PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024 Number 3591, 16 pp., 4 figures, 0 tables December 12, 2007

The Trilobite ‘‘Dalmanites’’ maecurua Clarke, 1890 (Middle Devonian, Amazon Basin, Brazil) and the New Genus Amazonaspis (Synphoriidae)

MARIA DA GLORIA PIRES DE CARVALHO1 AND VERA MARIA MEDINA DA FONSECA2

ABSTRACT The revision of ‘‘Dalmanites’’ maecurua Clarke, 1890, from the Upper Lontra Member, Eifelian, Maecuru Formation, Para´ State, Amazon Basin, Brazil is presented. The new genus Amazonaspis is assigned within the family Synphoriidae. Comparisons are made with other Silurian and Devonian synphoriids, and some paleobiogeographic inferences are presented.

RESUMO E´ apresentada a revisa˜o de ‘‘Dalmanites’’ maecurua Clarke, 1890, procedente da parte superior do Membro Lontra, Eifeliano, da Formac¸a˜o Maecuru, no estado do Para´, da Bacia do Amazonas, Brasil. O novo geˆnero Amazonaspis e´ atribuı´do a` famı´lia Synphoriidae. Sa˜o feitas comparac¸o˜es com outros synforiı´deos silurianos e devonianos, e apresentadas algumas infereˆncias paleobiogeo- gra´ficas.

INTRODUCTION the Maecuru and Erereˆ formations of the Amazon Basin, with ages varying from late Clarke (1890) originally described Dal- Eifelian to early Givetian. The material from manites maecurua among other trilobites from the Maecuru Formation (late Eifelian) was

1 Division of Paleontology, American Museum of Natural History ([email protected]). 2 Museu Nacional/Universidade Federal do Rio de Janeiro, Departamento de Geologia e Paleontologia (vmedina@ acd.ufrj.br).

Copyright E American Museum of Natural History 2007 ISSN 0003-0082 2 AMERICAN MUSEUM NOVITATES NO. 3591 collected in 1876 by the Imperial Geological borders and doublures; the presence or absence Commission, from several localities along the of a genal spine and an epiborder furrow; the Maecuru River, north of the Teaupixuna thoracic and pygidial architecture, especially rapids, in the state of Para´. Additional the morphology of the pleural furrows and the material was collected in the same area, by tips of the thoracic pleurae; and the position of the Orville Adalbert Derby Expedition, con- the pygidial apodemes. Thus, according to ducted by Petrobras in 1986. Campbell (1977), diagnostic features of the Clarke (1900: 15) proposed Synphoria as Synphoriinae are: gap between S1 and S2 more a subgenus of Dalmanites, for his new species than 1.5 times that between the occipital Dalmanites (Synphoria) stemmatus from the apodeme and S1; S1 slightly oblique to the Oriskany fauna, Lower Devonian of New sagittal line and S2 tends to be equidimensional York State. He considered that the Brazilian in dorsal view (the shape of the apodeme in S2 species Dalmanites maecurua was closely re- produces the transversally short furrow); ce- lated to that form as he included it in the phalic border narrow, convex, and poorly same subgenus as Dalmanites (Synphoria) defined; lateral parts of the cephalic doublure maecurua though he did not comment on the correspondingly narrow and rather evenly assignment (Clarke, 1900: 68). Delo (1935) rolled, or with a slight vincular furrow; genal reviewed the phacopid trilobites and divided spine (if present) short, lacking an epiborder the family Dalmanitidae into two subfamilies: furrow, tending to be oval in cross section; Dalmanitinae Reed, 1905, and Synphoriinae thoracic pleurae have a wide (exsag.) and well- Delo, 1935. He considered the Synphoriinae to rounded posterior pleural band, a thin anterior be characterized by more or less complete band, and pleural furrow not dominating; distal fusion and elevation of L2 and L3 in pygidial pleurae have well-developed anterior the glabella, the reduction of S1 and S2 to and posterior bands, and the posterior band submesial pits, and the presence in some forms does not fade toward the border; pygidial of spines, denticles, or crenulations on the apodemes discrete and positioned away from cephalon and pygidium. He also elevated the axial furrow. Synphoria to generic status. Holloway (1981) revised some Silurian In the classification of Richter et al. (1959), dalmanitaceans from North America and the subfamily Synphoriinae was not recog- also supported this concept of the family nized as valid and Synphoria was placed Synphoriidae. His diagnosis of the sub- in the subfamily Dalmanitinae. However, family Synphoriinae was adapted from Lespe´rance and Bourque (1971) resurrected Campbell’s (1977) work with slight modifica- the Synphoriinae and Lespe´rance (1975) sub- tions to accommodate stratigraphically earlier sequently elevated it to family status (Silurian) taxa. The modifications included the (Synphoriidae), characterized inter alia by shape of the cephalic border (flattened dor- the fusion of L2 and L3. Lespe´rance recog- sally but rolled downward and inward along nized two subfamilies within Synphoriidae: the outer margin) and the shape of the pleural the Synphoriinae and the Trypaulitinae. tips on the thoracic segments (which can be Campbell (1977) accepted Lespe´rance’s rounded, pointed, or deflected backward into (1975) subdivision of the Synphoriidae into a spine). the Synphoriinae and Trypaulitinae, but modified his concept of the family by rejecting MATERIAL AND METHODS the distal coalescence of L3 and L2 as diagnostic and arguing that this feature is Specimens studied and figured herein are widespread throughout the Dalmanitoidea. housed in the Museu Nacional/UFRJ (MN), New criteria proposed by Campbell for Rio de Janeiro and Universidade Federal of recognizing the Dalmanitinae, Synphoriinae, Rio de Janeiro (UFRJ). For conservational and Trypaulitinae included the spacing of the reasons it was not possible to coat these cephalic apodemes (i.e., the gap between the specimens for photography. occipital and S1 apodemes and between S1 Comparisons were made with another and S2); the development of the cephalic Malvinokaffric synphoriinid, Dalmanitoides 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 3

