Identification of the Bones of the Snout in Fossil Lower Actinopterygians—A New Nomenclature Scheme Based on Characters
Currently, there is no standardized nomenclatural scheme for identifying and naming the bones of the snout in lower actinopterygians and specifically those fishes referred to as palaeoniscoids. A literature review reveals that the same bone names are used by different researchers to identify very different bones. This lack of homogeneity is problematic because it makes comparisons among taxa difficult, impairs our understanding of the morphological diversity of lower actinopterygians, and presents potential pitfalls when building character matrices for phylogenetic analyses. Because of the problems the absence of a standardized nomenclature scheme presents, a new set of rules for the identification of the bones of the snout of lower actinopterygians is proposed. These definitions are based on characters that are commonly preserved such as the presence of sensory canals, location of bones in relation to other bones, and whether or not the bones contribute to forming nasal openings. When numerous characters are present in a single bone, this bone is considered to be a complex bone, and the name reflects this. The new definitions are based on Remane’s criteria of similarity in position and detail and are an attempt at identifying homologous structures. The snout bones of various Devonian and Carboniferous palaeoniscoids are re-identified using this new nomenclature scheme. After that, patterns regarding the makeup of the snout in Devonian and Carboniferous palaeoniscoids emerged. The snouts of Carboniferous palaeoniscoids show much more morphological diversity than those of the Devonian forms. The phylogenetic signal of these new characters was tested with their inclusion in a phylogenetic matrix constructed to investigate the relationships of lower actinopterygians. The phylogenetic tree that resulted from analysis of this matrix has clades supported by some of the new snout characters. These results suggest that in-depth investigations into such characters are necessary to form a stronger understanding of the morphological diversity of lower actinopterygians and have implications for phylogenetic studies.
Schematic showing Gardiner’s (1963) macromeric hypothesis regarding the evolution of the snout in actinopterygians. The term Neopterygii is used in the broad sense of the group. Illustration modified from figure 3 of Poplin and Lund (1997).
Schematic showing Poplin and Lund’s (1997) mesomeric hypothesis regarding the evolution of the snout in actinopterygians. The term Neopterygii is used in the broad sense of the group. Illustration modified from figure 4 of Poplin and Lund (1997).
Illustrations of various Devonian and Carboniferous actinopterygians in lateral (A–D, F, H–L) and dorsal (E, G) views. Illustrations depict how the term ‘rostral’ (A–D), ‘postrostral’ (E–I), and ‘premaxilla’ (J–L) have been used to identify very different bones. Bones shaded in dark gray represent how the bone name has been applied in the literature. (A) †Mimipiscis toombsi (modified from Gardiner, 1984); (B) †Rhadinichthys canobiensis (modified from Moy-Thomas and Bradley Dyne, 1938); (C) †Kentuckia deani (modified from Rayner, 1951); (D) †Kalops monophrys (modified from Poplin and Lund, 2002); (E) †Cheirolepis canadensis (modified from Arratia and Cloutier, 1996); (F) †Gogosardinia coatesi (modified from Choo et al., 2009); (G) †Cheirolepis trailli (modified from Pearson and Westoll, 1979); (H) †Kalops monophrys (modified from Poplin and Lund, 2002); (I) †Canobius ramsayi (modified from Moy-Thomas and Bradley Dyne, 1938); (J) †Kalops monophrys (modified from Poplin and Lund, 2002); (K) †Mimipiscis toombsi (modified from Gardiner, 1984); (L) †Howqualepis rostridens (modified from Long, 1988). Dashed lines represent paths of sensory canals in this and subsequent figures.
Dorsal view of the skull roof and snout region of †Dialipina salgueiroensis depicting a snout pattern of rostral and postrostral bones. Illustration modified from Schultze and Cumbaa (2001). Bone identifications are the same as the original identifications by Schultze and Cumbaa (2001).
Lateral and dorsal views of the snout region of †Cheirolepis depicting a snout pattern of rostral and postrostral bones. (A) Lateral and (B) dorsal views of †Cheirolepis canadensis; (C) lateral and (D) dorsal views of †Cheirolepis trailli. Illustrations and bone identifications are modified from Arratia and Cloutier, 1996 (A, B) and Pearson and Westoll, 1979 (C, D). Bone names in bold are reinterpretations using the new nomenclature scheme, bone names in parentheses and italics are original names that have been changed in this paper, and bone names in Roman type set are unchanged from the original descriptions.
Hypotheses regarding the trajectory of canal lines in †Cheirolepis trailli and subsequent bone identifications. (A) Known paths of the sensory canals of the snout; (B) hypothesized trajectory of the sensory canals, with the ethmoid commissure fully enclosed in the ventral rostral-premaxillo-antorbital; (C) hypothesized trajectory of the sensory canals, with no ethmoid commissure, extension of the supraorbital canal, and three nasal bones; (D) hypothesized trajectory of the sensory canals with no ethmoid commissure and the sensory canals extending into a rostral membrane from the premaxillo-antorbital to the nasal bones. Illustrations and bone identifications are modified from Pearson and Westoll (1979). The hypothesis in B is favored here.
