfi group zone, Digermul peninsula, Tanafjord, Finnmark, and the Holmia Shale, questionably in the Holmia kjerulfi group zone, Tomten Farm, Ringsaker. GENUS SCHMIDTIELLUS MOBERG IN MOBERG AND SEGERBERG 1906 Type species: Olenellus mickwitzi Schmidt 1888. Assigned taxa: Olenellus torelli Moberg 1899; Schmidtiellus reetae Bergstrom 1973; Schmidtiellus sp. Diagnosis: Anterior cephalic border prominently separated from extraocular area by furrow; frontal lobe contacts anterior border furrow, does not expand prominently dorsally; antero- lateral margins of glabella not prominently separated from extraocular area by furrow; length (sag.) of frontal lobe roughly 1.5 times length (sag.) of LO and LI; lateral margins of L4 directly anterior of lateral margins of LO; $3 conjoined medially; lateral margins of glabella opposite L2 subparallel; distal tip of S2 further forward than medial tip; posterior margin of ocular lobe opposite medial part of distal margin of LO; line from posterior mar- gin of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 20 to 25 degree angle with sagittal line; LO with spine, anterior of midline of LO; lateral lobes on LO present; intergenal angle developed directly behind distal tip of ocular lobe; intergenal angle forms roughly 10 degree angle with transverse line; extraocular region gently sloping from eye to lateral border furrow; genal spine angle opposite medial part of distal margin of LO; lateral margin of genal spine near spine angle forms roughly 10 to 15 degree angle with sagit- tal line; posterior margins of first four thoracic segments near spine roughly transverse; tho- racic pleural spines on first four segments extend back only one thoracic segment, lateral margins of spines inclined at roughly 10 to 20 degree angle relative to sagittal line. Discussion: Bergstrém (1973) gives an excellent discussion of the genus. There he states that the original material of O. mickwitzi, which was fragmentary, is now completely lost. 82 PEABODY MusEUM BULLETIN 45 Bergstrom (1973) went on to demonstrate that subsequent assignments of species to this genus by past authors were incorrect. He further claimed that S. torelli is almost indistin- guishable from S. mickwitzi, although there may be some differences between the two taxa. Bergstrom (1973) treated these taxa as two varieties of a single species, and his treatment is followed here. Schmidtiellus? sp., discussed in Ahlberg et al. (1986), is poorly known and could not be obtained for study in this analysis. If it does indeed belong to this genus, this would extend its range into the Holmia kjerulfi zone. Additional species that have been re- ferred to Schmidtiellus (such as S. panowi [Samsonowicz] and S. nodosus Orlowski) are from Polish sections, and are known from extremely limited and poorly illustrated material. These could not be obtained for the purposes of this analysis, and were not considered here. Species of this genus are, along with Elliptocephala lundgreni (discussed above), the earliest undoubted trilobite taxa in the Baltic fossil record, occurring in the Schmidtiellus mickwitzi zone. Because additional character evidence bearing on this genus was recovered, a diagnosis for its members is provided. SCHMIDTIELLUS MICKWITZI TORELLI (MOBERG 1899) Olenellus torelli; Moberg 1892:3. Schmidtia? torelli; Moberg 1899:330, pl. 15, figs. 1-17. Mesonacis torelli (Moberg); Walcott 1910:264, pl. 26, figs. 5-18. Schmidtiellus mickwitzi torelli; Bergstr6m 1973:296, figs. 9-14; Ahlberg et al. 1986:41, fig. 1; Palmer and Repina 1993:26. Types: Lectotype, PIL LO 1411T, see Bergstrom (1973). Discussion: Bergstrém’s (1973) treatment of this taxon is followed here. Occurrence: Sweden: Upper part of the Norretorp Formation, Schmidtiellus mickwitzi zone, at Hardeberga, South Sandby, E of Lund, and at Bjorkelunda and Brantevik, S of Simr- ishamn, Scania. SCHMIDTIELLUS REETAE BERGSTROM 1973 Schmidtiellus reetae; Bergstr6m 1973:301, figs. 15, 16; Ahlberg et al. 1986:41, fig. 1; Palmer and Repina 1993, fig. 6.2. Types: Holotype, Geological Institute of Tallinn, Estonia, Tr 290a. Occurrence: Estonia: The Lukati Beds, Volborthella zone, treated as coeval with the Schmidtiellus mickwitzi zone by Ahlberg et al. (1986), at the Saviranna outcrop, 15 km E of Tallinn. SCHMIDTIELLUS SP. Holmia cf. mickwitzi (Moberg); Skjeseth 1963:40. Schmidtiellus cf. mickwitzi (Moberg); Martinsson 1974:231. Holmia cf. mobergi (Bergstrém); Ahlberg and Bergstr6m 1978:6; Ahlberg and Bergstr6m in Ahlberg et al. 1986:47, fig. 6, ?fig. 7, non fig. 8, non Holmia mobergi Bergstrom 1973. Discussion: Because this taxon is known from such limited material (a single, incomplete cephalon) it is not described as a new species; however, its phylogenetic position was eval- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 83 uated and it was found to be nested within species referable to the genus Schmidtiellus. One of the specimens Ahlberg et al. illustrated (1986, fig. 8) as Holmia cf. mobergi is not con- specific with this new species. Instead, it belongs to Kjerulfia lata, as discussed in Nikolaisen and Henningsmoen (1990). Originally, Skjeseth (1963) and Martinsson (1974) held that this specimen was closely related to species of the genus Schmidtiellus, and this conclusion is upheld here. However, Ahlberg et al. (1986) treated this taxon as more closely related to species within the genus Holmia, commenting on the affinity between this specimen and H. mobergi, here treated as the type of the genus Baltobergstroemia. An apparently new species of Schmidtiellus does bear some affinity to B. mobergi, but can be distinguished from that taxon based on the condition of the following characters: 1. the anterior cephalic border is prominently separated from the extraocular area by a furrow; 2. the anterior cephalic bor- der is developed as a rounded ridge, rather than being an angled, flattened plane; 3. the frontal lobe is relatively shorter (sag.); 4. S3 is evenly versus jaggedly convex; 5. the distal margins of L3 are straight, rather than convex outward; 6. the lateral margins of the glabella opposite L2 are subparallel, rather than diverging when proceeding anteriorly; 7. S2 is not conjoined medially; 8. the extraocular area is relatively narrower (tr.); 9. the occipital spine is developed anterior of the midline on LO. Characters 1, 6, 7 and 9 are important in influ- encing the phylogenetic placement of this specimen within the genus Schmidtiellus rather than with Baltobergstroemia. Occurrence: Norway: | to 2 m above the top of the Ringsaker Quartzite Member, upper Vangsas Formation, Schmidtiellus mickwitzi zone, near Brennsaetersag at the Flagstadelva river, E of Lake Mjosa. GENUS PALMETTASPIS FRITZ 1995 Type species: Palmettaspis consorta Fritz 1995. Assigned taxa: Palmettaspis parallela Fritz 1995; Esmeraldina lidensis Fritz 1995; Esmerald- ina? cometes Fritz 1995. Diagnosis: Anterior cephalic border prominently separated from extraocular area by fur- row, border developed as rounded ridge; frontal lobe does not contact anterior cephalic border furrow (variably developed in “P.” cometes); frontal lobe expands prominently dor- sally; anterolateral margins of glabella prominently separated from anterior extraocular area; length (sag.) of frontal lobe equal to 1 to 1.1 times length (sag.) of LO and L1; S3 evenly convex; notch in ocular lobes where they contact frontal lobe (not in “P.” cometes); S1, S2 and $3 conjoined medially; posterior margin of ocular lobe opposite distal tip of SO or me- dial part of distal margin of LO; line from posterior margin of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 0 to 15 degree angle with sagittal line; LO with spine, posterior of midline; intergenal angle with small spine directly behind distal tip of ocular lobe; extraocular region prominently vaulted, width (tr.) at L1 45% to 55% width of glabella at L1; genal spine extends back roughly eight thoracic segments; thoracic pleural spines very short (exsag.), extend back only one segment, lateral margins inclined at 10 to 20 degree angle relative to sagittal line (see Fritz 1995 for additional characters). Discussion: A new diagnosis of this genus is presented, building on that of Fritz (1995), since in the course of this analysis several additional characters were recognized. The genus Palmettaspis consists of a clade of three species, plus one species, “Palmettaspis” cometes, that sits outside of this clade. Although this makes the genus paraphyletic it seemed preferable to erecting a new monotypic genus for this species. This could have been done; however, there were very few characters excluding this species from a monophyletic Palmettaspis. In 84 PEABODY MusEUM BULLETIN 45 addition, “P.” cometes is known from relatively limited material. Character differences sep- arating “P.” cometes from the monophyletic genus Palmettaspis include: “P.” cometes some- times has the frontal lobe in contact with the anterior border furrow; “P.” cometes has the lateral margins of L4 distal to the lateral margins of LO (as does P. lidensis); “P.” cometes does not have a notch in the ocular lobe where it meets the frontal lobe; and it does not have the straight distal margins of L3 (neither does P. consorta). Fritz (1995) suggested that this genus was closely related to either Mummaspis or Schmidtiellus. Phylogenetic analysis of the Olenelloidea (Lieberman 1998) suggested that this genus is distantly related to Mummaspis and is much more closely related to Schmidtiel- lus, though the two do not share a sister group relationship. This analysis predicts that this genus shares a more recent common ancestry with Esmeraldina and Holmiella, two other taxa known from western Laurentia. Species of this genus are confined to the “Nevadella” zone of the Great Basin of western Laurentia. PALMETTASPIS CONSORTA (FRITZ 1995) Figure 15.2 Palmettaspis consorta; Fritz 1995:720, figs. 5.4, 8.1-8.6, 10.1, 10.2 (see for more complete synonymy); Palmer and Repina 1997:416, fig. 262.5. Types: Holotype, USNM 476024, see Fritz (1995). Material examined: YPM 72908, 72930. Occurrence: Nevada: near the base of the Montenegro Member, Campito Formation, above the base of the “Nevadella” zone, Barrel Spring Section, Lida Wash, Silver Peak Range. PALMETTASPIS PARALLELA FRITZ 1995 Palmettaspis parallela; Fritz 1995:720, figs. 5.5, 5.6, 9.1-9.5 (see for more complete syn- onymy). Types: Holotype, USNM 476027, see Fritz (1995). Discussion: Fritz (1995) provisionally assigned a partial thorax to this species, and his as- signment was accepted for the purposes of coding thoracic characters in this species. Occurrence: Nevada: near the base of the Montenegro Member, Campito Formation, above the base of the “Nevadella” zone, Barrel Spring Section, Lida Wash, Silver Peak Range. PALMETTASPIS LIDENSIS (FRITZ 1995) Esmeraldina lidensis; Fritz 1995:716, figs. 5.2, 7.4—7.9. Types: Holotype, USNM 476020, see Fritz (1995). Discussion: Fritz (1995) provisionally assigned a lone pygidium to this species, and this as- signment was accepted for the purpose of coding pygidial characters for phylogenetic analy- sis. The posterior pygidial margin was treated as being roughly transverse. Fritz (1995) assigned this species to the genus Esmeraldina, but phylogenetic analysis indicates it is more closely related to a clade that includes the type of the genus Palmettaspis than it is to the type of Esmeraldina, E. rowei. Therefore, it is reassigned to the genus Pal- mettaspis. Characters placing this species with Palmettaspis rather than with Esmeraldina in- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 85 clude: 1. the frontal lobe does not contact the anterior border furrow; 2. the intergenal angle forms a greater angle with a transverse line; and 3. the genal spine angle is opposite the me- dial part of the distal margin of L1, rather than LO. However, interesting convergences be- tween this species and E. rowei include: 1. the anterodistal margins of L3 are formed by the axial furrows; 2. the posterior margin of the ocular lobe is opposite the distal tip of SO; and 3. the genal spine near the spine angle has a similar orientation relative to a sagittal line. Occurrence: Nevada: near the base of the Montenegro Member, Campito Formation, above the base of the “Nevadella” zone, Barrel Spring Section, Lida Wash, Silver Peak Range. “PALMETTASPIS” COMETES (FRITZ 1995) Esmeraldina? cometes; Fritz 1995:716, figs. 5.3, 9.8, 10.3-10.5. Types: Holotype, USNM 479804, see Fritz (1995). Discussion: Fritz (1995) questionably assigned this species to the genus Esmeraldina, but a clade including several species of Palmettaspis is more closely related to the type of the genus Esmeraldina than is “P.” cometes, based on phylogenetic analysis of the Holmiidae. Therefore, it cannot be placed within that genus without substantially changing the taxon- omy of the family. Occurrence: Nevada: near the base of the Montenegro Member, Campito Formation, above the base of the “Nevadella” zone, Barrel Spring Section, Lida Wash, Silver Peak Range, south- ern Nevada. GENUS ESMERALDINA RESSER AND HOWELL 1938 Type species: Holmia rowei Walcott 1910. Assigned taxa: Holmia? sp. of Fritz (1973). Diagnosis: Anterior cephalic border developed as rounded ridge, prominently separated by furrow from extraocular area; frontal lobe contacts anterior border furrow, lateral margins distal of lateral margins of LO, expands prominently dorsally; anterolateral margins of frontal lobe prominently separated from extraocular area by furrow (not anterior border furrow); length (sag.) of L4 equal to 1 to 1.1 times length of LO and L1; dorsal notch in ocular lobe where it contacts lateral margins of frontal lobe; S3 evenly convex, medial and distal tips as far forward; anterodistal margins of L3 formed by axial furrows; distal margin of L3 straight; lateral margins of glabella opposite L2 diverging anteriorly; medial and distal tips of S2 as far forward; $1, $2 and $3 conjoined medially; posterior edge of ocular lobe opposite distal tip of S0; line from posterior margin of ocular lobe to junction of lobe with glabella forms roughly 5 to 15 degree angle relative to sagittal line; LO with spine posterior of midline; lat- eral lobes absent from LO; intergenal angle with small spine directly behind distal tip of oc- ular lobe; intergenal angle forms roughly 10 to 15 degree angle with transverse line; extraoc- ular area prominently vaulted, width (tr.) at L1 45% to 55% width of glabella at L1; genal spine angle opposite distal margin of LO; lateral margin of genal spine near spine angle forms roughly 30 to 40 degree angle with sagittal line; genal spine extends back roughly 8 thoracic segments; thoracic pleural spines on first four segments extend back only one segment; lat- eral margins of spines inclined at 40 to 50 degree angle relative to sagittal line; posterior mar- gins of first four thoracic segments near spines flex slightly anteriorly; length (sag.) and width (tr.) of pygidium roughly equal; posterior margin of pygidium roughly transverse. Discussion: Fritz (1995) assigned a few additional taxa to this genus, but these had to be 86 PEABODY MuseEUM BULLETIN 45 placed within a revised “Palmettaspis.” The character evidence for this is discussed above. Resser and Howell (1938) included two taxa in this genus, Mummaspis occidens (Walcott) and M. macer (Walcott), which also can no longer be considered congeneric with E. rowel. Both of these are clearly different from the type of the genus Esmeraldina, and character ev- idence supporting their separation and distinctness from that genus is given below under the genus Mummaspis. Fritz (1973) figured a partial cephalon from the Sekwi Formation (probably in the “Nevadella” zone) of the Mackenzie Mountains, Northwest Territories, Canada that bears prominent similarity to the type of the genus Esmeraldina. For those characters that could be coded, this specimen appears to almost completely match the di- agnostic characters of the genus, and thus this species is tentatively placed within Esmeral- dina. The only significant difference between this species and E. rowez is in the condition of $2, which is less prominently incised in the new species. Thus, this genus probably ranges throughout western Laurentia during the “Nevadella” zone. Because additional characters of this genus were discovered in the course of this analysis, a revised diagnosis of the genus is presented. ESMERALDINA ROWEI (WALCOTT 1910) Fiparesds 15s Esmeraldina rowei (Walcott); Fritz 1995:714, figs. 5.1, 6.1-6.12, 7.1—7.3, 10.10, 10.11; Palmer and Repina 1997:414, fig. 262.1a, b; Lieberman 1998:71, fig. 3.4 (see for more complete synonymy). Types: Lectotype, USNM 56801c, see Fritz (1995). Material examined: LACMIP 6751, 6753, 6756-6758 (all lots containing a total of over 100 individuals); MCZ 2523 (lot containing 4 specimens), 7680; YPM 35865, 72909, 72929. Occurrence: Nevada: near the base of the Montenegro Member, Campito Formation, above the base of the “Nevadella” zone, Barrel Spring Section, Lida Wash, Silver Peak Range. GeENusS HOLMIELLA FRITZ 1972 Type species: Holmiella preancora Fritz 1972. Assigned taxa: Holmiella falcuta Fritz 1972, and Holmiella sp. of Nelson 1976. Diagnosis: Anterior cephalic border developed as flattened plane, prominently separated from extraocular area by furrow; frontal lobe contacts anterior border furrow, lateral mar- gins distal to lateral margins of LO, expands prominently dorsally; anterolateral margins of glabella prominently separated from extraocular area by furrow; length (sag.) of frontal lobe 1 to 1.1 times length (sag.) of LO and L1; dorsal notch-like truncation in ocular lobes where they contact frontal lobe of the glabella; anterodistal margins of L3 formed by axial furrows; distal margins of L3 straight; $3 evenly convex, medial and distal tips as far for- ward; lateral margins of glabella opposite L2 subparallel; $1, 2 and $3 not conjoined me- dially; medial and distal tips of $2 as far forward; line from posterior edge of ocular lobe to junction of lobe with glabella forms roughly 10 to 15 degree angle with sagittal line; poste- rior margin of ocular lobe opposite distal tip of SO; lateral lobes present on LO; node or spine on LO posterior of midline; intergenal angle with small spine directly behind distal tip of ocular lobe; intergenal angle forms 80 to 90 degree angle with transverse line; extraocu- lar region prominently vaulted, very narrow, width (tr.) at L1 10% to 15% width of glabella at L1; genal spine angle opposite distal tip of L3 or L4; lateral margin of genal spine near SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 87 spine angle forms roughly 60 to 70 degree angle relative to sagittal line; pygidium twice as wide (tr.) as long (sag.), posterior margin notched. Discussion: The node including this genus is the most strongly supported in the phylogeny given in Figure 13, and the two taxa definitively assigned to this genus share a large number of synapomorphous traits. We can be extremely confident that these two species are closely related. Nelson (1976) figured a spectacular, but poorly preserved, specimen from the “Nevadella” zone of the Great Basin that he referred to the genus Holmiella. This material was examined by the author and Nelson’s (1976) generic assignment appears correct. How- ever, it was too poorly preserved to incorporate into phylogenetic analysis. The chief differ- ence between this new species and the taxa from northwestern Canada is that the Laurent- ian taxon appears to have a somewhat broader extraocular area. Until this new taxon can be studied in detail, the generic diagnosis of Holmiella was treated as including a very narrow extraocular area. Including this species, exponents of the genus are known from through- out western Laurentia in the “Nevadella” zone. Ahlberg et al. (1986) figured a complete py- gidium from the Holmia kjerulfi zone of Sweden that they referred to as Holmiella sp. It does agree in one important respect with the pygidium of the type of the genus H. preancora: the width (tr.) of the pygidium is roughly twice its length (sag.). However, the posterior pygidial margin of Ahlberg et al’s (1986) Holmiella sp. is roughly transverse, or even weakly convex posteriorly, whereas in H. preancora it has a median notch. Because pygidial material for all the holmiids is so limited, and only two informative characters of the pygidium could be recognized in this analysis, it cannot be definitively stated whether or not this pygidium be- longs to Holmiella or to some other holmiid genus. This pygidium differs in as many char- acters from those known for Holmiella as it does from those known for Esmeraldina, Schmidtiellus, Holmia and “Baltobergstroemia.” Therefore, at this time the pygidium is treated as belonging to an indeterminate holmiid. If it does indeed belong to a species of Holmiella, it would extend the geographic range of that genus into Baltica. HOLMIELLA PREANCORA FRITZ 1972 Holmiella preancora; Fritz 1972:25, pl. 4, figs. 1-13; Palmer and Repina 1993:26, figs. 6.4, 6.65 Lieberman 1998:67. Types: Holotype, GSC 27241, see Fritz (1972). Material examined: GSC 27240, a paratype. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, 348 ft to 668 ft above the base of the formation, “Nevadella” zone, following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. HOLMIELLA FALCUTA FRITZ 1972 Figure 15.4 Holmiella falcuta; Fritz 1972:26, pl. 7, figs. 7-17. ivpessilolotypeGs€ 2727 7, see Fritz (1972). Material examined: GSC 27273, a paratype. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation 88 PeaBopY MusEUM BULLETIN 45 (Handfield 1968), 0.2 mi SE of June Lake, 668 ft to783 ft above the base of the formation, “Nevadella” zone, following Fritz (1972), S end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. Phylogenetic Analysis of Fritzolenellus Fritzolenellus is a small clade distributed throughout part of Laurentia (northwestern Canada and northwestern Britain) in the lower and middle Olenellus zone (according to the stratigraphic correlations of Fritz [1992] and Palmer and Repina [1993]). It includes three species, all of which were subjected to phylogenetic analysis here. Two outgroup taxa were employed in phylogenetic analysis, Wanneria walcottana and Holmia kjerulfi. These species are appropriate outgroups to the genus Fritzolenellus based on the higher level phylogenetic analysis of the Olenelloidea presented in Lieber- man (1998). Phylogenetic patterns were determined by parsimony analysis of four ho- laspid exoskeletal characters. Only a limited number of characters could be recognized because of the paucity of taxa assigned to the genus, as well as the poor state of preserva- tion of two of the species assigned to Fritzolenellus. However, since these characters seemed to contain some phylogenetic information content, cladistic analysis was pursued. These characters and character states are given in Table 9. The codings for the taxa ana- lyzed are given in Table 10. All characters were treated as unordered (nonadditive). These data were subjected to an exhaustive search on PAUP v. 3.1.1 (Swofford 1993). One most parsimonious tree (Figure 16) was recovered of length 5 steps, consistency index = 0.80, and retention index = 0.75 The g, statistic, was —0.94, and this value differs from those values derived from distributions using random data at the 0.01 level of confidence (Hillis 1991). This suggests that there is a strong and robust phylogenetic signal to the character data given in Table 10. The bootstrap confidence value for the node of the most parsimo- nious tree duplicated in the bootstrap analysis is given in Figure 16. Using the method of Bremer (1994), three trees of length less than or equal to 6 steps and eight trees of length less than or equal to 7 steps were recovered before the analysis was terminated because the consensus cladogram was a complete polytomy. The branch support value is given in Fig- ure 16. The total tree support for the cladogram is 0.2, a moderate value according to Bre- mer (1994). All the taxa within this genus are known from the lower and middle Olenellus zone, implying fairly good concordance between stratigraphic and phylogenetic patterns (ac- cording to the correlations of Fritz [1992] and Palmer and Repina [1993]). The species that appears first in the fossil record is cladistically subordinate to the other two species of Frit- zolenellus, which appear relatively later in the fossil record. Systematic Paleontology Famity “Laudoniidae” Harrington 1956 SUBFAMILY “Laudontinae” Harrington 1956 Lieberman (1998) argued for the designation of a paraphyletic family “Laudoniidae” and subfamily “Laudoniinae” (paraphyly designated using the shuttermark convention of Wiley 90 PEABODY MUSEUM BULLETIN 45 Wanneria walcottana Fritzolenellus truemani Fritzolenellus lapworthi Fritzolenellus reticulatus Holmia kjerulfi Figure 16. The most parsimonious cladogram of length 5 steps produced from analysis of character data in Table 10 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The retention index is 0.75, and the consistency index is 0.80. The following nodes in the text were supported by the following boot- strap confidence values (see text for bootstrapping procedure used): Node 1] = 0.52; Node 2 = 0.79. Node 2 has a branch support value (Bremer 1994) of 1. Character states are placed at nodes, using MacClade v. 3.04 (Mad- dison and Maddison 1992), with the characters given in Table 9. The apomorphic state is given