1. Introduction
The unique morphology of flatfishes (Pleuronectiformes) makes their evolutionary origins and systematic position an intriguing question for researchers. The lineage has been recognized as a derived clade of Acanthopterygii for many years, but within that group, their relationships long have been essentially unknown (Nelson Reference Nelson2006). In recent large-scale molecular analyses (Betancur-R et al., Reference Betancur-R, Broughton, Wiley, Carpenter, López, Li, Holcroft, Arcila, Sanciangco, Cureton II, Zhang, Buser, Campbell, Ballesteros, Roa-Varon, Willis, Borden, Rowley, Reneau, Hough, Lu, Grande, Arratia and Ortí2013, Reference Betancur-R., Wiley, Arratia, Acero, Bailly, Miya, Lecointre and Ortí2017), Pleuronectiformes is positioned close to Centropomidae (which is considered order incertae sedis), with the two together forming the sistergroup to Carangiformes + Istiophoriformes. It seems likely that flatfish origins lie with the Carangimorpha (e.g., Campbell et al., Reference Campbell, López, Satoh, Chen and Miya2014), but the relationships are yet to be satisfactorily determined.
Relationships among flatfish lineages are also unclear, with research ongoing and some studies suggesting the order may not be monophyletic. Within the order Pleuronectiformes, the suborder Psettodoidei (containing only one genus with three species) is often considered the sistergroup to all the rest, which are placed in the suborder Pleuronectoidei (Nelson et al., Reference Nelson, Grande and Wilson2016; Betancur-R et al., Reference Betancur-R., Wiley, Arratia, Acero, Bailly, Miya, Lecointre and Ortí2017). However, in some molecular studies, Psettodoidei has also been placed outside the Pleuronectiformes, resulting in the monophyly of flatfishes being questioned (see reviews in Campbell et al., Reference Campbell, Chen and López2013, Reference Campbell, López, Satoh, Chen and Miya2014). The consequent interpretation is that the unique morphology of flatfishes evolved twice; however, this has yet to be satisfactorily determined.
Extant flatfishes are laterally compressed fish in which one eye migrates early in ontogeny resulting in both eyes being on the same side of the body. The highly compressed fish are generally benthic, lying on the blind side with the eyes oriented upwards. The fossil record of pleuronectiforms sheds some light on the origin of their unusual morphology, with intermediate forms having been found. Heteronectes chaneti Friedman, Reference Friedman2008 is considered to be a stem-flatfish (Friedman Reference Friedman2008, Reference Friedman2012). This animal is from Eocene marine beds of Monte Bolca, Italy, and Friedman (Reference Friedman2012) gave reasons for assuming it comes more specifically from late early Eocene aged sediments of the Monte Postale locality. Heteronectes chaneti has incomplete migration of the eye, although it has an asymmetrical skull; it is considered to be family incertae sedis within Pleuronectiformes (Friedman Reference Friedman2012), or a stem-pleuronectiform (Friedman Reference Friedman2008).
A second intermediate form, Amphistium paradoxum Agassiz, Reference Agassiz1835, is also from the Monte Bolca deposits. This species is similar to Heteronectes chaneti in having an asymmetrical skull with incomplete migration of the eye (Friedman, Reference Friedman2008). In Friedman’s (Reference Friedman2008) analysis, Amphistium was also recovered as a stem-pleuronectiform but considered slightly more derived than Heteronectes. The presence of these two intermediate forms in Eocene deposits supports an early Cenozoic timing for the origination of the pleuronectiform lineage. This age is also supported by otoliths, which commonly predate osteological remains, that indicate the presence of pleuronectiforms in the early Eocene epoch (Munroe Reference Munroe, Gibson, Nash, Geffen and van der Veer2015; Schwarzhans Reference Schwarzhans1999).
