2 Bull. Southern CaliforniaAcad. Sci. 105(1).2006,pp. 30-42 i SouthernCaliforniaAcademyofSciences.2006 A New Late Miocene Species of Sciaenid Fish, Based Primarily on an in situ Otolith from California Richard W. Huddleston1 and Gary T. Takeuchi2 1 Scientific Research Systems, 11044 McGirk Avenue, El Monte, California 91731 2 Department of Vertebrate Paleontology, Natural History Museum ofLos Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007 — Abstract. A new species of sciaenid, Seriphus lavenbergi sp. nov., is described from the late Miocene Yorba Member of the Puente Formation, southern Cali- fornia. The holotype is represented by an incomplete disarticulated skull with the right saccular (= sagitta) otolith in situ, an extremely rare occurrence. This is the earliest known geological occurrence of Seriphus and represents a second species within the genus. It is hypothesized that the ancestor of Seriphus emigrated from the Gulf Stream of the Atlantic into the Pacific via the Panama Seaway and the genus evolved entirely in the eastern Pacific. During the spring of 1969, students from Rio Hondo College (Whittier, Cali- fornia) collected fossils from a late Miocene marine site referred to locally as "Chalk Hill" in the western Puente Hills, City ofIndustry, California. Mr. Melvin Caskey, former Associate Professor of Geology at Rio Hondo College, allowed one of us (RWH) to examine some of the material collected. Among the speci- mens was a small block of diatomaceous shale, which contained a crushed, in- complete neurocranium ofa fossil fish that still contained both saccular (= sagitta) otoliths in situ. We describe a new species of the sciaenid genus Seriphus based on this specimen. Previous to this report, Seriphus was considered monospecific (see discussion) with its earliest geological occurrence in the late Pliocene (Fitch and Reimer 1967). The family Sciaenidae represents a strongly provinicalized worldwide group of nearshore fishes containing some 270 extant species (Nelson 1994). This family is confined to shallow, coastal, and estuarine environments with a few freshwater taxa. These fishes are most abundant in tropical and subtropical areas, diminishing in more temperate zones. Sciaenids are absent from the Ryukyu Islands, New Zealand, and Pacific Oceanic Island groups (Sasaki 1989). They are also absent from the North Sea and only a single species is recorded in southern Australia (Schwarzhans 1993). This strongly suggests that the open ocean regions and cold extreme temperate zones were effective barriers to the distribution of this group. The fossil record of Seriphus is confined to the extant species S. politus Ayres, 1860, and is represented only by isolated otoliths from the Plio-Pleistocene of California. The earliest geological occurrence of Seriphus is from the Fernando Formation, late Pliocene (Fitch and Reimer 1967). Seriphus politus has also been reported from the Timms Point Silt, early Pleistocene (Fitch 1968), from the late Pleistocene Sand Pedro Sand, and overlying Palos Verdes Sand (Kanakoff 1956; Fitch 1964, 1966, 1967, 1970). Fitch (1970) reported S. politus from the Crannell 30 A NEW SPECIES OF SERIPHUS FROM CALIFORNIA 31 Fig. 1. Map of the Los Angeles basin indicating type locality (LACM locality 6907) ofSeriphus lavenbergi, sp. nov. (LACM 55484), holotype, late Miocene, Yorba Memberofthe Puente Formation, California. Road locality near Areata, northern California and referred this site to be age- equivalent to the late Pleistocene Palos Verdes Sand of southern California. We have determined these deposits to represent the late Pleistocene Battery Formation and constitutes the northern most fossil occurrence of Seriphus. Locality and Geology The locality, Natural History Museum ofLos Angeles County (LACM) locality 6907, was an approximately 10 m vertical exposure on a small west-facing hill in the western Puente Hills, City of Industry, Los Angeles County, California (Fig. 1). This site is located approximately 31 km east