in parentheses. Node 1, 3(1); Node 2, 1(1), 2(1). [1979]), modifying Laudoniinae Harrington 1956. The newly defined paraphyletic family and subfamily includes the genera Laudonia, Mummaspis and Fritzolenellus, and that supra- generic classification is followed for the purposes of this analysis. Included Taxa GENUS FRITZOLENELLUS LIEBERMAN 1998 Type species: Olenellus truemani Walcott 1913. Assigned taxa: Olenellus reticulatus Peach 1894; O. lapworthi Peach 1894; Fritzolenellus sp. aff. truemani; possibly Olenellus intermedius Peach 1894. Diagnosis: Anterior cephalic border developed as raised ledge; anterior border promi- nently separated from extraocular area by furrow; plectrum absent; anterior part of glabella prominently separated from anterior extraocular area by furrow (not anterior bor- der furrow); prominent parafrontal band not visible in dorsal view; anterior margins of frontal lobe at each side of midline deflected posteriorly at roughly 40 degree angle rela- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 91 tive to transverse line; length (sag.) of L4 equal to roughly 1.5 times length (sag.) of LO and L1; L4 expands prominently dorsally; lateral margins of L4 distal to lateral margins of LO; pre-ocular furrow on frontal lobe directed inward and forward from glabellar margin; an- terolateral margins of ocular lobe separated from extraocular area by prominent furrow; ocular lobe with prominent ocular furrow; line from posterior tip of ocular lobe to junc- tion of posterior margin of lobe with glabella forms roughly 10 to 15 degree angle with sagittal line; transverse profile of ocular lobes convex dorsally; surface of interocular area arched; width (tr.) of interocular area approximately equal to maximum width of ocular lobe; anterodistal margins of L3 formed by ocular lobe; distal margins of L3 convex out- ward; S3 jaggedly convex, conjoined medially; line between ends of S2 transverse; L2 and L3 do not merge; S2 convex anteriorly; distal margins of L2 when proceeding anteriorly di- verging; distal sector of SO convex anteriorly, well anterior of proximal end; posterior mar- gin of LO roughly transverse; lateral lobes on LO present; extraocular area flattened, width (tr.) roughly equal to 75% width of glabella at L1; length (exsag.) of genal spine roughly equal to length (sag.) of first four to five thoracic segments; genal spine angle developed opposite medial part of distal margin of LO; intergenal angle developed posterior of point half way between ocular lobes and genal spine; intergenal angle roughly parallel to trans- verse line; medial part of posterior border between intergenal angle and LO roughly trans- verse; thorax broken up into pro- and opisthothorax; anterior margin of T3 before flexing strongly posteriorly roughly parallel to transverse line; T3 macropleural, projects posteri- orly 6 to 8 thoracic segments; anterior margin of thoracic pleural furrow on T3 when pro- ceeding from proximal to distal edge directed weakly posteriorly, before flexing strongly posteriorly; posterior margin of thoracic pleural furrow on T3 directed evenly posterolat- erally; thoracic pleural spines on T5 to T8 developed as broad sweeping projections ex- tending roughly four segments back; prothoracic axial rings without prominent lateral fur- rows; single nodes very faint, but present on median portion of some thoracic axial rings; thoracic pleural furrows extend width of inner pleural region; boundary between thoracic pleural furrow and anterior band sharp; thoracic pleural furrows short (exsag.), length equal to roughly half length (exsag.) of posterior band; thoracic pleural spines T5 to T8 at spine midlength with width (tr.) one half to two-thirds length (exsag.) of medial part of inner pleural region; prominent spine on 15th thoracic axial ring of length (sag.) of entire prothorax, base of spine nearly as wide (tr.) as axis of segment; segments of opisthothorax with prominent pleurae; pygidium with length (sag.) equal to width (tr.); posterior mar- gin of pygidium weakly convex (only some of these thoracic characters, and none of the pygidial characters, could be determined in F. lapworthi; none of the pygidial characters could be determined in F. reticulatus). Discussion: Lieberman (1998) introduced the genus Fritzolenellus and provided a diagnosis for it. A slightly refined diagnosis is presented here because all species within the genus could be incorporated into this analysis. One of the characters listed in the above diagnosis that mer- its additional discussion is the condition of L2 and L3; these do not appear to merge distally due to $2, which contacts the axial furrows. On one specimen of FE. truemani figured by Fritz (1992, pl. 7, fig. 1) this character state is not clearly visible. In this figure, there are actually two cephala. One, GSC 99007, the inverted specimen, clearly displays the diagnostic character state. The other, GSC 99006, clearly displays the character state on the left lateral lobes, but the right lateral lobes are a different matter. Here, S2 is very weak, but appears to contact the axial furrow, faintly bisecting L2 and L3 such that they do not merge. Based on this character in- terpretation, this species was treated as having L2 and L3 not merging laterally. As previously defined (Cowie and McNamara 1978; Fritz 1972, 1992; Palmer and Q2 PEABODY MUSEUM BULLETIN 45 Table 9. Description of characters and character states used in phylogenetic analysis of Fritzolenellus; (0) does not always represent the primitive state as two outgroup taxa were used in this analysis; however, for convenience, one taxon, Holmia kjerulfi was coded as having all (0) character states; (1) can represent the derived state. Primitive state Derived states 1. Anterior cephalic border between (0) length (sag.) of LO (1) half the length (sag.) frontal lobe and genal spine angle of LO with length (exsag.) roughly equal to 2. Frontal lobe (0) contacts the anterior (1) does not contact the border furrow anterior border furrow 3. Anterolateral margins (0) not prominently (1) prominently separated of glabella separated from from extraocular area extraocular area by by marginal furrow marginal furrow 4. Posterior margins of (0) distal tip of SO (1) medial part of distal ocular lobes opposite margin of L1 Repina 1993) this taxon included a large number of phylogenetically disparate units. Some of these are more closely related to the type of the genus Olenellus, whereas others, including the taxa discussed below under Fritzolenellus, are more closely related to a clade that includes Laudonia, Mummaspis, Bristolia, Nephrolenellus and several other genera. A new genus was necessary for the latter species, otherwise Olenellus would be polyphyletic. This has implications for stratigraphic correlations in the late Early Cambrian (e.g., Palmer and Repina 1993) that rely on a genus concept (in particular the genus Olenellus of the eponymous Olenellus zone) to correlate strata. Certainly genera should be used to correlate strata only with caution, though such methods appear to predominate the stratigraphy of the Early Cambrian. If the taxonomic status of the genus Olenellus is changed fundamentally, it could potentially cause a concomitant change in stratigraphic patterns. A few species formerly assigned to the genus Olenellus need to be reassigned to Frit- zolenellus. In particular, what Cowie and McNamara (1978), McNamara (1978) and sources cited therein referred to as Olenellus reticulatus and O. lapworthi, two species from the mid- dle Olenellus zone of northwestern Scotland, actually appear to belong to the genus Frit- zolenellus. They bear all of the diagnostic characters of the genus given above. The thoracic material of F. lapworthi and F. reticulatus are very similar to the type of the genus, F. true- mani. However, one difference evident between the thorax of F. reticulatus and that of F. truemant is the relatively narrower (tr.) pleural spines in the former. This generic assignment for these two species challenges a central tenet of McNamara’s (1978) paper, as he placed these species into a paedomorphic evolutionary pathway along with Olenelloides armatus and what is viewed here as Mesonacis hamoculus (discussed in SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 93 Table 10. Character state distributions for Fritzolenellus and outgroups used in phylogenetic analysis. Characters and alternative states are as listed in Table 9. P23 4 Holmia kjerulfi 0000 Wanneria walcottana OF0501 Fritzolenellus truemani OFO=1 Fritzolenellus lapworthi a 170 Fritzolenellus reticulatus ites et leo greater detail above under the genus Mesonacis). Mesonacis and Olenelloides are distantly re- lated to Fritzolenellus, and to one another, based on a phylogenetic analysis of the Olenel- loidea by Lieberman (1998), and thus it is inconceivable that they are connected by a pae- domorphic evolutionary pathway. Character differences separating species of Mesonacis from species of Fritzolenellus include: 1. the anterior part of the glabella is prominently separated from the anterior ex- traocular area by a furrow in Fritzolenellus (not the anterior border furrow), it is not in Mesonacis; 2. the anterolateral margins of the ocular lobe are separated from the extraoc- ular area by a prominent furrow in Fritzolenellus, they are not in Mesonacis; 3. the inte- rocular area is arched dorsally in Fritzolenellus, it is developed as a flattened shelf in Mes- onacis; 4. the distal sector of SO is convex anteriorly in Fritzolenellus, it is straight in Mesonacis; 5. lateral lobes are present on LO in Fritzolenellus, they are absent in Mesonacis; 6. the intergenal angle roughly parallels a transverse line in Fritzolenellus, it is directed an- teriorly between 30 to 50 degrees relative to a transverse line in Mesonacis; 7. the anterior margin of T3 in Fritzolenellus parallels a transverse line medially, in Mesonacis medially it is directed anteriorly at least five degrees relative to a transverse line; 8. the pleural spine of T3 projects posteriorly 6 to 8 thoracic segments in Fritzolenellus, in Mesonacis it either projects posteriorly the entire length of the prothorax, or 3 to 4 segments; 9. the anterior margin of the pleural furrow on T3, when proceeding from the proximal to distal edge is directed weakly posteriorly before flexing strongly posteriorly in Fritzolenellus, in Mes- onacis it parallels a transverse line, before flexing strongly posteriorly; 10. in Fritzolenellus the thoracic pleural furrows extend the width of the inner pleural region, in Mesonacis they extend onto the spines; 11. in Fritzolenellus the thoracic pleural furrows (excluding those of T3) are very short (exsag.), equal to half the length (exsag.) of the posterior band, in Mesonacis they are relatively longer (exsag.), equal to 1 to 1.5 times the length of the posterior band; 12. in Fritzolenellus the pygidium is relatively broader, having length (sag.) equal to the width (tr.), whereas in Mesonacis the pygidium has the length (sag. ) equal to 1.5 times the width. Olenellus intermedius Peach is another species from the middle Olenellus zone of Scot- land. It is known from two poorly preserved specimens, both early ontogenetic stages. Mc- Namara (1978) also placed this species into the paedomorphic evolutionary pathway, along with F. reticulatus, F. lapworthi and M. hamoculus. Olenellus intermedius appears to be con- specific with, or belongs to a species closely related to, either F. reticulatus or F. lapworthi. Specimens of F. intermedius have the following characters, found in these species and not in 94 PEABODY MUSEUM BULLETIN 45 Figure 17. 1. Fritzolenellus truemani (Walcott), 495 ft to 575 ft above the base of the Mural Fm. (formerly Type Tah Fm.), Cinnamon Peak-Whitehorn Mountain section, just N of the Mount Robson Provincial Park boundary, western Alberta, Canada, GSC 99006 (top) and 99007, x1.4. 2-5. ?7The Mural Fm., in the talus slope immediately W of Mumm Peak, just N of the Mount Robson Provincial Park boundary, western Alberta, Canada. 2. Mummaspis occidens (Walcott), MCZ 110675, x1.5. 3. Mummaspis truncatooculatus (Fritz), MCZ 110677, x1.5. 4, 5. Mum- maspis muralensis (Fritz). 4. MCZ 110676, x1.5. 5. MCZ 110673, x1.5. 6. Mummaspis macer (Walcott), Kinzers Shale, Fruitville, Pennsylvania, USNM 60092, x1.7. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 95 M. hamoculus: the lateral margins of L4 appear to be distal to, rather than anterior to, the lateral margins of LO; a prominent pre-ocular furrow directed inward and forward is visi- ble; a furrow appears to be present in front of the anterolateral margins of the ocular lobe and the anterolateral margins of the glabella; and the lateral margins of LO and L1 are sub- parallel. The chief similarities between F. intermedius and M. hamoculus are the condition of the genal spine angle, relatively advanced in both taxa (opposite SO in the former, oppo- site the medial part of the distal margin of L1 in the latter) and the state of the intergenal angle, directed anteriorly at a roughly 30 to 35 degree angle relative to a transverse line. The bulk of the character evidence suggests that this species is closely related to F. lapworthi and F. reticulatus, but as it is known from poorly preserved, early ontogenetic stage material, it is only tentatively placed within the genus Fritzolenellus. FRITZOLENELLUS TRUEMANI (WALCOTT 1913) Figure 17.1 Olenellus truemant; Walcott 1913:316, pl. 54, figs. 2, 6, and 8; Walcott 1916, pl. 17, figs. 2, 6, and 8; Fritz 1992:15, pl. 6, figs. 1-4, pl. 7, figs. 1-6, and text fig. 6b. Fritzolenellus truemani (Walcott); Lieberman 1998:72. Types: Lectotype, USNM 60084, designated in Fritz (1992). Discussion: As discussed in Fritz (1992) and Lieberman (1998), the specimens that Lochman in Cooper et al. (1952) illustrated from the Caborca region of Mexico and as- signed to Olenellus truemani are distinct from Walcott’s (1913) original material of that species. The illustrated cephala from late ontogenetic stage individuals differ in the condi- tion of several characters from F. truemani: they lack the convex (tr.) ocular lobes; have the lateral margins of the frontal lobe medial to the lateral margins of LO; have L2 and L3 merg- ing distally; have a straight SO; and have LO with a strongly convex posterior margin. On the basis of these and other characters the material cannot be treated as conspecific, or even congeneric, with F. truemani. Instead, these specimens appear to be referable to Olenellus, and will be discussed in a subsequent publication. As discussed in Fritz (1992) and Lieberman (1998), the specimens Fritz (1972) illus- trated and assigned to O. truemani are not closely related to F. truemani when we consider on what Walcott (1913) originally based his species truemani. The material Fritz (1972) il- lustrated is actually a new species, Elliptocephala walcotti (discussed in greater detail above under the genus Elliptocephala). Fritz (1992) figured several cephala representing early ontogenetic stages that he as- signed to Mummaspis oblisooculatus Fritz. These specimens (1992, pl. 16, figs. 4-7, and pl. 17, figs. 1, 2) are similar, if not identical, to F. truemani. They are referred to as F. sp. aff. truemani. The chief difference between these specimens and F. truemanz is that in the for- mer the frontal lobe extends relatively farther forward. Differences between these speci- mens of Fritzolenellus and those of species belonging to the genus Mummaspis, such as true M. oblisooculatus, include the following: in Fritzolenellus the anterior margins of the frontal lobe at each side of the midline are deflected posteriorly at a roughly 40 degree angle relative to a transverse line, whereas in Mummaspis the margins are deflected poste- riorly at a roughly 10 degree angle relative to a transverse line; in Fritzolenellus S1 is not conjoined medially, at least in later ontogenetic stages, and this appears to be the case for EF. sp. aff. truemani, whereas in Mummaspis it is conjoined medially; in Fritzolenellus the distal sector of SO is straight, at least in later ontogenetic stages, and this appears to be the 96 PEABODY MUSEUM BULLETIN 45 case for F. sp. aff. truemani, whereas in Mummaspis it is convex; and in Fritzolenellus the posterior margin of LO is more transverse than it is in Mummaspis. Therefore, these spec- imens can no longer be assigned to the genus Mummaspis. Instead, they are treated as com- prising a species closely related to FE. truemani and are accordingly referred to as F. sp. aff. truemani. These specimens were not described as comprising a new species and will not be subjected to phylogenetic analysis until additional material comprising later ontogenetic stages is recovered. As defined originally by Walcott (1913), FE. truemani was held to be closely related to the genus Olenellus. However, a large number of characters separate this species from the types of O. (Olenellus) and what was formerly the type of O. (Paedeumias), and therefore F. truemani must be excluded from those genera. In particular: 1. in FE. truemani the antero- lateral margins of the frontal lobe are prominently separated from the anterior extraocular area by a furrow (not the anterior border furrow), whereas in Olenellus they are not; 2. in F. truemani L4 expands more prominently dorsally than it does in Olenellus; 3. in F. truemanti the lateral margins of L4 are distal to LO, whereas they are proximal to or directly anterior to the lateral margins of LO in Olenellus; 4. in F. truemani the pre-ocular furrow is directed inward and forward from the glabellar margin, whereas when visible it is transverse in Olenellus; 5.in F. truemani the anterolateral margins of the ocular lobe are prominently sep- arated from the extraocular area by a furrow, in Olenellus they are not; 6. in F. truemani a transverse profile of the ocular lobes is convex dorsally whereas it is flattened in Olenellus; 7.in F. truemani the interocular area is arched dorsally, whereas it is developed as a flattened shelf in Olenellus; 8. in F. truemani S3 is conjoined medially whereas it is not in Olenellus; 9. in F. truemani a line between the ends of S2 is transverse, whereas it is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line in Olenellus; 10. in F. truemani L2 and L3 do not merge distally, in Olenellus they do; 11. in F. truemani 82 is convex anteriorly, in Olenellus it is straight; 12. in F. truemani the distal sector of SO is straight whereas in Olenellus it is convex anteriorly; 13. in FE. truemant lateral lobes are pre- sent on LO, they are absent in Olenellus; 14. in F. truemani the intergenal angle is developed posterior of a point half way between the ocular lobes and the genal spine angle, whereas it is developed adjacent to the genal spine angle in Olenellus; 15. in F. truemani the medial part of the posterior border between the intergenal angle and LO is transverse, in Olenellus it flexes posteriorly; 16. in F. truemani the thoracic pleural furrows extend the width of the inner pleural region, in Olenellus they extend onto the spines; 17. in F. truemanti the thoracic pleural furrows (excluding those of T3) are very short (exsag.), with length equal to half the length (exsag.) of the posterior band, in Olenellus they are relatively longer (exsag.), with length equal to roughly 1.5 times the length (exsag.) of the posterior band; 18. in FE. true- mani the opisthothorax has prominent pleurae whereas these are absent in Olenellus; 19. in F. truemani the pygidium is relatively broad, with the length (sag.) roughly equal to the width (tr.), whereas it is relatively narrow in Olenellus, with the length (sag.) roughly equal to 1.5 times the width (tr.). Material examined: GSC 99002, 99003, 99006, 99007, 99010, 99011; USNM 60084, the lec- totype, and 60085-60091. Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately west of Mumm Peak, and from the middle Mural Formation, 412 ft to 439 ft above the base of that formation, at the Mumm Peak section, and 495 ft to 575 ft above the base of the Mural For- mation (formerly Type Tah Formation), Cinnamon Peak-Whitehorn Mountain section, in the lower Olenellus zone, Early Cambrian, just N of the Mount Robson Provincial Park boundary, western Alberta (see discussion in Fritz 1992). SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 97 FRITZOLENELLUS LAPWORTHI (PEACH AND HORNE 1892) Olenellus lapwortht; Peach and Horne 1892:227, pl. 5, figs. 2-4, 6; Cowie and McNamara 1978:620, pl. 69, figs. 1-6 (see for complete synonymy); McNamara 1978:635, text figs. 3, A 527. Types: Lectotype, GSE 5364, see Cowie and McNamara (1978). Discussion: Character evidence that this species can no longer be assigned to the genus Olenellus is congruent with the evidence that indicates the separation of F. truemani from Olenellus, and is discussed above under that species. Material examined: GSE 5364, the lectotype, and GSE 13301 and 13310, both paralecto- types. Occurrence: Scotland: The “Fucoid” Beds, middle Olenellus zone (according to Palmer and Repina 1993), Allt nan Righreon near the hill-track 4 mi S of Dundonnell, Ross and Cro- marty, northwestern Scotland (see Cowie and McNamara 1978). FRITZOLENELLUS RETICULATUS (PEACH 1894) Olenellus reticulatus; Peach 1894:665, pl. 30, figs. 1-5, pl. 31, figs. 1-7; Cowie and McNamara 1978:624, pl. 69, figs. 7-15, pl. 70, figs. 1, 2, 12 (see for complete synonymy); McNamara 1978:635, text figs. 4g, h, 5—7. Types: Lectotype, GSE 5343, see Cowie and McNamara (1978). Discussion: One of the specimens of F. reticulatus illustrated in Cowie and McNamara (1978, pl. 69, fig. 10) has a slightly narrower extraocular area than that typical of the lecto- type and other material of the species. As this specimen appeared to resemble typical F. reticulatus in the condition of all other morphological characters, at this time it was treated as conspecific with F. reticulatus. Character evidence that this species can no longer be as- signed to the genus Olenellus is congruent with the evidence that indicates the separation of F, truemani from Olenellus, and is discussed above under that species. Material examined: GSE 5343, the lectotype, and GSE 5372 and 13295. Occurrence: Scotland: The “Fucoid” Beds, middle Olenellus zone (according to Palmer and Repina 1993), the northern slopes of Meall aGhiubhais, 3 mi WNW of Kinlochewe, and Allt nan Righreon near the hill-track 4 mi south of Dundonnell, Ross and Cromarty, north- western Scotland (see Cowie and McNamara 1978). : Leemcy a af 1.0 % mg . oriliscnmees napa , or Se ee igre: “ peariatn Tiree spe Gijine &-te ee iencsae enn ‘9 panies? we gaily’ enruyte easiest a eae i agit te iecesdins 7 SOTO DS eset berp eee) ae <9 | vu Davtingy tid ay laces ae sydney atale, taulld eae Seabees ca i" e Dip geome nlc duiseere alls chanting =i stores vie esha elo ieee alk Bree of 0). ior tle? wider nly euborsds Gove geht) arial ward © phonlcanhy ithe? HECK Soh AMEE RR adn: ni wire) ttabae Oh pee ni laineienl 1 igitea) Merg ie, as aval ead , me ew owe sc apa ied Tare £204 Qi) Whe edly pero vite ~nucnno CRieh Liye ett lead Dao) leah hare ge? iS) Chae paeay wi git} aunty Ath cee ae : : » yy J sur Laxeset . PuleWpectayres by nome! ral abd vineielnenti tia thus jou Tee A, ’ ea Pepe dh eure wy Gee int Paes' e Pa 4) ) ora ts bP ee BE eT thet be tet a the ga woarnaattly: Hemet lua wavy { A . 4 DO! age ula wh eee ATH } witine’s hh Legit eteg? Hic” i dls Cuarpenh ome i . wit f so tity ee 2 44s . ) “i } | i fee ie ’ : Ree: L byii buy Awe oR) une iy anewe @) 34 ; 3 i ioe PTR anes OTe. MiG) Hae gert Pe ea pile ‘uel wt pile . @beterpetd ata Me | , ie ude ih etledd sada . boven taera dy =e ; Siiee, elle me ey Mi ron! ey weheTeR ees | mide \ptewoort a alia Pr pas > peridot inpegas A je tee Oy te \ a vgamein Yek quran AR aren ey eT a >< al ©) - ' a) ‘ ; f ‘ ar rll hitey ta Pat Ma vue i ‘7 , i. : pst be ) ‘ cl, > we be hn) von eG 4 Ae hee caper eat hermes } . henwl- lee a vw 1s | hare i oil weet Hite betucsrunrtrolwn* hb Jeltnideay i? sacs aremenedd OR nn: ? Aiea eDiw arg Sai) 49 2 Sere te ei db vlan hla ole aia waning . a ; Dever! Lada et eee) oe new ra ; 72, ; a of ic | rru woG! | A F, f a ue wait phew Ul | y) . oi) is payee | | , live | 1 ‘ ; / o-¢ é =A > «Ss vey * é , ii Eg eMaiitin i ao “ba = = nt ae ruil biuret Bement? Fe “TT s Lifer Fea a ih Ot S72 rte Sv otoi ue Ry: ae : 2% Jet Sofiia di eneneat ay fo gh Wh ee “a ibe Te ead aS d | nn eee C04 ame myske an wi. 4 —. Phylogenetic Analysis of Mummaspis Mummaspis is a small clade distributed throughout part of Laurentia (northwestern Canada and eastern Pennsylvania) in the lower middle and perhaps upper Olenellus zone. It includes five species, all of which were subjected to phylogenetic analysis here. A single outgroup taxon was employed in phylogenetic analysis, Fritzolenellus truemani (See Chapter 5). This species is an appropriate outgroup to the genus Mummaspis based on the higher level phylogenetic analysis of the Olenelloidea presented in Lieberman (1998). Phylogenetic patterns were determined by parsimony analysis of five holaspid exo-skeletal characters, given in Table 11. The codings for the taxa analyzed are given in Table 12. All char- acters were treated as unordered (nonadditive). These data were subjected to an exhaustive search on PAUP v. 3.1.1 (Swofford 1993). One most parsimonious tree (Figure 18) was re- covered of length 9 steps, consistency index = 0.78, and retention index = 0.67. The g, statistic, which is used to measure tree length skewness distributions, was —0.44. However, this value does not differ from those values derived from distributions using random data at the 0.05 level of confidence (Hillis 1991) (minimum value for g, in- dicating six taxa differ significantly from random data is —0.51). The bootstrap confidence values for the nodes of the most parsimonious tree duplicated in the bootstrap analysis are given in Figure 18. Using the analysis of Bremer (1994), 9 trees of length less than or equal to 10 steps were recovered before the analysis was terminated because the consensus clado- gram was a complete polytomy, implying limited total support (Bremer 1994). All the taxa within this genus are known from the lower and middle-upper Olenellus zone, implying fairly good concordance between stratigraphic and phylogenetic patterns. Systematic Paleontology Famity “Laudoniidae” Harrington 1956 SUBFAMILY “Laudoniinae” Harrington 1956 Included Taxa GENUS MumMaAsSPIS FRITZ 1992 ‘Type species: Wanneria occidens Walcott 1913. Assigned taxa: Mummaspis oblisooculatus Fritz 1992; Olenellus truncatooculatus Fritz 1992; Olenellus muralensis Fritz 1992; Holmia? macer Walcott 1913. Diagnosis: Anterior cephalic border between frontal lobe and genal spine angle with length (exsag.) equal to length of LO; anterior border prominently separated from extraocular area by furrow; anterolateral margins of frontal lobe separated from extraocular area by furrow (not anterior border furrow); anterior margins of frontal lobe at each side of midline de- flected posteriorly at roughly 10 degree angle relative to transverse line; prominent 100 PEABODY MUSEUM BULLETIN 45 Mummaspis truncatooculatus Mummaspis oblisooculatus Mummaspis muralensis Mummaspis macer Mummaspis occidens [= o = S = +2) = — o c 2 S) x nu w Figure 18. The most parsimonious cladogram of length 9 steps produced from analysis of character data in Table 12 with PAUP vy. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The retention index is 0.67, and the consistency index is 0.78. The following nodes in the text were supported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node 2 = 0.72; Node 3 = 0.52; Node 4 = 0.47. Character states are placed at nodes, using MacClade v. 3.04 (Maddison and Maddison 1992), with the characters given in Table 11. The apomorphic state is given in parentheses. Node 2, 3(1); Node 3, 2(1); Node 4, 5(1). parafrontal band not visible in dorsal view; L4 expands dorsally; lateral margins of L4 dis- tal to lateral margins of LO; pre-ocular furrow on frontal lobe directed inward and forward from glabellar margin; transverse profile of ocular lobes convex dorsally; posterolateral tip of ocular lobe directly behind anterolateral tip or rotated slightly laterally; vertical shelf of ocular lobe separated from extraocular area by furrow (except in M. truncatooculatus); S3 jaggedly convex or carat shape, conjoined medially; line between ends of S2 transverse; S2 convex anteriorly; S1 conjoined medially; distal sector of SO straight; lateral lobes on LO pre- sent; posterior margin of LO convex; extraocular area gently flattened or convex; intergenal angle developed posterior of point halfway between ocular lobes and genal spine; extraoc- ular region opposite L1 with width (tr.) equal to roughly 75% of width of glabella at L1; in- tergenal ridge visible as trace; intergenal area with small pointed spine (could not be deter- mined in M. macer); intergenal angle roughly parallels a transverse line; genal spine angle developed opposite medial part of distal margin of LO; T3 macropleural; boundary between thoracic pleural furrow and anterior band sharp; thoracic pleural furrows extend onto SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 101 spines; length (exsag.) of thoracic pleural furrows (excluding those on T3) at medial part of segment equal to half length (exsag.) of posterior band at similar point; prominent antero- lateral lobes variably present or absent on thoracic axial rings; thoracic pleural spines on segments T5 to T8 extend back approximately four segments; spine on T15 short, length (sag.) of five thoracic segments (could not be determined in M. macer or M. muralensis); and base of spine on T15 nearly as wide (tr.) as axis of segment (could not be determined in M. macer); pleurae present on opisthothorax (could not be determined in M. macer or M. muralensis). Discussion: When Fritz (1992) first created this genus he assigned only two species to it, M. occidens and M. obisooculatus. As three more species are assigned to it here, and additional defining characters for the genus are recognized, an expanded diagnosis is provided. Species of the genus are known from the lower Olenellus zone of western Laurentia and the ?mid- dle-upper Olenellus zone of eastern Laurentia (based on correlations in Fritz 1992 and Palmer and Repina 1993). MUMMASPIS OCCIDENS (WALCOTT 1913) Figure 17.2 Wanneria occidens; Walcott 1913:314, pl. 53, fig. 2. Esmeraldina occidens (Walcott); Resser and Howell 1938:229. Mummaspis occidens (Walcott); Fritz 1992:17, pl. 9, figs. 2-5, pl. 10, figs. 1-5, text fig. 6a; Palmer and Repina 1993:23, fig. 3.5; Lieberman 1998:72, fig. 4.1. Types: Holotype, USNM 60080, see Walcott (1913). Material examined: MCZ 110671, 110675, 110681; USNM 60080, the holotype. Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately W of Mumm Peak, and from the middle Mural Formation, 410.5 ft above the base of that for- mation, in the lower Olenellus zone, Early Cambrian, Mumm Peak section, just N of the Mount Robson Provincial Park boundary, western Alberta (see discussion in Fritz 1992). MUMMASPIS TRUNCATOOCULATUS (FRITZ 1992) Figure 17.3 Olenellus truncatooculatus; Fritz 1992:16, pl. 14, figs. 3-6, pl. 15, figs. 1-9, pl. 16, figs. 1-3, text fig. 5a. Types: Fritz (1992) failed to designate a type for this species. Accordingly, the well-preserved complete specimen Fritz (1992, pl. 15, fig. 6), USNM 443777, is designated the lectotype. It is from the Mural Formation, in the talus slope immediately west of Mumm Peak, in the lower Olenellus zone, Early Cambrian, just north of the Mount Robson Provincial Park boundary, western Alberta, Canada (see discussion in Fritz 1992). The specimens USNM 443769-443776 and 443778443782 become paralectotypes. Discussion: There is some variability in this species in the relative position of the ocular lobes. The posterior tip of the ocular lobes extends back to the medial part of the distal margin of L1, but one specimen Fritz figures (1992, pl. 15, fig. 4) has the posterior tip of the ocular lobes developed somewhat anterior of this, though it is still opposite the distal margin of L1. Originally, Fritz (1992) assigned this species to the genus Olenellus; however, there is much character evidence suggesting that it should be excluded from that genus and 102 Table 11. PEABODY MUSEUM BULLETIN 45 Characters and character states used in phylogenetic analysis of Mummaspis; (0) represents the primitive state, and (1) and (2) represent derived states. —_ . Posteriormost edge of ocular lobes 2. Posterior part of ocular lobes extend back to 3. Extraocular area 4.S1 5. Width (tr.) of thoracic pleural spines T5 to T8 at spine midlength Primitive state (0) deflected slightly laterally relative to anterior part of ocular lobes (0) medial part of margin of L1 (0) roughly flattened (0) convex (0) more than 2/3 length (exsag.) of medial part of inner pleural region Derived states (1) directly behind anterior edge of ocular lobes (1) SO (2) medial part of margin of LO (1) gently convex (1) transverse (1) less than half length (exsag.) of medial part of inner pleural region instead be assigned to the genus Mummaspis. M. truncatooculatus differs from the type species of Olenellus, and what was formerly referred to as the type species of O. (Paedeu- mias), in the condition of the following characters: 1. in M. truncatooculatus the anterior cephalic border between L4 and the genal spine angle has length (exsag.) equal to length (sag.) LO, in Olenellus it has length equal to one-half the length of LO; 2. in M. truncatooc- ulatus the anterolateral margins of the frontal lobe are prominently separated from the an- terior extraocular area by a furrow (not the anterior border furrow), in Olenellus they are not prominently separated from the anterior extraocular area; 3. in M. truncatooculatus the anterolateral margins of the frontal lobe at each side of the midline are deflected posteri- orly at a roughly 10 degree angle relative to a transverse line, in Olenellus they are directed posteriorly at a roughly 40 degree angle; 4. in M. truncatooculatus L4 expands prominently dorsally, in Olenellus it does not; 5. in M. truncatooculatus the lateral margins of L4 are dis- tal to the lateral margins of LO, in Olenellus they are proximal or directly anterior to the lateral margins of LO; 6. in M. truncatooculatus the pre-ocular furrow on L4 is directed in- ward and forward from the glabellar margin, in Olenellus, when visible it is transverse; 7. in M. truncatooculatus a transverse profile of the ocular lobes is convex dorsally, in Olenel- lus it is flattened; 8. in M. truncatooculatus the surface of the interocular area is arched dor- sally, in Olenellus it is developed as a flattened shelf; 9. in M. truncatooculatus S3 is con- joined medially, in Olenellus it is not; 10. in M. truncatooculatus a line between the ends of S2 is transverse, in Olenellus such a line is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line; 11. in M. truncatooculatus L2 and L3 do not merge, in Olenellus they do; 12. in M. truncatooculatus S2 is convex anteriorly, in Olenellus it is straight; 13. in M. truncatooculatus the posterior margin of LO is strongly convex pos- teriorly, in Olenellus it is roughly transverse; 14. in M. triuncatooculatus lateral lobes on LO SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 103 are present, in Olenellus they are absent; 15. in M. truncatooculatus the intergenal angle is developed posterior of a point halfway between the ocular lobes and the genal spine angle, in Olenellus it is practically directly behind the genal spine angle; 16. in M. truncatoocula- tus the medial part of the cephalic posterior border between the intergenal angle and LO flexes posteriorly, in Olenellus it is transverse; 17. in M. truncatooculatus the length (exsag.) of the thoracic pleural furrows (excluding T3) at the medial part of the segment relative to the length (exsag.) of the posterior band of the pleural segment is equal to half the length of the posterior pleural band, in Olenellus it is equal to 1.5 times the length of the poste- rior pleural band; 18. in M. truncatooculatus the spine on T15 is short, with length (sag.) of roughly five thoracic segments, in Olenellus the spine is roughly the length (sag.) of the entire prothorax; 19. in M. truncatooculatus the pleurae of the opisthothorax are present and continue the orientation of the prothoracic pleurae, in Olenellus they are absent. On the basis of these character differences M. truncatooculatus is placed in the genus Mum- maspis rather than in the genus Olenellus. Some of these characters may be plesiomorphic for the genus Mummaspis, but serve to differentiate a larger clade from the genus Olenel- lus. Therefore they were included in this list. Material examined: MCZ 110677, 110678 and many unnumbered specimens on a large slab. Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately W of Mumm Peak, and from the middle Mural Formation, 453 ft above the base of that forma- tion, type section of the Mural Formation, both in the lower Olenellus zone, Early Cam- brian, just N of the Mount Robson Provincial Park boundary, western Alberta (see discus- sion in Fritz 1992). MUMMASPIS OBLISOOCULATUS FRITZ 1992 Mummaspis oblisooculatus; Fritz 1992:19, text fig. 5b, pl. 17, figs. 3-5, non 1, 2, non pl. 16, figs. 4—7. Types: Fritz (1992) did not designate a type specimen for this species. Here USNM 443790 (Fritz 1992, pl. 17, fig. 4), a well-preserved, complete specimen, is designated the lectotype, and USNM 4437839, Fritz’s (1992, pl. 17, fig. 3), a well-preserved cephalon, becomes a para- lectotype. The other material Fritz (1992) illustrated as M. oblisooculatus no longer is refer- able to that species, and is discussed below. Discussion: Fritz (1992) suggested that, based on their relative stratigraphic position, M. ob- lisooculatus evolved from M. occidens. As these species are not sister taxa (Figure 8) this an- cestor-descendant relationship is unlikely. Several of the specimens that Fritz illustrated as M. oblisooculatus (1992, pl. 16, figs. 4-7, and pl. 17, figs. 1, 2), all small cephala, instead ap- pear to belong to a species closely related to F. truemani. These specimens share several char- acters in common with E truemani, to the exclusion of M. oblisooculatus. In particular, they have the anterolateral margins of the frontal lobe at each side of the midline deflected poste- riorly at a roughly 40 degree angle relative to a transverse line instead of at a roughly 10 de- gree angle; the posterior margins of the ocular lobes rotated laterally relative to the anterior margins; the distal sector of SO convex anteriorly rather than straight; and the posterior mar- gin of LO transverse rather than convex. (In one of the specimens of M. oblisooculatus [Fritz 1992, pl. 17, fig. 3] the posterior margin of LO is partly deformed as it is truncated by the an- terior margin of the first thoracic segment, so its condition is difficult to discern.) These specimens do differ from typical F. truemani and resemble M. oblisooculatus in having the frontal lobe contacting the anterior border furrow, and the anterior ocular line and genal 104 PEABODY MUSEUM BULLETIN 45 Table 12. Character state distributions for Mummaspis and outgroup used in phylogenetic parsimony analysis. Characters and alternative states are as listed in Table 11. Missing data are indicated by ?. Character states listed as X are polymorphic, where X=(1&2). id HAS Fritzolenellus truemant 00000 Mummaspis occidens 1 Aaa Mummaspis truncatooculatus 00010 Mummaspis oblisooculatus 10100 Mummaspis macer OX 1.0 k Mummaspis muralensis OX 100 ridge prominently developed; however, these characters are not crucial for the diagnosis of either Fritzolenellus or Mummaspis. These cephala are also discussed above under the head- ing of the genus Fritzolenellus, and are tentatively treated as F. sp. aff. truemant. Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately W of Mumm Peak, and from the middle Mural Formation, 490 ft and 806 ft above the base of that formation, in the lower Olenellus zone, Early Cambrian, Mumm Peak section, just N of the Mount Robson Provincial Park boundary, western Alberta (see discussion in Fritz 1992). MUMMASPIS MURALENSIS (FRITZ 1992) Figures 17.4, 17.5 Olenellus muralensis; Fritz 1992:13, pl. 8, figs. 1-7, pl. 9, fig. 1, text fig. 5c (see for additional synonymy). Types: Fritz (1992) failed to designate a type specimen for this species. Therefore, USNM 443742, the nearly complete specimen illustrated in Fritz (1992, pl. 8, fig. 5) is designated the lectotype. This specimen is from the Mural Formation, in the talus slope immediately west of Mumm Peak, in the lower Olenellus zone, Early Cambrian, just north of the Mount Robson Provincial Park boundary, western Alberta, Canada (see discussion in Fritz 1992). The specimens USNM 443738-443741, 443743 and 443744 become paralectotypes. Discussion: Originally this species was assigned to the genus Olenellus. However, analysis conducted here indicates that it should be excluded from that genus and instead be assigned to the genus Mummaspis. The character differences demonstrating this are identical to those listed above for M. truncatooculatus. However, the condition of two of these characters, the length of the thoracic axial spine on T15 and the presence or absence of the opisthothoracic pleurae, could not be determined in M. muralensis because the posterior region of the tho- rax is not adequately preserved. Material examined: MCZ 5575 (a lot of 5 specimens), 110672—110674, 110676 Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately W of Mumm Peak, in the lower Olenellus zone, Early Cambrian, just N of the Mount Robson Provincial Park boundary, western Alberta (see discussion in Fritz 1992). SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 105 MUMMASPIS MACER (WALCOTT 1913) Figure 17.6 Holmia? macer; Walcott 1913:313, pl. 54, fig. 1. Esmeraldina macer (Walcott); Resser and Howell 1938:229, pl. 8, figs. 10-12. Wanneria macer (Walcott); Fritz 1973:13; Fritz 1995:712. Wanneria walcottanus (Wanner); Walcott 1910:302, pl. 30, figs. 3, 4. Types: Holotype, USNM 60092, by monotypy, see Walcott (1913). Discussion: Walcott (1913) assigned this species to the genus Holmia. These taxa are not that distantly related within the Olenelloidea, based on Lieberman (1998), but they are not sister taxa. Differences between the type of Holmia, H. kjerulfi and M. macer include the fol- lowing characters: 1. in M. macer the anterior cephalic border is prominently separated from the extraocular area by a furrow, in H. kjerulfi it is not; 2. in M. macer the anterolat- eral margins of the frontal lobe are prominently separated from the extraocular area by a furrow (not the anterior border furrow), in H. kjerulfi they are not; 3. in M. macer the an- terolateral margins of the frontal lobe at each side of the midline are deflected posteriorly at a roughly 10 degree angle relative to a transverse line, in H. kjerulfi they are directed pos- teriorly 40 degrees; 4. in M. macer a line from the posterior tip of the ocular lobe to the junc- tion of the posterior margin of the lobe with the glabella is parallel to a sagittal line, in H. kjerulfi it forms a 10 to 20 degree angle; 5. in M. macer S1 is conjoined medially, in H. kjerulft it is not; 6. in M. macer the distal sector of SO is straight, in H. kjerulfi it is convex anteriorly; 7.in M. macer the extraocular area is gently convex, in H. kjerulfi it is prominently vaulted; 8.in M. macer the extraocular region opposite L1 has width (tr.) equal to 75% of the width of the glabella at L1, in H. kjerulfi it has width equal to 40% to 50%; 9. in M. macer the in- tergenal angle roughly parallels a transverse line, in H. kjerulfi it is directed anteriorly at a roughly 45 degree angle relative to a transverse line; 10. in M. macer T3 is macropleural, in H. kjerulfi it is not; 11.in M. macer the thoracic pleural spines are relatively longer than they are in H. kjerulfi; 12. in M. macer the prothoracic axial rings lack prominent anterolateral lobes, in H. kjerulfi they are present; 13. in M. macer the thoracic pleural furrows extend onto the spines, in H. kjerulfi they only extend the width of the inner pleural region; 14. in M. macer the boundary between the thoracic pleural furrow and the anterior band is sharp, in H. kjerulfi it is gradational; 15. in typical Mummaspis, the length (exsag.) of the thoracic pleural furrows is equal to half the length of the posterior pleural band, this character is dif- ficult to determine in M. macer but appears to have this state, in H. kjerulfi the thoracic pleural furrows are equal in length to the posterior band. On the basis of these character dif- ferences, M. macer must be treated as not closely related to H. kjerulft. Convergences between these taxa that may explain the misassignment of M. macer to the genus Holmia are that both have the posterior tips of the ocular lobes developed oppo- site the medial part of the distal margin of LO, and both have the genal spines extending back roughly 4 to 5 thoracic segments. Mummaspis macer also cannot be assigned to the genus Esmeraldina, which is closely related to Holmia, on the basis of character evidence similar to that discussed above. In the past this genus has also been assigned to Wanneria. However, there are several character differences separating M. macer from W. walcottana, and it seems inconceivable that they could be assigned to the same genus except for the fact that they occur in rocks of the same age and from the same locality. In particular: 1. in M. macer the anterior cephalic border is prominently separated from the extraocular area by a furrow, in W. walcottana it 106 PEABODY MUSEUM BULLETIN 45 is not prominently separated; 2. in M. macer the anterolateral margins of the frontal lobe are prominently separated from the extraocular area by a furrow (not the anterior border furrow), in W. walcottana they are not; 3. in M. macer L4 expands prominently dorsally, in W. walcottana it does not; 4. in M. macer a line from the posterior tip of the ocular lobe to the junction of the posterior margin of the lobe with the glabella parallels a sagittal line, in W. walcottana it forms a 10 to 20 degree angle with a sagittal line; 5. in M. macer S1 and S3 are conjoined medially, in W. walcottana they are not; 6. in M. macer the distal sector of SO is straight, in W. walcottana it is convex anteriorly; 7. in M. macer the proximal sector of SO is well posterior of the distal end, in W. walcottana the proximal and distal ends are on a transverse line; 8. in M. macer the posterior margin of LO is convex posteriorly, in W. wal- cottana it is roughly transverse; 9. in M. macer lateral lobes are present on LO, in W. walcot- tana they are not; 10. in M. macer the extraocular area is gently convex, in W. walcottana it is flattened; 11. in M. macer T3 is macropleural, in W. walcottana it is not; 12. in M. macer the anterior margin of the thoracic pleural furrow on T3, when proceeding from the prox- imal to the distal edge, is directed weakly posteriorly before flexing strongly posteriorly, in W. walcottana it parallels a transverse line before flexing strongly posteriorly; 13. in M. macer the thoracic pleural furrows extend onto the spines, in W. walcottana they extend only half of the width of the inner pleural region; 14. in M. macer the boundary between the tho- racic pleural furrow and the anterior band is sharp, in W. walcottana it is gradational; 15. in M. macer the length (exsag.) of the thoracic pleural furrows (excluding those of T3) at the medial part of the segment are short, equal to roughly half the length (exsag.) of the poste- rior pleural band, in W. walcottana they are equal to 1.5 times the length of the posterior band; 16. in M. macer the width (tr.) of the thoracic pleural spines T5 to T8 at the spine midlength is less than half the length (exsag.) of the medial part of the inner pleural region, whereas they are roughly thicker in W. walcottana, more than two-thirds the length of the medial part of the inner pleural region. The bulk of the character evidence suggests that this species belongs with Mummaspis. It matches the diagnostic characters of that genus and shares many more characters with the type of this genus than it does with any other genus. Material examined: USNM 60092, the holotype. Occurrence: Pennsylvania: the Kinzers Formation, in the middle upper Olenellus zone, ac- cording to Palmer and Repina (1993), from 2 mi N of York, from Fruitville, 3 mi N of Lan- caster, and 0.5 mi S of East Petersburg. ok VEN Phylogenetic Analysis of the Bristoliinae The subfamily Bristoliinae is a moderately diverse clade nested within the Olenelloidea that contains the genera Bristolia, Fremontella and Lochmanolenellus. All of the component taxa bear prominent genal spines and advanced genal spine angles. Bristolia consists of seven de- scribed species and one fragmentary cranidium of an undescribed species, and is restricted to the Lower Cambrian Olenellus zone of western Laurentia. All but one species occurs in the Great Basin of eastern California and western Nevada. One species is known from the Mackenzie Mountains, Northwest Territories, Canada. One new species of Bristolia is de- scribed here. Lochmanolenellus consists of a single species known from southwestern Lau- rentia, and Fremontella consists of a single species known from eastern Laurentia. A total of 11 taxa was subjected to phylogenetic analysis, including nine ingroup taxa. All species within the genus were analyzed, including what was termed Laudonia? sp. by Fritz (1972; later reassigned to Bristolia by Fritz 1992), here Bristolia sp., an extremely poorly preserved taxon, known from an external mold of about one-half of a cephalon. It is cer- tainly a species of Bristolia, bearing the hallmark characteristics of the genus. On the basis of phylogenetic analysis, it is most closely related to B. anteros Palmer and Halley and is dis- cussed in greater detail under that species. Because of its poor state of preservation, it was not described as a new species. The two known species of the genus Laudonia, L. bispinata and L. amputata, were used as outgroups in phylogenetic analysis. Laudonia is the sister taxon of a clade that includes Fremontella, Lochmanolenellus and Bristolia, based on higher level phylogenetic patterns within the Olenelloidea (Lieberman 1998). For the purposes of rooting the phylogeny of Fremontella, Lochmanolenellus and Bristolia, the two outgroup species of Laudonia were treated as monophyletic. Evolutionary relationships within Bristolia were determined by parsimony analysis of 17 exo-skeletal characters (only cephalic characters could be used be- cause of the dearth of thoracopygidia known for members of Fremontella, Lochmanolenel- lus and Bristolia), given in Table 13. The codings for the taxa analyzed are given in Table 14. These data were subjected to an exhaustive search using PAUP