The earliest known flatfish fossils assigned to extant families are also from Eocene deposits, but most are somewhat younger than the two stem fossils (Heteronectes chaneti and Amphistium paradoxum), which are considered to be Ypresian in age (Friedman, Reference Friedman2008). Rhombus minimus Agassiz, Reference Agassiz1839 (a preoccupied generic name) was removed to Eobothus by Eastman (Reference Eastman1914). This species is also from the Monte Bolca locality, but from deposits considered to be Lutetian in age (Chanet, Reference Chanet1994). Eastman (Reference Eastman1914) considered Eobothus minimus to be similar to the extant Bothus (now in Bothidae), but Chanet (Reference Chanet1994) considered it to be Pleuronectoidei incertae sedis. Two other species were assigned to the same genus, Eobothus singhi Sahni and Choudhary, Reference Sahni and Choudhary1971 from Ypresian deposits of India and Eobothus vialovi Berg, Reference Berg1941 from the Eocene of Uzbekistan. Both of these were listed as dubious pleuronectiforms by Chanet (Reference Chanet1997:tab 1). Similarly, Imhoffius lutetianus Chabanaud, Reference Chabanaud1940 from the Lutetian of France was listed as incertae sedis in Pleuronectiformes by Chanet (Reference Chanet1997:tab. 1). Eobothus vialovi and Imhoffius lutetianus are apparently the same taxa that Gaudant and Gaudant (Reference Gaudant and Gaudant1969) noted (but did not identify by name) from Lutetian deposits of Tian Shan, Uzbekistan and Nanterre, France, which they included in Bothidae. Friedman (Reference Friedman2008: fig. 2a) recovered Eobothus as the sister to Citharus (Citharidae) in his cladogram.
Other early fossil representatives of extant families include three monotypic genera of Lutetian fossils from the lower Mokattam at Jbel Turah in Egypt, southeast of Cairo (Murray Reference Murray2000). One of these, Eobuglossus eocenicus (Woodward, Reference Woodward1910) was originally described in the genus Solea but was given the new generic name by Chabaneau (1931), who also placed it in its own family Eobuglossidae. The generic name has also been misspelled as Eubuglossus with the incorrect form used by Chanet (Reference Chanet1994) in the title of his paper, but the correct spelling used throughout the same paper. Chanet (Reference Chanet1994) moved this flatfish to Soleidae, the family to which the second Egyptian specimen, Turahbuglossus cuvillieri Chabanaud, Reference Chabanaud1937, also has been assigned (Woodward Reference Woodward1910; Chabanaud Reference Chabanaud1937; Chanet Reference Chanet1994). The third Egyptian species, Joleaudichthys sadeki Chabanaud, Reference Chabanaud1937, was named for a specimen lacking the head. It was originally given its own family Joleaudichthyidae Chabanaud, Reference Chabanaud1937. Gaudant and Gaudant (Reference Gaudant and Gaudant1969) later placed it in suborder Psettodoidei, and Friedman (Reference Friedman2008) assigned it to the extant family Psettodidae. A reassessment of this fish by Chanet (Reference Chanet1995) identified it as the basalmost Pleuronectoidei. Gaudant and Gaudant (Reference Gaudant and Gaudant1969) reported and named a fourth Eocene flatfish from Africa, Numidopleura enigmatica, from an unknown specific collection locality, but around Thala, Gouvernorat de Kasserine. This district is in west-central Tunisia bordering Algeria. The lack of an age for this fossil makes it difficult to comment on its likely importance. Additional Eocene fossils that have been allied with flatfishes are of questionable lineage (see summary in Chanet, Reference Chanet1997:tab.1). Despite the various placements of these fossil taxa, it is clear that by the middle of the Eocene epoch, flatfishes had evolved and diversified into several different lineages in the Tethys area, potentially with representatives of Bothidae, Soleidae and Psettodidae all being present.