of downtown Los Angeles, but was largely destroyed by construction activity in the 1980's. The stratigraphic thickness of the section is 100 m and is composed of finely laminated diatoma- ceous shale interbedded with gray siltstone and fine dark mudstone deposited on a submarine fan at bathyal depths (Durham and Yerkes 1964; Yerkes 1972; Critelli et al. 1995). Durham and Yerkes (1964) and Yerkes (1972) have assigned this interval to the lower part of the Yorba Member of the Puente Formation. The 32 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Puente Formation was named by Eldridge and Arnold (1907) for exposures in the Puente Hills in the eastern Los Angeles basin, and they recognized a lower shale member, an intermediate sandstone member, and an upper shale member. Daviess and Woodford (1949) informally divided the formation into four mem- bers, and subsequently, Schoellhamer et al. (1954) formally named the members: The La Vida Member, the Soquel Member, the Yorba Member, and the Sycamore Canyon Member. Recently, some confusion about the naming ofthe Puente Formation has arisen by the assignment of the La Vida, Soquel, and Yorba members to the Monterey Formation and by the elevation of the Sycamore Canyon Member to formational rank (Dibblee 1995; 2001). The Puente Formation is considered by most workers equivalent in age to the upper portion of the Monterey Formation; however, it is distinctly different in lithology and depositional environment. The Monterey For- mation consists of siliceous hemipelagic shale deposited at bathyal depth (Pis- ciotto and Garrison 1981), whereas the Puente Formation consists of more terrig- enous sediments derived from the southeastern San Gabriel Mountains deposited as turbidites on submarine fans at bathyal depths (Critelli et al. 1995). Thus, we use Yerkes' (1972) terminology for the purposes of this study. Based upon benthic foraminiferans, Durham and Yerkes (1964) placed a late Miocene (upper Mohnian) age to the Yorba Member of the Puente Formation. Benthic foraminiferan ages have been found to be time transgressive in the Ce- nozoic rocks of the Pacific coast of North America and are not, by themselves, reliable age indicators relative to ages derived from planktonic microfossils (Prothero 2001). Micropaleontological studies at the locality have yielded diag- nostic diatom floras that confirm the late Miocene age. The diatoms are assignable to the Thalassionema schraderi Zone, which indicates an age of 8.6-7.6 Ma (J. D. Stewart, pers. comra.). The Chalk Hill locality has produced a large, as yet undescribed, ichthyofauna dominated by mesopelagic forms with additional epipelagic as well as nearshore taxa (Table 1). A single specimen of the deep sea anglerfish Acentrophryne has been reported from the Chalk Hill area (Pietsch and Lavenberg 1980). Other megafossils recovered include an odontocete, an unidentified bird, algal remains, and terrestrial spermatophyte leaves (L. G. Barnes, pers. comm.; J. D. Stewart, pers. comm.). Invertebrates collected include Delectopecten pedroana Trask, My- tilus sp., Lepas sp., and an unidentified shrimp. Probable marine mammal cop- rolitic material has also been collected. Material and Methods Recent comparative materials used in this study are from the collections ofthe Department of Ichthyology, LACM, and includes skeletons and otoliths ofRecent sciaenids. Additional comparative materials came from an otolith collection of one of the authors (RWH). The holotype of Seriphus lavenbergi, sp. nov. (LACM 55484), is catalogued and housed in the collections of the Department of Vertebrate Paleontology, LACM. The holotype was discovered by splitting the diatomaceous shale along bedding planes using small hand tools in the field and was damaged in the process. This method ofrecovery often damages the specimens because the skeleton splits between the two sides of the shale. Additional preparation, using a fine needle A NEW SPECIES OF SERIPHUS FROM CALIFORNIA 33 Table 1. Late Miocene