Delo, 1935, from the Devonian of Argentina; SYSTEMATIC PALEONTOLOGY Fenestraspis Branisˇa and Vaneˇk, 1973, from the Devonian of Bolivia; the North American ORDER PHACOPIDA SALTER, 1864 Devonian synphoriinids Synphoria stemmata SUBORDER PHACOPINA STRUVE, 1959 (Clarke, 1900), Odontocephalus selenourus (Eaton, 1832), Anchiops anchiopsis (Green, SUPERFAMILY DALMANITOIDEA VOGDES, 1832), and Synphoroides biardi (Clarke, 1907); 1890 and with two Silurian synphoriinids, FAMILY SYNPHORIIDAE LESPE´ RANCE, 1975 Lygdozoon collatum Holloway (1981) from North America and Delops obtusicaudatus SUBFAMILY SYNPHORIINAE DELO, 1935 (Salter 1849) from Britain. Terminology used here for the descriptions follows Whittington Amazonaspis, new genus and Kelly (1997). TYPE AND ONLY SPECIES: Dalmanites mae- curua Clarke, 1890, from the Upper Lontra GEOLOGICAL SETTING Member, Eifelian, Maecuru Formation, Para´ The fossils described here were collected State, Brazil. from fossiliferous sandstones that belong to DIAGNOSIS: Glabella wide, with sculpture the upper part of the Maecuru Formation of coarse tubercles. Frontal lobe rhombic in (Lontra Member). These sandstones crop out outline, moderately inflated, glabellar furrows along the Maecuru and Curua´ rivers, in the reaching axial furrows; S1 very close to northern border region of the Amazon Basin occipital ring and oblique, S2 subparallel to in Para´ State, Brazil (fig. 1), and are overlain S1, inner ends of S3 and S1 joined by shallow longitudinal depression separating inflated by the Erereˆ Formation (Melo, 1988). The L2 and L3 from depressed medial part of Lontra Member consists mainly of fluvial glabella. Anterior branch of facial suture sandstones and conglomerates, but its upper- running outside frontal glabellar lobe, parallel most part includes shallow marine tempestites to preglabellar furrows. Anterior cephalic with hummocky cross-stratified sandstones border increasing slightly in length (sag., that contain an important Devonian inverte- exsag.) medially, margin without spines or brate fauna recorded by pioneer workers, crenulations. Pygidium with 16–17 axial rings, among whom were Rathbun (1879), Clarke first three weakly convex (tr.); first four (1890, 1899), and Katzer (1903). The un- interpleural furrows weakly impressed, sub- derlying Jatapu Member of the Maecuru sequent ones line-like; 11 wide and well- Formation consists of bioturbated silty/shaly developed pleural furrows; narrow border; layers and argillaceous sandstones with chit- median keel or mucro absent. inozoans and acritarchs, but does not contain ETYMOLOGY: In reference to the Amazon marine macrofossils. Basin, where the type species occurs, together There is some palynological evidence that with the feminine Greek word aspis, ‘‘shield’’. the underlying fluvio-deltaic and neritic parts STRATIGRAPHIC RANGE: Eifelian, Middle of the Lontra Member are of Emsian to Devonian. Eifelian age, while the overlying Erereˆ REMARKS: Clarke (1890) did not provide Formation is of Eifelian-Givetian ages (Melo a diagnosis for maecurua and his description and Loboziak, 2003). Thus, the age of the was based upon a syntype series that consists fossiliferous marine sandstones forming the only of a few disarticulated specimens. uppermost part of the Lontra Member is Although the available material is fragmen- constrained to the Eifelian (fig. 2). tary and many morphological details are not The marine invertebrate fauna of the very clear, some statements about the affinities Maecuru Formation is considerably more of the species can be made. Amazonaspis diverse than other Devonian faunas from displays at least some of the cranidial char- Brazil; besides trilobites, it includes brachio- acters of synphoriinids, which include: dis- pods, bivalves, gastropods, bryozoans, corals, tance between the cephalic apodemes S1 and crinoids, and tentaculitids. S2 more than 1.5 times the gap between the 4 AMERICAN MUSEUM NOVITATES NO. 3591

Fig. 1. Location map of the studied fossiliferous outcrops. 1. Probable collecting site of the Imperial Geological Commission. 2. Collecting site of the Orville A. Derby Expedition. 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 5