v. 3.1.1 (Swofford 1993). Four most parsimonious trees of length 35 steps were recovered. A strict consensus of these trees is shown in Figure 19. The retention index of the tree is 0.80, and the consistency index is 0.67 when uninformative characters are excluded. Uncertainty in resolution in this con- sensus cladogram concerns the relationships of some of the species of Bristolia, which es- sentially only differ in the relative position of the genal spine angle and also in the angle the intergenal angle forms with a transverse line. The g, statistic, used to measure tree length skewness distributions, was —0.70, and this value differs from those values derived from dis- tributions using random data at the 0.01 level of confidence (Hillis 1991). This suggests that there is a strong and robust phylogenetic signal to the character data given in Table 14. The confidence values for the nodes of the consensus tree duplicated in the bootstrap analysis are given in Figure 19. Using the analysis presented in Bremer (1994), 30 trees of length less than or equal to 36 108 PEABODY MUSEUM BULLETIN 45 Lochmanolenelius mexicana Laudonia bispinata Laudonia amputata Bristolia harringtoni Bristolia mohavensis Bristolia bristolensis Fremontella halli Bristolia anteros Bristolia insolens Bristolia fragilis Bristolia sp. Figure 19. A strict consensus of the four most parsimonious cladograms of length 35 steps produced from analysis of char- acter data in Table 14 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The retention index is 0.80, and when uninformative characters are excluded the consistency index is 0.67. The following nodes in the text were supported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node 2 = 0.94; Node 3 = 0.76; Node 4 = 0.64; Node 5 = 0.66; Node 6 = 0.96. The following branch support values (Bremer 1994) were recovered for the following nodes: Node 2 = 1; Node 3 = 1; Node 6 = 1. Character states are placed at nodes, using MacClade v. 3.04 (Maddison and Maddison 1992), with the characters given in Table 13. The apomorphic state is given in parentheses. Square parentheses indicate equivocal character states that are ambiguous because of missing data, polymorphisms or multiple equally par- simonious resolutions. Equivocal characters are placed only at their basal phylogenetic position, and only un- ambiguous reversals are shown. Node 1, 1(4), 3(1), 5(1), 9[0, 2], 14[0, 1]; Node 2, 10(1), 11(1); Node 3, 7(1), 8(1), 12(1), 15[0, 1]; Node 4, 1[1, 2, 3, 4], 9[0, 1, 2], 13[0, 1]; Node 5, 1(3), 4[0, 1], 6(1), 9(2), 14(0), 15(0), 17(1); Node 6, 4(0), 5(0), 7(0), 16(1). steps, and 117 trees of length less than or equal to 37 steps were recovered before the analysis was terminated because the consensus cladogram was a complete polytomy. The amount of branch support for the various nodes is given in Figure 19. The total support index (Bremer 1994) for the tree is 0.09, towards the low end of the examples given in Bremer (1994). If the stratigraphic correlations of Fritz (1972) are correct for the Canadian sections, then Bristolia sp. (Laudonia sp. of Fritz 1972), nested within the Bristolia clade, predates the other members of the genus in the fossil record (lower versus upper Olenellus zone). Its phy- logenetic position indicates that the lineages leading to B. insolens and the other species of SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 109 Bristolia (and also to B. anteros, if Bristolia sp. is not ancestral to that species) must have dif- ferentiated by the lower Olenellus zone. This would imply at least a moderate gap in the his- tory of the lineages leading to the different species of Bristolia, equivalent in duration to the difference in ages of the early Olenellus zone and Bristolia zonule of the upper Olenellus zone. However, it is also conceivable that either the Canadian sections may have been in- correctly correlated with older strata, or the Great Basin sections may have been incorrectly correlated with younger strata. As mentioned above, phylogenetic analysis conducted here retrieved relatively little phylogenetic structure within the genus Bristolia. The evolution of the genus is depicted (Figure 19) as containing a basal polytomy due to lack of phylogenetically informative char- acter differences between most species of Bristolia. The bulk of differentiating criteria be- tween species center around the relative position of the genal spine angle and the angle the intergenal angle forms with a transverse line. These allow for easy discrimination between individual species of Bristolia, but basically do not provide phylogenetic resolution within the genus, unless they were to be ordered using some additional criterion. As no reasonable criterion could be found to order these characters, they were treated as unordered in phy- logenetic analysis. The only taxa that group together within the genus are B. insolens, B. an- teros and B. sp. These all possess distinctive genal spines, which loop anterolaterally before deflecting posteriorly, and short (exsag.) anterior cephalic borders. Additional morphome- tric data were gathered to settle the relative relationships of taxa within the genus Bristolia. Unfortunately, suitable sample sizes of measurable specimens exist for only four species: B. harringtoni n. sp. (43 specimens measured), B. bristolensis (Resser) (34 specimens mea- sured), B. insolens (6 specimens measured) and B. mohavensis (Hazzard and Crickmay) (30 specimens measured). All available specimens in the collections of the UCR and the LACMIP were measured from each of these species. The following 10 measurements were taken on cephala of each of these species: cephalic length (sag.); cephalic width (tr.) (ex- cluding the genal spines); length (sag.) from posterior margin of LO to genal spine angle; width (tr.) of LO; width (tr.) of L1; maximum width of L4 (tr.); maximum width (tr.) be- tween the distalmost edges of the ocular lobes; length (sag.) of LO and L1; length (sag.) of L4; and width (tr.) of the extraocular area measured from distal most edge of ocular lobe to the genal spine angle. (All data are available from the author on request.) A series of uni- variate and multivariate statistical analyses were conducted on these data, corresponding to those conducted in Lieberman et al. (1994, 1995), to determine whether or not differences in individual morphological variables between species were statistically significant, and also to determine how these species would be grouped based on multivariate morphometric data. None of the procedures for eliminating size-based differences seemed to be efficacious in the case of these morphometric data, so the data were not corrected for differences in size. Therefore, morphometric differences between species represent some combination of both size and shape. A principal components analysis using Minitab v. 10Xtra (1995) was conducted on the data to determine which variables or combinations of variables contributed the great- est amount of variance to the data set. The covariance matrix was used to emphasize po- tential differences in variance. The first principal component explained 92.9% of the vari- ance in the data and the second principal component explained 5.5%. All subsequent principal components contributed less than 0.7% of the variance and shall not be dis- cussed further. The loadings of the variables for principle components | and 2 are given in Table 15. The first has higher negative loadings on variables B and J, indicating that these con- 110 Table 13. Description of characters and character states used in phylogenetic analysis of Bristoliinae; (0) does PEABODY MUSEUM BULLETIN 45 not always represent the primitive state as two outgroup taxa were employed in phylogenetic analysis, but for simplicity one of the outgroups, Laudonzia bispinata, was coded with all “0” character states. Py i) nn ON “NI In adult genal spine angle opposite . Angle formed between transverse line and intergenal area Posterior edge of ocular lobe opposite Lateral and medial tips of S2 . Line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms following angle relative to sagittal line . Medial margins of genal spine . Faint ventral depression across entire region where ocular lobe hits frontal lobe . Prominent intergenal spine in adult . Length (exsag.) of anterior cephalic border at point midway between anterior tip of L4 and genal spine angle (0) medial tip of $4 (0) 85° to 95° (0) medial part of margin of L2 (0) roughly as far forward (0) approximately 40° (0) deflect posteriorly (0) absent (0) present (0) long, length (sag.) of L1 and L2 (1) distal tip of SO (2) distal tip of S2 (3) medial part of margin of L4 (4) distal tip of S3 (5) medial part of margin of Ll ()i3a4to 55° CQ)itTS" to: 120% (1) medial part of margin of L1 (1) medial edge declined posteriorly (1) approximately 10° (1) loop anteriorly before deflecting posteriorly (1) present (1) absent (1) very short, length less than or equal to half the length of L1 (sag.) (2) short, length equal to the length of L1 (sag.) Continued SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) Tat Table 13 continued. 1S: 16. 17. nS3 . Anterior cephalic border developed as . Intergenal ridge . Approximate length (exsag.) of genal spine . Lateral margins of glabella between posterior margin of LO and L1 Width (tr.) of genal field Lateral and medial tips of S3 S2 Primitive state (0) deepest medially (0) flattened ledge (0) developed as prominently expanded lineament (0) 4 to 5 thoracic segments (sag. ) (0) constricting anteriorly (0) short, equal to one-third width between distal most tips of ocular lobes (0) at same position (0) convex Derived states (1) same depth laterally as medially (1) elevated ridge (1) visible as faint trace (1) 8 thoracic segments (sag. ) (1) roughly parallel (1) moderate, equal to one-half width between distal most tips of ocular lobes (2) broad, equal to two-thirds width between distal most tips of ocular lobes (3) very broad, equal to three-quarters width between distal most tips of ocular lobes (1) medial tip deflected further anteriorly (1) transverse 112 PEABODY MUSEUM BULLETIN 45 Figure 20. 1. Laudonia bispinata Harrington, locality is the same as Figure 17.2, MCZ 110679, x1.5. 2—4. Locality is the same as Figure 3.4. 2. Bristolia insolens (Resser), UCR 10/2003, x1.7. 3. Bristolia harringtoni, new species, plaster cast of UCR 10/7, the holotype, <1.4. 4. Bristolia mohavensis (Hazzard and Crickmay), UCR 10/1185, x1.5. 5. Bristolia mohavensis (Hazzard and Crickmay), locality is the same as Figure 3.2, UCR 7002/6, x1.3. 6. Nephrolenellus jasperensis, new species, 10 m above the top of the Gog Group, about 2 mi SW of Mount Simla, Jasper Park, West- ern Alberta, Canada, GSC 16858, plaster cast of the holotype, x1.5. 7. Bolbolenellus groenlandicus (Poulsen), Cape Kent Fm., Cape Kent, NE end of Inglefield Land, NW Greenland, latex cast of MGUH 2235, x1.7. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 113 tribute a disproportionate amount of the variance explained by this component. The sec- ond has a high negative loading on variable J, and somewhat elevated, positive loadings on variables B and C. From this we can conclude that the bulk of the variance in the data set is associated with differences in variables B, C and J, essentially involving the overall width of the extraocular area and the cephalon, and the relative position of the genal spine angle. A discriminant analysis was also conducted using Minitab (1995) to calculate Maha- lanobis differences between the centroids of each of the four species of Bristolia. To perform a discriminant analysis, the number of specimens for each group should be equal and the data should be multivariate normal with a homogeneous variance-covariance structure. All available specimens of the four species were measured. For three of these species roughly equivalent amounts of specimens were available; however, for one, B. insolens, far fewer specimens were available. Assessing multivariate normality can be difficult. If samples are not univariate normal, then they will not be multivariate normal. Using the Anderson-Dar- ling test for normality on Minitab (1995), the following variables for the following groups were found to not be normal at the 0.05 level of confidence: Bristolia insolens for variable E; B. mohavensis for variable A; and B. harringtoni for variable C. Thus most, but not all, vari- ables for all of the species were normal, indicating that the assumption of multivariate nor- mality may not strictly hold. The results of a discriminant analysis are presented as a classification matrix (Table 16) and as a matrix displaying the squared Mahalanobis distance between groups (Table 17). The classification matrix gives the proportion of specimens that can be assigned cor- rectly to the group they are presumed to belong to, in this case the different species of Bristolia. Specimens of B. insolens could always be classified correctly, indicating that it is a well-constrained group morphologically. Very high percentages of the other species were also classified correctly, indicating that they too are very well-constrained groups morphologically. Specimens of B. mohavensis were occasionally confused with B. harring- toni, as were specimens of B. bristolensis. A single specimen of B. bristolensis was incor- rectly classified with B. insolens, specimens of B. harringtoni were occasionally confused with B. mohavensis, and one specimen of B. harringtoni was incorrectly classified with B. bristolensis. These results, along with the Mahalanobis distance values in Table 17, suggest several patterns. First, B. insolens is clearly distinct from the other species of Bristolia for which morphometric data exist. The phylogenetic analysis conducted here and the cladogram given in Figure 19 confirm this, as B. insolens is more closely related to B. anteros and B. sp., whereas the other species considered in morphometric analysis are part of a basal polytomy within the genus Bristolia. Second, these results suggest that in terms of their overall mor- phology, B. mohavensis and B. harringtoni are more closely similar to one another than ei- ther is to B. bristolensis. Finally, specimens of B. bristolensis are more similar to specimens of B. harringtoni than they are to specimens of B. mohavensis. These results are intuitively appealing and generally match the overall patterns of morphological similarity. Since phylogenetic analysis failed to provide resolution for the evolutionary relation- ships of B. mohavensis, B. bristolensis and B. harringtoni due to a paucity of informative characters, these morphometric data will be used as a proxy for relationship. Therefore, within the clade of unresolved Bristolia, on the basis of morphometric data B. mohavensis and B. harringtoni are sister species, and B. bristolensis is sister to the clade they form. This hypothesis of course leaves the relationships of the unsampled taxon Bristolia fragilis Palmer in Palmer and Halley still unresolved. 114 PEABODY MUSEUM BULLETIN 45 Systematic Paleontology Famiry “Laudoniidae” Harrington 1956 SUBFAMILY “Laudoniinae” Harrington 1956 Included Taxa GENUS LAUDONIA HARRINGTON 1956 ‘Type species: Laudonia bispinata; Harrington 1956. Assigned taxa: L. amputata; Fritz 1992. Discussion: This genus is treated as monophyletic and restricted to the lower Olenellus zone of western Alberta. Its two component taxa are united by their shared possession of the following characters: anterior cephalic border between L4 and genal spine angle with length (exsag.) of L1 and L2 (sag.), developed as flattened ledge, not prominently sepa- rated from extraocular area by furrow; plectrum absent; anterolateral margins of glabella prominently separated from extraocular area by furrow; frontal lobe intersects anterior border furrow, expands dorsally, anterolateral margins relative to transverse line deflected at about a 10 degree angle; intergenal angle developed posterior of genal spine angle; in- tergenal angle deflected at 85 to 95 degrees relative to transverse line; lateral margins of L4 distal to lateral margins of LO; distal tip of pre-ocular furrow directed inward and backward from glabellar margin; $3 convex, line between ends transverse, conjoined me- dially; S2 convex, contacts axial furrows; line between ends of S2 transverse; S1 conjoined medially; ocular lobe with prominent furrow, not dorsally flattened; interocular area arched dorsally; axial tubercle on LO; extraocular area flattened; prominent anastomosing ridges absent from extraocular area; genal spine length (exsag.) of four to five thoracic segments (sag.), deflected at roughly 45 degree angle relative to sagittal line; intergenal ridge developed as prominently expanded lineament; prominent intergenal spine present; T3 macropleural, spine projecting posteriorly six to eight thoracic segments; anterior margin of T3 transverse before flexing posteriorly; nodes present on medial part of tho- racic axial rings; anterior and posterior margins of thoracic pleural furrow on T3 directed weakly posterior laterally. LAUDONIA BISPINATA HARRINGTON 1956 Figure 20.1 Laudonia bispinata; Fritz 1992:26, pl. 12, figs. 4-8, pl. 13, figs. 1-8, pl. 14, figs. 1, 2, text fig. 6c (see for more complete synonymy); Palmer and Repina 1993:24; Lieberman 1998:73, fig. 4.2. Types: Holotype, KUMIP 32400, designated in Harrington (1956). Material examined: KUMIP 32400, the holotype; MCZ 110679, 110680. Occurrence: Canada: Alberta, the Mural Formation, in the talus slope immediately west of Mumm Peak, from the middle Mural Formation, 453 ft and 555 ft above the base of that formation, in the lower Olenellus zone, Early Cambrian, immediately W of Mumm Peak, just N of the Mount Robson Provincial Park boundary, and between Cin- namon Peak-Whitehorn Mountain, at the western end of Mount Robson Provincial Park (Fritz 1992). SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 115 Table 14. Character state distributions for Bristoliinae and outgroups used in phylogenetic parsimony analysis. Characters and alternative states are as listed in Table 13. Missing data are indicated by ?. Character states listed as W, X, Y and Z are polymorphic, where W=(0&1), X=(2&4), Y=(1&5) and Z=(2&5). 1 Leste 19293-4516 72809) 0 2S e456 a pond No — Laudonia bispinata OHOF0 020500) 0. 0°0.59 (0),0),0):0,0,040 Laudonia amputata AOE TO Me OvO0K000 ©. L020 L0R0ROF0N0 Lochmanolenellus mexicana 4010100021 OO MF ORO 0 Fremontella halli Ole Or Ort Or Oe 0 Bristolia bristolensis De Om le cleslsQ)sleste alba OOF a Bristolia insolens Se lem ee te Dll TESORO a Bristolia anteros S20 eOcOu Ot 2a Hest ORO seat Bristolia sp. SEO PIF OLO Mi Ones Dail es I Te dik ORO paling Bristolia mohavensis Vein sbOr aires LF AI SOTIOF0 Bristolia harringtont ep em ieNNaleO™ iets et SOON Bristolia fragilis ete Ot et iy Omi On LAUDONIA AMPUTATA FRITZ 1992 Laudonia amputata; Fritz 1992:26, pl. 11, figs. 1-5, pl. 12, figs. 1-3, text fig. 6d; Palmer and Repina 1993, fig. 4.4; Palmer and Repina 1997:412, fig. 259.2. Types: Fritz (1992) figured and described this species, but never designated types. All his fig- ured material (USNM 443752—443756) thus become syntypes. Here the most well-pre- served of his syntype specimens, USNM 443754, is designated the lectotype. The specimens USNM 443752, 443753, 443755 and 443756 become paralectotypes. Occurrence: Canada: Alberta, the Mural Formation, in the lower Olenellus zone, Early Cam- brian, immediately west of Mumm Peak, just north of the Mount Robson Provincial Park boundary (Fritz 1992). FaMILy Biceratopsidae Pack and Gayle 1971 SUBFAMILY Bristoliinae Harrington 1956 Discussion: Lieberman (1998) presented results of a phylogenetic analysis of the Olenel- loidea that recognized a clade of olenelloids including the genera Fremontella, Lochmanolenellus and Bristolia, which he assigned to the family Biceratopsidae Pack and Gayle. A paraphyletic family within the Biceratopsidae, the “Bristoliinae,” was recognized that included the genera Bristolia, Fremontella, Lochmanolenellus, Nephrolenellus, Bol- bolenellus and Olenelloides. Lieberman (1998) treated the “Bristoliinae” as paraphyletic. Now that detailed phylogenetic patterns are better understood within this part of the olenelloid tree and detailed species level analyses have been conducted including the genera Laudonia, Fremontella, Lochmanolenellus and Bristolia, the Bristoliinae is redefined as a 116 PEABODY MUSEUM BULLETIN 45 monophyletic subfamily that includes only the genera Fremontella, Lochmanolenellus and Bristolia. This subfamily can be defined by the possession of the following diagnostic char- acters (thoracic characters only found in a few species of Bristolia, the only members of the subfamily for which thoracic remains exist): 1. anterior cephalic border developed as an el- evated ridge, and prominently separated from extraocular area by furrow; 2. anterolateral margins of frontal lobe not prominently separated from anterior extraocular area by furrow (not anterior border furrow); 3. prominent parafrontal band not visible in dorsal view; 4. anterolateral margins of frontal lobe at each side of midline deflected posteriorly at roughly 10 degree angle relative to transverse line; 5. S3 same depth laterally as medially; 6. an- terodistal margins of L3 formed by ocular lobes; 7. surface of interocular area slopes evenly from tip of ocular lobe to glabella; 8. genal spine angle developed opposite SO or position anterior of S0; 9. posterior edge of ocular lobe developed anterior of or opposite medial part of distal margin of L1; 10. lateral lobes absent from LO; 11. distal sector of SO transverse; 12. intergenal angle relative to transverse line directed anteriorly at least 35 degrees; 13. T3 macropleural, pleural spine of T3 projects further posteriorly than length of entire protho- rax; 14. anterior margin of thoracic pleural furrow on T3, when proceeding from proximal to distal edge, parallels a transverse line before flexing strongly posteriorly; 15. posterior margin of thoracic pleural furrow on T3 with medial part directed strongly posteriorly, dis- tal part parallel to a transverse line or weakly flexing anterolaterally; 16. thoracic pleural spines on segments 5 to 8 developed as short projections extending two thoracic segments back; 17. sharp boundary between thoracic pleural furrow and anterior band; 18. promi- nent spine present on 15th thoracic axial ring. GENUS LOCHMANOLENELLUS LIEBERMAN 1998 Type species: Wanneria mexicana prima Lochman in Cooper at al. 1952. Assigned taxa: Monotypic. Diagnosis: See Lieberman (1998). LOCHMANOLENELLUS MEXICANA (LOCHMAN IN COOPER ET AL. 1952) Wanneria mexicana prima; Lochman in Cooper et al. 1952:96, pl. 18, figs. 1-3. Laudonia mexicana (Lochman in Cooper et al.); Fritz 1992:12. Laudonia sp.; Nelson 1976:31, pl. 5 (upper right hand corner). Lochmanolenellus mexicana (Lochman in Cooper et al.); Lieberman 1998:74, fig. 4.4. Types: Holotype, USNM 115681, paratypes USNM 115682, 115683, designated by Lochman in Cooper et al. (1952). Discussion: Lochman (Cooper et al. 1952) assigned this species to the genus Wanneria. However, based on phylogenetic topology within the Olenelloidea, this species belongs to Lochmanolenellus, a genus distantly related to the type of the genus Wanneria, W. walcot- tana. Character differences separating Lochmanolenellus from Wanneria include: 1. in Lochmanolenellus the anterior cephalic border is developed as a raised ridge, in Wanneria it is a flattened ledge; 2. in Lochmanolenellus the anterior cephalic border is prominently separated from the extraocular area by a furrow, in Wanneria it is not; 3. in Lochmanolenel- lus a plectrum is developed, in Wanneria it is absent; 4. in Lochmanolenellus the length (sag.) of L4 is approximately equal to the length of LO and L1 (sag.), in Wanneria it is ap- proximately equal to 1.5 times the length of LO and L1 (sag.); 5. in Lochmanolenellus, L4 SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 117 expands prominently dorsally, in Wanneria it does not; 6. in Lochmanolenellus the surfaces of the interocular area slope evenly from the tip of the ocular lobe to the glabella, in Wan- neria it is developed as a flattened shelf; 7. in Lochmanolenellus the posterior margins of the ocular lobes are developed opposite the distal tips of S1, in Wanneria they are developed opposite the medial part of the distal margin of L1; 8. in Lochmanolenellus the width (tr.) of the interocular area is approximately equal to about half the width of the ocular lobe at its midlength, in Wanneria it is equal to at least the width of the ocular lobe; 9. in Lochmanolenellus SO, $1, S2 and $3 are conjoined medially, in Wanneria they are not; 10. in Lochmanolenellus L1, L2 and L3 are all strongly transversely convex, in Wanneria they are only gently convex; 11. in Lochmanolenellus the distal sector of SO is straight, in Wan- neria it is convex anteriorly; 12. in Lochmanolenellus the proximal end of SO is well poste- rior of the distal end, in Wanneria the proximal and distal ends of SO lie on a transverse line; 13. in Lochmanolenellus the posterior margin of LO is convex posteriorly, in Wanneria it is roughly transverse; 14. in Lochmanolenellus the extraocular area is prominently vaulted, in Wanneria it is flattened; 15. in Lochmanolenellus the intergenal ridge is promi- nently expanded, in Wanneria it is only visible as a trace; 16. in Lochmanolenellus the ex- traocular region is about 40% to 50% of