ichthyofauna at LACM locality 6907 Pomona Freeway Chalk Hill, Yorba Member ofthe Puente Formation, California. Data from Fitch (1969), Pietsch and Lavenberg (1980), J. D. Stewart pers. comm., and personal observation. Chondrichthye* Lamnidae Isurus oxyrinchus Rafinesque, 1810 Cetorhinidae Cetorhinus sp. Osteichthyes Clupeidae Ganolytes cameo Jordan in Jordan and Gilbert, 1919 Ganolytes sp. Xyne grex Jordan and Gilbert, 1919 Atherinidae Atherinops sp. Athehnopsis sp. Argentinidae Argentina sp. Bathylagidae Genus and species not determined Gonostomatidae Cyclothone sp. Sternoptychidae Argyropelecus sp. Chauliodontidae Chauliodus eximius (Jordan and Gilbert in Jordan, 1925) Myctophidae Lampanyctus sp. Genus and species not determined Linophrynidae Acentrophryne longidens Regan, 1926 Moridae Genus and species not determined Serranidae Paralabrax sp. Carangidae Decapterus sp. Sciaenidae Lompoquia sp. Seriphus lavenbergi sp. nov. (this paper) Zaphlegidae Thyrsocles kriegeri (Jordan and Gilbert, 1919) Scombridae Sarda sp. Scomber sp. Pleuronectiformes Family not determined and a dissecting microscope, was required to carefully remove the fragmented left saccular otolith. This otolith had been crushed, inner face down, on top of the right otolith, and it was removed to expose the underlying diagnostic inner surface of the right otolith for study. Morphological terms used in the general description follow Nolf (1985), Sasaki (1989), and Schwarzhans (1993). Linear measurements were made on an EPOI Shopscope optical micrometer. Otolith measurements for the Sciaenidae (Fig. 2) follow Schwarzhans (1993), and characters and their definitions for features of the saccular otolith are described below. Measurements used for proportional ra- tios on the inner face include the following: otolith length (L) is the greatest anterior to posterior length; otolith height (H) is the greatest dorsal to ventral height; ostium length (OL) is taken from the anterior edge of the ostium to the posterior most extension of the ostium, including the postostial lobe if present; ostium height (OH) is the greatest dorsal to ventral height of the ostium with the otolith in natural position; caudal length (CL) is measured from the dorsal edge of the cauda, at the dorsal caudal joint and extending to the posterior most ex- tension of the dorsal caudal margin; length of horizontal portion of the cauda (X) is measured from the caudal joint of the ventral caudal margin and extending to the posterior most extension of the ventral caudal margin; length of down turned portion of cauda (Y) is from the highest point of the ventral caudal margin to the 34 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES H Fig. 2. Innerface ofgeneralized sciaenid saccularotolith with measurements usedforproportional ratios. Abbreviations: cl, cauda length; h, otolith height; 1, otolith length; oh, ostium height; ol, ostium length; x:y, caudal curvature index (cci). Follows Schwarzhans (1993). posterior most point of the cauda termination. The measurement used for pro- portional ratios from lateral view is otolith thickness (T), which is the greatest inner face to outer face thickness taken in dorsal view. Systematic Paleontology Class Actinopterygii (sensu Nelson, 1994) Division Teleostei (sensu Nelson, 1994) Order Perciformes (sensu Johnson and Patterson. 1993) Family Sciaenidae Cuvier, 1829 Subfamily Sciaeninae Gill, 1861 Genus Seriphus Ayres, 1860 Sehphus lavenbergi sp. nov. Figs. 3-4; Table 2 — LACM Holotype. 