Fig. 2. The Devonian miospore zonation schemes of the Amazon Basin and its relationships with regional rock units, the standard zonal schemes in Western Europe and Petrobras operational palynozonation. A. Chronostratigraphy. B. Palynozonal scheme in Western Europe. C. Palynozonal scheme in Amazon Basin. D. Lithostratigraphy. occipital and S1 apodemes (these apodemes Comparison of Amazonaspis with Syn- are more nearly equidistant in dalmanitinids); phoria: As Clarke (1900) recognized, Amazo- S1 directed obliquely forward adaxially; S2 naspis maecurua and Synphoria stemmata are also oblique; L2 and L3 inflated above the very similar, particularly in the general shape median part of the glabella adjacent to them; of the frontal glabellar lobe (with a large and genal spine lacks a longitudinal furrow. In rounded anterior margin); the smooth curva- addition, the pygidium of Amazonaspis lacks ture of the posterior cephalic border toward a well-defined border; the apodemes are the blunt and rounded genal spines; the discrete and isolated from the axial furrow; smooth occipital ring, without a central spine and the pleural furrows are deep and wide or tubercles; and the third and second (although it is not clear whether these are glabellar lobes, which have similar ornamen- symmetrical in cross section). It is concluded tation. The pygidial margins are also similar, that the Amazonian material can be placed with a slight outward curvature terminating within the family Synphoriidae rather than the posteriorly at a rounded angle, which does not Dalmanitidae. form a mucro or spine; furthermore, the axis 6 AMERICAN MUSEUM NOVITATES NO. 3591 does not reach the posterior margin of the convex (sag., tr.) with the highest point on pygidium. the posterior part of the frontal glabellar lobe. Despite these similarities, these two species The anterior cephalic border has a rounded differ in a number of characters: the axial and smooth margin. The axial furrow is broad furrows are shallower in A. maecurua; glabel- and shallow. It diverges gently forward to lar lobes L3 and L2 are not fused abaxially in about the middle (exsag.) of L3, thereafter A. maecurua; S2 and S3 are wider (tr.) and converges to the outer end of S3, from where touch the axial furrows in A. maecurua; the it becomes deeper and diverges again anteri- occipital furrow is distinct in both species, orly. The lateral glabellar furrows are well deep and almost transverse in Synphoria developed. S3 is broad, shallow, slightly stemmata, but slightly convex forward medi- narrower and deeper proximally, lengthening ally in A. maecurua; the central area of the (exsag.), and diverging forward abaxially. S2 glabella in A. maecurua bears tubercles but is is deeper than S3, directed slightly obliquely smooth in S. stemmata; the anterior cephalic backward abaxially, shallowing as it reaches margin of A. maecurua lacks crenulations like the axial furrow. S1 is situated very close to those of S. stemmata, and the pygidium the occipital ring, curved in a gentle forwardly includes 10–11 axial rings and 10–11 pleural concave arc parallel to the occipital furrow furrows in S. stemmata against 16–17 axial and becomes shallower distally where it meets rings and 13 pleural furrows in A. maecurua. the axial furrow. The occipital furrow (S0) is bowed slightly forward medially, becoming Amazonaspis maecurua (Clarke, 1890) narrower and deeper as it approaches the axial furrows. L3 and L2 are inflated above figure 3A–H the median part of the glabella adjacent to them. L3 is subtriangular with rounded edges; Dalmanites maecurua Clarke; Clarke, 1890: 23–29, pl. II, L2 is roughly rectangular with the edges figs. 1–3; 6–7; 10; 15. similarly rounded. L1 is shorter (sag., exsag.) Dalmanites (Synphoria) maecurua Clarke; Clarke, 1900: 68. than the occipital ring and extends across the Dalmanites maecurua Clarke; Katzer, 1903: 212, pl. 15, entire width of the glabella, slightly curved fig.12. forward medially. The occipital ring is longer Dalmanites maecurua Clarke; Lima, 1989: 12. than L1 (sag.) with a weakly concave anterior Dalmanites maecurua Clarke; Silva and Fonseca, 2005: 76, margin and convex posterior margin. The fig. 3C–D. median area of the glabella is ornamented TYPE MATERIAL: From Clarke’s syntypes, with tubercles. The frontal glabellar lobe is MN 3383-I, internal mold of glabella plus relatively large, rhombic in outline, with its small part of left fixigena (figured by Clarke, anterior margin broad and rounded, and its 1890: pl. II, fig. 2; refigured here as fig. 3C–E) posteromedian extremity forms a subacute is here designated as lectotype; this designa- and prominent angle. There is a very faint tion is made in order to establish a single elongate posteromedian depression on the name-bearing type for the species. Paralecto- sagittal line and a circular pit near each of types are: MN 3382-I, internal mold of part of the lateral angles. The frontal glabellar lobe, the left librigena including genal spine; MN L3, and L2 are all covered with low tubercles. 3384-I, incomplete internal mold of hypo- Those on the frontal glabellar lobe are stome; MN 3385-I, internal mold of an approx. 