Encrinurus" Variolaris Plexus (Trilobita)."

Home , Encrinuridae, Encrinurus

AMERICAN MUSEUM NornItates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2968, 22 pp., 9 figs. April 25, 1990

Mackenziurus, a New Genus of the Silurian "Encrinurus" variolaris Plexus (Trilobita)

GREGORY D. EDGECOMBE' AND BRIAN D. E. CHATTERTON2

ABSTRACT

Mackenziurus reimeri n. gen. n. sp., from Wen- common ancestry with Balizoma Holloway is a lock strata in the Mackenzie Mountains, North- competing hypothesis. Paraphyletic "Nucleurus" west Territories, Canada, is designated type species Ramsk6ld groups the sister taxa of Balizoma, ofa new genus ofthe "Encrinurus" variolaris plex- Mackenziurus + Fragiscutum, and Frammia Hol- us. Mackenziurus n. gen. also includes Wenlock tedahl. A silicified growth series for Mackenziurus species from Wisconsin-Illinois and Arkansas. A reimeri provides new data on ontogenetic trans- sister-group relationship with Fragiscutum Whit- formations for the variolaris plexus. tington and Campbell is suggested; more recent INTRODUCTION The systematics ofSilurian encrinurine tri- whereas E. (Nucleurus) Ramsk6ld, 1986a, lobites were set in an explicit genealogical represents the plesiomorphic Llandovery core framework by Strusz (1980) who distin- of the variolaris plexus. The latter group was guished three "phylogenetic plexi" within envisioned as ancestral to restricted Wen- Encrinurus Emmrich, 1844. Ramskold's lock?-Ludlow offshoots (Fragiscutum Whit- (1986a) subgeneric revision of Encrinurus tington and Campbell, 1967; Frammia Hol- formalized these relationships: E. (Encrinu- tedahl, 1914) as well as to geographically rus) and E. (Pacificurus) Ramsk6ld, 1986b, widespread and morphologically conserva- comprise the Euramerican punctatus and tive Wenlock-Ludlow Balizoma Holloway, Australasian mitchelli plexi, respectively, 1980.

' Graduate student, Department of Invertebrates, American Museum of Natural History; and Department of Geological Sciences, Columbia University. 2Professor, Department of Geology, University of Alberta, Edmonton, Alberta, Canada T6G 2E3.

a b c d e f k I m a b c d e f h i i k I m

3 j k I m A

0 0 I ICI

Fig. 1. Cladograms expressing component information retrieved from encrinurine classification (components analysis as outlined by Nelson and Platnick, 1981). A. Derivative cladogram from unsequenced generic list classification (Strusz, 1980: 8; cf. Hamada, 1961; Evitt and Tripp, 1977) predicts no component information additional to general component 0 (Encrinurinae). B. Cladogram expressing hypothesis of phylogenetic relationship for the same terminal taxa (from Strusz, 1980: fig. 9) is more informative, predicting three components. Unresolved polytomies within components 0, 1, and 3 result partly from paraphyletic ancestral taxa (0-Encrinuroides; 1 -Coronocephalus; 3-Encrinurus; see C for example). C. Hypothesis of relationships between terms of component 3 (from Strusz, 1980: fig. 9); Encrinurus ("i") is paraphyletic. Terminal taxa abbreviated as follows: a, Encrinuroides; b, Physema- taspis; c, Cromus; d, Coronocephalus; e, Kailia; f, Senticucullus; g, Erratencrinurus; h, Celtencrinurus; i, Encrinurus; j, Paraencrinurus; k, Batocara; 1, Fragiscutum; m, Frammia.

The present survey offers an assessment of Terminology applied to encrinurine mor- variolaris plexus relationships and taxonomy phology follows Evitt and Tripp (1977), with incorporating methods of phylogenetic sys- the exception that the conventional "axis" is tematics. We have thus attempted to recog- used rather than "rachis." Description of en- nize monophyletic taxa (sensu Hennig, 1966) crinurine early ontogeny is as outlined by based on synapomorphies, and to integrate Edgecombe et al. (1988). The preglabellar lat- these components into the hierarchical struc- eral lobe is termed PL following Howells ture of classification. The information con- (1982), and description of glabellar furrows/ tent (Mickevich and Platnick, 1989) of cur- lobes is as recommended for the revision of rent encrinurine classifications is adversely the Treatise on Invertebrate Paleontology, affected by paraphyletic ancestral taxa (see Part 0 (H. B. Whittington and S. R. A. Kelly, fig. 1). A distinctive cluster of North Amer- unpubl.). Ramskold's (1986a) notations for ican Wenlock species of the variolaris plexus fixigenal "circumocular tubercles" and ter- is herein recognized as a monophyletic tax- minal pygidial pleural ribs, as well as pygidial on and are described as the new genus Mack- "R/P ratio," are adopted. "Major row" and enziurus, with type species M. reimeri n. sp., "inter-row" glabellar tubercles are as intro- from the Wenlock of the Mackenzie Moun- duced by Edgecombe and Chatterton (1987). tains, Northwest Territories, Canada. Certain Orientations use the typical abbreviations for Wenlock species from Illinois, Wisconsin, and sagittal (sag.), exsagittal (exsag.), and trans- Arkansas previously referred to Encrinurus verse (tr.). and Balizoma are reassigned to Mackenziu- rus n. gen. Type and figured specimens of ACKNOWLEDGMENTS Mackenziurus reimeri are housed in the pa- leontological type collections of the Univer- This study was funded by operating grants sity of Alberta (UA). from the Natural Sciences and Engineering 1 990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 3

Research Council of Canada, and Energy, and an adaxially narrowing cranidial anterior Mines and Resources (Canada) to Chatterton border. Although the usefulness of "a vario- and from the Boreal Institute for Northern laris group" has been questioned on phenetic Studies to Edgecombe. We thank Dr. Lars grounds (Temple and Tripp, 1979), mono- Ramsk6ld and Mr. R. P. Tripp for helpful phyly of the variolaris plexus is supported by comments on a preliminary version of this these unique characters; we hold that mono- paper, as well as formal reviews. Thorough phyly is both a necessary and sufficient cri- criticism from Dr. D. J. Holloway forced us terion for a taxon's utility. to reevaluate many of our arguments. Pho- Enlarged L2-L4 lateral glabellar and adax- tographs of a well-preserved Mackenziurus ial fixigenal tuberculation, and reduced genal cephalon from the Racine Dolomite at Wau- spines, are derived characters uniting this watosa, Wisconsin, were kindly provided by plexus with Pacificurus Ramsk6ld. It must be L. Ramsk6ld. These, as well as loan of spec- noted, however, that the occurrence of these imens in the Milwaukee Public Museum col- states in certain Llandovery species with sev- lections (arranged by Dr. R. Watkins) and eral diagnostic characters of the punctatus correspondence with Mr. K. C. Gass, aided plexus (e.g., E. squarrosus Howells, 1982) greatly in the generic diagnosis. cautions that they may be symplesiomorphies, and thus specify a more inclusive group SYSTEMATICS OF THE (Encrinurus sensu Ramsk6ld, 1986a). (That VARIOLARIS PLEXUS hypothesis would demand reversals in derived parts of the punctatus plexus.) Predict- Strusz's (1980) Encrinurus variolaris plex- ing the component variolaris plexus + Pa- us included species referred by Ramskold cificurus at the exclusion of the punctatus (1986a) to Encrinurus (Nucleurus) and Bali- plexus, Encrinurus (cf. Strusz, 1980: 54), zoma Holloway. Fragiscutum and Frammia questions the grouping of Encrinurus, Paci- were perceived as "offshoots," i.e., descen- ficurus, and "Nucleurus" as a unique taxon dants excluded from the plexus. This dis- (i.e., as subgenera of Encrinurus). Although tinction between ancestral and descendant that grade group might reflect the equivocal taxa (Haeckelian ancestry of Rieppel, 1988) relationships of certain Llandovery species, is not endorsed here; the variolaris plexus is it predicts that "Nucleurus" shares indepen- expanded to include all descendants of a dent history with Pacijicurus and Encrinurus unique common ancestor which would itself rather than with other parts of the variolaris be part ofthe group. "Nucleurus, " Balizoma, plexus. This statement [as retrieved from the Fragiscutum, Mackenziurus n. gen., and classification; derivative cladogram (Nelson Frammia inclusively form a monophyletic and Platnick, 1981) in fig. 2a] is incongruent group for which the informal name "vario- with the phylogenetic hypothesis endorsed by laris plexus" is retained. Restricting this group Ramskold (1986a) and herein that "Nucleu- to an ancestral/primitive core (e.g., "Nucleu- rus" is most closely related to other genera rus," with or without Balizoma) but exclud- of the variolaris plexus (fundamental clado- ing Frammia and Fragiscutum (based on their gram in fig. 2b). Following Ramskold's uniquely derived characters) results in a para- (1986a) diagnoses, Encrinurus (the punctatus phyletic group. Schrank's (1972) synonymy plexus) and Pacificurus are treated as separate of Fragiscutum and Reed's (1928) Encrinu- generic taxa herein to eliminate the paraphy- rus variolaris group with Frammia would letic "Encrinurus" created by grouping them eliminate several generic names widely used with plesiomorphic parts of the variolaris by subsequent workers. It seems most infor- plexus (the "Nucleurus" grade). This problem mative to retain existing names for express- of paraphyly is only compounded when we ing relationships between parts of the more entertain excluded (autapomorphic) sister inclusive variolaris plexus. groups for the other "subgenera" ofEncrinu- Characteristic of the variolaris plexus are rus. For example, "E." (Encrinurus) is held reduction ofLI by adaxial merging ofS I with to share most recent common ancestry with the occipital furrow, a forwardly positioned Distyrax Lane, 1988 (Edgecombe and Chat- 1-1 tubercle pair, a relatively broad glabella, terton, in prep.), while monotypic Batocara 4 AMERICAN MUSEUM NOVITATES NO. 2968