55484, incomplete disarticulated skull with right saccular otolith in situ (Fig. 3). — Additional spe—cimens. The holotype is the only known specimen. Type locality. LACM locality 6907, Pomona Freeway Chalk Hill, City of Industry, Los Angeles County, California, USA; 17° 54' 41" W. longitude and 1 33° 59' 28"—N. latitude, La Habra 7.5 quadrangle (U.S.G.S). Horizon. The Yorba Member ofthe Puente Formation, late Miocene described by Durham an—d Yerkes (1964) and Yerkes (1972). Etymology. The holotype is named in honor of Robert J. Lavenberg, Curator Emeritus, Department of Ichthyology, LACM, who along with the late John E. Fitch and one of the authors (RWH) collected and initiated the first research on the Chalk Hill ichthyofauna. We acknowledge his friendship, assistance, and con- tributions to —the study of fossil and Recent fishes. Diagnosis. Seriphus lavenbergi, is distinguished from S. politus in the follow- ing combination of characters for the saccular otolith: 1) a more elongate overall A NEW SPECIES OF SERIPHUS FROM CALIFORNIA 35 Fig. 3. Photograph (A) and illustration (B) ofdisarticulated skull with right saccularotolith in situ ofSeriphus lavenbergi, sp. nov. (LACM 55484), holotype, late Miocene, Yorba Memberofthe Puente Formation, California. Abbreviations: acp, ascending process; arp, articular process; bo, basioccipital; fr, frontal; le, lateral ethmoid; mx, maxilla; ns, neural spine; p, pterygiophore; pmx, premaxilla; pro, prootic; psph, parasphenoid; pv, prevomer; rb, rib; saco, saccular otolith; sc, scale; v, vertebra. Scale bar equals 2 cm. 36 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES Fig.4. DrawingofrightsaccularotolithofSeriphuslavenbergi, sp. nov. (LACM55484),holotype, late Miocene, Yorba Member ofthe Puente Formation, California. Gray areas represent missing por- tions ofthe otolith. Scale bar equals 5 mm. shape (Fig. 4); 2) a low, nearly horizontal dorsal rim (Fig. 4); 3) a more blunted, nearly vertical posterior rim (Fig. 4); 4) a longer more anteriorly tapering ostium (Fig. 4); 5) a lower caudal curvature index (2.2 vs. 3.7 in S. politus; Table 2); and 6) the posterior cauda is more gently flexed posteroventrally (Fig. 4). The new species also differs from S. politus in the following: a premaxilla with a much longer alveolar process; and a maxilla that is strongly constricted between the anterior condyle and the posterior extension of the maxillary. — Discussion. The most diagnostic feature ofthe holotype is the saccularotolith, which has a broad, flat, shallow ostium. The cauda is narrow and horizontal with a curved posterior section. The sulcus is strongly homosulcoid with both the rostrum and antirostrum absent. These features are sufficient to place this speci- men within the family Sciaenidae. The sub-ovate otolith with a subangular ventral margin, ellipitical ostium, and posteroventrally-bent posterior cauda is character- istic of the gen—us Seriphus. Description. The holotype of Seriphus lavenbergi (Fig. 3-4) is a disarticu- lated incomplete skull with the right saccular otolith in situ. The specimen is preserved as brown bone in white diatomaceous shale. mm mm The neurocranium measures 48.20 in length and 21.24 at greatest width. The preserved elements of the skull are difficult to interpret because of damage. The frontals and lateral ethmoids appear to be preserved in ventral view. The paired triangular frontals are by far the largest bone in the skull roof and account for the anterior halfofthe neural cranium. The lateral ethmoids are paired, moderately thick and projecting lateroanteriorly, forming the anterior wall of the orbits. The edentulous prevomer is short and expands laterally. It is detached and preserved lateroanteriorly of the neurocranium. The nasals do not appear to be preserved in the holotype. The thick, flattened parasphenoid is detached and pre- served in ventral view anteriorly of the neurocranium. The anterior parasphenoid is attached to the prevomer. The paired prootics are large and broken, attached posteroventrally to the incomplete enlarged basiocceptals. The upper jaw consists of a maxilla and a premaxilla. The bone surfaces are all smooth without noticeable ornamentation. The premaxilla bears well-devel- oped ascending, articular, and postmaxillary processes. The ascending process is a thin blade-like structure and is higher than the articular process. The articular : 1 ! A NEW SPECIES OF SERIPHUS FROM CALIFORNIA 37 J—Z. —3 O 0o33 crr*-o«; osr^o ror^n^ rr-f: e©n A >^ -.C• M GW) "r03 h Oin vO _u 