1–1.5 mm across and scattered un- incomplete thoracic segment; MN 3386-I and evenly (2–3 mm apart) over its surface. MN 3387-I, external and internal molds Tubercles on L3 and L2 are slightly smaller respectively of an almost complete pygidium. and closer together. A small portion of the left OTHER MATERIAL: UFRJ-DG 115-Tr, in- fixigena is preserved, with a small and ternal and external molds of a cranidium with elevated palpebral lobe situated opposite the cheeks missing. anterior half of L2 and the posterior half of DESCRIPTION: The cephalon is represented L3. The anterior branch of the facial suture only by two very incomplete specimens (MN delimits a very narrow strip of the gena 3383-I, UFRJ 115-Tr). It is moderately between the facial suture and the axial furrow, 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 7 along of the frontal lobe, running close to the angle. The axial furrow is well developed, glabella. The lateral border is beveled; the shallow, and narrow. The axis is gently convex lateral and posterior furrows are broad and in transverse profile, tapered evenly backward, quite deep. The posterior furrow curves and its anterior width is slightly less than 30% backwards and tapers distally toward the of the maximum pygidial width. The axis genal spine but does not meet the lateral includes 16–17 axial rings plus a blunt termi- cephalic furrow (which is wide and also tapers nal piece that does not reach the posterior posteriorly). The furrows are separated by margin of the pygidium. The first three axial a small, narrow convex portion of the free rings are arched gently forward medially, but gena. The genal spine is small, diverging subsequent ones are transverse. The pleurae laterally, its internal margin forms a smooth are gently arched (tr.) and almost flat proxi- concave curve. The hypostome is elongate- mally, but curve more conspicuously toward linguiforme in outline, with a convex anterior the margin. There are 11 well-developed margin. The lateral margins converge in pleural furrows plus an inconspicuous final a gently concave curve from the anterior one. The first three pleural furrows curve wings to the shoulder, situated opposite the smoothly to the margin, but subsequent posterior lobe of the middle body; behind the furrows are almost straight and are directed shoulder the posterior margin is parabolic in successively more strongly backward, with the outline. The anterior lobe is convex, separated last one running very close to the axis. The from the posterior lobe by a well-developed pleural furrows are wide and deep, shallow furrow that is deep and narrower abaxially distally, and probably symmetrical in cross and very shallow medially. The macula is section, but this is not clear from the located approximately midway between the specimen. The pleural ribs become obsolete lateral margin and the midline. It is unknown before reaching the margin. Interpleural fur- whether the posterior margin was spinose, rows are narrow and weakly defined, and the although Clarke (1890: pl. II, fig. 3) restored subsequent ones are line-like. There are in- its outline with three small spines. distinct nodes scattered over some ribs in an The thorax is represented only by an apparently random pattern. The marginal area incomplete internal mold of a single segment is smooth, scarcely developed into a faint lacking the pleural termination. The axial ring border. Terminal spine or mucro is absent. is broad (tr.), rounded and thickened near the Comparison of A. maecurua with other axial furrow, which is shallow and moderately North American synphoriinids: wide. The posterior margin of the axial ring is Odontocephalus selenourus (Eaton, 1832), gently convex posteriorly while its anterior from the Lower Devonian of New York part is depressed medially and merges with the (Onandaga Limestone) is readily distinguished articulation furrow. The depression expands from A. maecurua by having an anterior toward the sagittal line, where it is approxi- cephalic border that bears a series of distally mately equal in length to the remainder of the broadened and coalesced processes, and the ring. The lateral part of the articulation pygidium bears a distinct pair of pointed or furrow/apodemal pits has a markedly oblique blunt spines posteriorly (see Whiteley et al., orientation. The posterior edge of the articu- 2002: pl. 125). lating half ring is transverse and sharply Anchiopsis anchiopsis (Green, 1832), from defined over its medial two-thirds. The pleural Lower Devonian of New York (Schoharie furrow is well developed, deep, wide, and Formation) differs from A. maecurua by symmetrical in cross section. The anterior having L2 and L3 highly elevated and more portion of the pleura adjacent to the axial ring or less completely fused; S1 and S2 with is almost transverse, then deflected gently adaxial pits connected by transglabellar backward and outward. furrows; the occipital ring bears a median The pygidium is triangular in outline, spine; and the pygidium has a long posterior approximately 125% as wide as long (sag.) process (see Lespe´rance and Bourque, 1971: and moderately convex (tr.), with gently pl. 27, figs. 1–3, 10–11; Whiteley et al., 2002, rounded margins and an obtuse posterior pl. 120). 8 AMERICAN MUSEUM NOVITATES NO. 3591 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 9