E P N Bz Fg Fm E P N Bz Fg Fm

E P Bz Fg Fm E D P Bt Bz Fg M Fm

V3 \00-10' Fig. 2. Cladograms expressing alternative hypotheses of relationships for the variolaris plexus and allied taxa. Terminal taxa are abbreviated as follows: Bz, Balizoma; Bt, Batocara; D, Distyrax; E, Encrinurus; Fg, Fragiscutum; Fm, Frammia; M, Mackenziurus; N, Nucleurus; P, Pacificurus. a. Com- ponent information retrieved from Ramsk6ld's (1986a) classification; component 1 is Encrinurus (with three subgenera). b. Hypothesis of phylogenetic relationships, after Ramskold (1986a); component 1 in the derivative cladogram a (fig. 2a) is incongruent with component 2 (variolaris plexus) in this fundamental cladogram. c. Hypothesis ofrelationships discussed herein; both components are congruent with cladogram b; Nucleurus is paraphyletic. d. Hypothesis of relationships including new (Distyrax Lane, 1988; Mackenziurus n. gen.) and previously excluded (Batocara Strusz, 1980) taxa.

Strusz, 1980, is interpreted as sister group to known. The diagnostic characters of "Nu- Pacificurus (if not part of the ingroup) (see cleurus" are primitive for the more inclusive fig. 2d). Pacificurus and Australian Ludlow variolaris plexus as recognized herein; they Batocara bowningi (Foerste, 1888) share apo- are either shared with outgroups (e.g., Paci- morphic hypostomal characters [long, tri- ficurus, Encrinurus) or approximate the state angular posterior border and conical rhyn- at the outgroup node more closely than do chos, e.g., P. mitchelli (Foerste, 1888)] and other genera ofthe variolaris plexus. It is thus an enlarged pygidium. Similarities to partic- inferred that assigned species are more closely ular Pacificurus species [e.g., coarse tuber- related to excluded genera than all are to each culation as in P. silverdalensis (Etheridge and other. Revision of this paraphyletic group Mitchell, 1916); constricted glabellar stalk as must search for monophyletic ingroup com- in P. rothwellae (Etheridge and Mitchell, ponents and reassign other species to the "de- 1916)] are noteworthy. scendant" genus with which they share apo- The systematics of species assigned by morphies. Ramsk6ld (1 986a) to Nucleurus are problem- The name Nucleurus remains available for atic, in part because many are inadequately a monophyletic group including the desig- 1 990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 5 nated type species, Nucleurus abyssalis (Man- dially flattened middle body of subcircular nil, 1977; see Ramskold and Bassett, 1990, outline; broad rhynchos and maculae incon- for revision). Morris (1988), however, syn- spicuous; posterior border short. Pygidium onymized Nucleurus and Aristobeggia La- subhemispherical in outline, with 6-8 pairs mont, 1978 (type species A. bargainensis La- of pleural ribs and 7-9 axial rings (R/P ratio mont, 1978) with Trippia Lamont, 1978 (type 1.1-1.3); sagittal groove narrow; coarse sag- species T. penkillensis Lamont, 1978). This ittal tubercles typically present on every sec- action was based on Howells' (1982) syn- ond ring; inner margin ofdoublure with pos- onymy of T. penkillensis and A. bargainensis teromedian notch broad and shallow or with Encrinurus mullochensis Reed, 1931, a lacking. species which Ramskold (1986a) referred to ETYMoLoGY: Mackenzie, and Greek oura, Nucleurus. Morris thus accepted the validity tail; referring to the Mackenzie Mountains, of Lamont's taxa, although Howells (1982: where the type species occurs. 27) did not. While Howells' synonymy of T. DISCUSSION: Mackenziurus n. gen. com- penkillensis appears plausible (although not prises a morphologically distinctive, tem- certain until Lamont's types are identified), porally and geographically restricted clade the paraphyletic composition of"Nucleurus" within the variolaris plexus. Uniquely diag- sensu Ramskold (1986a) renders the simple nosing the genus requires that several ce- substitution of Trippia inadequate. Pending phalic characters ofMackenziurus sp. (Tripp revision ofequivocal species, "Nucleurus" is et al., 1977) are identified as plesiomorphies used throughout the following parts of this (i.e., shared with each ofthe outgroups "Nu- paper as a paraphyletic grouping ofprimitive cleurus," Fragiscutum, and Balizoma); these parts of the variolaris plexus. serve to emphasize the many autapomorphies of M. reimeri n. sp. The type species differs from the Wisconsin-Illinois taxon in SYSTEMATIC PALEONTOLOGY its broad rostral plate and frontal glabellar FAMILY ENCRINURIDAE ANGELIN, 1854 lobe, enlarged adaxial fixigenal tubercles op- SUBFAMILY ENCRINURINAE ANGELIN, 1854 posite S I -S2 in large holaspides, and exsagittally aligned L4 tubercle pair (other specific MACKENZIURUS N. GEN. differences are noted below in discussion of TYPE SPECIES: Mackenziurus reimeri n. gen. M. reimeri). Mackenziurus sp. (Tripp et al.) n. sp. from the Delorme Formation (Wen- has a more gradually widening glabella (MPM lock), approximately 10 km east of Ava- 26515, figured by Tripp et al., is distorted), lanche Lake, Mackenzie Mountains, North- four subequal-sized fixigenal tubercles over- west Territories, Canada. hanging the axial furrow, and apparently a OTHER SPECIES: Encrinurus sp. of Tripp et typical variolaris plexus wedge-shaped/nar- al., 1977, from the Middle Wenlock of Illi- row trapezoidal rostral plate. The well- nois and Wisconsin; Balizoma sp. of Hollo- rounded cranidial anterior margin (versus the way, 1980, from the Wenlock of Arkansas. medially flattened margin of M. reimeri n. The specific name Encrinurus laurige, re- sp.) suggests that the anterior margin of the ferred to Tripp et al. (1977) by Emielity and hypostome is probably rounded and less pen- Bradbury (1986) for Niagaran material from tagonal in outline than is that of the type Brookfield, Wisconsin, is a nomen nudum species. These plesiomorphies, however, pro- based on the species left in open nomencla- vide no evidence that either Mackenziurus ture by Tripp et al. (1977). This species of species should be grouped with another genus Mackenziurus will be formally diagnosed in of the variolaris plexus rather than with each a forthcoming work by K. C. Gass, G. D. other. Edgecombe, and L. Ramskold. It might be argued that the distinctive au- DIAGNOSIS: Genus of the variolaris plexus tapomorphies of M. reimeri could warrant of relatively small size. Librigenal field sub- the erection of a monotypic genus, whereas equal in length (exsag.) and height (tr.) to pre- Mackenziurus sp. (Tripp et al.) could be re- cranidial lobe, with a row ofabout six coarse tained in Balizoma or some "ancestral tax- tubercles. Hypostome with inflated but me- on" based on symplesiomorphy. This has 6 AMERICAN MUSEUM NOVITATES NO. 2968