3- so t^- 00 SO o £ 'Jc"". <o+* o c —' — o ^ C « r- 065 ft x""sus55 ^GUc-3 1a(N; d(mN ta-tn orio -rhi X<! $Z -soa;sj >„ ^*>-nh ©T1tj0rc0ts. r°-^i —t .kg, Tejg <^ 13 'to+o- «4oon—, od di—n QSr^O- H— df- <O/J9 c "uo 'rt CCf1/>3t 2•cf—t ££ (asN. Cr—Oo< (©(NN 0s^0ot einn ©CO g c e X N 6 _c X a 00 qco cioo *coo h q c H CO "*f in © © »n Tt 3 _c £u j 3 CcJj 3 ^t O©N ON 0u3 JO srOrJj <SNO csso © csoo rojq ciin -3 U 3 09 co in Tf Tt ©(N SO OS CN «* in rt rt X 03 OJO <-H Tt 00 h O (N (©N 0©0 CN CN CN Oa 0o0 o o. u ^oo c—©o; m"* r-; rn in (N in in i-H in in CO so oq oq os © © © Os od O 03 £ -o O c u U o < Q> £o T3 u £3 i I. So3- w"C733 Uo 2£r33 S£cIX£) co CO 38 SOUTHERN CALIFORNIA ACADEMY OF SCIENCES process is large and rounded, while the alveolar process is narrow and incomplete. The postmaxilla process is also incomplete. The ventral edge of the alveolar process has a narrow row of small, uniform conical teeth. The maxilla is incomplete, and only the anterior portion is exposed. The maxilla has a distinct head with two condyles and a laterally flattened shaft. The dorsal- ventral width of the shaft is strongly constricted just posterior of the condyles. Other associated skeletal material includes four disarticulated abdominal ver- tebral centra with the parapophysis either poorly preserved or broken. Neural spines are either obscured by matrix or incomplete. Two pleural ribs are present. An isolated partially articulated segment of four dorsal fin pterygiophores are preserved near the posterior end of the neurocranium. The pterygiophores are bisegmental with well-ossified stays. Pterygiophores are broad, blade-like, and increase in size posteriorly. Two additional detached pterygiophores are present, anterior and posterior of the neurocranium. Numerous isolated scales and scale impressions on the slab are scattered around the neurocranium and are of the typical sciaenid type. They are ctenoid with the focus positioned posterior to the center of the scale. There are 12-13 anterior (basal) radii. The right saccular otolith (Fig. 4) is moderately elongate with a total length of 10.3 mm and a greatest height of 5.1 mm. Measurements and proportional ratios of the otolith are given in Table 2 and Fig. 2. The dorsal rim is low, flat, and nearly horizontal with a slight undulation. The anterior end is sharply rounded, lacking both rostrum and antirostrum, and the ventral rim is deep and strongly angular, reaching greatest depthjust posterior ofthe vertical midline ofthe otolith. The posteroventral margin is short and angled posterodorsally, with the posterior margin blunted. The ostium is large, broad, shallow, subelliptical, and elongated, tapering anteriorly. The area of the precaudal depression is not adequately pre- served for description. The cauda is narrow, with its anterior long and nearly horizontal. The posterior cauda is greatly flexed posteroventrally and terminates near the posteroventral margin. Discussion and Conclusions The genus Seriphus is generally acknowledged as monospecific, with S. politus as the only included species. Recently, Schwarzhans (1993) incorporated an ad- ditional extant species, Cynoscion striatus, within Seriphus. This assignment was based solely on similarities in otolith morphology and excluded the differences in osteology and soft body anatomy upon which the relationships of the modern sciaenids are based. A phylogenetic analysis for the Sciaenidae by Sasaki (1989) that used morphological, osteological, and myological evidence concluded that Seriphus is broadly separated from Cynoscion by possessing the following char- acters: an enlarged and anteriorly located toothplate on the pharyngobranchial 2, the flexor ventralis externus fades into the flexor ventralis, the basiphenoid is separate from the parasphenoid ventrally, a dentary foramen is present, there is a secondary reversal from an enlarged and anteriorly located tooth plate on phar- yngobranchial 2, the posterior dorsal fin spines are not exposed, and the soft dorsal and anal fin