Few Oligocene flatfishes have been reported. Most are unnamed remains that have been assessed as not belonging to Pleuronectiformes, or being incertae sedis in the order (Chanet, Reference Chanet1997:tab. 1), including some more recently reported from early Oligocene (Rupelian) aged deposits of Poland which were also assessed as indeterminate Pleuronectiformes (Přikryl et al, Reference Přikryl, Kania and Krzemiński2016). Named Oligocene-aged flatfishes are Oligobothus pristinus (Bothidae) and Scophthalmus stamatini (Scophthalmidae) from middle Oligocene deposits of Rumania (Baciu and Chanet, Reference Baciu and Chanet2002), and two Rupelian species, Oligoscophthalmus weissi and Oligopleuronectes germanicus (Pleuronectidae) from Germany (Sakamoto et al., Reference Sakamoto, Uyeno and Micklich2003, Reference Sakamoto, Uyeno and Micklich2004). Flatfish fossils are more numerous in Miocene deposits, including a variety of species from the West Coast of the United States found in late Miocene localities of Lompoc, California (Chanet Reference Chanet1997: tab. 1). We here report on an early Oligocene flatfish from Oregon, USA. An articulated specimen of a flatfish was donated to the University of British Columbia (UBC) Geology Museum collections and its presence was made known to one of us (DEC) in the late 1980s. The specimen was later transferred to the Beaty Museum of Biodiversity at UBC. The specimen has apparently languished since then, and it is only now with this paper that a description is presented. The specimen is from the Keasey Formation, from a locality near Mist, Oregon well known for its crinoid fauna. The fossil represents a small individual, so it was either a very young fish or represents a species that has a small maximal adult size. Maximal adult size for flatfishes ranges from 220 to 2500 mm (Cooper and Chapleau, Reference Cooper and Chapleau1998). To our knowledge, the flatfish described here represents the only articulated fish remains known from the formation.
2. Geology
The Keasey Formation is known as a rare Cenozoic Lagerstätte for Isocrinus oregonensis crinoids (Burns et al. Reference Burns, Campbell and Mooi2005). The specimen was reported to have been recovered during excavation to uncover this crinoid-rich layer. The Keasey Formation is dated as late Eocene through early Oligocene and was deposited in a deep-water marine environment (Burns et al., Reference Burns, Campbell and Mooi2005), with the Mist deposits considered to be earliest Oligocene in age, and more specifically magnetozone Chron C13n (33.0-33.5 Ma) in age (Prothero and Hankins, Reference Prothero and Hankins2000). In addition to the crinoids and associated asteroids and echinoids (Burns et al. Reference Burns, Campbell and Mooi2005), the formation contains numerous molluscs as well as a crab (Nyborg et al. Reference Nyborg, Nyborg, Garassino and Vega2016).
Very few fish remains have been reported from the Keasey Formation. Welton (Reference Welton1979) reported squalomorph teeth from the locality near Mist. He identified these as representing the extant seven-gill shark Heptranchias howellii (now in family Hexanchidae) and a possibly unique species of Centrophorus (a squaliform shark now in the family Centrophridae).
David (Reference David1956) studied fish scales from the formation and considered them to represent modern groups of gadiform fishes. She named a new genus and species, Probathygadus keaseyensis in Bathygadidae (Gadiformes), for one scale morphology. David (1959) noted the presence of a herring (Wisslerius), several species of the acanthomorph Beryx, two macrurids and the lanternfish Lampanyctus as being common in the sediments, as well as the rare presence of a merlucciid, a serranid and a percomorph. David’s (1959) study was an analysis of fish scale assemblages, indicating these identifications are based only on scales; no articulated fish material has previously been reported from the Keasey Formation.
3. Material and methods
3.a. Examined material
The fossil is permanently catalogued in the University of British Columbia Beaty Biodiversity Museum, as specimen #3959. It is an almost complete articulated fish preserved as a two-dimensional specimen in matrix. At some point in the past, the specimen was surrounded by plaster within a wooden box frame. The fish is missing the anterior tip of the head, and many of the bones are preserved as natural moulds. Comparative material examined is from the University of Alberta Museum of Zoology (UAMZ) and the Royal Ontario Museum (ROM).
3.b. Phylogenetic analysis
Hoshino (Reference Hoshino2001b) provided a phylogenetic analysis for the members of Pleuronectoidei, using Psettodidae (the only family in the suborder Psettodoidei) as the outgroup, and coding for 45 morphological characters. Two families of Soleoidea (Achiropsettidae and Paralichthodidae) were not included in that analysis, but all other flatfish families were included, with multiple genera of Citharinidae being included as separate taxa, as that family was the main focus of the study. We used the published characters and coding (Hoshino, Reference Hoshino2001b:tab. 1) to create a data matrix in Mesquite (Madison and Madison Reference Maddison and Maddison2021) to which we added the data for the fossil from the Keasey Formation (online Supplementary Material at http://journals.cambridge.org/geo). The resulting matrix for our analysis (http://morphobank.org/permalink/?P5600) contains 18 taxa and 45 characters. Characters were coded as unordered and of equal weight. Characters that are inapplicable in a given taxon are coded as ‘?’. The data matrix was run in a parsimony analysis in PAUP* v. 4.0a151 (Swofford Reference Swofford2002), using an heuristic search with tree bisection and reconnection (TBR) swapping algorithm. The consistency (CI) and retention (RI) indices were calculated in Mesquite v. 3.03 (Maddison and Maddison 2016).
