Three New Palaeoniscoid Fishes from the Bear Gulch Limestone
Total Page:16
File Type:pdf, Size:1020Kb
Three new palaeoniscoid fi shes from the Bear Gulch Limestone (Serpukhovian, Mississippian) of Montana (USA) and the relationships of lower actinopterygians Kathryn E. MICKLE University of Kansas, Department of Ecology and Evolutionary Biology, Natural History Museum and Biodiversity Research Center, 1345 Jayhawk boulevard, Lawrence, Kansas 66045 (USA) [email protected] Richard LUND 18 Hillside road, Mount Holly, New Jersey 08060 (USA) [email protected] Eileen D. GROGAN Saint Joseph’s University, Department of Biology, 5600 City Line avenue, Philadelphia, Pennsylvania 19131 (USA) [email protected] Mickle K. E., Lund R. & Grogan E. D. 2009. — Three new palaeoniscoid fi shes from the Bear Gulch Limestone (Serpukhovian, Mississippian) of Montana (USA) and the relationships of lower actinopterygians. Geodiversitas 31 (3) : 623-668. ABSTRACT Th ree new palaeoniscoid fi shes (Osteichthyes, Actinopterygii), representing two new genera, Lineagruan judithi n. gen., n. sp., L. snowyi n. gen., n. sp., Beagias- cus pulcherrimus n. gen., n. sp., are described from the Bear Gulch Limestone Member of the Heath Formation (Serpukhovian) of Montana, a 318 million year old lagerstätte. Morphological, morphometric, and meristic data were ana- lyzed and compared to data for other Paleozoic actinopterygians. Diff erences among the species were noted in character complexes that may have played a role in feeding or propulsive regimes and fi ne-scale niche partitioning. A matrix of 111 characters and 40 taxa was constructed using relatively complete taxa ranging from the Devonian to the Recent. Cladistic analysis using Hennig86 KEY WORDS Bear Gulch Limestone, and Winclada resulted in two trees. Branch-and-bound treatment generated Palaeoniscoids, one tree, in which the palaeoniscoids were paraphyletic. In all instances, the lower actinopterygians, Carboniferous, relationships of the three new species were consistent. It is noted that a number Mississippian, of the palaeoniscoid clades proposed by the current cladistic analysis have also USA, been generated in earlier analyses of Bear Gulch fi shes and other independent phylogeny, new species, analyses. Th is recurring pattern implies some robustness to these associations new genera. in spite of variations between the matrices generating them. GEODIVERSITAS • 2009 • 31 (3) © Publications Scientifi ques du Muséum national d’Histoire naturelle, Paris. www.geodiversitas.com 623 Mickle K. E. et al. RÉSUMÉ Trois nouveaux poissons palaeoniscoïdes du calcaire de Bear Gulch (Serpukhovian, Mississippien) du Montana et rapports entre actinoptérygiens inférieurs. Trois nouveaux poissons palaeoniscoïdes (Osteichthyes, Actinopterygii), repré- sentant deux nouveaux genres, Lineagruan judithi n. gen., n. sp., L. snowyi n. gen., n. sp., Beagiascus pulcherrimus n. gen., n. sp., sont décrits provenant du Calcaire de la Bear Gulch Formation de Heath (Serpukhovien) du Mon- tana, gisement vieux de 318 millions d’années. Les données morphologiques, morpho métriques, et méristiques sont analysées et comparées aux données d’autres actinoptérygiens paléozoïques. Les diff erences entre les espèces sont exprimées en complexes de caractères qui peuvent avoir joué un rôle dans les modes de nutrition et de propulsion, et dans la répartition en des niches de petite taille. Une matrice de 111 caractères et 40 taxons est élaborée en utilisant des taxons relativement complets allant du Dévonien à l’époque actuelle. Une MOTS CLÉS analyse cladistique utilisant Hennig 86 et Winclada aboutit à deux arbres. Le Calcaire de Bear Gulch, paléoniscoïdes, traitement “branch-and-bound” a généré un arbre, dans lequel les paléoniscoïdes Actinoptérygiens sont paraphylétiques. Dans tous les cas, les relations des trois nouvelles espèces inférieurs, Carbonifère, sont cohérentes. Il est à noter qu’un certain nombre de clades de paléoniscoïdes Mississippien, résultant de la présente analyse cladistique ont également été générés par des USA, analyses antérieures des poissons de Bear Gulch et par d’autres analyses indé- phylogénie, espèces nouvelles, pendantes. Ce modèle récurrent accorde quelque robustesse à ces associations genres nouveaux. en dépit des diff érences des matrices qui les ont générées. INTRODUCTION hindered by few and/or poorly preserved specimens. In contrast, the Bear Gulch Limestone of Montana Th e term palaeoniscoid refers to a group of more than (Serpukhovian, Namurian E2b) preserves a diverse 250 extinct genera of lower actinopterygian fi shes assemblage of well-preserved actinopterygians. ranging from the Late Silurian to the Cretaceous Th e Lower Carboniferous Bear Gulch Limestone that share a similar body shape and arrangement of beds (c. 318 million years old) are one of a number