the width (tr.) of the glabella at L1, in Wanneria it is greater than 75% of the width of the glabella at L1; 17. in Lochmanolenellus the genal spine near where it hits the cephalic border is directed posterolaterally at a roughly 65 to 80 degree angle relative to a sagittal line, in Wanneria it is directed posterolaterally at roughly a 10 to 20 degree angle; 18. in Lochmanolenellus the genal spine angle is developed opposite the medial part of the distal margin of L3, in Wanneria it is developed opposite the medial part of the distal margin of LO; 19. in Lochmanolenellus the intergenal angle is developed directly behind the genal spine angle, in Wanneria it is developed posterior of a point halfway between the ocular lobes and the genal spine angle; 20. in Lochmanolenellus the intergenal spine is prominently developed, in Wanneria it is represented only by a slight dorsally directed swelling; 21. in Lochmanolenellus the intergenal angle is directed anteri- orly at a roughly 80 to 90 degree angle relative to a transverse line, in Wanneria it roughly parallels a transverse line. On the basis of this large number of character differences it is clear that Lochmanolenellus is distantly related to Wanneria, and the assignment of this species to the genus by Lochman (Cooper et al. 1952) appears to be untenable. More recently, Nelson (1976) and Fritz (1992) suggested that this species belonged in the genus Laudonia. Al- though Lochmanolenellus is much more closely related to Laudonia than it is to Wanneria, there are a fair number of character differences, and a few key characters, that separate the species of these genera. Conceivably Lochmanolenellus could be lumped with the genus Laudonia; however, because of phylogenetic topology within the Olenelloidea (Lieberman 1998), this would necessitate either lumping the genera Nephrolenellus, Bolbolenellus, Olenelloides, Peachella, Biceratops, Fremontella and Bristolia within Laudonia, or making Laudonia paraphyletic. Because Bristolia, Laudonia and the other aforementioned genera as they are currently construed represent good monophyletic genera, some of them contain- ing several species, it seems preferable to treat Lochmanolenellus as a monotypic genus rather than to significantly alter the taxonomic status of Laudonia, Bristolia and the rest of the Biceratopsidae. Differences separating Lochmanolenellus and Laudonia include: 1. in Lochmanolenel- lus the anterior cephalic border between L4 and the genal spine angle has length (exsag.) equal to length (sag.) of LO, in Laudonia the length equals approximately 1.5 times the length of LO; 2. in Lochmanolenellus the anterior cephalic border is developed as a raised 118 PEABODY MUSEUM BULLETIN 45 ledge that is prominently separated from the extraocular area by a furrow, in Laudonia the anterior cephalic border is flattened and is not prominently separated from the extraocular area by a furrow; 3. in Lochmanolenellus a plectrum is present, in Laudonia it is absent; 4. in Lochmanolenellus the anterolateral margins of the glabella are not prominently separated from the extraocular area by a furrow (not the anterior border furrow), in Laudonia it is; 5. in Lochmanolenellus L4 is of length (sag.) equal to the length (sag.) of LO and L1, in Laudo- nia it is approximately equal to 1.5 times the length of LO and L1; 6. in Lochmanolenellus the pre-ocular furrow on L4 is directed inward and forward from the glabellar margin, in Laudonia it is directed inward and backward; 7. in Lochmanolenellus S3 is of equal depth medially and laterally, in Laudonia it is deeper medially; 8. in Lochmanolenellus the ocular lobe smoothly merges into the extraocular area, in Laudonia it is separated from the ex- traocular area by a furrow; 9. in Lochmanolenellus the surface of the interocular area slopes evenly from the tip of the ocular lobe to the glabella, in Laudonia the interocular area is arched; 10. in Lochmanolenellus the width (tr.) of the interocular area is approximately equal to the width of the ocular lobe, in Laudonia it is equal to about half the width of the ocular lobe; 11. in Lochmanolenellus L1, L2 and L3 are strongly transversely convex, in Laudonia they are gently convex; 12. in Lochmanolenellus the lateral margins of the glabella are roughly parallel between the posterior margin of LO and L1, in Laudonia the glabella con- stricts anteriorly in this region; 13. in Lochmanolenellus SO is conjoined medially, in Laudo- nia it is not; 14. in Lochmanolenellus the extraocular area is prominently vaulted, in Laudo- nia it is flattened; 15. in Lochmanolenellus the width of the extraocular area is equal to the width (tr.) of 40% to 50% of the glabella at L1, in Laudonia it is approximately equal to 75% of the width of the glabella at L1; 16. in Lochmanolenellus the genal spine near where it hits the cephalic border is directed posterolaterally at a roughly 65 to 80 degree angle relative to a sagittal line, in Laudonia it is directed posterolaterally at a roughly 45 degree angle; 17. in Lochmanolenellus the medial part of the posterior border between LO and the intergenal angle flexes posterolaterally, in Laudonia it is transverse. Thus, on the basis of these charac- ter differences, Lochmanolenellus is treated as distinct from Laudonia, though the two are held to be closely related. Phylogenetic analysis of the Olenelloidea in Lieberman (1998) suggested that Lochmanolenellus was more closely related to the genus Bristolia than Fremontella. However, when all the species of Laudonia and Bristolia are considered it actually appears that Fre- montella is sister to Bristolia to the exclusion of Lochmanolenellus. This is only a slight change in phylogenetic topology, but the basic premise of Lieberman (1998) that Laudonia is sister to a clade including Fremontella, Lochmanolenellus and Bristolia is upheld by this analysis. Material examined: USNM 115683 and 115681, the holotype. Occurrence: Mexico: 590 ft above the base of the Puerto Blanco Formation section at the W side of the Proveedora Hills on the N side of Puerto Blanco, 6 to 7 mi W of Caborca. Cali- fornia: the upper part of the Poleta Formation, White-Inyo Mountains/Death Valley region, both treated as in the lower Olenellus zone, following Nelson (1976) and Fritz (1992). GENUS FREMONTELLA HARRINGTON 1956 Type species: Wanneria halli Walcott 1910. Assigned taxa: Monotypic. Diagnosis: Length (exsag.) anterior cephalic border near but not directly anterior to frontal lobe very long, equal to length (sag.) of L1 and L2; length (sag.) L4 equal to 1.5 times length SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 119 (sag.) of LO and L1; distal margins of L2 subparallel; SO, S1 not conjoined medially; LO smooth, convex posteriorly; genal spine near cephalic border directed posterolaterally at 35 to 45 degree angle relative to sagittal line, length (exsag.) equal to four times length (sag.) LO; genal spine angle opposite medial part of distal margin of L3; cephalic posterior border transverse between intergenal angle and LO. FREMONTELLA HALLI (WALCOTT 1910) Wanneria halli; Walcott 1910:301, pl. 31, figs. 1-11. Olenellus halli (Walcott); Resser 1938:52, pl. 5, figs. 7, 8, 18. Olenellus alabamensis; Resser 1938:53, pl. 5, figs. 16, 17. Fremontella halli (Walcott); Harrington 1956:58, pl. 15, figs. 1-3, 8-9, text fig. 1c; Harring- ton et al. 1959:0192, fig. 133.1; Palmer and Repina 1993:23, fig. 3.4; Palmer and Repina 1997:408, fig. 256.2; Lieberman 1998:74. Types: Lectotype, USNM 56806b, designated by Resser (1938), not USNM 56806c, which was erroneously designated as the lectotype in Harrington (1956) and subsequently re- ported as such in Palmer and Repina (1993, 1997). Paralectotypes USNM 56808c-k were designated in Resser (1938) (he called them paratypes). The specimen USNM 56806a was part of Walcott’s (1910) original type series for his Wanneria halli. Resser (1938) assigned this to his new species, Olenellus alabamensis, which is identical to F. halli and a junior sub- jective synonym. Thus USNM 56806a is also a paralectotype of F. halli. Discussion: This species is the type and only known species of the genus. Harrington (1956) suggested that this genus belonged to the Olenellinae Walcott 1890 and was closely related to the genus Olenellus. He also added that it was easily distinguishable from the genus Wan- neria, to which it was originally assigned by Walcott (1910). However, clear differences exist between Fremontella and the genus Olenellus (including what was formerly referred to as the subgenus O. [Paedeumias]), and phylogenetic patterns within the Olenelloidea (Lieber- man 1998) suggested that these taxa are distantly related. These taxa differ in the condition of several characters, including: 1. in Fremontella the anterior cephalic border between L4 and the genal spine angle has length (exsag.) roughly equal to 1.5 times length (sag.) of LO, in Olenellus it is roughly equal to half the length of LO; 2. in Fremontella the anterolateral margins of L4 at each side of the midline are deflected posteriorly at a roughly 10 to 15 degree angle relative to a transverse line, in Olenellus they are directed posteriorly at a roughly 40 degree angle; 3. in Fremontella the pre-ocular furrow on L4 is directed inward and forward from the glabellar margin, when visible in Olenellus, it is transverse; 4. in Fremontella the surface of the interocular area slopes evenly from the tip of the ocular lobe to the glabella, in Olenellus it is developed as a flattened shelf; 5. in Fremontella the posterior tips of the ocular lobes are developed op- posite the medial part of the distal margin of L1, in Olenellus they extend further posteri- orly; 6. in Fremontella the interocular area is relatively much narrower than it is in Olenel- lus; 7. in Fremontella SO, $2, and S3 are conjoined medially, in Olenellus they are not; 8. in Fremontella a line between the ends of S2 is transverse, in Olenellus it is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line; 9. in Fremontella $2 is convex anteriorly, in Olenellus it is straight; 10. in Fremontella L2 and L3 do not merge, in Olenellus they do; 11. in Fremontella the lateral margins of L2 are subparallel, in Olenellus they diverge anteriorly; 12. in Fremontella the posterior margin of LO is convex posteriorly whereas in Olenellus it is more transverse; 13. in Fremontella the genal spine 120 PEABODY MUSEUM BULLETIN 45 near where it hits the cephalic border is directed posterolaterally at a roughly 35 to 45 de- gree angle relative to a sagittal line, in Olenellus it is directed posterolaterally at a roughly 10 to 20 degree angle; 14. in Fremontella the genal spine angle is developed opposite the medial part of the distal margin of L3, in Olenellus it is developed opposite the medial part of the distal margin of LO; 15. in Fremontella the intergenal spine is not developed, whereas in Olenellus it is; 16. in Fremontella the medial part of the posterior border between LO and the intergenal angle is transverse, in Olenellus it flexes posteriorly. Based on these and other character differences it is clear that Fremontella is distantly related to Olenellus, contra Har- rington’s (1956) and Palmer and Repina’s (1993) contention and Resser’s (1938) generic assignment of halli to the genus Olenellus. Whatever characters these taxa share are either primitive retentions or convergences based on phylogenetic topology within the Olenel- loidea (Lieberman 1998). Occurrence: Alabama: the Rome Formation, NE of Helena (Resser 1938), treated as in the upper part of the Olenellus zone, following Barnaby and Read (1990). GENUS BRISTOLIA HARRINGTON 1956 Type species: Mesonacis bristolensis Resser 1928. Assigned taxa: Mesonacis bristolensis Resser 1928; Mesonacis insolens Resser 1928; Bristolia harringtoni n. sp.; Bristolia anteros Palmer in Palmer and Halley 1979; Paedeumias mo- havensis Hazzard and Crickmay 1933; Bristolia fragilis Palmer in Palmer and Halley 1979; Bristolia sp. Diagnosis: Anterior cephalic border between L4 and genal spine angle relatively short, length (exsag.) less than or equal to length (sag.) of L1; anterior cephalic border developed as elevated, flattened ridge, prominently separated from extraocular area; frontal lobe con- tacts anterior border furrow; anterolateral margins of frontal lobe deflected posteriorly at roughly 10 degree angle relative to transverse line; frontal lobe moderately long (sag.), length equal to length of LO and L1 medially; glabellar furrows prominently incised; S3 same depth laterally as medially; lateral margins of L2 when proceeding anteriorly bulging laterally relative to LO; lateral margins of glabella between posterior margin of LO and L1 constrict; distal tips of SO straight; extraocular region flattened; intergenal ridge visible as faint trace; intergenal spine absent in adult; genal spine length (exsag.) of approximately first eight thoracic segments; thorax divided into pro- and opisthothorax; nodes present on medial part of thoracic axial rings; T3 macropleural, spine length (exsag.) greater than length of thoracopygidium; anterior margin of T3 deflects anteriorly before flexing poste- riorly; anterior margin of pleural furrow on T3 parallels transverse line before flexing pos- teriorly; anterior margin of thoracic pleural furrow separated from body of segment by prominent ridge; thoracic pleural spines, behind T4, developed as short sweeping projec- tions extending two to three thoracic segments back, lying in roughly same dorso-ventral plane as pleural segments; prominent spine on axial ring of T15 (see Harrington 1956 and Lieberman 1998 for additional characters). Discussion: This genus comprises a moderately diverse radiation. Described species are re- stricted to roughly coeval deposits of the Carrara Formation, in the Great Basin of the United States. All of the Great Basin species occur in the Bristolia zonule, which lies in the upper part of the Olenellus zone (Nelson 1976; Palmer and Halley 1979). A single fragmen- tary cranidium, referable to Bristolia, Fritz’s (1972) Laudonia? sp., is known from the Sekwi Formation, lower Olenellus zone, the Mackenzie Mountains, Northwest Territories, Canada. This species is too poorly preserved to be described as a new species. However, it is placed SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 11 within the genus Bristolia and subjected to phylogenetic analysis. The first occurrence of this species, if the stratigraphic correlations for the Canadian sections are correct, extends the range of Bristolia throughout the Olenellus zone. Poulsen (1927) figured two new taxa from the Lower Cambrian Cape Kent Formation, upper Olenellus zone (Poulsen 1964), Cape Kent, NW Greenland, that he referred to as Olenellus groenlandicus and O. kentensis. These were later reassigned to Bristolia by Poulsen (1964). However, as discussed in detail below under the genus Bolbolenellus, these taxa must be removed from the genus Bristolia and reassigned to the genus Bolbolenellus. BRISTOLIA BRISTOLENSIS (RESSER 1928) Mesonacis bristolensis; Resser 1928:7, pl. 2, figs. 5-8. Bristolia bristolensis (Resser); Harrington 1956:59, text fig. 1d; Riccio 1952:30, pl. 7, figs. 1, 2, 5, non pl. 8; Harrington et al. 1959:0192, fig. 133.3; Mount 1976:175, fig. 12; Palmer in Palmer and Halley 1979:64, pl. 1, figs. 14, 16, 17 non 18, 19; Mount 1980:27, fig. 12; Palmer and Repina 1993:24, non figs. 4.5, 13; Palmer and Repina 1997:409, non fig. 258.1; Lieberman 1998:73, fig. 4.3. Olenellus gilberti (Meek); Walcott 1910, pl. 37, figs. 16, 18, 19. ?Fremontella sp.; Nelson 1976:31, pl. 8 (center). Types: USNM 78390, the lectotype (defined in Palmer and Repina 1993); paralectotypes USNM 78391 and 78392. Discussion: Bristolia bristolensis had originally been recognized (e.g., Riccio 1952; Palmer and Halley 1979; Palmer and Repina 1993) as a highly variable species, particularly in the condition of the intergenal and genal spine angles. However, Lieberman (1998) used uni- variate measurements and statistical analyses to define a well-constrained B. bristolensis based on Resser’s (1928) lectotype, and demonstrated that what had traditionally been treated as a broadly variable B. bristolensis was actually two different species. The species re- ferred to as Bristolia n. sp. in Lieberman (1998) is described here as B. harringtoni n. sp. Bris- tolia bristolensis has the intergenal angle deflected at an 80 to 95 degree angle relative to a transverse line and the genal spine angle is variably developed opposite and/or between the distal tips of $2 and S3. In B. harringtoni the intergenal angle is deflected at a 50 to 65 de- gree angle relative to a transverse line and the genal spine angle is developed variably op- posite the distal tip of S2 or the medial part of L1. Originally, Resser (1928) assigned this species to the genus Mesonacis. However, species of Bristolia differ from species of Mesonacis in the condition of several characters, including: 1. in Bristolia the anterolateral margins of the frontal lobe at each side of the midline are deflected posteriorly at roughly a 10 degree angle relative to a transverse line, whereas in Mesonacis the margins are deflected posteriorly at a roughly 40 degree angle; 2. in Bristolia the length (sag.) of L4 is roughly equal to the length (sag.) of LO and LI, in Mesonacis it is roughly equal to 1.5 times the length of LO and L1, except in M. hamocu- lus (Cowie and McNamara 1978), a derived member of the genus Mesonacis, where the length of L4 is similar to that found in Bristolia; 3. in Bristolia the lateral margins of L2, when proceeding anteriorly, bulge laterally relative to LO, in Mesonacis they do not bulge laterally relative to LO; 4. in Bristolia the surface of the interocular area slopes evenly from the tip of the ocular lobe to the glabella, whereas in Mesonacis it is arched or developed as a flattened shelf; 5. in Bristolia the width (tr.) of the interocular area is about half the width of the ocular lobe at its midlength, whereas in Mesonacis it is as wide as the width 122 PEABODY MUSEUM BULLETIN 45 Table 15. The first two principal components generated from analysis of morphometric data from species of the genus Bristolia using Minitab v. 10Xtra. The covariance matrix was used. Character variables are abbreviated, and explained in the text. Variable PGi. P@2 A — 0.335 0.066 B — 0.608 0.463 \C, — 0.158 0.483 D — 0.167 0.088 E —0.135 0.045 F — (0.176 0.068 G — 0.251 — 0.052 H — 0.096 0.037 I — 0.148 0.056 J — 0.567 — 0.725 of the ocular lobe; 6. in Bristolia the posterior margin of LO is convex posteriorly, whereas in Mesonacis it is roughly transverse; 7. in Bristolia the genal spine near where it hits the cephalic border is directed posterolaterally at a roughly 35 to 45 degree angle relative to a sagittal line, or loops prominently anteriorly before deflecting posteriorly, whereas in Mesonacis it is directed posterolaterally at a roughly 10 to 20 degree angle relative to such a line, except in M. cylindricus (Palmer in Palmer and Halley 1979), a highly derived mem- ber of the genus Mesonacis; 8. in Bristolia an intergenal spine is not developed in the adult, whereas in Mesonacis it 1s; 9. in Bristolia the posterior margin of the thoracic pleural fur- row on the third segment medially deflects strongly posteriorly before distally flexing an- teriorly to parallel a transverse line, whereas in Mesonacis the posterior margin of this fur- row is directed evenly posterolaterally (this thoracic character not preserved in all species of Mesonacis and/or Bristolia). On the basis of these and other characters, B. bristolensis, and all other species of Bristolia, must be excluded from the genus Mesonacis. These two genera are distantly related based on phylogenetic topology within the Olenelloidea (Lieberman 1998). Material examined: LACMIP 4908/8, 4917/5 and several unnumbered specimens in the gen- eral stratigraphic collections; SDSNH 16785, 16829, 16832, 16834, 16835, 16840, 16843 (2 specimens), 16899, 17014, 17017; UCR 10 (2 specimens), 10/8 (3 specimens), 10/10, 10/81 (3 specimens), 10/87, 10/127 (3 specimens), 10/128, 10/509, 10/531, 10/1100, 7270, 7967, 7968 (2 specimens), 7969; USNM 78390, the lectotype. Occurrence: California: the Carrara Formation, in the lower Bristolia zonule, upper part of the Olenellus zone, Early Cambrian, following Nelson (1976) and Palmer and Halley (1979), in the Grapevine, Funeral and Resting Spring ranges, the White-Inyo/Death Valley region, and in the Latham Shale, Marble Mountains, 190 m W of the limestone quarry, 0.5 mi E of Cadiz, in the Mojave Desert portion of San Bernardino County, possibly equivalent to Haz- zard’s (1933) locality M-5, treated as in the Bristolia zonule. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 123 BRISTOLIA INSOLENS (RESSER 1928) Figure 20.2 Mesonacis insolens; Resser 1928:8, pl. 2, figs. 1-4. Olenellus insolens (Resser); Riccio 1952:30, pl. 5, figs. 1-13, pl. 6, figs. 1-3. Bristolia insolens (Resser); Mount 1976:175, fig. 14; Mount 1980:27, fig. 14. Types: Lectotype, designated here, USNM 78387 (Resser 1928, pl. 2, fig. 2), from “Bristol Mountain, near Cadiz, California, on the Santa Fe Railroad, about 100 mi east of Barstow” (Resser 1928:1). Possibly equivalent to Hazzard and Crickmay’s (1933) locality M-5. Para- lectotypes USNM 78386, 78388, 78389, from the same horizon as the lectotype. Discussion: Mount (1976, 1980) correctly recognized that this species should be assigned to the genus Bristolia rather than to the genera Olenellus or Mesonacis. All species of Bristolia, including B. insolens, differ from species of Mesonacis in the condition of characters dis- cussed above under B. bristolensis. Species of Bristolia differ from species of Olenellus (in- cluding what was formerly referred to as the subgenus O. [Paedeumias]) in the condition of the following characters: 1. in Bristolia, the length (sag.) of L4 is approximately equal to the length of LO and LI, in Olenellus it is typically equal to approximately 1.5 times the length of LO and L1; 2. in Bristolia the lateral margins of L4 are distal to the lateral margins of LO, in Olenellus they are either proximal or directly anterior of the lateral margins of LO; 3. in Bristolia the pre-ocular furrow on L4 is directed inward and forward from the glabellar mar- gin, in Olenellus, when visible, the furrow is transverse; 4. in Bristolia, the ocular lobes are convex dorsally in transverse profile, in Olenellus they are flattened; 5. in Bristolia the sur- face of the interocular area slopes evenly from the tip of the ocular lobe to the glabella, in Olenellus it is developed as a flattened shelf; 6. in Bristolia the posterior tips of the ocular lobes are developed opposite the medial part of the distal margin of L1, in Olenellus they are developed opposite the medial part of the distal margin of LO; 7. in Bristolia, the width (tr.) of the interocular area is equal to about half the width of the ocular lobe at its midlength, in Olenellus the width of the interocular area is greater than or equal to the width of the oc- ular lobe; 8. in Bristolia, S2 and S3 are conjoined medially, in Olenellus they are not con- joined; 9. in Bristolia the genal spine near where it hits the cephalic border is either directed posterolaterally at a roughly 35 to 45 degree angle relative to a sagittal line, or loops strongly anteriorly before deflecting posteriorly, in Olenellus the genal spine is directed posterolater- ally at a roughly 10 to 20 degree angle relative to such a line; 10. in Bristolia an intergenal spine is absent, whereas in Olenellus it is present; 11. in Bristolia, the intergenal angle is di- rected anteriorly at a 30 to 120 degree angle relative to a transverse line, in Olenellus it is de- flected at a roughly 0 to 10 degree angle relative to a transverse line; 12. in Bristolia the genal spine angle is always further forward than in Olenellus; 13. in Bristolia the posterior margin of the thoracic pleural furrow on T3 medially deflects strongly posteriorly before distally flexing anteriorly to parallel a transverse line, whereas in Olenellus the posterior margin of this furrow is directed evenly posterolaterally. On the basis of these and other characters, B. insolens, and all other species of Bristolia, must be excluded from the genus Olenellus. These two genera are distantly related based on phylogenetic topology within the Olenelloidea (Lieberman 1998). Material examined: LACMIP 200-E and 22 unnumbered specimens in the general strati- graphic collections; UCR 10 5/22, 10/2003, 10/2011, 10/2014, 10/2016, 10/2024, 2836/2, 70025727 Wi 1b) 727 1(Se)5.7 3113/6. Occurrence: If the type locality is equivalent to Hazzard and Crickmay’s (1933) locality M- 124 PEABODY MUSEUM BULLETIN 45 Table 16. A summary of classification produced from a linear discriminant analysis using Minitab (1995). Rows represent the species a specimen was assigned to, and columns represent the true species it belongs to based on qualitative character analysis. The proportion correctly assigned is shown underneath each species. | NM eS) — Bristolia insolens (1) 6 0 0 Bristolia mohavensts (2) Oy 24 0 3 Bristolia bristolensis (3) 0 D7, ] Bristolia harringtoni (4) 0 6 Gua oo Proportion correct MOV O:SS LOST IO:9 5, then the species would occur in California: the Latham Shale, Marble Mountains, 190 m W of the limestone quarry, 0.5 mi E of Cadiz, in the Mojave Desert portion of San Bernardino County, treated as in the Bristolia zonule, upper Olenellus zone, Early