3.c. Abbreviations used in figures
aa, anguloarticular; cl, cleithrum; cor, coracoid; den, dentary; ect, ectopterygoid; ep, epural; fapt, first anal-fin pterygiophore; fr, frontal; hy1+2, fused hypurals one and two; hy3+4, fused hypurals three and four fused with compound centrum; hy5, fifth hypural; iop, interopercle; l, left; le, lateral ethmoid; mx, maxilla; pal, palatine; pcl, postcleithrum; pg, pelvic girdle; ph, pharyngeal bone remains; phy, parhypural; pop, preopercle; ps, parasphenoid; pu2, second preural centrum; qu, quadrate; r, right; sca, scapula; soc, supraoccipital; vo, vomer.
4. Familial designation
Following the taxonomic hierarchy and features collated by Nelson et al. (Reference Nelson, Grande and Wilson2016), the specimen is easily recognizable as belonging to Pleuronectiformes (the order containing all flatfish) by its compressed body, elongate fins and an asymmetrical skull with both eyes on the same side of the head. There are several other features proposed for the order (Chapleau, Reference Chapleau1993; Friedman, Reference Friedman2012), which, although they are convergent with other fish lineages, can be identified in the Oregon fossil. These include absence of supraneurals, absence of well-developed membranous extensions on the shafts of pterygiophores in the dorsal and anal fins, full neural spine on second preural centrum and two or fewer epurals.
The two pleuronectiform suborders can be distinguished from one another by whether the dorsal fin extends onto the head, as it does in Pleuronectoidei, or does not extend onto the head, as in Psettodoidei (e.g., Nelson et al., Reference Nelson, Grande and Wilson2016). The fossil clearly preserves the dorsal fin extending onto the head, reaching the level of the eye. In addition, it shares with other pleuronectoids the lack of fin spines in the dorsal and anal fins (e.g., Hensley and Ahlstrom, Reference Hensley, Ahlstrom, Moser, Richards, Cohen, Fahay, Kendall and Richardson1984) and the enlarged, curved first anal pterygiophore (e.g., Chanet et al., Reference Chanet, Mondéjar-Fernández and Lecointre2020).
Pleuronectoidei contains three superfamilies, Citharoidea, Soleoidea and Pleuronectoidea. Following our analysis using the data of Hoshino (Reference Hoshino2001b) and including the fossil, the Oligocene fossil is recovered as the earliest-diverging member of the superfamily Pleuronectoidea (see below). Pleuronectoidea contains four families, Bothidae, Scophthalmidae, Paralichthyidae and Pleuronectidae. This is similar to a grouping that has been referred to as the ‘bothoid group’ (Chapleau Reference Chapleau1993; Chanet et al. Reference Chanet, Mondéjar-Fernández and Lecointre2020). The ‘bothoid group’ was recognized on the basis of the caudal fin having the fused hypural 1+2 closely articulating with the ventral surface of the first preural centrum and an upper plate (hypural 3+4) fused to the centrum. This condition is present in the fossil. Hoshino (Reference Hoshino2001b) found the ‘bothoid group’ was monophyletic after the removal of the genus Brachypleura. The Oligocene fossil is recovered as the sister group to this revised ‘bothoid group’, i.e., a redefined Pleuronectoidea excluding Brachypleura (Hoshino, Reference Hoshino2001b, as “Clade F’). Consequently, we include it in Pleuronectoidea, but leave it incertae sedis in that superfamily.