cranial bones (Janvier 2002; Poplin & Lund 2002). of limestone lenses that make up the Bear Gulch Th ey are typically characterized by the presence of an Limestone Member of the Heath Formation of immovable maxilla-preoperculum-hyomandibula- Montana and North Dakota (Grogan & Lund palate structure, a fusiform or elongate fusiform or 2002). Th e Heath Formation and underlying Otter deep body shape, a single triangular or elongated and Kibbee Formations are Serpukhovian (Namu- dorsal and anal fi n, a heterocercal caudal fi n, and rian E2b) in age (Lund & Poplin 1997; Grogan & articulated rhombic scales (Schaeff er & Rosen 1961; Lund 2002; Poplin & Lund 2002). Th e Bear Gulch Moy-Th omas & Miles 1971; Janvier 2002). Limestone was deposited in a shallow Paleozoic Moy-Th omas & Miles (1971) recognized the pal- tropical marine bay located 12°N of the equator aeoniscoids as representative of the ancestral actin- that contained a diverse marine fl ora and fauna opterygian condition. Gardiner (1984) and Zhu & (Williams 1983; Lund & Poplin 1999; Grogan & Schultze (2001) later proposed that the palaeoniscoids Lund 2002). Field excavations over 39 years have may provide a greater understanding of both the early revealed an ichthyofauna with over 5500 specimens actinopterygian condition and the original osteich- of fi shes from over 125 species (Lund 2000; Gro- thyan plan. Yet, most studies of the interrelation- gan & Lund 2002). Actinopterygians constitute ships of Paleozoic lower actinopterygians have been approximately 72% of recovered specimens but 624 GEODIVERSITAS • 2009 • 31 (3) Head-Bear Gulch Palaeoniscoids only 33% of putative species. Th e abundance and in the visualization of select features. Digitized diversity of actinopterygian material alone sets the images were prepared by scanning the fossils Bear Gulch apart from other Paleozoic deposits with a Microtek 4800 × 3200 PPI fl atbed scan- around the world (Lund & Poplin 1997). ner or using a Canon XSi digital camera with a Th ree previously undescribed Bear Gulch palaeo- macro lens. Meristic counts and morphometric niscoids are now described. Extensive morphometric measurements were taken directly from original parameters are explored for their potential utility specimens according to Poplin & Lund (2002). in identifi cation and objective description. Th e A series of potentially braincase-independent phylogenetic relationship of these fi shes to other lateral cranial measurements were added to this actinopterygians is also investigated. measurement scheme to maximize the recovery of objective data for analyses (Fig. 1). Cranial BONE NOMENCLATURE measurements were performed upon high-resolu- Traditional and topographic actinopterygian tion scans of specimen heads, using the measure bone nomenclature is used to limit contention function of Adobe Illustrator software. Where over inferred bone homologies. Frontal and possible, latex peels were prepared to provide parietal are used rather than parietal and post- three dimensional views of specimens preserved parietal, respectively. Th e terms posteroventral from negative impressions. Th ese were shadowed infraorbital and posterodorsal infraorbital are with magnesium oxide and examined in parallel used to refer to the bones bearing the infraorbital with original specimens. Original specimens and canal that are located in the posteroventral and latex peels were used to prepare drawings and il- posterodorsal corners of the orbit, respectively, lustrations using Adobe Photoshop and Adobe following the scheme presented by Poplin & Illustrator softwares. Lund (2002). Posterodorsal infraorbital is used Statistical analyses of measurements and graphs instead of dermosphenotic, because the former of morphometric and meristic data were prepared name pertains to a cheek bone rather than a using SPSS software. Th e classifi cation function skull roofi ng bone, as in these fi shes. Th e two of Discriminant Analysis was used to determine bones in the skull roof that carry the otic canal whether the variables supported the three proposed are referred to as the dermosphenotic and der- species. Principle Component Analysis identifi ed mopterotic rather than the intertemporal and what factors (combinations of variables) best de- supratemporal, respectively. (For another view fi ned the three species (Harris 1975). on the dermosphenotic and dermopterotic, see A cladistic matrix (111 characters, 40 taxa; Ap- Poplin [2004]). Th e terms basal fulcra, fringing pendix 1) was constructed to examine the inter- fulcra, and procurrent fi n rays are used follow- relationships of lower actinopterygian fi shes. Th is ing the scheme presented by Schultze & Arratia matrix of 83 cranial and 28 postcranial characters (1989) and Arratia (2008). includes Bear Gulch and non-Bear Gulch actinop- terygians. Trees were rooted