Cam- brian. BRISTOLIA ANTEROS PALMER IN PALMER AND HALLEY 1979 Bristolia anteros; Palmer in Palmer and Halley 1979:63, pl. 1, figs. 1-13; Mount 1980:27, fig. 15: Bristolia new species A; Mount 1976:175, fig. 15. ?Callavia ¢ nevadensis; Walcott 1910:285, pl. 38, fig. 13. Types: Holotype, USNM 177181, see Palmer and Halley (1979). Discussion: This species is closely related to a fragmentary cranidium from the Sekwi For- mation, lower Olenellus zone, of the Mackenzie Mountains, Northwest Territories, Canada, figured as Laudonia? sp. by Fritz (1972:27, pl. 9, fig. 21), later reassigned to Bristolia by Fritz (1992), and referred to here as Bristolia sp. This cranidium, GSC 27303, examined by the au- thor for this study, bears all of the diagnostic characteristics of the genus Bristolia listed above, and also shares many character states with B. anteros, including: 1. genal spine angle opposite medial part of distal margin of L4; intergenal angle forms roughly 95 degree angle with transverse line; 2. posterior edge of ocular lobe opposite medial part of distal margin of L1; 3. medial tip of S3 further anterior than lateral tip; 4. S2 transverse, contacts axial fur- row; 5. line from anterior to posterior edge of ocular lobe forms roughly 40 degree angle relative to sagittal line; 6. medial margins of genal spines loop anteriorly before deflecting posteriorly; 7. prominent furrow does not extend across entire margin of point where eye hits frontal lobe. These characters, and the possession of characters diagnostic of Bristolia, group this partial cephalon with the genus Bristolia rather than with the genus Laudonia. Because Bristolia sp. is so poorly preserved it is not described as a new species here; how- ever, it was coded for the characters given in Table 13 and was subjected to phylogenetic analysis. Complete character state codings for Bristolia sp. are given in Table 14. According SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 125 to the phylogeny presented in Figure 10, B. anteros, from the Great Basin, is sister to Bristo- lia sp. from the Mackenzie Mountains, rather than to other Great Basin taxa. Walcott’s (1910, pl. 38, fig. 13) Callavia ? nevadensis is poorly preserved, but appears to represent a specimen closely related to, if not conspecific with, B. anteros. In particular, it has the advanced genal spines that loop prominently forward. Because of the poor state of preservation of this specimen, it is only questionably assigned to B. anteros at this time. Material examined: LACMIP 4908-6 (6 specimens); UCR 10-3/1, 7271/8, 7271/14, 7271/16 (2 specimens). Occurrence: Nevada: the upper Bristolia zonule, uppermost Olenellus zone, Early Cambrian, in the Grapevine Mountains and Last Chance Range, the Funeral Mountains, Desert Range, and Nevada Test Site (Palmer and Halley 1979); California: the upper Bristolia zonule, White-Inyo Mountains/Death Valley region. BRISTOLIA MOHAVENSIS (HAZZARD AND CRICKMAY 1933) Figures 20.4, 20.5 Paedeumias mohavensis; Hazzard and Crickmay 1933:74, pl. 1, figs. c-e, non f; Riccio 119522335 pO figse7,.8: Paedeumias sp.; Riccio 1952, pl. 9, fig. 9. Olenellus mohavensis (Hazzard and Crickmay); Mount 1976:175, fig.9; Mount 1980:27, fig. 9. Bristolia new species B; Mount 1976:175, fig. 16. Bristolia new species; Mount 1980:27, fig. 16. Types: Holotype, see Hazzard and Crickmay (1933). Discussion: Hazzard and Crickmay’s (1933, pl. 1, fig. f), one of their paratypes for B. mo- havensis, appears to represent a specimen of Mesonacis fremonti, but their holotype and other paratypes are still valid. However, the holotype could not be located and may be miss- ing. At this time a neotype is not erected as it is not essential for the purposes of taxonomic stability. Originally this species was assigned to the genus Paedeumuas (here treated as equiv- alent to Olenellus); however, on the basis of its possession of certain characters diagnostic for the genus Bristolia and the absence of characters diagnostic of Olenellus, it is assigned to the genus Bristolia. A large number of character differences between Bristolia and Olenellus are listed above under the species B. insolens. Material examined: LACMIP 4917-15 (3 specimens) and 4 unnumbered specimens in the general stratigraphic collections; UCR 10 (10 specimens), 10/110, 10/316, 10/320, 10/709, 10/1185, 10/2025 (2 specimens), 2847/1, 4079/19, 4079/23, 4081/70, 7002 (2 specimens), 7002/2 (2 specimens), 7002/4, 7002/6, 7313/3, 7313/7. Occurrence: California: the Latham Shale, treated as in the Bristolia zonule, upper Olenellus zone, at the southern end of the Marble Mountains, near Chambless in the Mojave Desert portion of San Bernardino County, at the end of the dirt road on the opposite side of the hill from the limestone quarry (Chambless Limestone), about 1.25 mi N, 28 degrees E of Cadiz railroad station, Hazzard and Crickmay’s (1933) locality M-5. Locality lies just above the crossbedded Zabriskie Quartzite at the end of the dirt road. Fossils in gray-brown shale, equal to the upper Latham Shale, which is 6 to 10 m thick. It is 416 ft W, 216 ft S of NE cor- ner sec 11, T 5 N, R 14 E, San Bernardino baseline and meridian, on the USGS 15 ft Danby Quadrangle. Riccio (1952) also reports the species from the Latham Shale, Marble Moun- tains, 190 m W of the limestone quarry, 0.5 mi E of Cadiz, in the Mojave Desert portion of San Bernardino County, possibly equivalent to Hazzard and Crickmay’s (1933) locality M-5. 126 PEABODY MUSEUM BULLETIN 45 Table 17. The results of a linear discriminant analysis using Minitab (1995). Shown are the squared Mahalanobis distances between species. 1 2 3 4 Bristolia insolens (1) O. T4057 rald2is 29:0 Bristolia mohavensis (2) 40.7 Om Si5:5 2.9 Bristolia bristolensis (3) 15:2" 13:8 0 6.2 Bristolia harringtoni (4) 29.0 2.9 6.2 0 BRISTOLIA HARRINGTONI LIEBERMAN NEW SPECIES Figure 20.3 Olenellus bristolensis (Resser); Riccio 1952:30, pl. 8, figs. 1-11 non pl. 7, figs. 1, 2, 5. Bristolia bristolensis (Resser); Mount 1976:175, fig. 13; Palmer in Palmer and Halley 1979:64, pl. 1, figs. 18, 19, non 14-17; Mount 1980:27, fig. 13; Palmer and Repina 1993:24, figs. 4.5, 13; Palmer and Repina 1993:409, fig. 258.1. Bristolia sp.; Harrington 1956:59, pl. 15, fig. 7. Bristolia n. sp.; Lieberman 1998:73. Types: The holotype is UCR 10/7 (Figure 20.3) from UCR locality 10, the Latham Shale, Bristolia zonule, at the southern end of the Marble Mountains, near Chambless in the Mo- jave Desert portion of San Bernardino County, California, at the end of the dirt road on the opposite side of the hill from the limestone quarry (Chambless Limestone), about 1.25 mi N, 28 degrees E of Cadiz railroad station, equivalent to horizon of Hazzard and Crickmay’s (1933) locality M-5. Locality lies just above the crossbedded Zabriskie Quartzite at the end of the dirt road. Fossils in gray-brown shale, equal to the upper Latham Shale, which is 6 to 10 m thick. It is 416 ft W, 216 ft S of NE corner sec 11, T 5 N, R 14 E, San Bernardino base- line and meridian, on the USGS 15 ft Danby Quadrangle. Diagnosis: Anterior cephalic border between L4 and genal spine angle very short (exsag.), length equal one-half length (sag.) LO; genal spine angle opposite medial part of distal mar- gin of L1 or S2; intergenal angle forms roughly 50 to 60 degree angle with transverse line; slight ventral depression across entire margin of frontal lobe at point where ocular lobe in- tersects frontal lobe; lateral and distal tips of S3 as far forward; S2 convex or transverse, me- dial edge declined posteriorly relative to lateral edge, lateral edge generally not contacting axial furrows; line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms 10 degree angle relative to sagittal line; posterior edge of ocular lobe op- posite medial part of distal margin of L1; width (tr.) of extraocular area roughly equal to two-thirds width between distal most tips of ocular lobes; spine on T15 developed as long needle shaped projection. Description: Anterior cephalic border narrow, raised ridge, length (exsag.) at point halfway between lateral margin of L4 and genal spine angle equal to one-half length (sag.) LO; frontal lobe contacts anterior border furrow; anterolateral margins of frontal lobe deflected posteriorly at roughly 10 degree angle relative to transverse line; lateral margins of frontal lobe distal to lateral margins of LO; where ocular lobes contact frontal lobe slight depression SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 127 visible; lateral margins of glabella expand from posterior margin of L2 to midpoint of mar- gin of L3; lateral margins of glabella contract between midpoint of L3 and $3; $3 convex or transverse, medial edge declined posteriorly, conjoined medially; S2 convex or transverse, medial edge declined posteriorly; L2 and L3 typically merge distally; line from posterior edge of ocular lobe to junction of posterior margin of lobe with glabella roughly forms 10 degree angle relative to sagittal line; SO, S1 transverse, contact axial furrows, medial edges declined posteriorly; lateral margins of glabella constrict between LO and L1 when pro- ceeding anteriorly; posterior edge of ocular lobe opposite medial part of distal margin of L1; S1 conjoined medially; posterior margin of LO weakly convex posteriorly, with faint axial node; genal spines developed as long projections, length (exsag.) of 8 to 10 thoracic segments (sag.), sweeping backward at roughly 45 degree angle relative to sagittal line for first two-thirds of length; posterior third of genal spines roughly parallel sagittal line; genal spine angle opposite medial part of distal margin of L1 or S2; intergenal angle prominently developed, without spine, inclined at 50 to 60 degree angle relative to transverse line; ex- traocular region broad, width (tr.) approximately two-thirds width between distal most tips of ocular lobes; faint anastomosing ridges visible on extraocular region; faint intergenal and genal ridges visible; posterior margin of cephalic posterior border roughly transverse. Thorax divided into pro- and opisthothorax; faint nodes present on medial part of thoracic axial rings; axial rings at T4 30% width of pleural field, excluding spines; T3 macropleural, spines length (exsag.) greater than length of thoracopygidium; medially, an- terior margin of T3 deflects anteriorly before flexing posterolaterally; medially, pleural field of T3 length (exsag.) equal to length of segments four to seven; spines of T3 deflected pos- teriorly at roughly 30 degree angle relative to sagittal line; anterior margin of thoracic pleural furrow of T3 transverse before flexing posterolaterally; posterior margin of pleural furrow of T3 deflects strongly posteriorly medially, distally it is transverse; anterior margins of thoracic pleural furrows prominently separated from anterior band; thoracic pleural fur- rows extend onto spines; thoracic pleural spines behind T4 developed as short sweeping projections extending two to three thoracic segments back, lie in roughly same dorso-ven- tral plane as pleural segments; prominent spine on axial ring of T15 developed as long nee- dle shaped projection. Pygidium not known. Discussion: Specimens of this species have been confused with B. bristolensis; however, sta- tistical analyses given in Lieberman (1998) and above indicate that B. harringtoni must be viewed as distinct from B. bristolensis. Phylogenetic analysis indicates that these species are closely related and map as part of a basal polytomy within the genus Bristolia. However, morphometric data suggest that in terms of overall similarity, B. harringtoni more closely resembles B. mohavensis. Etymology: Named for H. J. Harrington, who did some of the important early research on Cambrian trilobites with advanced genal spines. Material examined: LACMIP 4875-0 and many unnumbered specimens in the general stratigraphic collections; MCZ 7371 (4 specimens); SDSNH 16949, 17004, 17007, 20711 (10 specimens), 20803 (5 specimens); UCR 10 (4 specimens), 10/3, 10/4, 10/5, 10/7, 10/8—28, 10/22 (2 specimens), 10/113 (4 specimens), 10/330, 10/855, 10/1104, 10/1173, 10/2013, 10/2014, 2836/1, 7002 (5 specimens), 7002/11, 7271, 7272, 7272/2. Occurrence: California: in addition to the type locality known from Riccio’s (1952) locality in the Latham Shale of the Marble Mountains (these horizons are possibly equivalent), also from the Carrara Formation, lower Bristolia zonule, Salt Spring Hills and Grapevine Moun- tains, White-Inyo/Death Valley region (Palmer and Halley 1979). 128 PEABODY MUSEUM BULLETIN 45 BRISTOLIA FRAGILIS PALMER IN PALMER AND HALLEY 1979 Olenellus fremonti; Walcott 1910:320, pl. 37, fig. 18. Fremontia fremonti (Walcott); Harrington 1956:57, pl. 15, fig. 6. Bristolia fragilis; Palmer in Palmer and Halley 1979:65, pl. 2, figs. 1-6. Types: Holotype, USNM 177190, see Palmer and Halley (1979). Material examined: UCR 2847/1. Occurrence: Nevada: the Carrara Formation, Grapevine and Funeral Mountains, California, and Nevada Test Site and Desert Range; California: White-Inyo/Death Valley region, upper Bristolia zonule (Palmer and Halley 1979). Phylogenetic Analysis of Bolbolenellus The genus Bolbolenellus is a small clade comprised of five species nested among several of the derived genera of the Olenelloidea. Its members occur throughout Laurentia in strata assigned to the middle-upper Olenellus zone of the Early Cambrian. The results of a phylo- genetic analysis of the five species within the genus are presented here. A total of seven taxa was subjected to phylogenetic analysis. This includes all species of the genus Bolbolenellus and two outgroup taxa: Nephrolenellus multinodus and N. jasperensis n. sp. These taxa are appropriate outgroups to the genus Bolbolenellus based on the higher level phy- logenetic analysis of the Olenelloidea presented in Lieberman (1998). The outgroups were treated as forming part of a basal polytomy. Species were assigned to the genus Bolbolenellus based on diagnostic characters elucidated below, recognized in part from a phylogenetic analy- sis of the Olenelloidea. Phylogenetic patterns were determined by parsimony analysis of seven holaspid exo-skeletal characters (Table 18). The codings for the taxa analyzed are given in Table 19. All characters were treated as unordered (nonadditive). These data were subjected to an ex- haustive search on PAUP v. 3.1.1 (Swofford 1993). One most parsimonious tree (Figure 21) was recovered of length 12 steps, consistency index = 0.67, and retention index = 0.60. The distrib- ution of all tree lengths was left-skewed relative to the distribution of tree lengths produced by random data, but the g, statistic, used to measure tree length skewness distributions, was —0.16. This value does not differ significantly frlom those obtained from distributions constructed using random character data. Thus, these data do not have a strong phylogenetic signal, proba- bly due to the limited number of character differences among these different species. The boot- strap confidence values for the nodes of the consensus tree duplicated in the bootstrap analysis are given in Figure 21. There is limited bootstrap support for all of the nodes in the phylogeny. When the analysis presented in Bremer (1994) was conducted, 29 trees of length less than or equal to 13 steps were recovered before the analysis was terminated because the consensus cladogram was a complete polytomy, and no branch support exists for the cladogram in Fig- ure 21. This is probably because these various taxa are separated by relatively few character dif- ferences and are presumed to be very closely related. This is attributed to the verisimilitude of the taxa within the genus Bolbolenellus and the fact that thoracic material for different species within the genus is not preserved, eliminating a potential line of exploration for character dif- ferences. All taxa within the genus Bolbolenellus are known from the middle lower Olenellus zone, implying good congruence between stratigraphic and phylogenetic information. Systematic Paleontology Famity Biceratopsidae Pack and Gayle 1971 SUBFAMILY Biceratopsinae Pack and Gayle 1971 Lieberman (1998) assigned the genera Bristolia, Fremontella, Lochmanolenellus, Nephrolenel- lus, Bolbolenellus, Olenelloides, Peachella and Biceratops to the family Biceratopsidae. Within 130 PEABODY MUSEUM BULLETIN 45 Bolbolenellus groenlandicus Bolbolenellus sphaerulosus Bolbolenellus altifrontatus g $ 3 2 2) o = i“ = Q 3 n S & +2) +2) 3 3 _— — o a ¢ c 2 a S S & 5 is) ct) = = Bolbolenellus hermani Bolbolenellus euryparia Figure 21. The most parsimonious tree of length 14 steps produced from analysis of character data in Table 19 with PAUP v. 3.1.1 (Swofford 1993). The cladogram is constructed using an exhaustive search. The retention index is 0.60, and the consistency index is 0.67. The value of the gj statistic is—0.16. The following nodes in the text were sup- ported by the following bootstrap confidence values (see text for bootstrapping procedure used): Node | = 0.58; Node 2 = 0.44; Node 3 = 0.31; Node 4 = 0.25. Character states are placed at nodes, using MacClade v. 3.04 (Mad- dison and Maddison 1992), with the characters given in Table 18. The apomorphic state is given in parentheses. Square parentheses indicate equivocal character states that are ambiguous because of missing data or multiple equally parsimonious resolutions. Equivocal characters are placed only at their basal phylogenetic position, and only unambiguous reversals are shown. Node 1, 4(1); Node 2, 5(1), 6(1); Node 3, 7(1); Node 4, 3(1). that family, Lieberman (1998) placed two subfamilies, the Biceratopsinae, comprised of Peachella and Biceratops, and the paraphyletic “Bristoliinae,” which included the rest of the aforementioned genera. In Lieberman (1998), “Bristoliinae” contained basically the same genera as in Palmer and Repina (1993) since some of the supra-generic categories had been tentatively defined until all available species within these genera could be considered. This study presents data from all available species for several different olenelloid genera, and is thus an appropriate place to reconsider supra-generic classification within the Olenelloidea. Above, under the genus Bristolia, Bristoliinae was redefined to be monophyletic and restricted to the genera Bristolia, Fremontella and Lochmanolenellus. Consequently, all other taxa within the Biceratopsidae were reassigned to a monophyletic, expanded Bicer- atopsinae, including those assigned to that subfamily in Lieberman (1998) and to the gen- era Bolbolenellus, Nephrolenellus and Olenelloides. These taxa were placed within the Bicer- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 131 atopsinae rather than either erecting a series of monotypic subfamilies for each of these three genera or creating a new paraphyletic subfamily. The subfamily Biceratopsinae is de- fined here by the possession of the following characters (thoracic characters not verifiable for all species as thoracic remains do not always exist): 1. length (exsag.) of anterior cephalic border between the lateral margins of L4 and the genal spine angle roughly equal to half length (sag.) of LO; 2. anterior cephalic border developed as a raised ridge; 3. promi- nent parafrontal band not visible; 4. pre-ocular furrow not visible on L4; 5. ocular lobe smoothly merges with extraocular area (except in B. sphaerulosus |Fritz]); 6. line from pos- terior tip of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 10 to 20 degree angle relative to sagittal line; 7. distal margins of L3 convex outward; 8. line between ends of S2 transverse; 9. S1 conjoined medially; 10. distal sector of SO straight or convex anteriorly; 11. lateral lobes on LO absent (except in Olenelloides armatus Peach); 12. extraocular area prominently vaulted; 13. thorax broken up into pro- and opisthothorax; 14. T3 macropleural, pleural spine projects further posteriorly than length of entire pro- thorax; 15. anterior margin of thoracic pleural furrow on T3 when proceeding from prox- imal to distal edge parallels a transverse line before flexing strongly posteriorly; 16. tho- racic pleural furrows extend onto spines (except in Nephrolenellus); 17. boundary between thoracic pleural furrow and anterior band sharp; 18. pleurae of opisthothoracic segments present, with orientation distinct from that of prothoracic segments, flexing less strongly posterolaterally. Included Taxa GENUS NEPHROLENELLUS PALMER AND REPINA 1993 Type species: Olenellus multinodus Palmer in Palmer and Halley 1979. Included species: Nephrolenellus jasperensis new species. Diagnosis: Anterior cephalic border prominently separated from extraocular area by fur- row; plectrum absent; L4 contacts anterior border furrow; anterior margins of L4 at each side of midline deflected posteriorly at roughly 10 to 20 degree angle relative to transverse line; length (sag.) of L4 equal to 1.5 times length of LO (sag.); L4 expands prominently dor- sally; lateral margins of L4 distal to lateral margins of LO; region of anterior part of ocular lobe between visual surfaces and L4 broad (tr.), one-third width of glabella at L1; ocular lobe without prominent ocular furrow; line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 45 degree angle with sagittal line; transverse profile of ocular lobes convex dorsally; posterior tips of ocular lobes developed opposite medial part of distal margin of L1; width (tr.) of interocular area greater than width of ocular lobe; $3 jaggedly convex, not conjoined medially; L2 and L3 do not merge; S2 convex anteriorly, conjoined medially; distal margins of L2 when proceeding anteriorly diverging; $1 conjoined medially; SO convex anteriorly, not conjoined medially, proximal and distal ends on transverse line; LO, L1, L2 and L3 with axial spine; posterior margin of LO convex posteriorly; prominent anastomosing ridges absent on extraocular area; width (tr.) of extraocular region opposite L1 equal to 75% of width of glabella at L1; genal spine near cephalic border directed posterolaterally at roughly 35 to 45 degree angle relative to sagittal line; genal spine short (exsag.), length approximately equal to length of first 2 to 3 thoracic segments; genal spine angle developed opposite medial part of distal margin of LO; intergenal angle deflected anteriorly at least 30 degrees relative to transverse line; medial part of posterior border between intergenal angle and LO flexes posteriorly. 132 PEABODY MUSEUM BULLETIN 45 Discussion: This genus is known from the western part of Laurentia in strata equivalent to the upper Olenellus zone, late Early Cambrian. NEPHROLENELLUS MULTINODUS (PALMER IN PALMER AND HALLEY 1979) Olenellus multinodus; Palmer and Halley 1979:72, pl. 4, figs. 1-5, 7-9. Nephrolenellus multinodus (Palmer and Halley); Palmer and Repina 1993:24, fig. 4.6; Palmer and Repina 1997:411, fig. 258.4a. Types: Holotype, USNM 177225, see Palmer and Halley (1979). Discussion: This species is the type of the genus. It has a small intergenal spine at the inter- genal angle. The intergenal angle of N. jasperensis n. sp. is too poorly preserved to determine whether this character state is also present in this species, potentially making it an addi- tional diagnostic character of the genus. Occurrence: California: upper Carrara Formation and Pioche Shale, Nephrolenellus multin- odus zonule, uppermost Olenellus zone following Palmer and Halley (1979), southern Great Basin in the Funeral Mountains, Resting Springs Range, Eagle Mountains, Grapevine Range; Nevada: same formation, Desert and Delamar Ranges. NEPHROLENELLUS JASPERENSIS LIEBERMAN NEW SPECIES Figure 20.6 Olenellus multinodus; Palmer in Palmer and Halley 1979:72, pl. 4, fig. 6. Types: Holotype, GSC 16858, designated here (Figure 20.6), from 10 m above the top of the Gog Group, upper or ?uppermost Olenellus zone, following Fritz (1972), about 2 mi SW of Mount Simla, Jasper Park, western Alberta, Canada (locality described in detail in Mountjoy 1962). Diagnosis: Posteromedial part of L4 at same elevation as glabella at L3; entire margin of S3 arched furthest anteriorly at point midway between midline of glabella and axial furrows; medial and distal tips of S2 deflected as far forward anteriorly; S2 arching first anteriorly then posteriorly when going from medial to lateral tip; posterior margin of ocular lobes de- veloped opposite medial part of distal margin of L1; medial part of intergenal angle devel- oped at point about two-thirds of way between distal tip of ocular lobes and genal spine angle; intergenal angle forms 30 to 35 degree angle with transverse line; genal spine angle opposite distal edge of medial margin of LO. Description: Anterior cephalic border narrow, raised, rounded ridge, length (exsag.) be- tween L4 and genal spine angle equal to one-half length L1, prominently separated from ex- traocular area by furrow; frontal lobe contacts anterior border furrow; plectrum absent; an- terior margins of L4 at each side of midline deflected posteriorly at roughly 10 to 20 degree angle relative to transverse line; lateral margins of frontal lobe distal to LO; frontal lobe ex- panded dorsally, long, length (sag.) equal to roughly 1.5 times length of LO (sag.); lateral margins of glabella expand from posterior margin of L2 to midpoint of margin of L3; lat- eral margins of glabella contract between midpoint of L3 and $3; S3 jaggedly convex, con- joined medially; $2 convex, conjoined medially; L2 and L3 do not merge distally; line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 45 degree angle relative to sagittal line; LO, L1, L2 and L3 with axial node; SO, S1 straight, contact axial furrows, medial edges declined posteriorly; lateral margins of glabella SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 133 Table 18. Description of characters and character states used in phylogenetic analysis of Bolbolenellus; (0) does not always represent the primitive state as two outgroup taxa were employed in phylogenetic analysis, but for simplicity one of the outgroups, Nephrolenellus multinodus, was coded with all 0 character states. 