Synphoroides biardi (Clarke, 1907) from the forming suboval impressions that do not reach Lower Devonian of the Gaspe´ Peninsula, the axial furrows ( in A. maecurua S1 and S2 Canada, differs from Amazonaspis maecurua distinctly meet the axial furrow); the occipital in possessing a trifid anterior cephalic process ring bears three spines in F. amauta, but is and crenulated lateral cephalic margin. In smooth in A. maecurua; finally the exoskeleton addition, its pygidium ends in a blunt, tri- is densely and coarsely tuberculated in F. angular, slightly upturned spine (see Clarke, amauta (the tubercles are not as coarse in A. 1907: pl. 6, figs. 1–12). maecurua). The pygidium of F. amauta has at Comparison with Malvinokaffric synphor- least 20 rings and is extended posteriorly over iinids from South America: the pygidial margin to form a stout, upwardly These include Dalmanitoides drevermanni directed spine (A. maecurua has 16–17 rings (Thomas, 1906) from the Lower Devonian of and the axis ends in front of the posterior the Talacasto Formation, Ja´chal, Argentina margin, which is smooth and lacks a mucro or and Fenestraspis amauta Branisˇa and Vaneˇk spine); there are stout, upwardly directed (1973) from the Lower Bele´n Formation paired spines on several axial rings situated (Lower Devonian), of Chacoma, Bolivia. toward or at the distal ends of the pleural ribs Dalmanitoides drevermanni (see Thomas, (such spines are absent in A. maecurua); 1906: pl. 11, figs. 1–3) is readily distinguished and the pygidial apodemes are deeply im- from A. maecurua by having five very short pressed (exsag.) and connected with the axial frontal processes on the cephalon; S1–S3 furrow (they are comparatively shallow and connected by indistinct transglabellar furrows; are not connected to the axial furrow in A. well-developed longitudinal glabellar furrow; maecurua.) a slender pygidial axis reaching the posterior Lespe´rance (1975) excluded Fenestraspis end and extending into a short, slightly from the Synphoriidae despite the fact that it upturned point, as well as a higher number possesses at least one of the characters he of pygopleurae (14). regarded as diagnostic of the family (fusion of Branisa and Vanek (1973) suggested that L2 and L3). Significantly, it also has S2 Dalmanites maecurua may belong to their new reduced in width (tr.) and forming a suboval genus Fenestraspis, but Eldredge and impression; the distance between cephalic Ormiston (1979) considered the Brazilian apodemes S1 and S2 is more than 1.5 times species to be distinct and referred it to a new that between the occipital and S1 apodemes; genus aff. Fenestraspis. We are in agreement and the genal spine does not have a longitudi- with Eldredge and Ormiston (1979) that the nal furrow, all features now considered di- Brazilian and Bolivian species deserve separa- agnostic of the Synphoriidae (Campbell, 1977; tion at genus level. Holloway, 1981). Fenestraspis amauta (interpreted here large- Comparison with Silurian synphoriinids: ly on the basis of specimens in the AMNH Additionally, two Silurian synphoriinid collections) differs from Amazonaspis maecur- taxa were compared with A. maecurua, ua in the presence of small crenulations on the Lygdozoon collatum Holloway, 1981, and anterior cephalic margin (the anterior margin Delops obtusicaudatus (Salter, 1849). Like A. is smooth in A. maecurua); the shape of S1 and maecurua, these forms have a rounded anteri- S2, which in F. amauta are deeper adaxially or cephalic margin, a smooth, entire pygidial r