been avoided to minimize paraphyly and re- Campbell, 1967) forces a homoplasy; either dundancy (monotypy). It is more informative this resemblance is convergent, or F. rhy- to broaden the diagnosis ofMackenziurus (i.e., tium's more general (Balizoma/"Nucleu- exclude the unique specific characters of M. rus"-like) adult form is a reversal (both trans- reimeri) to recognize a more inclusive mono- formations are equally parsimonious); phyletic group indicated most clearly by py- 4. inconspicuous maculae; gidial synapomorphies. The diagnostic char- 5. low field of the librigena. Mackenziurus acters of Mackenziurus predict that M. reimeri and Mackenziurus sp. (Tripp et al.) reimeri, Mackenziurus sp. (Tripp et al.), and have a librigenal field about 70 percent as (most doubtfully, due to missing data) Mack- high (tr.) as the lateral border (measured be- enziurus sp. (Holloway) are each other's clos- neath the eye), Fragiscutum has 50-60 per- est relatives, sharing common ancestry ex- cent. Although some Ludlow Balizoma cluding all other encrinurines ofthe variolaris species have a comparably low genal field [B. plexus. Priority is placed on identifying obtusa (Angelin, 1851); specimens figured by monophyletic groups, rather than speculating Schrank (1972) and Ramskold (1986a)], this on whether or not a species group is distinc- is interpreted as a convergent similarity to tive enough to warrant generic ranking. Our Mackenziurus (and Fragiscutum), since it rationale for erecting a new taxon (at the ge- does not appear to be the plesiomorphic con- neric level by convention) is thus to identify dition for Balizoma. Wenlock B. variolaris a monophyletic group and discuss its histor- (Brongniart, 1822) (see Tripp et al., 1977) has ical relations to other taxa. If the evidential a librigenal field taller than the lateral border, (i.e., character) support for grouping Mack- as does Ludlow B. hyperborea (Thomas in enziurus species is judged insubstantial, we Thomas and Narbonne, 1979). Wenlock taxa would caution that the alternative (a lack of of the group most closely allied to B. obtusa any evidence for grouping each species with [following Ramsk6ld's (1986a) species syn- different clades) is surely more suspect. onymy] have a taller genal field than Gotland The search for Mackenziurus' sister group Ludlow samples; northern Canadian speci- within the variolaris plexus reveals conflict- mens figured by Perry and Chatterton (1977, ing character distributions, different charac- 1979, and unpubl.) have a field about 85 per- ters suggesting closest relations to either Fra- cent as high as the lateral border. Accepting giscutum or Balizoma. It is noted, however, the monophyly ofBalizoma and the "obtusa that identifying Mackenziurus as a hypoth- group" sensu Ramskold (1986a), a condition esized monophyletic group does not appear (e.g., tall genal field) shared by B. variolaris to render any existing genus paraphyletic (e.g., and primitive parts of the "obtusa group" is species ofBalizoma sensu Ramsk6ld, 1986a, observed in a doublet ofterminal taxa (sensu share more unique characters among them- Maddison et al., 1984) and is parsimoniously selves than with Mackenziurus). inferred to be the plesiomorphic state for the Mackenziurus is particularly comparable genus. It is observed that American Ludlow to Fragiscutum in the following characters: species allied to B. obtusa also have taller 1. Nonenlargement of PL relative to adax- genal fields (90-95% height of border) than ial tubercles on the cranidial anterior border. Baltic material [e.g., B. indianensis (Kindle Strusz (1980: 39) cited this character as di- and Breger, 1904), Milwaukee Public Mu- agnostic of Fragiscutum (in contrast to most seum 26014, reefal Racine Dolomite, Thorn- "Nucleurus" and Balizoma spp. with en- ton, Illinois; Balizoma n. sp., Yale Peabody larged PL), but it is more general (shared with Museum 16270, Brownsport Formation, near Mackenziurus); Decaturville, Tennessee]. 2. reduced number oftubercles in the adax- Outgroup comparison with inferred prim- ial fixigenal tubercle row (see discussion be- itive states for Balizoma and the Llandovery low); "Nucleurus" grade suggests that these simi- 3. rounded, medially flattened hypostomal larities are apomorphic. This hypothesis of middle body. This similarity between F. gle- unique common ancestry of Mackenziurus bale Campbell, 1967, and M. reimeri (but and Fragiscutum (i.e., sister-group relation- also juvenile F. rhytium Whittington and ship) identifies Laurentia as an area of ende- 1 990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 7 mism, and the two genera have similar tem- known, Fragiscutum is an eastern/southern poral distributions (see below). Ramsk6ld Laurentian endemic. (1986a) suggested that Fragiscutum may be Mackenziurus n. gen. may be most readily descended from an early Llandovery branch distinguished from Fragiscutum by its lower of the variolaris plexus including Encrinurus number of pygidial pleural ribs and, partic- rotundus Mainnil, 1977. This species shows ularly, axial rings, with a resultant low R/P cephalic features which are generally plesio- ratio (1.1-1.3 versus 1.7-2.0). In this char- morphic for the plexus, and pygidial form (16 acter, Fragiscutum more closely resembles the axial rings/8 pleural ribs; R/P ratio about 1.8) primitive condition in "Nucleurus" (while indeed comparable to Fragiscutum (R/P ratio Mackenziurus is similar to Balizoma, as dis- approximately 1.7-2.0). A unique common cussed below). ancestor of Mackenziurus and Fragiscutum The type species, M. reimeri n. sp., reveals may thus have differentiated in the early distinctive autapomorphies differing from Llandovery. There is, however, an apprecia- those of Fragiscutum: ble stratigraphic gap separating the known 1. Almost complete reduction of the hy- occurrence of Mackenziurus (mid-Wenlock) postomal rhychos, and associated (as coap- and Fragiscutum (late Wenlock or early Lud- tative structures) absence ofa U-shaped pos- low) from this hypothesized "stem species." teromedian notch in the inner margin of the We follow Tripp et al. (1977), Strusz (1980), pygidial doublure. Holloway (1980) observed and Ramsk6ld (1986a) in restricting Fragis- that this notch, permitting insertion of the cutum to the type species, F. rhytium from rhynchos into the pygidial doublure during Maine, and F. glebale from Oklahoma (and enrollment, is well developed in species here the Ludlow Brownsport Formation of west assigned to Balizoma and Fragiscutum, while central Tennessee; Edgecombe, in prep.). Edgecombe and Chatterton (1987) further These (late Wenlock?) Ludlow species are suggested a correspondence between these distinguished from other variolaris plexus taxa coaptative structures and the plesiomorphic by the development of only 10 thoracic seg- encrinurine enrollment pattern. Despite ments; low (nonpedunculate), enlarged eyes; modification (i.e., almost complete loss) of and marked adaxial displacement of lateral both ofthese structures in Mackenziurus rei- glabellar lobe tubercles. Figure 3 documents meri, the shallow vincular furrow and dis- allometric change in length of the palpebral tinct marginal flange beneath the anterior lobe relative to length of the cranidium for three or four pygidial pleural ribs are com- Fragiscutum, Mackenziurus, and Balizoma parable to those of Fragiscutum and Bali- species. These data suggest that the large-eyed zoma, and suggest similar cephalic-pygidial condition (a synapomorphy of Fragiscutum) interlocking mechanisms. Mackenziurus sp. resulted from accelerated rate of eye size in- (Tripp et al.) retains a broad, shallow pos- crease (relative to the primitive small-eyed teromedian notch, comparable to that ofFra- state of Balizoma and Mackenziurus; it is giscutum glebale. noted that some Ludlow species ofBalizoma 2. Broader (tr.) rostral plate (possibly an have enlarged eyes; see discussion below). The allometric byproduct ofbroadening the fron- taxonomic significance of the reduced num- tal glabellar lobe). In this character, Mack- ber of thoracic segments in F. rhytium and enziurus reimeri most closely resembles Lud- F. glebale has been questioned (Perry and low Frammia (see cephalon of F. arctica Chatterton, 1979; Holloway, 1980). How- figured by Tripp et al., 1977: pl. 115, fig. 16; ever, since the plesiomorphic state of 11 seg- text-fig. 4). The rostral plate ofF. arctica dif- ments is so fixed and widespread within En- fers from that ofM. reimeri, however, in nar- crinurinae, the uniquely shared state in this rowing ventrally, instead ofbeing subrectan- character (in congruence with the others list- gular. Broadening of the rostral plate is ed above) provides a reliable synapomorphy. regarded as a convergence between these taxa; Reports of Fragiscutum from Greenland a greater number of putative synapomor- (Lane, 1984) and Arctic Canada (Perry and phies suggest that Frammia shares unique Chatterton, 1977; Pojeta and Norford, 1987) history with part of "Nucleurus," such as E. are here recognized as Balizoma. As now ("N"). elegantula (Billings, 1866) (and pos- 8 AMERICAN MUSEUM NOVITATES NO. 2968