bases are of equal length. Sasaki's (1989) phylogenetic analysis placed Seriphus and related genera near Cheilotrema and Cillus. The saccular otolith of S. politus and Cynoscion striatus display some super- A NEW SPECIES OF SERIPHUS FROM CALIFORNIA 39 ficial similarities, however, a closer examination as part of this study revealed significant differences. The otoliths of C. striatus differ from Seriphus by being more elongate and subrectangular. They are considerably thicker with a small, more elliptical ostium. The anterior cauda is relatively longer, and the posterior cauda is extremely short and sharply curved. These features are more consistent with the interspecific variation seen in the species of Cynoscion. Based on these observations and the differences in the osteological and soft-bodied anatomy, we consider the placement of C. striatus within Seriphus untenable. Thus, we rec- ognize only a single extant species within Seriphus. The occurrence of S. lavenbergi in the late Miocene Yorba Member of the Puente Formation suggests that Seriphus appears to have evolved entirely in the subtropical eastern Pacific during, or prior to, the late Miocene. The Sciaenidae display pronounced provincialism with most genera confined to a specific bio- province. The greatest diversity and highest number of species within the Sciaen- idae primarily concentrate in two distinct areas of the world: Tropical America (on both Pacific and Atlantic sides) and the Indo-West Pacific, particularly along the coasts from India to China. A third more isolated endemic bioprovince occurs along the tropic West Africa coast. These areas may represent the evolutionary centers for sciaenid radiation and dispersal (Schwarzhans 1993). In his analysis of the monophyletic groups within the Sciaenidae, Sasaki (1989) noted that the family possibly originated in a restricted area ofthe New World (= eastern Pacific and western Atlantic) with successive dispersals extending the range ofthe family and increasing the diversity of the genera and species. The North American fossil record ofthe Sciaenidae strongly suggests that Trop- ical America was the possible major evolutionary center for this family. The earliest geologic occurrence of the family is from the middle to late Eocene Gulf Coast region of North America (Koken 1888; Frizzell and Dante 1965; Breard and Stringer 1999; Nolf 2003; Nolf and Stringer 2003). Eight fossil sciaenid species have been described with an additional specimen assigned to a Recent species. Of these, five have been recognized as valid by Schwarzhans (1993). The fauna is dominated by very plesiomorphic genera that includes ?Umbrina pseu- doradians, Frizzellithus gemma, Eokokenia epporrecta, Jefitchia claybornensis, and Jefitchia copelandi. None of the sciaenids from the Eocene through Miocene Gulf Coast region display any ancestral relationship to Seriphus. Dispersal of the family into the eastern Pacific from the Gulf Coast region occurred by way of the Panama Seaway sometime prior to the early Miocene. Assuming fossil sciaen- ids had similar ecological preferences as extant sciaenids, they probably used the existing coastline of North America to radiate into the eastern Pacific. Final clo- sure of the Panama Seaway with the uplift of the Isthmus of Panama occurred about 3.5 Ma (Coates et al. 1992) and divided a former single bioprovince in two. Otolith-rich Eocene deposits along the eastern Pacific of North America show a conspicuous absence of sciaenids. They first appear in the early Miocene (ca 23 Ma) of the west coast of North America within the Jewett Sand Formation in central California (Huddleston, pers. obs.). This small poorly preserved fauna includes several undescribed sciaenids. By the late early Miocene (ca 16.7 Ma), the family was well established in the eastern Pacific, and this is reflected in the occurrence of several thousand sciaenid otoliths representing more than a dozen different taxa from the late early Miocene Olcese Sand of central California