Fig. 3. A–H. Amazonaspis maecurua (Clarke, 1890), Upper Lontra Member, Eifelian, Maecuru Formation, Para´ State, Brazil. A. Paralectotype, MN 3387-I, internal mold of pygidium, dorsal view; B. Latex cast of external mold of an incomplete cephalon, UFRJ 115-Tr, dorsal view. C–E. Lectotype, MN 3383-I, internal mold of an incomplete cephalon; C, dorsal view; D, lateral view; E, frontal view. F. Paralectotype, MN 3384-I, internal mold of an almost complete hypostome, ventral view. G. Paralectotype, MN 3382-I, internal mold of part of the left librigena including genal spine, dorsal view. H. Paralectotype, MN 3385-I, internal mold of an incomplete thoracic segment, dorsal view. 10 AMERICAN MUSEUM NOVITATES NO. 3591 margin, and the pygidium is not mucronate. family Synphoriidae) into the Silurian, by However, both forms differ from A. maecurua including Lygdozoon and Delops. Here we in certain features. propose geographical extension of the family L. collatum from the St. Clair Limestone Synphoriidae into South America, including (Wenlock) of Arkansas differs from A. mae- species from Argentina, Bolivia and Brazil curua in having a longer, more slender genal (Amazon Basin). spine; an ornamented occipital ring (with Since the pioneering work of Katzer (1897) a small tubercle medially and weak nodes the affinities of the invertebrate fauna of the distally); a smaller pygidium (9–11 axial rings Maecuru Formation to both austral (Mal- and 9 pleural furrows); and the hypostome has vinokkafric) and boreal (Appalachian/Old a deeper posterior border furrow. World) faunas has become well-known Delops Rickards 1965 was originally as- (Loczy, 1966; Harrington, 1967; Boucot, signed to the subfamily Zeliskellinae within 1975, 1988; Carvalho, 1975; Machado, 1990; the Dalmanitidae, but Holloway (1981) con- Melo, 1988). In the Maecuru Formation this sidered it to be a synphoriinid. The type ‘‘mixed’’ faunal composition is particularly species Delops obtusicaudatus (Salter, 1849) is evident among the bivalves (e.g. Ptychopteria, readily distinguished from Amazonaspis mae- Leptodesma, and Pseudaviculopecten) and bra- curua by several features; the frontal glabellar chiopods (e.g., orthids including Platyorthis, lobe protrudes in front of the outline of the Discomyorthis, and Tropidoleptus; the stropho- cheeks; the genal spine is slender; L2 and L3 menids Megastrophia and Protoleptostro- are fused abaxially; S2 does not reach the axial phia; the chonetids ‘‘Chonetes’’ freitasi and furrows; the pygidium is smaller (11–13 axial Montsenets; the spiriferids Patriaspirifer, rings and 9 pleural furrows), and the posterior Mucrospirifer and Plicoplasia; and terebratu- extremity of the pygidial axis reaches the lids such as Amphigenia, Derbyina, and posterior margin of the pygidium. Paranaia). The high number of Appalachian Three other Devonian species, Gamone- genera among the brachiopods led Boucot daspis scutata, from the Devonian of Bolivia, (1975) and Boucot et al. (1969) to consider Gamonedaspis accola (Clarke, 1913) from the the Amazon Basin a southern paleogeographi- Ponta Grossa Formation of the Parana´ Basin, c extension of the Eastern Americas Realm, Brazil, and Gamonedaspis boehmi (Knod, designated the Amazon-Colombian Subpro- 1908) from the Voorstehoek Formation of vince. This faunal ‘‘mixing’’ is not as strongly South Africa were previously assigned to the expressed among the trilobites from the Subfamily Synphoriinae. However, Gamone- Maecuru Formation, since the majority of daspis Branisa and Vanek. 1963, is now these show greater affinity with Malvinokaffric considered a dalmanitinid rather than a syn- forms (Eldredge and Ormiston, 1979), espe- phoriinid (Edgecombe, 1993). cially with Bolivian calmoniids. Nevertheless, certain trilobites show some affinity to BIOGEOGRAPHIC CONSIDERATIONS Appalachian and Old World forms (e.g., homalonotids and synphoriinids). Lespe´rance (1975) suggested that the These Devonian Appalachian immigrants Family Synphoriidae was exclusively Devo- into the Amazon Basin (in the uppermost nian and almost restricted to cratonic North Maecuru Formation) may have migrated America and the Appalachian Region; how- during a widespread Eifelian transgression ever, he also included in the family specimens from the northeast (formerly regarded as late assigned by Haas (1969: pl. 83, figs. 5–8) Emsian-Eifelian by Melo, 1988; Carozzi et al., to Synphoroides biardi (Clarke, 1907), from 1973), most probably related to transgression- the upper Pragian of Central Chihuahua, regression cycle Id of Johnson et al. (1985), Northern Mexico, a region related to the which thoroughly flooded this basin. In Appalachian geosyncline. He excluded Fene- addition, a marine connection between Brazil straspis from this family. and Bolivia may have developed at that time Holloway (1981) extended the stratigraphic (Loczy, 1966; Melo, 1988; Isaacson and range of the synphoriinids (and thus the Sablock, 1990), as suggested by the presence 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 11