5. Systematic palaeontology
Sept ACANTHOMORPHA Rosen, Reference Rosen, Greenwood, Miles and Patterson1973
Order PLEURONECTIFORMES Bleeker, Reference Bleeker1859
Superfamily PLEURONECTOIDEA Rafinesque, Reference Rafinesque1815
Family incertae sedis
Keasichthys gen. nov.
Type species. Species Keasichthys oregonensis sp. nov.
Etymology. The generic name is derived from a combination of Keasey, for the Keasey Formation, and the Greek ending ‘ichthys’ (fish), resulting in the name meaning Keasey fish. Gender is masculine.
Photograph of the holotype of Keasichthys oregonensis gen. et sp. nov. University of British Columbia Beaty Biodiversity Museum, specimen #3959. Scale bar is 1 cm.
Photograph (above) and interpretive drawing (below) of the head and pectoral girdle of the holotype of Keasichthys oregonensis gen. et sp. nov. University of British Columbia Beaty Biodiversity Museum, specimen #3959. Labelled bones are from the left side unless indicated. Scale bar is 1 cm.
Photograph of the pharyngeal teeth of the holotype of Keasichthys oregonensis gen. et sp. nov. University of British Columbia Beaty Biodiversity Museum, specimen #3959. Arrows point to some of the more distinct teeth and tooth bases. Scale bar is 1 mm.
Photograph of the pelvic fins of the holotype of Keasichthys oregonensis gen. et sp. nov. University of British Columbia Beaty Biodiversity Museum, specimen #3959. Arrow points to distal tip of the unbranched, segmented first ray. Scale bar is 1 mm.
Photograph (above) and interpretive drawing (below) of the caudal skeleton of the holotype of Keasichthys oregonensis gen. et sp. nov. University of British Columbia Beaty Biodiversity Museum, specimen #3959. Scale bar is 5 mm.
Holotype. An almost complete articulated fish, missing the anterior tip of the head, preserved in right lateral view. UBC Beaty Biodiversity Museum specimen #3959.
Type locality and age. Keasey Formation, near Mist, Oregon, USA; earliest Oligocene in age (Prothero and Hankins, Reference Prothero and Hankins2000).
Etymology. The specific epithet is for the state of Oregon, where the type locality is found.
Diagnosis. A member of Pleuronectoidea differing from included families by lacking autapomorphies of each given by Hoshino (Reference Hoshino2001b). Further differs from other members of the Pleuronectidae, with the exception of a few individuals of Pleuronectes isolepis, P. asper and P. sakhalinensis, by a low number of abdominal centra (nine, compared to ten or more in the other species); distinguished from P. isolepis, P. asper and P. sakhalinensis by a higher number of branched plus unbranched caudal fin rays (19 compared to 18), and further distinguished from P. isolepis by having 29 caudal centra (compared to 31–33). Data for extant taxa are from Sakamoto (Reference Sakamoto1984).
6. Description
6.a. General body form
The specimen is almost complete, with only the anterior tip of the jaws missing (Fig. 1). It is preserved in the right lateral view, with the blind side visible; therefore, it is a left-eyed individual. The total length of the preserved portion is 142.5 mm, and the standard length is 117.8 mm; the missing portion, which includes the anterior half of the dentary, is estimated to have been about 4–6 mm. The head is about one quarter of the standard length and would have been about as long as it is deep. The head (excluding the missing anterior tip) measures 28.6 mm in length and 33.7 mm in depth. Overall body shape of the fossil agrees well with some members of Pleuronectoidea, in particular Pleuronectidae and Paralichthyidae, in having a relatively long caudal peduncle and body that is not greatly deepened (body depth = 50.0 mm, less than ½ standard length). The caudal peduncle is 16.1 mm long and 12.7 mm deep. Scales have not been preserved anywhere on the body.