3 - O MNackenzlurus relmeri

* Ballzoma spp. E . o Fragiscutum rhytium a) Q * F. glebale 0 0 2 a co 03a n0 a) a. 0 * c .0* -c0- 0 to 0 0) . CD 0 11 a0 a) s -J : 00 0 0 0

0* 0 I a a a 0 2 4 6 8 10 Length of cranidium (mm) Fig. 3. Scatter plot of maximum length of palpebral lobe versus sagittal length of cranidium in taxa of the variolaris plexus. Data for Fragiscutum rhytium and F. glebale from photographs in Whittington and Campbell (1967) and Campbell (1967), respectively; additional specimens of F. glebale measured are Milwaukee Public Museum 26523 (Henryhouse Formation, Oklahoma) and Yale Peabody Museum 16271 (Brownsport Formation, near Decaturville, Tennessee). Mackenziurus reimeri from section Av- alanche Lake Five, 58-60 m above base, and section Avalanche Lake Seven, 27 m above base. Balizoma spp. from section Avalanche Lake Four, 126 m above base [= Balizoma dimitrovi (Perry and Chatterton, 1979) of Edgecombe and Chatterton, 1987: fig. 7]. sibly E. globosus Maksimova, 1962), and also ering the outgroups Pacificurus and Encrinu- primitively had a narrower rostral plate. rus, five adaxial fixigenal tubercles would ap- Characters supporting this grouping include pear to be plesiomorphic for the variolaris a rounded genal angle, short and wide gla- plexus; species of "Nucleurus" have three to bella, lateral lobe tuberculation much en- five, and Balizoma usually has five or six. larged relative to low, dense glabellar tuber- Species of Mackenziurus and Fragiscutum cles, strong reduction of LI dorsally, and develop two (some M. reimeri), three (some reduction/loss of pygidial sagittal tubercula- M. reimeri; F. glebale), or four (Mackenziu- tion. The (primitive) presence of a typical rus sp. ofTripp et al.) tubercles in the adaxial narrow rostral plate in Mackenziurus sp. fixigenal row. F. rhytium is more variable, (Tripp et al., 1977), which shares several apo- with holaspides having three, four, or (rarely) morphic characters with M. reimeri, suggests five adaxial fixigenal tubercles. Growth series that broadening of the rostral plate occurred for Balizoma spp. and Mackenziurus reimeri independently within Frammia and Mack- indicate that development of this row is ini- enziurus. tiated in the meraspid period (fig. 4), although 3. Modification of the adaxial fixigenal tu- clusters of tiny spinose denticles along the bercle row, with two coarse tubercles over- axial furrow opposite S 1 and S2 in early me- hanging the axial furrow opposite S1 and S2 raspides are a precursor to tubercle devel- in largest holaspides (figs. 6: 8; 7: 1). Consid- opment. Edgecombe and Chatterton (1987) 1990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 9

0~~~~~~~~~~~~~~ U') ~ O 0c?

0~~~~~~ < 00~~~~~~~~~~~~ 1 0~~~~~~~~0

0 ~ 0 W 0

Fig. 4. Comparative ontogeny ofthe genal field and phylogenetic relationships in the variolaris plexus. Adaxial fixigenal tubercles in meraspides (bottom row, approximately x 1 1) and large holaspides (top row, approximately x 5.5) are shaded. Apomorphic character transformations: A. reduction of adaxial fixigenal row to 2-4 (rarely 5) tubercles; B. enlargement of eyes (accelerated rate of allometric growth); C. adaxial row in large holaspides dominated by coarse tubercles opposite SI -S2. Based on originals of figs. 6: 8, 7: 1, and 8: 5 (Mackenziurus reimeri); Whittington and Campbell, 1967: pl. 12, figs. 1, 15 (Fragiscutum rhytium); Campbell, 1967: pl. 8, fig. 8 (F. glebale); Edgecombe and Chatterton, 1987, fig. 6G and unpublished (Balizoma spp. cf. B. dimitrovi). noted that this tubercle row in Balizoma is composed of five or six (rarely seven) sub- formed as a unified series distinct from other equal-size tubercles in adults. Apomorphy genal tubercles in Ramskold's (1986a) "cir- "A" in figure 4 is conditional upon a second cumocular tubercle ring." Balizoma meras- outgroup having five or more adaxial fixi- pides show a rather uniform-size row ofsmall, genal tubercles; this applies to "Nucleurus" granulose tubercles opposite the glabellar fur- rotundus (Miinnil, 1977), which Ramsk6ld rows (and farther forward opposite the abax- (1986a) suggested was near the ancestry of ial end of the preglabellar furrow), although Balizoma and Fragiscutum. However, it earlier onset ofdevelopment results in slight- would be more parsimonious that Balizoma ly larger tubercles opposite S1 and S2. This is autapomorphic using four-tubercle "Nu- adaxial fixigenal row enlarges through ontog- cleurus" as second outgroup. eny relative to other fixigenal tubercles, and In contrast to the inferred plesiomorphic 10 AMERICAN MUSEUM NOVITATES NO. 2968

state, a reduced number of adaxial fixigenal genus is unknown. There is a possibility that tubercles in Fragiscutum and Mackenziurus reduction in the number ofthoracic segments is expressed in the meraspid period by prom- to 10 originated in a common ancestor ofthe inence of tubercle pairs opposite S1-S2, but Mackenziurus-Fragiscutum clade. tubercles opposite S3 being either absent (in Species here referred to Mackenziurus n. M. reimeri) or small (in F. rhytium). This gen. were assigned to Balizoma by Holloway includes meraspides at degrees certainly more (1980) and Ramsk6ld (1986a). Shared char- advanced than those at which Balizoma has acter states are partly plesiomorphic for the developed five tubercles in the row (see fig. variolaris plexus (or perhaps a Mackenziurus- 4). Further development of the adaxial row Fragiscutum-Balizoma clade), and reflect de- is often completely arrested in Mackenziurus scent of both genera from Llandovery lin- reimeri, in which tubercles opposite S1 and eages grouped as "Nucleurus." It is recog- S2 coarsely enlarge through the growth series, nized that Mackenziurus shows greater but anterior or posterior pairs are usually similarity in its low R/P pygidial form to Ba- either small or absent in large holaspides. This lizoma (notably B. variolaris) than to Fra- paedomorphic state differs from the more giscutum (e.g., see Schrank, 1972: pl. 13, fig. primitive condition of F. rhytium, in which 8; Tripp et al., 1977: pl. 113; Thomas, 1981: a row of three to five subequal-size tubercles pl. 18, fig. 2). As well, both taxa typically bear also includes pairs positioned opposite S3, sagittal tubercles on every second axial ring. and sometimes opposite the preglabellar and/ However, as Ramskold (1986a) noted, sag- or occipital furrows, more comparable to ittal tubercle spacing is correlated with the "Nucleurus" and Balizoma. Holaspides of number of axial rings; it is thus not feasible Fragiscutum glebale Campbell, 1967, have to recognize these as independent similari- three nodular tubercles opposite S1-S3. A ties. The criterion in postulating Fragiscutum correlation is apparent between expansion of as sister group to Mackenziurus is parsimony; the greatly enlarged eyes to immediately ad- a greater number of apparent synapomor- jacent to the axial furrow and reduction in phies unite these taxa than unite Mackenziu- the area ofthe genal fields and reduced adax- rus with Balizoma (e.g., unique hypostomal ial fixigenal tuberculation (i.e., absence ofan- characters). The similarity ofthe Mackenziu- terior tubercles sometimes developed in F. rus pygidium to that of B. variolaris, how- rhytium). It is thus plausible that accelerated ever, provides a competing hypothesis to test rate of eye size increase (fig. 3) imposed a with additional characters. spatial constraint on ontogenetic develop- Characters by which Mackenziurus may be ment oflate-forming anterior tubercles in the distinguished from Balizoma include the fol- adaxial fixigenal row. lowing: Mackenziurus may also be distinguished 1. relatively small size; from Fragiscutum by its primitive character 2. reduction of the adaxial fixigenal tuber- states of smaller eyes on short eye socles, cle row to two to four tubercles (versus, typ- symplesiomorphies with Llandovery "Nu- ically, five or six subequal-size tubercles in cleurus." These states are also retained in Balizoma); Wenlock species of Balizoma, such as the 3. subpentagonal hypostome with reduced type species, B. variolaris (Brongniart). They rhynchos and maculae, and rounding/flatten- are apparently plesiomorphic conditions for ing ofthe middle body (versus rhomboid hy- the group of species allied to B. obtusa (An- postomal outline, prominent rhynchos and gelin) [see fig. 3 for eye size of Wenlock B. maculae, and inflated, subovate middle body spp. cf. B. dimitrovi; see Perry and Chatter- in Balizoma). Maculae are bulbous in B. var- ton, 1977: pl. 4, fig. 26; 1979: pl. 74, fig. 21, iolaris (Tripp et al., 1977: pl. 13, figs. 5, 10); for eye socle]. The large, sessile eyes of some they are distinct in Wenlock B. dimitrovi? Ludlow "obtusa group" species (e.g., B. ob- (Perry and Chatterton, 1977: pl.4, fig. 8) and tusa Form B of Ramsk6ld, 1986a) are thus Ludlow B. dakon Snajdr, 1983 (Snajdr, 1985: inferred to be homoplastic with Fragiscutum. pl. 12, fig. 9), B. rosensteinae (Tripp et al., The number of thoracic segments in the new 1977: pl. 115, fig. 4), and B. obtusa (Schrank, 1 990 EDGECOMBE AND CHATTERTON: MACKENZIUR US I1I