Fig. 4. Paleogeographic map of Brazil and adjoining areas of South America and Western Africa (indicated by dashed lines) in the late Emsian-Eifelian. Arrows indicate inferred migration routes of the Appalachian fauna into Brazil (Amazon Basin), Bolivia, and Argentina. 1. Amazon Basin; 2. Bolivia; 3. Argentina. Ocean represented by wave symbols. Base map after Melo (1988). of calmoniids similar those from Bolivia in the 2. The genus Synphoria is restricted to Maecuru fauna (see fig. 4). North America. 3. Amazonaspis maecurua, Dalmanitoides drevermanni,andFenestraspis amauta CONCLUSIONS are considered to be Malvinokaffric synphoriinids from South America. 1. Although Amazonaspis maecurua from 4. The family Synphoriidae occurs in North Brazil is probably closely related to America, Europe (England, Wales, Czech Synphoria stemmata from the Devonian Republic), and also in South America. of New York, the two forms are consid- 5. Synphoriinids probably dispersed from the ered sufficiently distinct to merit separa- Appalachian Province into South America tion at genus level. during a major Eifelian transgression. 12 AMERICAN MUSEUM NOVITATES NO. 3591

6. The presence of Malvinokaffric, Eastern Campbell, K.S.W. 1977. Trilobites of the Haragan, American, and Old World trilobites in the Bois d’Arc and Frisco Formations (Early Amazon Basin corroborates its previous Devonian) Arbuckle Mountains Region, interpretation as a biogeographic bound- Oklahoma. Bulletin of the Oklahoma ary area. Geological Survey 123: 1–227. Carozzi, A.V., H.R.P. Pamplona, J.C. de Castro, 7. The occurrence of synphoriinids in the and C.J.A. Contreiras. 1973. Ambientes depos- Amazon Basin and Bolivia suggests icionais e evoluc¸a˜o tecto-sedimentar da sec¸a˜o a marine connection between these two cla´stica paleozo´ica da bacia do Me´dio areas during the Eifelian. This is also in Amazonas. In 27 Congresso Brasileiro de agreement with previous works. Geologia 3: 279–314. Aracaju: Sociedade 8. Within Gondwana, the Devonian syn- Brasileira de Geologia. phoriinids are restricted to its northwest- Carvalho, R.G. de. 1975. Braquiopo´des devonianos ern area (northern Argentina, Bolivia, da bacia do Amazonas: Orthida, Stropho- and the Amazon Basin of northern menida, Spiriferida e Terebratulida. Boletim Brazil) and do not occur outside the do Museu Paranaense Emilio Goeldi 21: 1–36. Clarke, J.M. 1890. As trilobitas do grez de Erereˆe Malvinokaffric Realm. Maecuru. Archivos do Museu Nacional (Rio de Janeiro) 9: 1–58. ACKNOWLEDGMENTS Clarke, J.M. 1899. Molluscos devonianos do Estado do Para´, Brazil. Archivos do Museu Nacional (Rio de Janeiro) 10: 49–174. We thank D. Holloway (Museum of Clarke, J.M. 1900. The Oriskany fauna of the Victoria, Australia) for comments and sugges- Becraft Mountain. New York State Museum tions on the manuscript, J. Maisey (Division Memoir 3(3): 1–128. of Paleontology, American Museum of Clarke, J.M. 1907. Some new Devonic fossils Natural History) for his encouragement and (Quebec, New Brunswick, and Maine). New for taking the photographs, J.H.G. Melo York State Museum Bulletin 107: 153–291. (Petrobras) for revising the section on geology Clarke, J.M. 1913. Fo´sseis devonianos do Parana´. and paleogeography, and S. Thurston Monographias do Servic¸o Geolo´gico e (American Museum of Natural History) for Mineralo´gico do Brasil 1: 1–353. Cooper, M.R. 1982. A revision of the Devonian making the maps and the stratigraphic col- (Emsian-Eifelian) Trilobita from the Bokkeveld umn. We also thank B. Chatterton and P. Group of South Africa. Annals of the South Budil for their valuable comments. African Museum 89: 1–173. Delo, D.M. 1935. A revision of the phacopid REFERENCES trilobites. Journal of Paleontology 9: 402–420. Delo, D.M. 1940. Phacopid trilobites of North Boucot, A.J. 1975. Evolution and rate controls. America. Geological Society of America Amsterdam: Elsevier, 427 pp. [Developments in Special Paper 29: 1–135. Palaeontology and Stratigraphy, 1] Derby, O.A. 1878. Contribuic¸o˜es para geologia da Boucot, A.J. 1988. Devonian biogeography: an regia˜o do baixo Amazonas. Archivos do Museu update. In N.J. McMillan, A.F. Embry and Nacional (Rio de Janeiro) 2: 77–107. D.J. Glass (editors), Devonian of the world: Eaton, A. 1832. Geological text-book for aiding the 211–227. Calgary: Canadian Society of Petro- study of North American geology: being leum Geologists. a systematic arrangement of facts collected by Boucot, A.J., J.G. Johnson, and J.A. Talent. 1969. the author for his pupils, under the patronage Early Devonian brachiopod zoogeography. of the Hon. Stephen Van Rensselaer. Albany: Geological Society of America Special Paper Webester and Skinners. 119: 1–43. Edgecombe, G.D. 1993. A revision of the Devonian Branisa, L. 1965. Los fo´ssiles guı´as de Bolivia I. dalmanitid trilobite Gamonedaspis. In R. Paleozo´ico. Boletin del Servicio Geologico de Suarez-Soruco (editor), Fosiles y facies de Bolivia 6: 1–282. Bolivia Vol. II: Invertebrados y paleobotanica. Branisa, L., and Vaneˇk, J.. 1973. Several new Revista Te´cnica de Yacimeientos Petroliferos trilobite genera of the superfamily Fiscales Bolivianos 13–14(1–4): 161–166. Dalmanitacea Vogdes, 1890 in the Devonian Eldredge, N., and A.R. Ormiston. 1979. of Bolivia. Veˇstnik U´ strˇednı´ Ustav Geologicky Biogeography of Silurian and Devonian trilo- 48: 97–102. bites of the Malvinokaffric Realm. In J. Gray 2007 CARVALHO AND DA FONSECA: REVISION OF TRILOBITE SPECIES 13