Illustrations of the snout region of Devonian and Carboniferous actinopterygians depicting a snout pattern of either a single rostral (A–B) or dorsal and ventral rostral bones (C–F). (A) Lateral and (B) dorsal views of the snout region of †Gogosardinia coatesi; (C) lateral and (D) dorsal views of the snout region of †Kalops monophrys; (E) lateral and (F) dorsal views of †Paratarrasius hibbardi. Illustrations and bone identifications modified from Choo et al., 2009 (A, B), Poplin and Lund, 2002 (C, D), and Lund and Melton, 1982; Lund and Poplin, 2002 (E, F). Bones infilled with light gray are reconstructed on the basis of other Devonian actinopterygians.
Illustrations of the snouts of Devonian actinopterygians depicting a snout pattern of a single rostral and a complex premaxillo-antorbital. (A) Lateral and (B) dorsal views of †Donnrosenia schaefferi; (C) lateral and (D) anterior views of †Mimipiscis toombsi, external view of premaxilla on the left, internal view on the right; (E) lateral and (F) anterolateral view of †Moythomasia durgaringa. Illustrations and bone identifications modified from Long et al., 2008 (A, B) and Gardiner, 1984 (C–F). Bones infilled with gray are reconstructed on the basis of other Devonian actinopterygians.
Illustrations of the snouts of Devonian and Carboniferous actinopterygians depicting a snout pattern of a single rostral bone. (A–C) Single rostral and a premaxillo-antorbital; (D, E) single rostral and a separate antorbital, premaxillae absent. (A) Lateral and (B) dorsal views of †Howqualepis rostridens; (C) close up of the snout of †Woodichthys bearsdeni with internal surface of a premaxillo-antorbital figured; (D) lateral view of †Cyranorhis bergeraci; and (E) lateral view of †Wendyichthys dicksoni. Illustrations and bone identifications modified from Long, 1988 (A, B), Coates, 1998 (C), and Lund and Poplin, 1997 (D, E).
Illustrations of the snouts of Carboniferous actinopterygians depicting a snout pattern of a single rostral and separate premaxillary and antorbital bones. (A) Lateral and (B) dorsal views of †Lineagruan judithi; (C) lateral and (D) dorsal views of †Beagiascus pulcherrimus; (E) lateral view of †Aesopichthys erinaceus; (F) lateral view of †Canobius ramsayi. Illustrations and bone identifications modified from Mickle et al., 2009 (A–D), Poplin and Lund, 2000 (E), and Moy-Thomas and Bradley Dyne, 1938 (F).
Illustrations of the snouts of Carboniferous actinopterygians depicting snout patterns of either a single rostral (A) or two rostral bones (B–D). Lateral views of (A) †Canobius elegantulus; (B) †Mansfieldiscus sweeti; (C) †Rhadinichthys canobiensis; and (D) †Gonatodus punctatus. Illustrations and bone interpretations modified from Moy-Thomas and Bradley Dyne, 1938 (A, C), Long, 1988 (B), and Gardiner, 1967 (D).
Problematic Devonian actinopterygians. (A–E) Illustrations of the snout of Devonian fishes. (A) Lateral view of †Kentuckia deani; (B) lateral view of †Tegeolepis clarki. (C–E) Hypotheses regarding the trajectory of canal lines in †Tegeolepis clarki and subsequent bone identifications. (C) Known paths of the sensory canals of the snout; (D) hypothesized trajectory of the sensory canals, with the ethmoid commissure exiting the antorbital and entering the rostral; (E) hypothesized trajectory of the sensory canals, with the ethmoid commissure exiting the antorbital and entering the ventral rostro-premaxilla. Illustrations and bone interpretations modified from Rayner, 1951 (A) and Dunkle and Schaeffer, 1973 (B–E).
Illustrations of the snout of some problematic Carboniferous deep-bodied actinopterygians. (A) Lateral view of †Platysomus superbus; (B) lateral view of †Platysomus schultzei; (C) lateral view of †Platysomus swaffordae. Illustrations and bone interpretations modified from Moy-Thomas and Bradley Dyne, 1938 (A), Zidek, 1992 (B), and Mickle and Bader, 2009 (C).
Phylogenetic hypothesis of relationships of lower actinopterygian fishes resulting from analysis of a character matrix updated to include characters and character states using the new snout nomenclature scheme. Phylogenetic tree modified from Mickle (2012:Ch. 6, fig. 9). Clade 1 is Actinopterygii, Clade 2 is †Palaeonisci, Clade 3 is †Palaeoniscomorpha, and Clade 4 is †Palaeonisciformes.
Contributor Notes
Associate Editor: G. Arratia.