1972: pl. 11, fig. 8a). As suggested above, coarse glabellar and pygidial sagittal tuber- some ofthese differentia may be restricted to culation. From the ambiguous evidence cur- M. reimeri; rently available, Langgonia is regarded as 4. a low, more sparsely tuberculate libri- Encrinuridae incertae sedis, with some con- genal field (typically bearing two "rows" of vergent similarity to Laurentian Mackenziu- tubercles in Balizoma); rus. 5. more round-margined pygidium, usually with fewer axial rings and pleural ribs (7-9 Mackenziurus reimeri n. sp. rings in Mackenziurus n. gen. versus 9-11 in Figures 5-8 Balizoma variolaris, 10-15 in B. obtusa and allied species; 6-8 ribs in Mackenziurus ver- TYPES: Holotype pygidium UA 7843 (fig. sus 7-8 in B. variolaris, 8-12 in B. obtusa and 6: 14, 16-18), and paratypes UA 7836-7842, allied species). As a result, the pygidial R/P 7844-7853, 7857, 7861. Type locality is sec- ratio is typically lower (see Ramskold, 1986a: tion Avalanche Lake Five, 58-60 m above text-fig. 3); and, base (Delorme Formation, Wenlock; see Over 6. posteromedian notch in inner margin of and Chatterton, 1987). pygidial doublure is shallower or absent. OCCURRENCE: Type locality; Avalanche The recognition of Mackenziurus n. gen. Lake Four, 136-138 m above base; Ava- thus restricts the degree of morphological lanche Lake Seven, basal 70 m of measured variation within Balizoma as diagnosed by section (occurrences at AV7-0, 27, 38, and Ramskold (1986a). 70 m above base). West Malaysian late Llandovery or early DIAGNOSIS: Species of Mackenziurus with Wenlock Langgonia Kobayashi and Hama- Ll present as low, discontinuous lobes; max- da, 1971, shows at least superficial resem- imum glabellar convexity forward on frontal blance to Mackenziurus. Holloway (1981) lobe, which bears small paired lateral lobe convincingly argued that this genus, origi- tubercles; I-1 present; genal spines bluntly nally described in a new monotypic subfam- pointed; axial furrow diverges abruptly in ily of Dalmanitidae, lacks synapomorphies front of enlarged fixigenal tubercle pair op- of the more-inclusive Phacopina and shares posite S2. Librigenal field bears several unique characters with Encrinuridae. These smaller adventitious tubercles in addition to include a longitudinal median glabellar furrow of 6-7 coarse tubercles. Lateral border row, straight (tr.) lateral glabellar furrows, tuberculation weak. Rostral plate wide, sub- short L1-L3, and librigenal precranidial lobes; rectangular. Eye opposite S3. Pygidial length derived characters of Encrinurinae (or 55-65 percent ofwidth, with 7-8 pleural ribs, subgroups) include a forward-expanding gla- 8-9 (rarely 10) axial rings; axial tuberculation bella, tuberculiform L2-L3, adaxial fixigenal includes alternating small and large sagittal lobes opposite the lateral glabellar furrows, tubercles and paired tubercles abaxial to sag- and more than five aspinose pairs ofpygidial ittal groove on several anterior rings; inner pleurae. Its pygidial morphology (e.g., hemi- margin of pygidial doublure broadly para- spherical shape with seven rib pairs, and sev- bolic, lacking posteromedian notch. en or eight axial rings) and overall confor- ETYMOLOGY: After Mr. Weldon Reimer of mation of the librigena (especially in small- the Canada Tungsten mining operation, eyed L. araiorachis Kobayashi and Hamada) Tungsten, Northwest Territories. are similar to those of Mackenziurus. The DESCRIurTION: Cranidial length 55-65 per- Malaysian species obviously possess many cent of width; L4 140-150 percent of width autapomorphies (e.g., the transverse furrow of L3, and 120-130 percent of width of oc- on the fixigenal field separating the adaxial cipital ring. Glabellar length 110-120 percent lobes), but these are uninformative in dis- ofwidth across L4, with maximum convexity cerning outgroup affinities. Plesiomorphies positioned forward on frontal lobe. S1 di- which would, however, exclude Langgonia rected posteromedially and merging with oc- from the variolaris plexus include its well- cipital furrow behind I-I tubercle pair. Lat- developed LI, transverse S1, and absence of eral lobe L4 bearing a pair of exsagittally 12 AMERICAN MUSEUM NOVITATES NO. 2968

Fig. 5. Dorsal view of a reconstruction of cephalon and pygidium of Mackenziurus reimeri n. gen. n. sp. aligned tubercles (small adaxially positioned pair (PL) not relatively enlarged; cranidial anterior tubercle, and larger posterior tuber- anterior margin straight sagittally, defining cle), contrasting with single rounded tuber- wide (tr.) rostral suture. Adaxial fixigenal cles on L2-L3 distinctly larger than more margin almost straight anterior to enlarged medial glabellar tubercles. Glabellar tubercu- subconical tubercles overhanging axial fur- lation includes I-1; II- 1, 2; iii-0; III-1, 2(F), row opposite SI -S2. Palpebral lobe with an- 3; frontal lobe bears distinct IV-1, 2, 3F/R; terior edge close to axial furrow, but sepa- v- I, 2; Vl- 1, and smaller inter-row tubercles. rated from it by one small tubercle. Fixigenal Occipital ring and cranidial posterior/pos- field pitted, densely tuberculate, including terolateral border tuberculate in all but larg- distinct CT 1 -CT4 ofsubequal size to coarsest est holaspides (see Ontogeny). Preglabellar glabellar tubercles; CT I adjacent to posterior furrow shallow; adaxially narrowing anterior limit of palpebral lobe (opposite front of L2 border bears 10-14 tubercles, abaxialmost to back of L3).

Fig. 6. Mackenziurus reimeri n. gen. n. sp., Delorme Formation, Mackenzie Mountains, Northwest Territories, Canada. All specimens from section Avalanche Lake Five, 58-60 m above base (Over and Chatterton, 1987) except 13 (from section Avalanche Lake Four, 138 m above base), x 10 except where noted otherwise. 1. UA 7836, dorsal view of librigena. 2-4, 6. UA 7837, dorsal, ventral, lateral, and anterior views of cranidium. 5. UA 7838, dorsal view of thoracic segment. 7-10. UA 7839, anterior, dorsal, ventral, and lateral views of incomplete cranidium and librigena. 11. UA 7840, dorsal view of cranidium. 12. UA 7841, lateral view of cranidium. 13. USA 7842, dorsal view of pygidium. 14, 16- 18. UA 7843, ventral (x 5), lateral, dorsal, and posterior views of holotype pygidium. 15. UA 7844, external view of librigena. 1990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 13

ZC_~~~~~~~~~1

Z .*l_l 14 AMERICAN MUSEUM NOVITATES NO. 2968 1 990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 15