and A.J. Bouco (editors), Historical biogeog- Anais do Congresso Brasileiro de Geologia 1: raphy, plate tectonics, and the changing envi- 425–435. ronment: 147–167. Corvallis: Oregon State Machado, D.M.C., V.M.M. Fonseca, and L.V.M. University Press. Rego. 1996. Estudos preliminares sobre a dis- Green, J.M. 1832. Synopsis of the trilobites of tribuic¸a˜o espacial da macrofauna da Formac¸a˜o North America. American Journal of Geology Maecuru (Devoniano Me´dio) da Bacia do and Natural History 1(12): 558–560. Amazonas, Estado do Para´. In Simpo´sio Sul Haas, W. 1969. Lower Devonian trilobites from Americano do Siluro-Devoniano: Estratigra- Central Nevada and Northern Mexico. Journal fia e Paleontologia: 239–245. Ponta Grossa: of Paleontology 43(3): 641–659. Gra´fica Planeta. Harrington, H.J. 1967. Devonian of South Melo, J.H.G. 1988. The Malvinokaffric Realm in America. In D.H. Oswaldo (editor), Inter- the Devonian of Brazil. In N.J. McMillan, A.F. national Symposium on the Devonian System Embry and O.J. Glass (editors), Devonian of 1: 651–671. Calgary, Alberta Society of Petro- the world 14: 669–802. Calgary: Canadian leum Geologists. Society of Petroleum Geologists. Holloway, D.J. 1981. Silurian dalmanitacean trilo- Melo, J.H.G., and S. Loboziak. 2003. Devonian– bites from North America and the origins of the Early Carboniferous miospore biostratigraphy Dalmanitinae and Synphoriinae. Palaeontology of the Amazon Basin, Northern Brazil. Review 24(4): 695–731. of Palaeobotany and Palynology 124: 131–202. Isaacson, P.E., and P.E. Sablock. 1990. Devonian Rathbun, R. 1879. The Devonian Brachiopoda of palaeogeography and palaeobiogeography of the Province of Para, Brazil. Proceedings of the the central Andes. In W.S. McKerrow and C.R. Boston Society of Natural History 20: 14–39. Scotese (editors), Palaeozoic, palaeogeography Reed, F.R.C. 1905. The classification of the and biogeography. Geological Society of Phacopidae. Geological Magazine 2: 172–178, America Memoir 12: 431–435. 224–228. Johnson, J.G., G. Klapper, and C.A. Sandberg. Richter, R., E. Richter, and W. Struve. 1959. 1985. Devonian eustatic fluctuations in Dalmanitinae. In R.C. Moore (editor), Euramerica. Geological Society of America Treatise on invertebrate paleontology. Part O, Bulletin 95: 567–587. Arthropoda 1: 468–473. Lawrence: Geological Katzer, F. 1897. Das Amazonas-Devon und seine Society of America and University of Kansas Beziehungen zu den anderen Devongebieten Press. der Erde. Sitzungsberichte der Ko¨niglichen Rickards, R.B. 1965. Two new genera of Silurian Bo˝hmischen Gesellschaft der Wissenschaften, Phacopid trilobites. Palaeontology 7(4): Mathematisch-naturwissenschaftliche Classe 541–551. 46: 1–50. Salter, J.W. 1849. Figures and descriptions illustra- Katzer, F. 1903. Grundzu¨ge der Geologie des tive of British organic remains, decade 2. unteren Amazonasgebietes (des Staates Para´ Memoir of the Geological Survey of United in Brazilien). Leipzig: Verlag von Max Weg. Kingdom 1849: 1–8. Knod, R. 1908. Devonische Faunen Boliviens. In Salter, J.W. 1864. A monograph of the Britsh G. Steinmann (editor), Beitra¨ge zur Geologie trilobites from the Cambrian, Silurian and und Pala¨ontologie von Su¨damerika. Neues Devonian Formations. Palaeontographical Jahrbuch fu¨r Geologie und Pala¨ontologie 25: Society of London Monographs: 1–80. 493–600. Silva, C., and V.M. Fonseca. 2005. Ha´bitos de vida Lespe´rance, P.J. 1975. Stratigraphy and paleontol- dos trilobitas das formac¸o˜es Maecuru e Erereˆ, ogy of the Synphoriidae (Lower and Middle Devoniano da Bacia do Amazonas, Brasil. Devonian dalmanitacean trilobites). Journal of Revista Brasileira de Paleontologia 8(1): 73– Paleontology 49(1): 91–137. 82. Lespe´rance, P.J., and P.A. Bourque. 1971. The Struve, W. 1959. Suborder Phacopina. In R.C. Synphoriinae: an evolutionary pattern of Moore (editor), Treatise on invertebrate pale- Lower and Middle Devonian trilobites. ontology, Part O, Arthropoda 1: 461–462. Journal of Paleontology 45: 182–208. Lawrence: Geological Society of America and Loczy, L. 1966. Contribuic¸o˜es a paleogeografia e University of Kansas Press. histo´ria do desenvolvimento geolo´gico da bacia Thomas, I. 1906. Neue Beitra¨ge zur Kenntniss der do Amazonas. Boletim da Divisa˜o de Devonischen Fauna Argentiniens. Zeitschrift Geologia e Mineralogia 222: 1–96. der Deutschen Geologischen Gesellschaft 57: Machado, D.M. 1990. Algumas considerac¸o˜es 233–290. estratigra´ficas e biogeogra´ficas acerca dos Vogdes, A.W. 1890. A bibliography of Paleozoic bivalvios devonianos da bacia do Amazonas. Crustacea from 1698 to 1889 including a list of 14 AMERICAN MUSEUM NOVITATES NO. 3591

North American species and a systematic Whittington, H.B., and S.R.A. Kelly. 1997. Terms arrangement of genera. United States applied to Trilobita. In R.L. Kaesler (editor), Geological Survey Bulletin 63: 1–177. Treatise on invertebrate paleontology, Part O. Whiteley, T.E., G.J. Kloc, and C.E. Brett. 2002. Arthropoda 1. Trilobita, revised: O313–O329. Trilobites of New York: an illustrated guide. Lawrence: Geological Society of America and Ithaca, NY: Cornell University Press. University of Kansas Press.

Complete lists of all issues of the Novitates and the Bulletin are available at World Wide Web site http://library.amnh.org/pubs. Inquire about ordering printed copies via e-mail from [email protected] or via standard mail from: American Museum of Natural History, Library—Scientific Publications, Central Park West at 79th St., New York, NY 10024. TEL: (212) 769-5545. FAX: (212) 769-5009.

This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).