6.b. Skull
The bones of the skull are not well preserved (Fig. 2), with most of them represented by natural moulds. Individual elements cannot be easily discerned. The two orbits are both on the left side of the head. A large triangular bone anterodorsal to the left eye is identified as the right lateral ethmoid, meeting the parasphenoid anteriorly but the two bones diverging posteriorly. Posterior to these elements, the bones are poorly preserved, and the delineation of the individual elements cannot be determined. The right frontal is posterior to the right lateral ethmoid, and the supraoccipital crest is low. Of the oral jaws, a small bone identified as part of the maxilla, and the anguloarticular and posterior part of the dentary are preserved. These are presumed to be from the left side; the jaws of the other side are not visible. The mandible appears to be longer than tall, but the coranoid process is not well developed. A bone identified as the head of the left palatine is present articulating with the parasphenoid anteriorly. Details of the opercular series and suspensorium are not clear, with the exception of most of the quadrate being partially preserved in articulation with the posteroventral part of the ectopterygoid.
A few teeth and tooth sockets are preserved in the middle of the head, in a position that indicates they were associated with the fifth ceratobranchial. The teeth are short and stout, peg-like, with a rounded tip (Fig. 3).
6.c. Paired fins and girdles
The pectoral and pelvic fins are preserved, as well as natural moulds of part of the girdles (Fig. 2). There are 14 rays (no spines) visible in the left pectoral fin. The left cleithrum is gently curved but fairly narrow. The coracoid has a long ventral arm forming a foramen with the posterior edge of the cleithrum. The scapula is smaller than the coracoid and supports the fin rays. There are two long, thin bones posterior to the right cleithrum, which we identify as postcleithra, one from the left and one from the right side.
The left and right halves of the pelvic girdle are elongate narrow bones that do not meet along their medial edges. The pelvic girdle is angled from posteroventral to anterodorsal, with the anterior tips reaching the level of the posterior edge of the left cleithrum. The left and right pelvic fins are both present, low on the body and have a similar position on the blind and eyed side; i.e., one fin is not positioned anterior to the other. There are five pelvic rays each fin. There is no indication of any spine; the outermost ray of the pelvic fin is clearly segmented (Fig. 4).
6.d. Dorsal and anal fins
The fossil clearly preserves the dorsal fin extending onto the head and reaching to the level of the eye (Figs. 1, 2). There are 47 dorsal-fin rays preserved, but some are clearly missing because there are at least 58 pterygiophores. There are no spines in the fin, and none of the dorsal-fin rays are significantly elongated compared to the rest. The majority of the pterygiophores are thin, lacking significantly expanded anterior and posterior bony flanges, with the exception of the anterior four or five, which are broader in lateral view. There are seven pterygiophores anterior to the first neural spine, of which the six anteriormost are located above the skull, and appear to originate in the midline of the skull, not on the blind side.
The anal fin also reaches far anteriorly, originating just posterior to the pelvic fin rays (Figs. 1, 2). The first anal-fin pterygiophore is large and robust, curving around the abdominal cavity. It appears to support the first four rays. Based on a combination of pterygiophore and ray counts, there would have been 52 rays (no spines) in the anal fin.
6.e. Caudal fin and axial skeleton
The caudal fin is narrow and rounded (Fig. 1). There are a total of 19 rays in this fin, which we interpret as a lower lobe with three unbranched (the lower two unsegmented and the third segmented) and six branched rays and the upper lobe with seven branched and three unbranched rays (the dorsal-most one or possibly two unsegmented). Following Reference HoshinoHoshino (2001a), the formula for the caudal fin would be 2+1+6+7+2+1 or 2+1+6+7+1+2.
Although the caudal skeleton is preserved predominantly as a natural mould, several details can be seen (Fig. 5). The hypurals exhibit reduction in numbers, with a single hypural plate formed from fused hypural 1 + hypural 2 in the ventral lobe (identification of elements follows Chanet and Wagemans, Reference Chanet and Wagemans2001). The dorsal lobe has two hypural plates, the more ventral being the fused hypurals 3+4, with the fifth hypural remaining separate. The larger plate (hypural 3+4) is fused with the first preural centrum (= terminal half centrum). Whether or not there is a ball and socket joint between the lower plate (hypurals 1+2) and the terminal centrum as would be expected for members of family Pleuronectidae (e.g., Chapleau, Reference Chapleau1993; Chanet and Wagermans, Reference Chanet and Wagemans2001) is difficult to discern. We cannot confirm its presence, but we cannot refute it either. The parhypural is broad distally and narrows to a pointed proximal tip that does not reach the terminal half centrum. There is no evidence of distal splitting of the hypurals and parhypural. There is a single epural in the caudal skeleton and a single full neural spine on the second preural centrum.