Librigenal field pitted; height (tr.) about 70 row and weakly flexed downward; anterior percent of lateral border, very slightly longer border furrow rather broad, shallow, and with (exsag.) than densely tuberculate precranidial elongate pits anterolaterally at change in cur- lobe. Anterior furrow shallow. Eye socle about vature of margin; anterior margin rather 60 percent as high as visual surface, defined straight sagittally, with abrupt posterolateral by distinctly incised furrows. reorientation on straight anterolateral mar- Cephalic apodemes expanded distally, di- gins so that outline of hypostome is subpen- rected inward at outer end of occipital fur- tagonal (particularly in small growth stages). row, SI, S2, and S3; SI -S2 apodemes large, Anterior wing large, positioned anterior to longitudinally ovate and subspherical re- midlength of middle body, shorter in height spectively; occipital and particularly S3 apo- than middle body, and with short wing pro- demes small. In ventral view, S3 short, an- cess near upturned anterior edge. Posterior teromedially directed; straight S2 weak but wing positioned almost midway between an- distinct across much of glabella. Cephalic terior wing and posterior margin of middle doublure uniformly broad beneath librigenal body. Doublure narrow, but widening slight- lateral border but narrowing gently beneath ly posteriorly and extending almost halfway anterior border - precranidial lobe; cranidial to posterior border furrow, inner margin doublure narrows abruptly adaxial to genal lacking anteriorly projecting posteromedian angle. Shallow vincular furrow along ventro- tongue (as present in Fragiscutum rhytium lateral margin of librigenal doublure corre- and many other encrinurines). Posterolateral sponding with marginal flange beneath an- border narrow, widening gradually postero- terior three or four pairs of pygidial pleural medially; border furrow deep, narrow, and ribs. Pygidial doublure broad, widening shallowing sagittally. Margin flexed back- slightly posteromedially. ward posterolaterally with distinct length- Rostral plate (shape inferred from course ening of broadly rounded posterior border, of rostral, connective, and hypostomal su- angled ventrally, comprising less than 10 per- tures, and from specimens with complete li- cent of sagittal length of hypostome. brigenae articulated with cranidium) subrect- Number of thoracic segments unknown. angular, narrowing very slightly ventrally, Axial ring comprising about 35 percent of width (tr.) approximately 150 percent of sag- width of thorax, bowing gently forward me- ittal length. dially and adjacent to rather deeply impressed Hypostome with large middle body, slight- axial furrow; articulating half-ring about 40 ly broader than long, overhanging lateral bor- percent of sagittal length of axial ring; ring der behind anterior wing. Rhynchos weakly moderately convex, rounded, or weakly flat- developed as faint swelling on anterior mar- tened sagittally. Inner half of pleurae sloping gin ofmiddle body, defined abaxially by very gently inward and fairly straight, outer half shallow, strongly posteriorly divergent fur- strongly curved downward; pleural band rows, and incorporating about 60 percent of gently widening abaxially toward abrupt width of middle body opposite front of an- change in slope, abaxially narrowing and terior wings. Maculae forming small, very low, gently flexing backward with a weakly con- ovate swellings on posterolateral margin of cave anterior margin; pleurae terminating in middle body. Anterior border uniformly nar- a bluntly rounded point. Axial articulating

Fig. 7. Mackenziurus reimeri n. gen. n. sp., Delorme Formation, Mackenzie Mountains, Northwest Territories, Canada. All specimens from section Avalanche Lake Five, 58-60 m above base (Over and Chatterton, 1987) except 3, 4, and 6 (from Avalanche Lake Seven, 27 m above base), x 10 except where noted otherwise. 1, 2, 5. UA 7845, dorsal, dorsolateral, and lateral views ofcranidium. 3,4,6. UA 7846, dorsal, lateral, and anterior views of cranidium. 7, 8. UA 7847, dorsal, and anterior views of thoracic segment. 9-11. UA 7848, lateral, ventral, and anterior views of hypostome. 12-14. UA 7849, anterior, ventral, and posterior views of thoracic segment, x 5. 15. UA 7850, dorsal view of cranidium. 16, 19- 21. UA 7851, posterior, dorsal, ventral (x 5), and lateral views of pygidium. 17. UA 7852, ventral view of hypostome. 18. UA 7853, dorsal view of pygidium. 16 AMERICAN MUSEUM NOVITATES NO. 2968 1 990 EDGECOMBE AND CHATTERTON: MACKENZIURUS 17 processes prominent, transversely ovate; an- tributed the small number of axial rings and terior articulating flange uniformly narrow pleural ribs to the small size of known spec- and about 30 percent length of pleural band imens, apparently implying that these were in inner half of pleurae, widening abaxially juveniles. This possibility is somewhat weak- and projecting forward abaxial to fulcrum to ened by comparison with similarly small adult form a prominent bluntly pointed facet. holaspides ofM. reimeri, in which a full com- Doublure of axial ring gently widening sag- plement of eight pairs ofpleural ribs may be ittally to almost 50 percent of length of seg- attained in pygidia of length as little as 25 ment; apodemes positioned slightly inward percent of the largest known specimens, and ofaxial furrow, directed strongly inward and by the ontogenetically advanced appearance weakly downward; posterior recess well de- of the smooth pygidial margin (lacking spi- veloped in inner half of pleural doublure. nose rib terminations characteristic of early Downturned outer part of the first pair of ontogeny in the type species and other taxa pygidial pleural ribs inclined at 70-80o; in- of the variolaris plexus). terrib furrows widened toward change in slope, then narrow distally at expanded tips ONTOGENY of pleural ribs; posterior two or three pairs ofinterrib furrows indistinctly continuous to Silicified material ofMackenziurus reimeri ventral margin of pygidium (distal part of n. gen. n. sp. includes a relatively complete pleural ribs fused). All but the anterior one growth series for most sclerites, recording on- or two axial rings with a narrow sagittal togenetic development from the protaspid groove; in lateral view, axis is gently convex through holaspid periods. and weakly sloping for most of length but A protaspis ofthis species figured by Edge- with abrupt steep slope posteriorly. combe et al. (1988) as "Balizoma sp." [var- DIscussIoN: Mackenziurus reimeri n. gen. iolaris plexus n. gen. n. sp.] is reillustrated n. sp. can be distinguished from Mackenziu- herein (fig. 8: 1, 2) to document ontogenetic rus sp. (Tripp et al., 1977), from the Middle changes across the protaspid-meraspid tran- Wenlock of Illinois and Wisconsin, by ce- sition. Trends in postprotaspid ontogeny are phalic characters listed in the species diag- generally comparable to those documented nosis above. Also, the new species has a taller, by Whittington and Campbell (1967) in Fra- more strongly sloping pygidial profile, lack- giscutum and by Edgecombe and Chatterton ing a posteromedian notch in the doublure. (1987) in Balizoma. These include the fol- Mackenziurus pygidia from the St. Clair lowing: Limestone of Arkansas described by Hollo- 1. Near isometry in glabellar length/width way (1980) as Balizoma sp. are more elongate (fig. 9A) and width ofthe frontal lobe relative and slope more strongly in lateral view (no- to L1-L3 (fig. 9B); slight increase in both ra- tably postaxially) than those ofM. reimeri n. tios with increasing size. Protaspis with sub- sp. The Arkansas species has only six pleural ovate frontal lobe, slightly broader than sub- ribs that are more strongly turned downward equal (in length and width) L2-L3, with and backward abaxially, the first rib being straight, transglabellar S1-S3. Growth series almost vertically inclined; seven axial rings; demonstrating relative lengthening offrontal and a shallower sagittal groove. Holloway at- lobe, becoming subhemispherical in outline,

Fig. 8. Mackenziurus reimeri n. gen. n. sp., Delorme Formation, Mackenzie Mountains, Northwest Territories, Canada. All specimens from section Avalanche Lake Five, 58-60 m above base (Over and Chatterton, 1987) except 6, 10, and 11 (from section Avalanche Lake Four, 138 m above base). All figures except 6 and 10-12 are scanning electron micrographs. 1, 2. UA 7829, dorsal and ventral views of protaspis, x 75. 3. UA 7854, dorsal view of small (meraspid) cranidium, x 50. 4. UA 7855, external view of small librigena, x 50. 5. UA 7856, dorsal view of small cranidium, x 30. 6, 10, 11. UA 7857, dorsal, lateral (both x 5), and ventral (x 10) views of hypostome. 7. UA 7858, ventral view of small hypostome, x 50. 8. UA 7859, ventral view of hypostome, x 30. 9. UA 7860, dorsal view of small cranidium, x30. 12. UA 7861, external view of librigena, x5. 13. UA 7862, dorsal view of small cranidium, x 30. 18 AMERICAN MUSEUM NOVITATES NO. 2968

5U-4

E cd

C',

N O

4~~~0 ~~~~~~~~~~~-~

* . . .~~~~~~~~~~~~~~~~~~~~C .