1. Genal spine angle developed (0) medial part of margin (1) medial part of margin opposite of LO of L1 (2) anteriormost tip of S3 2. Medial and distal tips of S2 (0) on transverse line (1) medial tip deflected further anteriorly 3. Medial most part of intergenal (0) directly behind genal (1) point two-thirds of the angle developed spine way between distal tip of ocular lobes and genal spine angle 4. Entire margin of $3 arched (0) at point midway between (1) at point three-quarters of furthest anteriorly midline of glabella and way between midline of axial furrows glabella and axial furrows 5. Postero-medial part of L4 (0) at same level as posterior (1) significantly elevated part of glabella between above posterior part lobes L1 to L3 of glabella between lobes Lito ks 6552 (0) arching first anteriorly, (1) roughly transverse then posteriorly, when going from medial to lateral tip (jaggedly convex) 7. Posterior margin of ocular lobes (0) medial part of margin (1) distal tips of SO developed opposite of LI contract between LO and LI; line from posterior tip of ocular lobe to junction of posterior margin of lobe with glabella forms roughly 45 degree angle with sagittal line; region of an- terior part of ocular lobe between visual surfaces and L4 broad (tr.), one-third width of glabella at L1; ocular lobes without prominent ocular furrow; transverse profile of ocular lobes convex dorsally; posterior edges of ocular lobes opposite medial part of distal margin of L1; SO and S1 conjoined medially; posterior margin of LO convex posteriorly; genal spines developed as short projections, length (exsag.) approximately equal to length (sag.) of two to three thoracic segments, sweeping posterolaterally at roughly 35 to 45 degree angle rela- tive to sagittal line; genal spine angle opposite medial part of distal margin of LO; intergenal angle inclined at 30 to 35 degree angle relative to transverse line; extraocular region broad, 134 PEABODY MUSEUM BULLETIN 45 width (tr.) approximately 75% of width of glabella at L1; faint anastomosing ridges not vis- ible on extraocular region; medial part of intergenal angle developed at point about two- thirds of way between distal tip of ocular lobes and genal spine angle; posterior margin of cephalic posterior border between LO and intergenal angle deflected posterolaterally. Thorax and pygidium unknown. Discussion: This species is unfortunately known from very limited material. Originally Palmer in Palmer and Halley (1979) suggested that material from western Alberta was con- specific with N. multinodus from the Great Basin. However, there are subtle differences be- tween the holotype specimen of N. jasperensis and all material from the Great Basin as- signed to N. multinodus that indicate these species are subtly distinct, and therefore a new species was described. Etymology: Species name derived from Jasper Park, in Canada, where the species is found. Material examined: GSC 16858, the holotype. Occurrence: Canada: Alberta, 10 m above the top of the Gog Group, upper or ‘uppermost Olenellus zone, following Fritz (1972), about 2 mi SW of Mount Simla, Jasper Park. GENUS BOLBOLENELLUS PALMER AND REPINA 1993 Type species: Olenellus euryparia Palmer in Palmer and Halley 1979. Assigned taxa: Olenellus altifrontatus Fritz 1972; O. cylindricus Palmer in Palmer and Halley 1979; O. groenlandicus Poulsen 1927; O. sphaerulosus Fritz 1991; O. hermani Kindle and Tasch 1948. Diagnosis: Anterior cephalic border prominently separated from extraocular area by fur- row, developed as raised, rounded ridge; plectrum absent; L4 contacts anterior border furrow; anterior margins of L4 at each side of midline deflected posteriorly at roughly 40 degree angle relative to transverse line; length (sag.) of L4 equal to 1.5 times length of LO and LI (sag.); L4 expands prominently dorsally; lateral margins of L4 distal to lateral margins of LO; region of anterior part of ocular lobe between visual surfaces and L4 nar- row (tr.); ocular lobe with ocular furrow; line from posterior tip of ocular lobe to junc- tion of posterior margin of lobe with glabella forms roughly 10 degree angle relative to sagittal line; transverse profile of ocular lobes convex dorsally; interocular area arched; posterior tips of ocular lobes developed opposite SO or medial part of distal margin of L1; width (tr.) of interocular area roughly equal to width of ocular lobe; S3 jaggedly convex, conjoined medially; entire margin of $3 arched furthest anteriorly at point three-quarters of way between midline of glabella and axial furrows; L1, L2, and L3 of strong transverse convexity; L2 and L3 do not merge; $2 conjoined medially; distal margins of L2 when proceeding anteriorly subparallel; $1 conjoined medially; distal sector of SO straight, con- joined medially, proximal end well posterior of distal end; axial part of LO with axial node; posterior margin of LO convex posteriorly; lateral lobes on LO absent; posterior cephalic border between LO and intergenal angle deflected posterolaterally; prominent anasto- mosing ridges present on extraocular area; width (tr.) of extraocular region opposite L1 equal to roughly 75% of width of glabella at L1; genal spine near cephalic border directed posterolaterally at angle greater than or equal to 25 degrees relative to sagittal line; genal spine length (exsag.) greater than or equal to length of first 4 to 5 thoracic segments; in- tergenal angle developed two-thirds of way between distal tips of ocular lobes and genal spine angle, or directly behind genal spine angle. Discussion: Originally, all species assigned to this genus had been assigned to the genus Olenellus. However, Palmer and Repina (1993) recognized that the type of the genus, B. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 135 Table 19. Character state distributions for Bolbolenellus and outgroups used in phylogenetic parsimony analysis. Characters and alternative states are as listed in Table 18. Missing data are indicated by “?.” ays Or Geolgorerd Nephrolenellus multinodus 03050;0,070.0 Nephrolenellus jasperensis O70 1070) 0':0 Bolbolenellus euryparia Leiria Ob. Bolbolenellus groenlandicus Zale OPA. 'OxO Bolbolenellus hermant Onell 1 70 Bolbolenellus sphaerulosus ORO Ot Tat Bolbolenellus altifrontatus OVO On euryparia, was distinct from members of the genus Olenellus and belonged in a new genus, Bolbolenellus. Based on phylogenetic topology within the Olenelloidea (Lieberman 1998), species of the genera Olenellus and Bolbolenellus are distantly related. Species of Bolbolenellus can always be distinguished from species of Olenellus (including what was formerly referred to as O. [Paedeumias]) by the condition of the following characters: 1. in Bolbolenellus the anterior cephalic border is developed as a rounded ridge, in Olenellus it is a flattened ledge; 2. in Bolbolenellus the anterolateral part of L4 is more prominently separated from the extraocular area than it is in Olenellus; 3. in Bolbolenellus L4 expands prominently dorsally, in Olenellus it does not; 4. in Bolbolenellus the lateral margins of L4 are distal to the lateral margins of LO, in Olenellus the lateral margins of L4 are directly anterior to or proximal to the lateral margins of LO; 5. in Bolbolenellus the ocular lobes are convex dorsally in transverse profile, in Olenellus the ocular lobes are flattened dorsally; 6. in Bolbolenellus the interocular area is arched dorsally, in Olenellus it is developed as a flattened shelf; 7. in Bolbolenellus the posterior tips of the ocular lobes are developed op- posite the medial part of the distal margin of L1 or opposite SO, in Olenellus they are de- veloped opposite the medial part of the distal margin of LO; 8. in Bolbolenellus $3, S2 and SO are conjoined medially, in Olenellus they are not; 9. in Bolbolenellus the middle sector of S3 is linear, in Olenellus it is convex anteriorly; 10. in Bolbolenellus a line between the ends of S2 is transverse, in Olenellus it is directed inward and posteriorly at a roughly 45 degree angle relative to a transverse line; 11. in Bolbolenellus L2 and L3 do not merge lat- erally, in Olenellus they do; 12. in Bolbolenellus the distal margins of L2 when proceeding anteriorly are roughly subparallel, whereas in Olenellus they expand; 13. in Bolbolenellus the posterior margin of LO is more convex than it is in Olenellus; 14. in Bolbolenellus the extraocular area is more prominently vaulted than it is in Olenellus; 15. in Bolbolenellus prominent anastamosing ridges are present on the extraocular area, in Olenellus they are not; 16. in Bolbolenellus the genal spine near where it hits the cephalic border is directed posterolaterally at an angle greater than or equal to 35 degrees relative to a transverse line (except for B. sphaerulosus), in Olenellus it is directed posterolaterally at a roughly 10 to 15 degree angle. On the basis of these and other characters, species of Bolbolenellus are easily distinguished from species of Olenellus. 136 PEABODY MusEUM BULLETIN 45 BOLBOLENELLUS EURYPARIA (PALMER IN PALMER AND HALLEY 1979) Olenellus fremonti; Walcott 1910:320, pl. 37, figs. 1, 4, 5. Olenellus euryparia; Palmer in Palmer and Halley 1979:69, pl. 2, figs. 15-18; Fritz 1991:15. Bolbolenellus euryparia (Palmer in Palmer and Halley); Palmer and Repina 1993:24, fig. 4.10; Palmer and Repina 1997:409, fig. 258.3. Olenellus sp.; Nelson 1976:31, pl. 8 (upper right hand corner of plate). Types: Holotype, USNM 177204, see Palmer and Halley (1979). Discussion: Palmer in Palmer and Halley (1979) was correct in recognizing the close affin- ity between this species and B. altifrontatus (Fritz). On the basis of phylogenetic analysis of the genus Bolbolenellus (Figure 10), these species appear to be sister taxa. Material examined: UCR 7271A. Occurrence: California: upper Carrara Formation, Bristolia zonule, upper Olenellus zone, following Nelson (1976) and Palmer and Halley (1979), Eagle Mountain, White/Inyo Mountain region; and the middle part of the Latham Shale, the Bristolia zonule, E slope of hill 1440 in southern end of the Marble Mountains, 834 ft E and 1,999 ft N of the SW cor- ner sec 12, T 5 N, R 14 E, San Bernardino County. BOLBOLENELLUS GROENLANDICUS (POULSEN 1927) Figures 20.7, 22.1 Olenellus groenlandicus; Poulsen 1927:254, pl. 24, figs. 28-31. Bristolia groenlandica (Poulsen); Poulsen 1964:10. @Olenellus kentensis; Poulsen 1927:255, pl. 24, fig. 32. ?Bristolia kentensis (Poulsen); Poulsen 1964:11. Types: Poulsen (1927) did not designate a particular specimen as the type of his O. groen- landicus. Poulsen (1964) designated MGUH 2233, the most well-preserved specimen of Poulsen’s (1927) suite of specimens, as the holotype. However, since several specimens were figured by Poulsen (1927), this actually qualifies as a lectotype designation. Poulsen’s spec- imen (1927, pl. 24, figs. 29, 30), MGUH 2233, is the lectotype of the species (Figure 22.1) and MGUH 2232 and 2234 (Poulsen 1927, pl. 24, figs. 28 and 31, respectively) become para- lectotypes. MGUH 2235 was designated the holotype of O. kentensis by monotypy in Poulsen (1964). This remains the holotype of the species, but O. kentensis is treated here as a junior subjective synonym of Bolbolenellus groenlandicus (Figure 20.7; see discussion below for explanation). Discussion: Poulsen (1927) figured and described two new taxa from the Lower Cambrian Cape Kent Formation, upper Olenellus zone (Poulsen 1964), Cape Kent, NE end of Ingle- field Land, NW Greenland, which he referred to as Olenellus groenlandicus and O. ken- tensis. These were later reassigned to Bristolia by Poulsen (1964) primarily on the basis of the possession of an advanced genal spine (sensu Harrington 1956) in Bolbolenellus groenlandicus (presumably what he referred to as Bristolia kentensis). However, there are several other genera within the Olenelloidea that bear advanced genal spines, including Fremontella, Lochmanolenellus, Laudonia, Elliptocephala and Mesonacis (for these genera see discussions above under the appropriate generic headings), and the condition of this character thus should not be the sole arbiter of the taxonomic status of a species. Indeed, Poulsen (1964) actually considered this and was cautious in his generic assignment, sug- SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 137 Figure 22. 1. Bolbolenellus groenlandicus (Poulsen), locality is the same as Figure 20.7, latex cast of the holotype, MGUH 2233, x1.5. 2. Bolbolenellus altifrontatus (Fritz), locality is the same as Figure 12.2, GSC 27435, the holotype, x1.6. gesting that what he called Bristolia groenlandica might be related either to Fremontella or to some other new genus. Poulsen (1964) marshaled additional character evidence to sug- gest a close relationship between Bristolia bristolensis (Resser) and Bolbolenellus groen- landicus. In particular, he argued that these species have a similar general outline of the cephalon, convexity of the border and position of the genal spines, though he admitted that they did differ in other respects. What Poulsen (1927) referred to as Olenellus ken- tensis is very poorly preserved, lacking the anterior and lateral cephalic borders, and is thus difficult to comment on. However, it appears very similar to, if not identical with, Bolbolenellus groenlandicus. Poulsen (1964) claimed that Olenellus kentensis and B. groen- landicus differed in the condition of L4, which he claimed attained its maximum width at the midline of L4, whereas B. groenlandicus attained its maximal width at the anterior margin of L4; and what he called the transglabellar furrow (S3), which he claimed was shallower in O. kentensis than in B. groenlandicus. Since the anterior margin of L4 and the cephalon in O. kentensis are not preserved, it is very difficult to ascertain whether it truly differs from B. groenlandicus in the condition of L4 described above. $3 appears to be equally incised in both taxa. Because O. kentensis is so poorly preserved and known from such limited material, and because it seems to show no significant differences from B. groenlandicus, at this time it is treated as conspecific with B. groenlandicus until additional material can be recovered. When considered in detail, Bolbolenellus groenlandicus has the following characters: 1. the intergenal angle forms a 90 degree angle with a transverse line; 2. there is a long (exsag.) genal spine, with the genal spine angle opposite the medial tip of S3; 3. there is a thin ante- rior cephalic border developed as a raised, rounded ridge; there is a prominently incised an- terior border furrow; 4. the lateral margins of LO and L1 when proceeding anteriorly con- strict; 5. the extraocular area is prominently arched and very narrow; 6. SO to S3 are all conjoined medially; 7. $2 is convex, contacts the axial furrow, and the medial tip is further anterior than the distal tip; 8. there is a prominent node on LO; 9. the posterior edge of the ocular lobe is nearly directly behind the anterior edge of the ocular lobe, rather than being 138 PEABODY MUSEUM BULLETIN 45 rotated laterally; 10. the posterior cephalic border, when proceeding laterally, first deflects posterolaterally and then anterolaterally. Bolbolenellus groenlandicus does bear a superficial resemblance to Bristolia bristolensis and other species of the genus Bristolia, as Poulsen (1964) implied by his assignment of this species to the genus Bristolia. In particular, it does have a prominent genal spine deflected far forward anteriorly, and the intergenal angle forms a roughly 90 degree angle with a transverse line. However, based on the possession of the diagnostic characters of the genus Bolbolenellus presented above, this species appears to be referable to that genus rather than to the genus Bristolia. In particular, B. groenlandicus has the following characters, which suggest it lies with the genus Bolbolenellus rather than with Bristolia: 1. the anterior cephalic border is developed as a narrow, raised, and rounded ridge (rather than raised and flattened as in Bristolia); 2. the anterolateral margins of L4 at each side of the midline are deflected posteriorly at a roughly 40 degree angle relative to a transverse line rather than at a roughly 10 degree angle as in Bristolia; 3. L4 is relatively longer (sag.) than it is in Bristolia; L4 ex- pands prominently dorsally, it does not in Bristolia; 4. a pre-ocular furrow is not visible, it is in Bristolia; 5. the interocular area is arched rather than sloping gently from the ocular lobe to the glabella as it does in Bristolia; 6. the interocular area is relatively broader than it is in Bristolia; 7. the extraocular regions are far more vaulted and narrow (tr.) than they are in Bristolia; 8. SO is conjoined medially, it is not typically in Bristolia; 9. S2 is jaggedly con- vex, rather than transverse or gently convex (as in Bristolia); 10. S2 has the medial end fur- ther anterior than the distal end, and this is never found in Bristolia. On the basis of this character evidence Bolbolenellus groenlandicus is excluded from the genus Bristolia and thus Olenellus gronelandicus, and by default O. kentensis, are referred to the genus Bolbolenellus. This expands the number of genera with highly advanced (sensu Harrington 1956) genal spines in the superfamily Olenelloidea. Poulsen (1927) figured a hypostome that he assigned to his O. groenlandicus. This hy- postome is treated as within that species until additional evidence can either confirm or deny this contention. Material examined: MGUH 2232, a paralectotype, 2233, the lectotype, and 2235 (formerly referred to O. kentensis). Occurrence: Denmark: Greenland, Cape Kent Limestone, upper Olenellus zone, NE end of Inglefield Land, NW Greenland, following Poulsen (1927, 1964). BOLBOLENELLUS HERMANI (KINDLE AND TASCH 1948) Olenellus hermani; Kindle and Tasch 1948:136, pl. 1, figs. 1-3; Tasch 1949:210; Shaw 1955:792; Tasch 1952:486, fig. 1-1; Fritz 1972:13; Fritz 1991:15. Esmeraldina hermani (Kindle and Tasch); Shaw 1962:333, pl. 50, figs. 30-33. Types: The holotype appears to be a specimen illustrated in Kindle and Tasch (1948, pl. 1, figs. 1, 2) and reported as in the private collection of C. H. Kindle. At present the where- abouts of this material has unfortunately not been ascertained, despite numerous attempts. Discussion: Shaw (1962) assigned this species to the genus Esmeraldina; however, that generic assignment appears untenable as Esmeraldina rowei is distantly related to B. eury- paria (Lieberman 1998). A fragmentary cephalon apparently identical to B. hermani was re- cently recovered from the upper Olenellus zone of the White/Inyo Mountain region in Cal- ifornia (Palmer, pers. comm. 1996). Although poorly preserved, as is the type of the species and other available material for B. hermani, this specimen appears to be almost identical to SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 139 B. hermani from Vermont, and thus these taxa at this point in time are treated as conspe- cific. They both possess the spinose projection on the expanded posterior part of L4. Material examined: MCZ 5029. Occurrence: Vermont: 30 m above the base of the Monkton Quartzite, middle upper Olenel- lus zone, following Rankin et al. (1989), 0.5 mi W of the junction of interstate highways 2 and 7, near Clay and Red Rock Points on Lake Champlain (Kindle and Tasch 1948), lower 10 ft of the Parker Slate, 2 mi W of Georgia Center; California: the upper Olenellus zone, White/Inyo Mountain region. Thus, this species was presumed to be distributed through- out much of Laurentia during the upper Olenellus zone. BOLBOLENELLUS ALTIFRONTATUS (FRITZ 1972) Figure 22.2 Olenellus altifrontatus; Fritz 1972:12, pl. 18, figs. 14-17; Palmer in Palmer and Halley 1979:70; Fritz 1991:15. Types: Holotype, GSC 27435, see Fritz (1972). Discussion: Fritz (1972) suggested that this species was closely related to Olenellus hermani Kindle and Tasch, a conclusion supported by the current analysis. Both species are reas- signed to the genus Bolbolenellus, and in the phylogenetic analysis conducted here (Figure 21) they appear to be closely related but not sister taxa. Both uniquely share the transverse condition of S2, treated as a convergent similarity in this analysis. Material examined: GSC 27435, the holotype. Occurrence: Canada: Northwest Territories, the type section of the Sekwi Formation (Handfield 1968), 0.2 mi SE of June Lake, uppermost Sekwi Formation, 2367 ft above the base of the formation, upper Olenellus zone, following Fritz (1972), south end of the Sekwi Range, Mackenzie Mountains, District of Mackenzie. BOLBOLENELLUS SPHAERULOSUS (FRITZ 1991) Olenellus sphaerulosus; Fritz 1991:15, pl. 10, figs. 1-13. ?Olenellus bufrontis; Fritz 1991:12 , pl. 12, figs. 6-8. Types: Holotype, GSC 91835, see Fritz (1991). Discussion: Early ontogenetic stages of this species in some cases have the posterior tips of the ocular lobes developed opposite the medial part of the distal margin of LO. However, in typical adult material the ocular lobes only extend back to SO, and the species was coded as having this character state in this analysis. Olenellus bufrontis Fritz appears almost identical to B. sphaerulosus, differing solely from that taxon in coming from a much earlier ontoge- netic stage. As the only known locality of O. bufrontis is identical to one of the localities of B. sphaerulosus, they are provisionally treated as conspecific. Fritz (1991) noted that B. sphaerulosus was closely related to B. altifrontatus and B. euryparia, and also to B. hermani, and this conclusion is supported by this analysis. Bolbolenellus sphaerulosus is the only species within the Biceratopsinae that has a prominent furrow or depression separating the ocular lobe from the extraocular area, and the ocular lobe also sits on a raised ledge so that it does not smoothly merge into the extraocular area. An ocular lobe smoothly merging with the extraocular area is a feature considered diagnostic of the Biceratopsinae. However, because B. sphaerulosus matches all other species of Bolbolenellus and the Biceratopsinae in 140 PEABODY MUSEUM BULLETIN 45 the condition of so many other characters, it is treated as belonging to the genus Bol- bolenellus, and the condition of this character is treated as a convergence in a relatively de- rived species within the genus. Other taxa that typically bear such a feature include Frit- zolenellus truemani and Laudonia bispinata, taxa not closely related to the genus Bolbolenellus (Lieberman 1998). Occurrence: Canada: Yukon Territory, the upper part of the Illtyd Formation, Fritz’s (1991) Unit 6, ?upper Olenellus zone, Early Cambrian, Wernecke Mountains. Integrating Phylogenetic and Stratigraphic Data These phylogenetic data from separate clades can be integrated along with the results from Lieberman (1998) to give a detailed picture of patterns of evolution and speciation in the Olenelloidea during the Early Cambrian. Such a phylogenetic tree for the Olenelloidea is shown in Figure 23. Phylogenetic topology is taken directly from this study and Lieberman (1998), except for the addition to the tree of Peachella brevispina Palmer in Palmer and Halley. This species is assumed to be the sister taxon of P. iddingsi (Walcott). It includes in- formation about the relative appearance of species in the fossil record based on the corre- lation schemes of Ahlberg et al. (1986), Ahlberg (1991), Kirschvink et al. (1991), Rozanov (1992), Bowring et al. (1993), Palmer and Repina (1993), Isachsen et al. (1994), Geyer and Palmer (1995), Brasier et al. (1996), Kaufman et al. (1996), Knoll (1996), Vidal and Moczydlowska-Vidal (1997) and Landing et al. (1998). The ghost lineage method elabo- rated in Edgecombe (1992) is used to determine the latest time that a lineage must have di- versified by, with dashed lines representing that part of a species’ range not preserved in the fossil record. This tree can be used to elaborate biogeographic patterns during the Early Cambrian using the method of Lieberman and Eldredge (1996), as was done in Lieberman (1997), and also to look at how speciation rates during the Early Cambrian may have dif- fered from those at subsequent time periods in the fossil record (Lieberman 1999). Based on patterns of speciation and the stratigraphic distribution of taxa, it appears that all the major groups within the Olenelloidea must have diversified prior to the Cambrian Radia- tion, though speciation events are also distributed throughout the interval traditionally consigned to the Radiation. This implies either that there is a gap in the fossil record of in- determinate duration that cloaks the history of the Olenelloidea, or that the group diver- sified with startling rapidity and in a single pulse, right at the base of the S. mickwitzi zone. Even assuming that a gap is present in the history of the Olenelloidea, levels of speciation appear to be quite intense. For instance, when compared with levels of speciation in trilo- bites from other time periods that have been well-studied in the fossil record using phylo- genetic methods, such as the Middle Devonian (see Lieberman 1994), it is apparent that speciation may have been occurring at a much higher rate in the Early Cambrian. 