There are 38 vertebrae in total, including the terminal half centrum. Of these, nine are in front of the first anal pterygiophore and so are considered abdominal, and the rest (29 centra) are considered caudal. There are no supraneurals bones. The haemal arches on the anterior caudal centra appear to be broad, extending from anterior to posterior of each centrum, but whether or not there was a lateral foramen in the arch cannot be determined.
7. Phylogenetic analysis
Our analysis of the data set of Hoshino (Reference Hoshino2001b) with the addition of the data for Keasichthys oregonensis recovered four equally-parsimonious trees with a tree length of 98 steps. The strict consensus tree (Supplementary data Fig. S1) shows the same resultant polytomies among the citharid genera as found by Hoshino (Reference Hoshino2001b:fig. 7) and branches fully resolved among the families. The addition of the data for the fossil did not change any previously recovered relationships. The resulting tree with the genera of citharids collapsed into a single familial lineage (Fig. 6) retains the three separate superfamilies as found by Hoshino (Reference Hoshino2001b:fig. 7) with Keasichthys oregonensis recovered as the sister-group to Hoshino’s clade ‘F’ which is equivalent to a redefined superfamily Pleuronectoidea (i.e., excluding Brachypleura). We include K. oregonensis within the Pleuronectoidea rather than naming a new group, to minimize proliferation of new grouping names; alternately, it could be considered a stem-Pleuronectoidea. The consistency index for the trees (with separate citharid genera) is 0.5306, and the retention index is 0.7527.
Cladogram of pleuronectiform relationships including the Oligocene fossil Keasichthys oregonensis gen. et sp. nov. based on our analysis using the data of Hoshino (Reference Hoshino2001b). The citharid genera are grouped as Citharidae in the image only; data were run for each genus separately.
The characters that support each lineage are those presented and discussed by Hoshino (Reference Hoshino2001b). The characters supporting the placement of Keasichthys oregonensis with the Pleuronectoidea are: fusion of hypurals 1 + 2 and 3 + 4 (character 15, state 1); fusion of hypurals 3 + 4 to preural centrum 1 (character 17, state 1) and second neural spine erect, not attached to the cranium (character 34, state 0, interpreted as a reversal by Hoshino, Reference Hoshino2001b). The families of Pleuronectoidea are united (to the exclusion of K. oregonensis) by having seventeen ‘principal rays’ (character 37, state 0), which was interpreted as a reversal by Hoshino (Reference Hoshino2001b) although he noted the number is variable in some included taxa. There are fewer than 17 rays in K. oregonensis (state 1).
8. Discussion
Keasichthys oregonensis joins a number of other named flatfishes from the early Cenozoic era that document the early evolution of the group. Pleuronectiformes likely arose during this time because the stem-pleuronectiformes Amphistium and Heteronectes, which have a morphology intermediate between flatfishes and other fish, are early Eocene in age (Friedman, Reference Friedman2008, Reference Friedman2012). Other fossil flatfishes known from Eocene deposits likely represent both suborders, with Psettodoidei (Psettodidae) and Pleuronectoidei (Bothidae and Soleidae) identified (see section 1.0 above). This indicates that the separation into the two different suborders likely occurred soon after the lineage arose.
In the Oligocene, three families of Pleuronectoidea are represented: Bothidae, Scophthalmidae and Pleuronectidae (Baciu and Chanet, Reference Baciu and Chanet2002; Sakamoto et al., Reference Sakamoto, Uyeno and Micklich2004) with the bothid and scophthalmid known from Rumania and the pleuronectid from Germany. The addition of Keasichthys oregonensis as an early-diverging member of Pleuronectoidea (or a stem-Pleuronectoidea) again indicates fairly rapid diversification among the families of Pleuronectiformes.