0 a U,~~~~~~~~~~~~~~~~~~~~~~t~C0 A 0 s0u Er.. in~~~~0 'u (N E E Lfpt lIPBA -0 ' CY 0 E Z~~~~~~. 4 U-- 0 0cid-' C')~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,

.0 4 0 K 0~~~ ~ ~ ~ ~ ~ ~ ~o.~~~~~ ~ ~ n0C - ri

* ~~~~~~~~~~~~~~~~~~00 O Ei.N 00 6~~~~~~~~~~~~~~~~~~~~~~~~~0

0~~~ ~~~~~~~' * .~~~~~~~~~~~~~~~~~~~cs

0 -~~~~~~~~-0 I!

00 0~~~~~~~~~~~*

0 1-~~~~~~~~4 * U~~~~~~~~~~~~~~~~~~C)ICd '

C')N C~~~~~~~0 C) C

04 1990 EDGECOMBE AND CHATTERTON: MACKENZIUR US 19 with S3 reduced to short anteromedially ori- part of thoracic axial rings are weakly devel- ented furrow; S2 remaining straight, but be- oped in other taxa of the variolaris plexus, coming shallow adaxially, SI reorienting pos- most prominently in Frammia. In the type teromedially to merge with occipital furrow species, F. arctica (Salter, 1852; see Bolton, behind I-I tubercle pair; LI greatly reduced 1965), depressed anterior band very distinct- to low, discontinuous lobes in large holas- ly separated from elevated posterior band and, pides (fig. 7: 1, 2), although continuous LI/ as noted by Tripp et al. (1977), simulating straight SI sometimes retained into holaspid L1 [cf. Gass and Mikulic's (1982) description period (fig. 6: 2). of LI fused to occipital ring]. 2. Glabellar tubercle development typical 4. Ten to fourteen tubercles developing on for the variolaris plexus, as documented for cranidial anterior border. Protaspis has two Balizoma by Edgecombe and Chatterton blunt spine pairs (fig. 8: 1, 2); these are elon- (1987). Protaspis with IV-1 tubercle pair po- gate spines in small meraspides (fig. 8: 3), sitioned posterolaterally on the frontal lobe; with small third pair developing sagittally. V-1 is present in early meraspides. A small Smaller tubercles inserting between these pairs (meraspid) cranidium (fig. 8: 3) showing dis- increasing in relative size through the holas- tinct I-i; 11- 1; 1II- 1; IV- I (enlarged); V- 1; VT- I pid period, resulting in more uniformly sized spine pairs. At this developmental stage, lat- row of flattened tubercles in adults. eral glabellar lobes nontuberculate (bearing 5. Stout, elongate anterior fixigenal spine short spines), and inter-row tubercles indis- ofprotaspis (subequally long as the posterior tinct. Additional major row tubercles added fixigenal spine) greatly reduced in early me- abaxially through the meraspid period, on L3 raspides, and forming rounded CT4 tubercle (111-2) and frontal lobe (IV-2; V-2). Small in holaspides. Midfixigenal spine also re- inter-row tubercles including iii-0 (and oc- duced, migrating posteriorly in front of pos- casionally ii-0), and abundant on frontal lobe. terior fixigenal spine. Posterior fixigenal spine Rounded tubercles developing on L2-L4 lat- remaining elongate in meraspides, reorient- eral lobes, finely denticulate in earlier stages. ing backward; progressive reduction occur- Large holaspides recording flattening of tu- ring through holaspid period, with genal an- bercles through this period (spinose in small gle in large individuals only bluntly pointed. growth stages), enlargement of tubercles on 6. Small, spinose fixigenal tubercles devel- L2-L3 lateral lobes relative to more medial oping along axial furrow opposite S1 and S2 tubercles, and allometric trend toward sub- (with clusters of tiny denticles in these po- equal size of major row pairs (IV-1 notably sitions in early growth stages; fig. 8: 3, 5), larger in early stages), and enlargement ofin- coarsening through the growth series, assum- ter-row tubercles. ing subconical shape, nearly buttressing axial 3. Occipital ring and cranidial posterior/ furrow (fig. 7: 1). posterolateral border prominently tubercu- 7. Small cranidia demonstrate primary de- late in meraspides and small holaspides, but velopment of CTI/CT2/CT3 fixigenal "cir- smooth or only faintly tuberculate in large hol- cumocular" spines, with torulus becoming aspides. Occipital tuberculation including indistinct in early meraspides. Fixigenal field large median tubercle and smaller pair pos- undergoing increased tuberculation; later- terolaterally. Occipital ring in large holas- forming tubercles enlarging to size only pides (fig. 7: 1, 2) broad and conspicuously slightly smaller than primary "circumocular" furrowed; broad, shallow furrow curving pos- tubercles. terolaterally from abaxial edge of anterior 8. Palpebral lobe, bearing short spines in margin and extending straight across mid- early stages, migrating backward from posi- length ofoccipital ring, separating slightly de- tion opposite L4 through much ofgrowth se- pressed anterior band from raised posterior/ ries to opposite S3 in large holaspides. posterolateral band. Longitudinal swellings 9. Hypostomal middle body (widest an- on abaxial part ofoccipital ring are compara- terolaterally in small stages and more bluntly bly developed adjacent to axial furrows on rounded anteriorly than posteriorly) flatten- thoracic segments. Such transverse furrowing ing, becoming rounder in outline, and inflat- of occipital ring and swelling ofanterolateral ing laterally to overhang lateral border. Small 20 AMERICAN MUSEUM NOVITATES NO. 2968 pair of denticles on lateral border and two rences of congenerics in other geographic pairs on posterolateral border present only in areas; Mackenziurus sp. from Illinois and early growth stages (fig. 8: 7). Wisconsin, for example, is known from in- 10. Librigenal lateral border bearing nu- terreef dolomites. These ecological differ- merous short spines in small stages (fig. 8: 4), ences might have been significant in effecting reducing to faintly tuberculate adult state. the spatial differentiation of Mackenziurus. Tuberculation on lateral border preceding that Further, they attest to the inherent limits in on precranidial lobe and field, more densely construing taxa (genealogical entities) as co- tuberculate in later ontogeny. herent ecological entities (cf. Eldredge, 1989). 11. Approximately rectilinear increase in The facies distribution of Balizoma species pygidial length/width occurring through hol- provides a similar example. Balizoma cf. B. aspid period (fig. 9C). Other trends through dimitrovi (Perry and Chatterton, 1979) is this period (early holaspides already possess- abundant in sympatry with Mackenziurus ing seven or eight pairs of pleural ribs and reimeri. Ramsk6ld (1985), however, ob- eight or nine axial rings) include the follow- served the morphologically similar B. obtusa ing: reduced spinosity of free rib termina- (Angelin, 1851) from Gotland to be associ- tions, with smoothing of pygidial margin; ated with high-energy reefal calcarenite fa- deepening and broadening ofsagittal groove; cies. This emphasizes that differing ecologies shortening of sagittal spines to rounded tu- are not necessarily correlated with notable bercles, and development of axial tubercle phenotypic differences [indeed, Ramsk6ld pairs lateral to sagittal groove; loss ofpleural (1986a) discussed the problem of separating tubercles; deepening and broadening of axial B. dimitrovi from B. obtusa on morphological and interrib furrows; and, steepening slope grounds]. ofdistal part ofpleurae. Change in axial width relative to pygidial width nearly isometric (fig. REFERENCES 9D). Angelin, N. P. ENVIRONMENT AND ECOLOGY 1851. Palaeontologia Svecica. I: Iconographia crustaceorum formationis transitionis. Variable stratigraphic occurrence ofMack- Fasc. 1. Holmiae. enziurus reimeri in sections representing a 1854. Palaeontologia Scandinavica. I: Crus- shelf-to-basinward depth gradient in the Av- tacea formationis transitionis. Fasc. 2. alanche Lake area (see Over and Chatterton, Lund. 1987) demonstrates facies control on the dis- Billings, E. tribution of this species. M. reimeri occurs 1866. Catalogues of the Silurian fossils of the through 70 m of thinly bedded, dark gray, Island ofAnticosti, with descriptions of some new genera and species. Geol. argillaceous micrite in Avalanche Lake Sev- Surv. Can.: 1-93. en, the most distal sampled section. This Bolton, T. E. species is represented through progressively 1965. Trilobites from Upper Silurian rocks of narrower stratigraphic intervals in equivalent the Canadian Arctic Archipelago: En- horizons in shelfward (i.e., shallower) sec- crinurus (Frammia) and Hemiarges. tions. It ranges through only 4 m in section Bull. Geol. Surv. Can. 134: 1-14. Avalanche Lake Five (thinly bedded micrite Brongniart, A. underlying a transgressive graptolitic shale), 1822. Histoire naturelle des Crustaces fossiles, through 2 m in Avalanche Lake Four (dark- sous les rapports zoologiques et geolo- colored micrites with coarse bioclastic al- giques, savoir les Trilobites par Alex- andre Brongniart. Les Crustaces propre- lochthonous interbeds),and is unrepresented ment dits par A.