142 PEABODY MUSEUM BULLETIN 45 Olenellus Mesonacinae eo an lepiana 1 11 11 1 | 1 | | 11 1 | 1 | Wanneria | Elliptocephala inal 11 =. Middle I Olenellus peas La zone er NVINWOLOd > --| S) > ee) > = > Z QUOZ B//A@DEAAN | euoz dnoib yinialy ‘H S. mickwitzi ope ee ae SS a heel ee en i af “NVILOWWOL Figure 23. A phylogenetic tree depicting topology and timing of patterns of speciation in the Olenelloidea. Tree built using information from phylogenetic analyses presented herein and in Lieberman (1998). Stratigraphic correlations and provisional radiometric dates from Ahlberg et al. (1986), Ahlberg (1991), Kirschvink et al. (1991), Rozanov (1992), Bowring et al. (1993), Palmer and Repina (1993), Isachsen et al. (1994), Geyer and Palmer (1995), Brasier et al. (1996), Kaufman et al. (1996), Knoll (1996), Vidal and Moczydlowska-Vidal (1997) and Landing et al. (1998). Solid lines represent approximate stratigraphic ranges of species. However, in many instances, based on SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 143 [ Nephrolenellus Bristoliinae Bolbolenellus levee t k= | , aes ill ee ely a ela eae I oA Fritzolenellus ete i li heer fowl | eu rigatet l l Laudonip’ yp M1 VNl ys ! ey ale po I Mummaspis | | “us Me Ne i Fea, nls as mS ach lh a l stilt gral he oem | Lali elle | ve LD tae fies a l aD le eae i fied glimigl oh l Fao 110 a al Ieee I alent pie | egy Rite) ae ie ena phylogenetic information, the ancestral lineages leading to these species must have appeared earlier. The ranges of these species were extended to make them square with information based on phylogenetic topology. Extended lineages are depicted as dashed lines that represent ghost lineages sensu Edgecombe (1992). The generic or sub- familial name is placed above each of the clades within the Olenelloidea. The Clade A refers to the following nested set of taxa: Olenelloides armatus Peach, Biceratops nevadensis Pack and Gayle, Peachella iddingsi (Walcott) and P. brevispina Palmer in Palmer and Halley. it rT ae he i iay ie 34 . t Sy i\lacuginw en” Bae i ‘ re bi? Zn " i helt t na, / decays d tbe § / ‘| vero” - : : at ' y eras ue ‘<= = > é i ; oi i j ‘a . iy F nme — aa poh | =| a : why my: i Bilt ped dee if oil “~ ( a ii? o e veg ! 5 i r - : t j | . \ , ce ee ee ee ee ee help rat ® ebeyit animes ie hy atucomar ten toro “ne hi cet, eet ee By Sih pe Gettin PAE nal, = RerGilal seem nee pall pi eee bedh tats PIGS eels ogi Oye Hiya! percent ) oh Mile ecm a0 te ding il | art thera nelitunse «ebay Weasley Se ee ' 7 eteee 1s 7 2 yy =i of Par ans “uy oe. ‘te < T i} ; : [ stole . a ii ; > . a = - 7 ey s* - REFERENCES AHLBERG, P. 1984. Lower Cambrian trilobites from the Laisvall area, northern Sweden. Geol. Foren. Stockholm Forh. 105:251—259. — 1985. Lower Cambrian trilobite faunas from the Scandinavian Caledonies—a review. In: D. G. Gee and B. A. Sturt, eds. The Caledonide Orogen-Scandinavia and Related Areas. New York: J. Wiley. pp. 339-346. — 1991. Trilobites in the Lower Cambrian of Scandinavia. Geol. Foren. Stockholm Forh. 113:74—75. AHLBERG, P. AND J. BERGSTROM. 1983. Lower Cambrian trilobites from southern Swedish Lapland. Geol. Foren. Stockholm Forh. 104:241—246. AHLBERG, P., J. BERGSTROM AND J. JOHANSSON. 1986. Lower Cambrian olenellid trilobites from the Baltic faunal province. Geol. Foren. Stockholm Forh. 108:39-56. BARNaBY, R. J. AND J. FE. READ. 1990. Carbonate ramp to rimmed shelf evolution: Lower to Middle Cambrian continental margin, Virginia Appalachians. Geol. Soc. Am. Bull. 102:391—404. BERGSTROM, J. 1973. Classification of olenellid trilobites and some Balto-Scandian species. Nor. Geol. Tidsskr. 53:283-314. Best, R. V. 1952. Two new species of Olenellus from British Columbia. Trans. R. Soc. Can. 46:13—22. BowRING, S. A., J. P. GROTZINGER, C. E. ISACHSEN, A. H. KNOLL, S. M. PELECHATY AND P. KOLosov. 1993. Calibrating rates of early Cambrian evolution. Science 261:1293-1298. Brasier, M. D., G. SH1ELDs, V. N. KULESHOV AND E. A. ZHEGALLO. 1996. Integrated chemo- and biostratigraphic calibration of early animal evolution: Neoproterozoic-early Cambrian of southwest Mongolia. Geol. Mag. 133:445—485. BreMER, K. 1994. Branch support and tree stability. Cladistics 10:295-304. Bupp, G. E. 1995. Kleptothule rasmusseni gen. et sp. nov.: an @olenellinid-like trilobite from the Sirius Passet fauna (Buen Formation, Lower Cambrian, north Greenland). Trans. R. Soc. Edinburgh, Earth Sci. 86:1—12. Bur ina, L. D. 1916. Paedeumias and the Mesonacidae, with description of a new species, having at least 44 segments, from the Lower Cambrian of British Columbia. Ottawa Nat. 30:53—58. Cowltk, J. W. 1968. Lower Cambrian faunas from Ellesmere Island, district of Franklin. Geol. Surv. Can., Bull. 163:1—127. Cowrrg, J. W. AND K. J. McNamara. 1978. Olenellus (Trilobita) from the Lower Cambrian strata of north-west Scotland. Palaeontology 21:615—634. Deo, D. M. 1935. A revision of the phacopid trilobites. J. Paleontol. 9:402—420. Dunsar, C. O. 1925. Antennae in Olenellus getzi, n. sp. Am. J. Sci. (5th ser.) 9:303-308. EDGECOMBE, G. D. 1992. Trilobite phylogeny and the Cambrian-Ordovician “event”: a cladistic reappraisal. In: M. J. Novacek and Q. D. Wheeler, eds. Extinction and phylogeny. New York: Columbia Univ. Pr. pp. 144-177. ELDREDGE, N. 1989. Macroevolutionary dynamics. New York: McGraw-Hill. 215 pp. ELDREDGE, N. AND J. CracrAFT. 1980. Phylogenetic patterns and the evolutionary process: method and theory in comparative biology. New York: Columbia Univ. Pr. 349 pp. 146 PEABODY MUSEUM BULLETIN NO. 45 Emmons, E. 1844. The Taconic System: based on observations in New York, Massachusetts, Maine, Vermont and Rhode Island. Albany, NY: Privately puiblished. 653 pp. Fortey, R. A. AND R. M. Owens. 1990. Evolutionary radiations in the Trilobita. In: P. D. Taylor and G. P. Larwood, eds. Major evolutionary radiations. Syst. Assoc. Spec. Vol. 42. Oxford: Clarendon. pp. 139-164. Fritz, W. H. 1972. Lower Cambrian trilobites from the Sekwi Formation type section, Mackenzie Mountains, northwestern Canada. Geol. Surv. Can., Bull. 212:1—90. ——— 1973. Medial Lower Cambrian trilobites from the Mackenzie Mountains, northwestern Canada. Geol. Surv. Can., Pap. 73-24, 43 pp. — 1991. Lower Cambrian trilobites from the Illtyd Formation, Wernecke Mountains, Yukon Territory. Geol. Surv. Can., Bull. 409:1—77. ——— 1992. Walcott’s Lower Cambrian olenellid trilobite collection 61K, Mount Robson area, Canadian Rocky Mountains. Geol. Surv. Can., Bull. 432:1—65. ——— 1995. Esmeraldina rowei and associated Lower Cambrian trilobites (1f fauna) at the base of Walcott’s Waucoban series, southern Great Basin, U.S.A. J. Paleontol. 69:708—723. Geyer, G. 1993. The giant Cambrian trilobites of Morocco. Beringeria 8:71—107. Geyer, G. AND A. R. Parmer. 1995. Neltneriidae and Holmiidae (Trilobita) from Morocco and the problem of Early Cambrian intercontinental correlation. J. Paleontol. 69:459—474. GROTZINGER, J. P., S. A. BowrRING, B. Z. SAYLOR AND A. J. KAUFMAN. 1995. Biostratigraphic and geochronologic constraints on early animal evolution. Science 270:598—604. HAL, J. 1859. Trilobites of the shales of the Hudson River Group. 12th Annual Report of the New York Cabinet for Natural History. Albany, NY: State of New York. pp. 59-62. —— 1862. Supplementary note to the thirteenth report of the Regents of the State Cabinet. 15th Annual Report of the New York Cabinet for Natural History. Albany, NY: State of New York. pp. 113-119. HANDFIELD, R. C. 1968. Sekwi Formation, a new Lower Cambrian formation in the southern Mackenzie Mountains, District of Mackenzie. Geol. Surv. Can., Pap. 68—47. 23 pp. HarrinctTon, H. J. 1956. Olenellidae with advanced cephalic spines. J. Paleontol. 30:56—61. HARRINGTON, H. J., G. HENNINGSMOEN, B. FE. HowELt, V. JAANUSSON, C. LOCHMAN-BaLk, RAC: Moore, C. Poutsen, F. Raserti, E. RICHTER, R. RICHTER, H. SCHMIpT, K. Spzuy, W. STRUVE, L. SrorMER, C. J. STUBBLEFIELD, R. Tripp, J. M. WELLER AND H. B. WHITTINGTON. 1959. In: R. C. Moore, ed. Treatise on invertebrate paleontology, Part O, Arthropoda 1. Boulder, CO and Lawrence, KS: Geol. Soc. Am. and Univ. Kansas Pr. pp. 560. Hazzarp, J. C. AND C. H. Crickmay. 1933. Notes on the Cambrian rocks of the eastern Mohave Desert, California. Univ. Calif. Publ., Bull. Dept. Geol. Sci. 23:57—80. Hixuts, D. M. 1991. Discriminating between phylogenetic signal and random noise in DNA sequences. In: M. Miyamoto and J. Cracraft, eds. Analysis of DNA sequences. New York: Oxford Univ. Pr. pp. 278-294. Hupgé, P. 1953. Contribution a l’etude du Cambrien inferieur et du Precambrien III de l’Anti-Atlas marocain. Notes Mem. Serv. Geol. (Morocco), 103. 402 pp. ISACHSEN, C. E., S. A. Bowrina, E. LANDING AND S. D. Samson. 1994. New constraint on the division of Cambrian time. Geology 22:496—498. KAUEMAN, A. J., A. H. KNott, M. A. SEMIKHATOV, J. P. GROTZINGER, S. B. JACOBSEN AND W. ADAMS. 1996. Integrated chronostratigraphy of Proterozoic-Cambrian boundary beds in the western Anabar region, northern Siberia. Geol. Mag. 133:509-533. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 147 KIAER, J. 1917. The Lower Cambrian Holmia fauna at Tomten in Norway. Vedenskaps. Skrifter, I. Matem. Nat. Skapelig Kl. 10, 140 pp. KigELAN, Z. 1960. On two olenellid trilobites from Hornsund, Vestspitsbergen. Stud. Geol. Pol. 4:83—92. KINDLE, C. H. AND P. Tascu. 1948. Lower Cambrian fauna of the Monkton Formation of Vermont. Can. Field Nat. 62:133-139. KIRSCHVINK, J. L., M. MARGARITZ, R. L. RIPPERDAN AND A. Y. ZHURAVLEV. 1991. The Precambrian/Cambrian boundary: magnetostratigraphy and Carbon isotopes resolve correlation problems between Siberia, Morocco, and south China. GSA Today 1:70-91. KNOLL, A. H. 1991. Environmental context of evolutionary change: an example from the end of the Proterozoic eon. In: L. Warren and H. Koprowski, eds. New perspectives on evolution. New York: Wiley-Liss. pp. 77-85. — 1992. Biological and biogeochemical preludes to the Ediacaran radiation. In: Lipps, J. H. and P. W. Signor, eds. Origin and early evolution of the Metazoa. New York: Plenum. pp. 53-84. —— 1996. Daughter of time. Paleobiology 22:1-7. KNOLL, A. H., A. J. KAUFMAN, M. A. SEMIKHATOY, J. P. GROTZINGER AND W. ADAMS. 1995. Sizing up the sub-Tommotian unconformity in Siberia. Geology 23:1139-1143. KNOLL, A. H. AND K. Swett. 1987. Micropaleontology across the Precambrian-Cambrian boundary in Spitsbergen. J. Paleontol. 61:898—926. Knott, A. H., K. Swett anp E. BuRKHARDT. 1989. Paleoenvironmental distribution of microfossils and stromatolites in the Upper Proterozoic Backlundtoppen Formation, Spitsbergen. J. Paleontol. 63:129-145. LANDING, E. AND K. E. BartTowskI. 1996. Oldest shelly fossils from the Taconic allochthon and the late Early Cambrian sea-levels in eastern Laurentia. J. Paleontol. 70:741—761. LANDING, E., S. A. BowRING, K. DaviDEK, S. R. WesTROP, G. GEYER AND W. HELDMAIER. 1998. New U-Pb zircon dates from Avalon and Gondwana and duration of the Early Cambrian. Can. J. Earth Sci. 35:329-338. LIEBERMAN, B. S. 1993. Systematics and biogeography of the “Metacryphaeus Group” Calmoniidae (Trilobita, Devonian), with comments on adaptive radiations and the geological history of the Malvinokaffric Realm. J. Paleontol. 67:549—570. — 1994. Evolution of the trilobite subfamily Proetinae Salter, 1864, and the origin, diversification, evolutionary affinity, and extinction of the Middle Devonian proetid fauna of eastern North America. Bull. Am. Mus. Nat. Hist. 223:1-176. — 1997. Early Cambrian paleogeography and tectonic history: a biogeographic approach. Geology. 25:1039-1042. — 1998. Cladistic analysis of the Early Cambrian olenelloid trilobites. J. Paleontol. 72:59-78. — 1999. A probabilistic analysis of rates of evolution during the Cambrian radiation. Abstracts with Programs, Geol. Soc. Am. Annual Meeting; October 1998; Toronto. LIEBERMAN, B. S., G. D. EDGECOMBE AND N. ELprEDGE. 1991. Systematics and biogeography of the “Malvinella group,” Calmoniidae (Trilobita, Devonian). J. Paleontol. 65:824—843. LIEBERMAN, B. S. AND N. ELDREDGE. 1996. Trilobite biogeography in the Middle Devonian: geological processes and analytical methods. Paleobiology. 22:66—79. Linnarsson, J. G. O. 1871. Om nagra forsteningar fran Sveriges och Norges “Primordialzon.” Ofver. KI. Sven. Vetenskaps. Akad. Forh. 6:789-796. 148 PeaBopy MuseuM BULLETIN NO. 45 LocHMAN, C. 1952. Trilobites. In: G. A. Cooper, A. R. V. Arellano, J. H. Johnson, V. Okulitch, A. Stoyanow and C. Lochman, eds. Cambrian stratigraphy and paleontology near Caborca, northwestern Sonora, Mexico. Smithson. Misc. Collect. 119:60—184. —— 1956. Stratigraphy, paleontology, and paleogeography of the Elliptocephala asaphoides strata in Cambridge and Hoosick quadrangles, New York. Bull. Geol. Soc. Am. 67:1331-1396. Mappison, W. P. AND D. R. MAppison. 1992. MacClade: Analysis of phylogeny and character evolution [computer program]. Version 3.04. Sunderland, MA: Sinauer. MartINsson, A. 1974. The Cambrian of Norden. In: C. H. Holland, ed. Lower Palaeozoic rocks of the world. Volume 2, Cambrian of the British Isles, Norden, and Spitsbergen. London: J. Wiley. pp. 185-283. MatTHEWw, G. F. 1890. On Cambrian organisms in Acadia. R. Soc. Can., Proc. Trans. 7:135-162. — 1897. What is the Olenellus fauna? Am. Geol. 19:396—407. McNamara, K. J. 1978. Paedomorphosis in scottish olenellid trilobites (Early Cambrian). Palaeontology 21:635—655. Meek, E. B. 1874. Preliminary report upon invertebrate fossils collected by the expeditions of 1871, 1872, and 1873, with descriptions of new species. In: C. A. White, U. S. Geographical and Geological Explorations and Surveys West of the One Hundredth Meridian. Washington, D.C.:U.S. Army, Engineering Dept. pp. 5-27. Minitas [computer program]. 1995. Version 10Xtra State College, PA: Minitab, Inc. Moser«, J. C. 1892. Om Olenellusledet i sydliga Skandinavien. Forh. Skand. Nat. 14:434—439. —— 1899. Sveriges alsta kanda trilobiter. Geol. Foren. Stockholm Forh. 21:309-348. MobeEr«, J. C. AND C. O. SEGERBERG. 1906. Bidrag till Kannedomen om Ceratopyge regionen med sarskild hansyn till des utveckling i Fogelsangtraken. Medd. Lunds Geol. Faltkubb (series B) 2:1—113. Mount, J. D. 1976. Early Cambrian faunas from eastern San Bernardino County, California. Bull. S. Calif. Paleontol. Soc. 8:173-182. —— 1980. Characteristics of Early Cambrian faunas from eastern San Bernardino County, California. S. Calif. Paleontol. Soc. Spec. Pub. 2:19-29. s Mountjoy, E. W. 1962. Mount Robson (southeast) map-area, Rocky Mountains of Alberta and British Columbia (83 E/SE). Geol. Surv. Can., Pap. 61-31. 114 pp. NELSON, C. A. 1976. Late Precambrian-Early Cambrian stratigraphic and faunal succession of eastern California and the Precambrian-Cambrian boundary. In: J. N. Moore and A. E. Fritsche, eds. Depositional environments of Lower Paleozoic rocks in the White-Inyo Mountains, Inyo County, California. Los Angeles: Soc. Econ. Paleontol. Mineral., Pacific Sect. pp. 31-41. NIKOLAISEN, FE. AND G. HENNINGSMOEN. 1990. Lower and Middle Cambrian trilobites from the Digermul peninsula, Finnmark, northern Norway. Norges Geol. Unders. Bull. 419:55—95. Ortowskt, S. 1974. Lower Cambrian biostratigraphy in the Holy Cross Mts., based on the trilobite family Olenellidae. Acta Geol. Pol. 24:1—-16. — 1985. Lower Cambrian and its trilobites in the Holy Cross Mts. Acta Geol. Polonica 35:231—250. Pack, P. D. AND H. D. Gaye. 1971. A new olenellid trilobite, Biceratops nevadensis, from the Lower Cambrian near Las Vegas, Nevada. J. Paleontol. 45:893-898. Pater, A. R. 1997. Suborder Olenellina. In: R. L. Kaesler, ed. Treatise of invertebrate paleontology Part O, Arthropoda 1, Trilobita, revised. Boulder, CO and Lawrence, KS: Geol. Soc. Am. and Univ. Kansas Pr. pp. 404—429. SYSTEMATIC REVISION OF THE OLENELLOIDEA (TRILOBITA, CAMBRIAN) 149 PALMER, A. R. AND R. B. HALtey. 1979. Physical stratigraphy and trilobite biostratigraphy of the Carrara Formation (Lower and Middle Cambrian) in the southern Great Basin. U. S. Geol. Surv., Prof. Pap. 1047:1-131. PALMER, A. R. AND J. S. PEEL. 1979. New Cambrian faunas from Peary land, eastern North Greenland. Gronl. Geol. Unders. Rapp. 91:29-36. PALMER, A. R. AND L. N. Repina. 1993. Through a glass darkly: taxonomy, phylogeny and biostratigraphy of the Olenellina. Lawrence, KS: Univ. Kans. Paleontol. Contrib. New Series 3. 35 pp. Peach, B. N. 1894. Additions to the fauna of the Olenellus-zone of the north-west highlands. Quart. J. Geol. Soc. London 50:661—675. Peacu, B. N. AND J. Horne. 1892. The Olenellus zone in the north-west highlands of Scotland. Quart. J. Geol. Soc. London 48:227—242. PouLsEN, C. 1927. The Cambrian, Ozarkian, and Canadian faunas of Northwest Greenland. Medd. Gronl. 70(2):1—343. — 1958. Contribution to the palaeontology of the Lower Cambrian Wulff River Formation. Medd. Gronl. 162:1—25. PouLSEN, V. 1964. Contribution to the Lower and Middle Cambrian paleontology and stratigraphy of Northwest Greenland. Medd. Gronl. 164(6):1—105. — 1974. Olenellacean trilobites from eastern North Greenland. Bull. Geol. Soc. Den. 23:79-101. RANKIN, D. W., A. A. DRAKE Jr., L. GLover III, R. Gotpsmiru, L. M. HALL, D. P. Murray, N. M. RATCLIFFE, J. F. READ, D. T. Secor JR. AND R. S. STANLEY. 1989. Pre-orogenic terranes. In: R. D. Hatcher Jr., W. A. Thomas and G. W. Viele, eds. The Appalachian-Ouachita Orogen in the United States. The Geology of North America, Volume F-2. Boulder, CO: Geol. Soc. Am. pp. 7-101. RaseTTI, F. 1948. Lower Cambrian trilobites from the Conglomerates of Quebec. J. Paleontol. 11:1—24. — 1966. New Lower Cambrian trilobite faunule from the Taconic sequence of New York. Smithson. Misc. Collect. 148:1—52. — 1967. Lower and Middle Cambrian trilobite faunas from the Taconic sequence of New York. Smithson. Misc. Collect. 152:1-111. Raw, F. 1936. Mesonacidae of Comley in Shropshire, with a discussion of classification within the family. Quart. J. Geol. Soc. London 92:236—293. Reprna, L. N. 1979. Dependence of morphologic features on habitat conditions in trilobites and evalution of their signficance for the systematics of the superfamily Olenelloidea. Akad. Nauk SSSR, Sib. Otd., Tr. Inst. Geol. Geof. 431:11-30. REssER, C. E. 1928. Cambrian fossils from the Mohave Desert. Smithson. Misc. Collect. 81(2):1-14. — 1938. Cambrian system (restricted) of the southern Appalachians. Geol. Soc. Am. Spec. Pap. 15. 140 pp. Resser, C. E. AND B. FE. Howe t. 1938. Lower Cambrian Olenellus zone of the Appalachians. Geol. Soc. Am. Bull. 49:195—248. Riccio, J. E 1952. The Lower Cambrian Olenellidae of the southern Marble Mountains, California. Bull. S. Calif. Acad. Sci. 51:25-49. RicuTER, R. 1933. Crustacea (Palaontologie). Handwort. Naturwiss. 2:840—864. Rosison, R. A. AND L. F. Hintze. 1972. An Early Cambrian trilobite faunule from Utah. Brigham Young Univ. Geol. Stud. 19:3-13. 150 PEABODY MuseuM BULLETIN NO. 45 Rozanoy, A. Y. 1992. Some problems concerning the Precambrian-Cambrian transition and the Cambrian faunal radiation. J. Geol. Soc. London 149:593-598. SAMSONOWICz, J. 1959. On the Holmia-Fauna in the Cambrian of the anticlinorium of Klimotow. Bull. Acad. Pol. Sci., Ser. Sci. Chim., Geol. Geogr. 7:447—451. ScHMIDT, F. 1888. Uber eine neuentdeckte unterkambrische Fauna in Estland. Mem. Acad. Imp. Sci. St.-Pétersbourg, VII Ser. 36:1—27. Spzuy, K. 1961. Das Kambrium Spaniens. Teil I: Trilobiten. 1. Abschnitt. Akad. Wiss. Lit. Mainz, Abh. Math. Naturwiss. Kl. 1961:217-312. SHaw, A. B. 1955. Paleontology of northwestern Vermont. V. The Lower Cambrian fauna. J. Paleontol. 29:775—805. — 1962. Paleontology of northwestern Vermont. IX. Fauna of the Monkton Quartzite. J. Paleontol. 36:322—345. SKJESETH, S. 1963. Contributions to the geology of the Mjgsa districts and the classical sparagmite area in southern Norway. Nor. Geol. Unders. 220:1—126. SwoFkorD, D. L. 1993. PAUP (Phylogenetic analysis using parsimony) [computer program]. Version 3.1.1. Champaign, IL: Illinois Natural History Survey. Tascu, P. 1949. A new fossil locality in the Lower Cambrian Monkton Formation of Vermont. Can. Field Nat. 63:210—211. —— 1952. Adaptive trend in eyeline development in the Olenellidae. J. Paleontol. 26:484—488. VipaL, G. AND M. MoczypLowskKa-VIDAL. 1997. Biodiversity, speciation, and extinction trends of Proterozoic and Cambrian phytoplankton. Paleobiology 23:230—246. Watcortrt, C. D. 1885. Paleozoic notes: new genus of Cambrian trilobites, Mesonacis. Am. J. Sci. (series 3) 29:328—330. — 1890. The fauna of the Lower Cambrian or Olenellus zone. U. S. Geol. Surv. 10th Annual Report. pp. 509-763. — 1910. Olenellus and other genera of the Mesonacidae. Smithson. Misc. Collect. 53(6):231—422. — 1913. Cambrian geology and paleontology, No. 11. New Lower Cambrian subfauna. Smithson. Misc. Collect. 57(11):309-326. ; Wanner, A. 1901. A new species of Olenellus from the Lower Cambrian of York County, Pennsylvania. Proc. Wash. Acad. Sci. 3:267—272. WHITFIELD, R. P. 1884. Notice of some new species of primordial fossils in the collections of the Museum, and corrections of previously described species. Bull. Am. Mus. Nat. Hist. 1:139-154. WHITTINGTON, H. B. 1988. Hypostomes and ventral cephalic sutures in Cambrian trilobites. Palaeontology 31:577-609. —— 1989. Olenelloid trilobites: type species, functional morphology and higher classification. Philos. Trans. R. Soc. London Ser. B 324:111—147. — 1990. Articulation and exuviation in Cambrian trilobites. Philos. Trans. R. Soc. London Ser. B 329:27—-46. Wiey, E. O. 1979. An annotated Linnaean hierarchy, with comments on natural taxa and competing systems. Syst. Zool. 28:308—337. YALE PEABODY MUSEUM PUBLICATIONS The Yale Peabody Museum Publications Office supports the communication mandate of the Yale Peabody Mu- seum’s mission by making the Peabody’s research, collections, exhibits and educational activities accessible to the academic community and to the general public. An important part of the Publications Office’s responsibil- ities is publication of the work of curators, staff and research associates, and their colleagues, with an emphasis on research based on the collections of the Yale Peabody Museum. The Publications Office administers the Yale Peabody Museum’s scholarly publishing program, and produces the following titles: BULLETIN OF THE PEABODY MUSEUM oF NATURAL HISTORY In publication since the opening of the Yale Peabody Museum building in 1925, the Bulletin publishes monographs of original research in the natural science disciplines represented by the collections of the Peabody Museum’s curatorial divisions. Volumes are numbered consecutively and appear at irregular inter- vals. The Bulletin incorporates the Bulletin of the Bing- ham Oceanographic Collection, which ceased indepen- dent publication after Volume 19, Article 2 (1967). POSTILLA Published at frequent intervals since 1950, Postilla pre- sents short papers based on original scientific research in the natural science disciplines represented by the col- lections of the Peabody Museum’s curatorial divisions. YALE UNIVERSITY PUBLICATIONS IN ANTHROPOLOGY YUPA embodies the results of research in anthropol- ogy directly conducted or sponsored by the Yale Uni- versity Department of Anthropology and the Peabody Museum Division of Anthropology. Occasionally, other manuscripts of outstanding quality on subjects of special interest to the Faculty of the Department of Anthropology may also be included. Titles range in size from brief papers to extensive monographs, are numbered consecutively as independent contribu- tions, and appear at irregular intervals. SPECIAL PUBLICATIONS The Peabody Museum on occasion produces special publications based on exhibitions, conferences and other Museum projects as appropriate. Publications available in each series, along with ordering information and forms, are available from our website at http://www.peabody.yale.edu/publications, or contact the Publications Office at the address below. GUIDELINES FOR AUTHORS Priority is given to manuscripts by Yale Peabody Museum curators and staff, and to research associates using the Museum’s collections or research facilities. For the YUPA series, members of the Department of Anthropology are given priority over outside submissions. All manuscripts are sent out for external peer review by one or more referees, and are also reviewed internally at the Peabody Museum by the Executive Editor in consultation with the Curatorial Editor-in-Chief for the series, and with one or more members of the appropriate curatorial divi- sion. Manuscripts may go through a single or several revision cycles; revision and major changes are not per- mitted once page proofs are set. At present there are no page charges for accepted manuscripts. Detailed submission guidelines with manuscript preparation instructions are available on our website at http:// www.peabody.yale.edu/publications/Guide.html, or contact the Publications Office at the address below. Interested authors should contact the Executive Editor before submission of manuscripts for possible publica- tion. Address correspondence to: Lawrence F. Gall, Executive Editor Peabody Museum of Natural History Yale University P.O. Box 208118 170 Whitney Avenue New Haven, CT 06520-8118 Phone: (203) 432-9892 Fax: (203) 432-9816 E-mail: lawrence.gall@yale.edu URL: http://www.peabody. yale.edu/publications ISBN 0-912532-47-5