A number of Miocene species have been described, in particular from California, USA, and from Pliocene and Pleistocene deposits of Japan (Chanet Reference Chanet1997: tab. 1). Of those he reviewed (16 species), Chanet (Reference Chanet1997) only considered one (the Pliocene Pleuronectes platessa Newton, Reference Newton1882 from the UK) to be properly assigned to the family, and considered the rest to be Pleuronectiformes or Pleuronectoidei incertae sedis, or not to be flatfish at all. However, Chanet (Reference Chanet1997) noted that those he considered incertae sedis were designated as such because of a lack of preserved apomorphies that could be used to assign them to a specific family. While apomorphies may be necessary to definitively assign fossils to extant taxa, often they are not preserved (e.g., soft tissue characters) and the overall features of the fossils must be used to determine relationships, whether or not they are apomorphic. Many of these fossils, therefore, may indeed belong to families within Pleuronectoidea despite a lack of familial synapomophies.
Many of the fossil Pleuronectoidea previously reported are younger than the early Oligocene fossil described here and the pleuronectid from Germany (Sakamoto et al, Reference Sakamoto, Uyeno and Micklich2004). This includes Hippoglossoides macroptera Smirnov, Reference Smirnov1936 and Protopsetta parvula Smirnov, Reference Smirnov1936, both of which were reported from the Northern Caucasus. These two species were both considered to be dubious pleuronectiforms by Chanet (Reference Chanet1997:tab. 1) and listed as being Oligocene in age. Bannikov (Reference Bannikov2001) reviewed both these fish and noted H. macroptera was moved to the bothid genus Arnoglossus, and P. parvula was moved to Platichthys in the Pleuronectidae. However, Bannikov (Reference Bannikov2001) also reported these fish as coming from lower Miocene deposits, thus much younger than the early Oligocene Keasichthys oregonensis from Oregon and Oligopleuronectes germanicus from Germany. The other fossil species listed by Chanet (Reference Chanet1997), whether or not they are truly in the Pleuronectoidea, are from the late Miocene of California, middle Miocene to late Pleistocene of Japan and Pliocene of Sakhalin. Additional material not covered by Chanet (Reference Chanet1997) includes Pleuronectes asperoides Nazarkin, Reference Nazarkin1997 from Sakhalin, also of Miocene age.
The paucity of Oligocene fossil flatfish is likely a result of the type of fossiliferous environments known and the life history of these animals. Bannikov (Reference Bannikov2001) noted that benthic organisms such as flatfish are not often preserved as fossils, since these fish must live in oxygenated areas, but anoxic waters are best for fossil preservation. He further observed that those fossil flatfish skeletons that are known are rare in their respective locality and usually represent juvenile (i.e., metamorphosed but not fully adult) individuals since young flatfish often retain a more pelagic lifestyle compared to the benthic adults. These observations fit well with the new Keasey Formation fossil, which is the only articulated fish known from the formation, and is a small individual. The other fish material recovered from the formation comprises shark teeth and osteichthyan scales (Welton Reference Welton1979; David 1959). The fragmentary preservation of these other fishes, along with the reconstruction of the environment as being deep-water (Burns et al., Reference Burns, Campbell and Mooi2005), supports the idea that the locality was not one likely to be favourable to benthic flatfishes. The presence of Keasichthys oregonensis in the area may indicate that it was an immature member of the species that would not normally have inhabited the crinoid beds.
9. Conclusions
The articulated flatfish reported here is the only known articulated fish specimen from the Oligocene Keasey Formation. It can be included in the superfamily Pleuronectoidea based on a combination of features of the caudal fin, overall body shape and meristic characters. Articulated fossil flatfishes are generally fairly small individuals, which Bannikov (Reference Bannikov2001) noted was likely caused by the more pelagic nature of younger individuals compared to the benthic adults. Because of their oxygenated habitat, it is less likely that benthic adults will be recovered from fossiliferous sediments preserving articulated skeletons that are often deposited in anoxic environments. Because of this, rare benthic fishes such as flatfish are all the more intriguing when encountered.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0016756825100125
Acknowledgements
Our thanks to B. Archibald, Beaty Museum of Biodiversity, UBC, for lending the specimen and being enthusiastic about our working on it. The manuscript was greatly improved with comments from two reviewers and the Associate editor; we are grateful for their very helpful suggestions and thorough reviews. Funding for this research is from the Natural Sciences and Engineering Research Council of Canada Discovery Grant 327448 (AMM).
Competing interests
The authors declare none.
Open access