-G. Desmarest. Paris. in more proximal Avalanche Lake One and Campbell, K. S. W. Two. The restricted spatiotemporal distri- 1967. Trilobites of the Henryhouse Forma- bution and lithofacies association of M. rei- tion (Silurian) in Oklahoma. Bull. Okla. meri attest to relatively deep-water ecology Geol. Surv. 115: 68 pp. ofconstituent organisms, in outer shelf/slope Edgecombe, G. D., and B. D. E. Chatterton muds. This provides a contrast with occur- 1987. Heterochrony in the Silurian radiation 1 990 EDGECOMBE AND CHATTERTON: MACKENZIUR US 21

of encrinurine trilobites. Lethaia 20: Howells, Y. 337-351. 1982. Scottish Silurian trilobites. Palaeontogr. Edgecombe, G. D., S. E. Speyer, and B. D. E. Soc. Monogr. (London) 561: 1-76. Chatterton Kindle, E. M., and C. L. Breger 1988. Protaspid larvae and phylogenetics of 1904. The stratigraphy and paleontology ofthe encrinurid trilobites. J. Paleontol. 62: Niagara ofnorthern Indiana. Ann. Rep. 779-799. Indiana Dept. Geol. Nat. Res. 28: 428- Eldredge, N. 486. 1989. Macroevolutionary dynamics: species, Kobayashi, T., and T. Hamada niches, & adaptive peaks. New York: 1971. Silurian trilobites from the Langkawi Is- McGraw-Hill. lands, West Malaysia, with notes on the Emielity, J., and D. P. Bradbury Dalmanitidae and Raphiophoridae. 1986. List of Silurian trilobites of the United Geol. Palaeont. Southeast Asia 9: 87- States, Canada and Greenland. Spec. 134. Pap. So. Cal. Paleontol. Soc. 5: 39 pp. Lamont, A. Emmrich, H. F. 1978. Pentlandian miscellany: Mollusca, Tri- 1844. Zur naturgeschichte der Trilobiten. lobita, etc. Scottish J. Sci. 1: 245-299. Meiningen. Lane, P. D. Etheridge, R., Jr., and J. Mitchell 1984. Silurian trilobites from Hall Land and 1916. The Silurian trilobites of New South Nyeboe Land, western North Green- Wales, with references to those of other land. Rapp. Gr0nl. Geol. Unders0gelse parts of Australia. Part V. The Encri- 121: 53-75. nuridae. Proc. Linn. Soc. N. S. W. 40 1988. Silurian trilobites from Peary Land, (for 1915): 646-680. central North Greenland. Rapp. Gr0nl. Evitt, W. R., and R. P. Tripp Geol. Unders0gelse 137: 93-117. 1977. Silicified Middle Ordovician trilobites Maddison, W. P., M. J. Donoghue, and D. R. from the families Staurocephalidae and Maddison Encrinuridae. Palaeontographica, ser. 1984. Outgroup analysis and parsimony. Syst. A, 157: 109-174. Zool. 33: 83-103. Foerste, A. F. Maksimova, Z. A. 1888. Notes on Paleozoic fossils. Bull. Sci. 1962. Trilobity ordovika i silura sibirskoy Labs. Denison Univ. 3: 117-139. platformy. Trudy. Vses. Nauch.-Issled. Gass, K. C., and D. G. Mikulic Geol. Inst. [VSEGEI] 76: 1-214. 1982. Observations on the Attawapiskat For- Miinnil, R. mation (Silurian) trilobites of Ontario 1977. Novye enkrinuridy (Trilobita) Llando- with description of a new encrinurine. veri Pribaltiki. Eesti NSV Tead. Akad. Can. J. Earth Sci. 19: 589-596. Toim., Keem. Geol. 26: 46-56. Hamada, T. Mickevich, M. F., and N. I. Platnick 1961. Evolution ofencrinurid cephalons. Jap- 1989. On the information content of classifi- anese J. Geol. Geogr. 32: 205-218. cations. Cladistics 5: 33-47. Hennig, W. Morris, S. F. 1966. Phylogenetic systematics. Urbana: Univ. 1988. A review of British trilobites, including Illinois Press. a synoptic revision of Salter's mono- Holloway, D. J. graph. Palaeontogr. Soc. Monogr. (Lon- 1980. Middle Silurian trilobites from Arkan- don) 574: 316 pp. sas and Oklahoma, U.S.A. Part 1. Pa- Nelson, G., and N. Platnick laeontographica, ser. A, 170: 1-85. 1981. Systematics and biogeography: cladis- 1981. Silurian dalmanitacean trilobites from tics and vicariance. New York: Colum- North America and the origins of the bia Univ. Press. Dalmanitinae and Synphoriinae. Pa- Over, D. J., and B. D. E. Chatterton laeontology 24: 695-731. 1987. Silurian conodonts from the southern Holtedahl, 0. Mackenzie Mountains, Northwest Ter- 1914. On the fossil faunas from Per Schei's ritories, Canada. Geol. Palaeontol. 21: Series B in South western Ellesmere- 1-49. land. Report of the 2nd Norwegian Ex- Perry, D. G., and B. D. E. Chatterton pedition in the "Fram" 1898-1902, 32: 1977. Silurian (Wenlockian) trilobites from 1-48. Baillie-Hamilton Island, Canadian Arc- 22 AMERICAN MUSEUM NOVITATES NO. 2968

tic Archipelago. Can. J. Earth Sci. 14: Straits, in the years 1850-1851, vol. 2. 285-317. London. 1979. Wenlock trilobites and brachiopods Schrank, E. from the Mackenzie Mountains, north- 1972. Proetacea, Encrinuridae und Phacopina western Canada. Paleontology 22: 569- (Trilobita) aus silurischen Geschieben. 607. Geologie 21: 1-117. Pojeta, J., Jr., and B. S. Norford Snajdr, M. 1987. A Bohemian-type Silurian (Wenlock- 1983. New Silurian trilobites from Bohemia. ian) pelecypod faunule from Arctic Vest. Ustr. Ust. Geol. 58: 175-178. Canada. J. Paleontol. 61: 508-520. 1985. Bohemian representatives of the Ramsk6ld, L. Subfamily Encrinurinae (Trilobita). 1985. Studies on Silurian trilobites from Got- Sbor. Geol. VWd, Paleontol. 27: 9-46. land, Sweden. Ph.D. dissertation, Univ. Strusz, D. L. Stockholm and Swed. Mus. Nat. Hist. 1980. The Encrinuridae and related trilobite 1986a. Silurian encrinurid trilobites from Got- families, with a description of Silurian land and Dalarna, Sweden. Palaeontol- species from southeastern Australia. ogy 29: 527-575. Palaeontographica, ser. A, 168:1-68. 1986b. Pacificurus, new name for Australurus Temple, J. T., and R. P. Tripp Ramskold (not Jell and Duncan). Geol. 1979. An investigation of the Encrinurinae Foren. Stockholm Forh. 108: 380. (Trilobita) by numerical taxonomic Ramskold, L., and M. G. Bassett methods. Trans. R. Soc. Edinburgh 70: 1990. A middle Llandovery (Silurian) shelly 223-250. fauna from Ostergotland, Sweden. Geol. Thomas, A. T. Fbren. Stockholm Forh. (in press). 1981. British Wenlock trilobites. Part 2. Pa- Reed, F. R. C. laeontogr. Soc. Monogr. (London) 559: 1928. Notes on the Family Encrinuridae. Geol. 57-99. Mag. 65: 51-77. Thomas, A. T., and G. M. Narbonne 1931. The lower Paleozoic trilobites of the 1979. Silurian trilobites from arctic Canada. Girvan District, Ayrshire. Supplement Geol. Mag. 116: 1-19. No. 2. Palaeontogr. Soc. Monogr. (Lon- Tripp, R. P., J. T. Temple, and K. C. Gass don) 1-30. 1977. The Silurian trilobite Encrinurus var- Rieppel, 0. iolaris and allied species with notes on 1988. Fundamentals of comparative biology. Frammia. Palaeontology 20: 847-867. Basel, Boston, Berlin: Birkhiiuser Ver- Whittington, H. B., and K. S W. Campbell lag. 1967. Silicified Silurian trilobites from Maine. Salter, J. W. Bull. Mus. Comp. Zool. Harvard Univ. 1852. Geology. In P. C. Sutherland. Journal 135: 447-483. of a voyage in Baffin's Bay and Barrow

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER.