Remarkable dominance of myctophid otoliths in Upper Miocene Chagres Formation, Caribbean Panama

Sampling

The fossil site is located about ∼500 m northeast of Piña town, Colón, Panama (Fig. 1; 9°17′09.11″N, 80°02′41.28″W). This locality corresponds to the Piña Norte site described by Stiles et al. (2022) and the STRI site 650009 in Pyenson et al. (2015). The Chagres Sandstone Member of the Chagres Formation is exposed along the Caribbean shoreline, especially at low tide (Fig. 2B). Fragments of echinoids, mollusks, and fish otoliths, as well as trace fossils, are easily visible on the surface of the siltstone (Stiles et al., 2022). Surface sediments are yellowish to brownish, contrasting with fresh sediments located several centimeters below the surface, which are darker and have a distinct blue-gray color. We collected thirty-three bulk sediment samples laterally from the same stratigraphic level in this fresher, blue-grey material in 2018 and 2024 (Fig. 2A). Samples weighed on average 0.6 kg each (Table S1). An additional larger bulk sample (unweighted but less than 2 kg, CH18-1-1) was also collected from this fresh sandstone layer for otolith extraction (Tables S1, S2). Permits for collecting and exporting paleontological samples were issued by the Ministerio de Comercio e Industrias (MICI, SE/AO-4-18) in Panamá.

Otolith preparation, imaging, and identification

Bulk sediment samples were disaggregated using freeze-thaw cycles and Glauber’s Salt (saturated sodium sulfate solution) methods (Hanna & Church, 1928; Herrig, 1966), then wet-sieved through a 500-µm mesh. After sieving, sediments were dried overnight in an oven at 40 °C. Otoliths larger than this 500-µm mesh were carefully hand-picked under a stereomicroscope. In this study, the term “otolith” refers to the saccular otolith (sagitta). Representative otoliths were photographed using a digital camera adapted to a Nikon SMZ1270 stereomicroscope, and image stacking was performed using Helicon Focus software. Final figures were prepared using Adobe Photoshop.

For otolith identification and terminology, we followed key references, including Rivaton & Bourret (1999), Lin & Chang (2012), Schwarzhans (2013b), Schwarzhans & Aguilera (2013); Schwarzhans & Aguilera (2016), Nolf (2013), and Haimovici et al. (2024). In addition, direct comparisons were made with a reference collection of extant otoliths housed at the Biodiversity Research Museum, Academia Sinica, Taiwan (BRCAS) under the code CHLOL. Whenever possible, otoliths were identified to the species level. All collected specimens are stored at BRCAS, and figured specimens are archived under the registration code ASIZF.

Sampling completeness and biodiversity analysis

Due to the overwhelming number of otoliths in each sample, we conducted biodiversity analysis to assess sampling completeness and diversity. Otolith abundances were quantified by dividing otolith counts by the corresponding dry sediment weight (kg) for each sample (except CH18-1-1). Family-level abundances were calculated by summing otolith counts across all samples, and families were ranked by total abundance. To evaluate statistical uncertainty in these abundance estimates, binomial 95% confidence intervals were computed using Wilson’s method. We computed these with and without the unweighted sample CH18-1-1. Diversity was estimated using Hill numbers (Hill, 1973) calculated at three different orders: q = 0 (⁰D, species richness), q = 1 (¹D, Shannon diversity), and q = 2 (²D, Simpson diversity), representing total (alpha) species richness, abundant species diversity, and dominant species diversity, respectively (Chao, Chiu & Jost, 2014; Chao et al., 2020; Lin et al., 2023a). Rarefaction and extrapolation methods were applied to create species accumulation curves, with 1,000 bootstrap resampling iterations used to estimate 95% confidence intervals. Specimen-based abundance data was analyzed to evaluate sample coverage comprehensively. All analyses were conducted using the R package iNEXT (Chao, Chiu & Jost, 2014; Hsieh, Ma & Chao, 2016).

Nomenclatural acts for new species

The electronic version of this article in Portable Document Format (PDF) will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is: urn:lsid:zoobank.org:pub:996A25D1-9CB7-4AAD-9041-0ABCF49710C5. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central and CLOCKSS.

Systematic paleontology

A list of identified taxa and their abundances is presented in Table 1. Classification scheme follows Nelson, Grande & Wilson (2016). Morphometrics and measurements include otolith length (OL), otolith height (OH), sulcus length (SuL), ostium length (OsL), and cauda length (CaL). Descriptions and discussions for common or previously described taxa are provided briefly under remarks, while new species include detailed descriptions and diagnostics.

Holotype: ASIZF 0100943 (Fig. 3A), Piña Norte, Panama. Upper Miocene, Chagres Formation. OL = 5.48 mm, OH = 4.23 mm.

Paratype: One specimen: ASIZF 0100944 (Fig. 3B), same data as holotype. OL = 1.85 mm, OH = 1.52 mm.

Etymology: The species name aflorens is derived from the Spanish word “afloramiento”, meaning “upwelling”. It refers to the flourishing productivity and dynamic marine conditions of the coastal upwelling system in which this species lived. It also symbolically reflects the scientific blossoming of paleontological research in Panama.

Diagnosis: OL/OH = 1.20–1.30, OL/SuL = 1.55–1.80. Otoliths oval with thick profile. Dorsal rim dome-shaped, evidently elevated anterior to midline; ventral rim smoothly curved. Sulcus moderately wide, poorly differentiated into ostium and cauda. Cauda short with an obtuse posterior tip.

Description: Otoliths oval to elliptic, thick; inner and outer faces highly convex. Anterior and posterior rims pointed in holotype, blunt to nearly vertical in juvenile paratype. Dorsal rim dome-shaped, elevation just anterior to midline. Ventral rim smoothly curved. Sulcus median, very slightly inclined (∼15°), with ostium and cauda only faintly differentiated due to the indistinct collum, resulting in a nearly continuous sulcus. Ostial channel nearly indiscernible in holotype, but narrow, vertical, and ending just before dorsal elevation in paratype. Cauda short, extending slightly posterior to dorsal elevation; bears obtuse, truncated posterior end.

Remarks: Chiloconger is distinguished from other congrids by the notably short cauda, a character state considered plesiomorphic (Schwarzhans, 2019; Schwarzhans & Nielsen, 2021). Although based on small and not fully morphologically mature specimens, the new species differs from the two extant species, C. dentatus (Garman, 1899) and C. philippinensis Smith & Karmovskaya, 2003, by having a less pronounced, more anteriorly positioned dorsal elevation (Schwarzhans, 2019). Additionally, compared to the Early Miocene Chiloconger chilensis Schwarzhans & Nielsen, 2021 from Chile (Schwarzhans & Nielsen, 2021), C. aflorens exhibits a more compact and rounded shape.

Occurrence: Currently known only from the Piña Norte locality, Panama (Upper Miocene, Chagres Formation).

Remarks: A single, fragmentary otolith exhibiting key Rhynchoconger characteristics is identified to the genus level. The preserved portion shows a discernible ostial channel and anterior ostium outline, diagnostic for Rhynchoconger (Schwarzhans, 2019). Due to the incomplete preservation, further species-level identification is not possible.

Remarks: A single thin otolith is assigned to the genus Argentina based on a strong, nearly orthogonal posterodorsal angle, a straight posterior rim, a curved ventral rim, and a median, horizontally oriented sulcus. The anterior portion of the specimen is missing, preventing confident assignment to species level.

Remarks: Polyipnus otoliths are distinctive by their tall, compressed shape, indistinctive sulcus outline, thin, slender rostrum bearing most of the ostium, elevated colliculum crest along the crista inferior, and a considerable thickness in the posterior rim. However, species-level identification is challenging due to extensive interspecific overlap and the limited availability of modern otolith reference material from the region, and the rostrum is very fragile and usually broken off in the fossil record, seriously hampering species definitions and recognition. Additionally, most specimens from the Chagres Formation are poorly preserved, with only the thick posterior rims preserved. The better-preserved specimens are depicted here.

Remarks: The abundance of myctophid otoliths in the Chagres Panama is extraordinary. Most identifications are based on large subadult to adult individuals; tentative assignments for juvenile or poorly preserved specimens were conservative, resulting in a significant number of otoliths classified as Myctophidae indet. Consequently, the true myctophid diversity is likely underestimated in this collection. Significant advances in myctophid otolith taxonomy, sourced by the development of comprehensive global reference collections (Rivaton & Bourret, 1999; Schwarzhans, 2013b), have greatly improved identification in fossil assemblages (Schwarzhans et al., 2022; Lin et al., 2023b). We primarily follow Schwarzhans & Aguilera (2013) for taxonomic treatment and species-level identification of the myctophid otoliths in this study.

Remarks: Otoliths of B. pluridens are relatively common in the collection. They are characterized by a sub-rectangular outline, a relatively flat dorsal rim, and multiple ventral denticles. However, juvenile myctophid otoliths often display intermediate outlines between rounded and sub-rectangular forms. To ensure accuracy, only specimens closely matching those illustrated by Schwarzhans & Aguilera (2013: pl. 1, figs. 8–12) were assigned to B. pluridens; other, less definitive specimens were conservatively classified under Myctophidae indet. Therefore, the true abundance of B. pluridens may be underestimated.

Remarks: A single, juvenile otolith is assigned to Bolinichthys based on its prominent rostrum and gently curved, oblique posterior rim, consistent with diagnostic features of the genus (Rivaton & Bourret, 1999). Notably, Bolinichthys otoliths have not been previously reported from the Neogene of tropical America (Schwarzhans & Aguilera, 2013).

Remarks: Following the molecular phylogenetic revision by Martin et al. (2018), seven species previously assigned to Myctophum (M. asperum, M. brachygnathos, M. lychnobium, M. obtusirostre, M. orientale, M. selenops, and M. spinosum) were reallocated to Dasyscopelus. We adopt this updated classification herein, although we note that, in our view, otolith morphology among these species does not consistently show clear differentiation from other Myctophum species or internally within Dasyscopelus itself (see Schwarzhans & Aguilera, 2013: pl. 3). Otoliths attributed to Dasyscopelus are typically characterized by a pronounced posterior extension and a pointed ventral rim, giving them a more angular appearance compared to the generally rounded and often deeper-bodied otoliths of Myctophum. Exceptions include D. brachygnathos and D. selenops, which possess a much shorter posterior rim and a taller overall shape (Schwarzhans & Aguilera, 2013; Ng et al., 2024a).

Three closely related species—Dasyscopelus degraciai (Schwarzhans & Aguilera, 2013), Dasyscopelus jacksoni (Aguilera & Rodrigues de Aguilera, 2001), and the newly described Dasyscopelus inopinatus sp. nov. (see below)—are here allocated to the genus Dasyscopelus due to their otoliths having a protruding posterior part, a relatively flat dorsal rim, and general similarity to extant Dasyscopelus species, such as D. asper and D. lychnobius.

Remarks: The otoliths of D. degraciai are distinguished by their compact outline, less pronounced posterior extension, gently curved dorsal rim, and narrower sulcus. They differ from the co-occurring D. inopinatus sp. nov. and the Pliocene D. jacksoni by their shorter posterior tip, more compact, less angular shape, narrower ostium, and the more delicate crenulation of the otolith rims.

Holotype: ASIZF 0100962 (Fig. 5I), Piña Norte, Panama. Upper Miocene, Chagres Formation. OL = 3.61 mm, OH = 2.43 mm.

Paratypes: Five specimens: ASIZF 0100957–0961 (Figs. 5G–5H, 5J–5L), same data as holotype. OL = 3.28–4.06 mm, OH = 2.31–2.66 mm.

Additional material: 57 specimens, unfigured, same data as holotype.

Etymology: From Latin inopinatus (feminine inopinata) = unexpected, alludes to the surprising and remarkable discovery of this new species, which exhibits a mosaic of morphological features shared with closely related congeners.

Diagnosis: OL/OH = 1.40–1.65 (mean = 1.48, n = 10), OL/SuL = 1.15–1.25 (mean = 1.23, n = 6), OsL/CaL = 1.45–2.10 (mean = 1.76, n = 10). Elongate otoliths with thin profile. Dorsal rim flat, nearly horizontal, with slight or no posterior elevation and a postero-dorsal angle. Ventral rim curved, bearing around eight lobes or denticles. Posterior rim short, nearly vertically straight. Sulcus very wide, with large, rectangular ostium and squarish cauda.

Description: Otoliths elongate, thin; both inner and outer faces are relatively flat. Anterior rim bearing large, protruding rostrum and shorter but conspicuous antirostrum, separated by clearly defined notch (excisura), varies from shallow to deep. Dorsal rim nearly horizontally flat, occasionally slightly elevated posteriorly, forming postero-dorsal angle just anterior to marked postero-dorsal concavity. Posterior rim short and nearly vertically straight. Ventral rim broadly curved, bearing ∼eight lobes or denticles. Sulcus wide, median-positioned. Ostium large, deep, rectangular; cauda short, squarish.

Remarks: The otoliths of the new species exhibit an intermediate morphology between D. degraciai and D. jacksoni. They share with D. degraciai a compact, deep-bodied outline and a less elongate posterior part, but resemble D. jacksoni in having a horizontal dorsal rim, wide sulcus, and a postero-dorsal concavity. Notably, the original illustrations of D. jacksoni (as Lampadena jacksoni in Aguilera & Rodrigues de Aguilera, 2001: figs. 7.15–7.22) suggest a more elongated and posteriorly extended shape compared to the ones documented in Schwarzhans & Aguilera (2013: pl. 5, figs. 6–10), as also indicated by differences in OL/OH ratios (1.60–1.85 vs. 1.50–1.65), although this may reflect ontogenetic variation.

Schwarzhans & Aguilera (2013) proposed a species turnover from the Late Miocene D. degraciai to the Pliocene D. jacksoni at the boundary between Messinian and Zanclean and further suggested a linear relationship between the two species. However, the finding of D. inopinatus suggests a more complex evolutionary history, with this new species likely representing a transitional or closely related lineage to D. jacksoni.

Occurrence: Currently known only from the Piña Norte locality, Panama (Upper Miocene, Chagres Formation).

Remarks: Diaphus represents the most abundant and diverse taxon in our collection. A total of seven species, D. aequalis, D. apalus, D. barrigonensis, D. dumerilii, D. multiserratus, D. pedemontanus, and D. rodriguezi, are recorded. Our identifications are primarily based on larger, more mature specimens, which exhibit sufficient diagnostic characters to allow confident assignment (see remarks under the family). The otolith taxonomy of Diaphus has been extensively revised by Schwarzhans & Aguilera (2013), whose detailed descriptions and high-quality images serve as the primary references for this study. Therefore, only brief remarks distinguishing among similar co-occurring species are presented here.

Remarks: The species is common but never abundant within the Piña assemblage. Its otoliths are morphologically most similar to smaller individuals of the co-occurring D. apalus and, to a lesser extent, D. barrigonensis (see below). It differs from D. apalus by its very rounded and compact outline, and by possessing only a subtle postero-dorsal concavity (vs. a pronounced, deeply indented concavity). Compared to D. barrigonensis, D. aequalis is distinguished by its shorter rostrum and gently curved dorsal and posterior rims (vs. steeply inclined dorsal rim and sharp posterior rim).

Remarks: See remarks under D. aequalis.

Remarks: See remarks under D. aequalis.

Remarks: The otoliths of D. dumerilii are most recognizable by their slightly elevated antero-dorsal rim and the narrow antero-ventral part of the rostrum. However, as noted by Schwarzhans & Aguilera (2013), these otoliths are morphologically inconspicuous and, in our view, confident identification is generally limited to larger, well-preserved specimens. Some otoliths assigned to D. dumerilii by Nolf (1976) from Neogene deposits in Trinidad display a more compact outline (e.g., pl. 3, figs. 8, 11–12), suggesting they may actually belong to different species. Nevertheless, distinguishing such variations based solely on the available figures remains difficult.

Remarks: Otoliths of D. multiserratus are readily distinguished by their elongate outline, wide sulcus, and, most conspicuously, the presence of numerous minute denticles along the ventral rim. Although not a frequent species in the collection, it is typically represented by larger, well-preserved specimens.

Remarks: Diaphus pedemontanus closely resembles the co-occurring and most abundant species D. rodriguezi but can be distinguished by a high, more undulating dorsal rim and a short, nearly vertically straight posterior rim, whereas D. rodriguezi has a gently curved dorsal rim, a deeper and wider excisura, and a larger rostrum (see below). Otoliths of D. pedemontanus are widely recorded from the Miocene and Pliocene of the Mediterranean (Girone, Nolf & Cavallo, 2010; Lin, Girone & Nolf, 2015; Lin et al., 2017a) and have also been documented in the coeval Caribbean assemblages (Schwarzhans & Aguilera, 2013). Ontogenetic variation in D. pedemontanus is considerable (Brzobohatý & Nolf, 2000), although we note that Caribbean specimens tend to be smaller than their European counterparts.

Remarks: See remarks under D. pedemontanus.

Remarks: A single otolith, characterized by a rounded outline (OL/OH = 1.05) and a thick profile, is assigned to the genus Diogenichthys (see Schwarzhans & Aguilera, 2013). However, due to partial damage along the anterior rim and the absence of additional material for comparison, species-level identification was not attempted.

Remarks: Otoliths of this species are very small, but highly diagnostic within the assemblage. They are particularly recognized by a subtle ventral inflection along the ostial crista inferior, although this feature is variably preserved among specimens. These otoliths also resemble much to those of Lampanyctus latesulcatus from the Tortonian Mediterranean; however, L. latesulcatus displays a more compact outline and lacks the ventral inflection characteristic of L. inflectus (Nolf & Steurbaut, 1983).

Remarks: Lobianchia johnfitchi is readily distinguished from other myctophid otoliths by its wide sulcus, prominently elevated antero-dorsal rim, and strongly depressed postero-dorsal rim. A closely related extant congener, L. dofleini (Zugmayer, 1911), exists during the Miocene–Pliocene in the NE Atlantic and Mediterranean (Lin et al., 2017a). We follow Schwarzhans & Aguilera’s (2013) interpretation that L. johnfitchi persisted in the Caribbean until the Middle Pliocene, whereas L. dofleini continued its presence into the modern Atlantic and Mediterranean.

Remarks: Myctophum affine is recognized by its very rounded and relatively flat otolith outline. The specimens from the Piña assemblage agree well with extant M. affine otoliths (Nolf & Stringer: pl. 10, figs. 16–17; Schwarzhans & Aguilera: pl. 3, figs. 17–18). It differs from the closely related fossil species Myctophum arcanum Schwarzhans & Aguliera, 2013 by having a shorter posterior extension and a slightly curved dorsal rim (see below).

Remarks: See remarks under M. affine.

Remarks: Two incomplete otoliths are assigned to Coelorinchus based on the presence of the characteristic pince-nez-shaped sulcus (homosulcoid-type) and the presence of a collicular crest at the collum. However, due to the fragmentary preservation, further identification beyond the genus level is not possible.

Remarks: Bregmaceros otoliths are representatives of the third most common family in the Piña assemblage, although nearly all specimens are poorly preserved. They typically show a heavily abraded inner surface, such that the ostial and caudal depressions are not preserved (Figs. 9C–9D), while the general outline remains intact, complicating detailed taxonomic assignment. The best-preserved specimen is illustrated in Fig. 9E. Due to their preservation state, the specimens are conservatively identified only to the genus level.

Holotype: ASIZF 0101006 (Fig. 10A), Piña Norte, Panama. Upper Miocene, Chagres Formation. OL = 5.54 mm, OH = 5.38 mm.

Paratypes: Three specimens: ASIZF 0101007–1009 (Figs. 10B–10D), same data as holotype. OL = 2.60–6.24 mm, OH = 2.52–5.67 mm.

Etymology: Named in honor of Brígida De Gracia (Boya in Ngäbere, the language of the Ngäbe-Buglé people) for her outstanding contributions to scientific research, public communication, and outreach activities in Panamá. The Ngäbe and their ancestors have inhabited the Isthmus of Panama for millennia, developing traditional ecological knowledge deeply connected to marine productivity cycles. Historical records demonstrate the Ngäbe’s reliance on seasonal fish abundance driven by upwelling systems along Panama’s Caribbean coast (Cybulski et al., 2025), creating a meaningful temporal bridge between the ancient upwelling ecosystem preserved in the Chagres Formation and the traditional knowledge systems that have recognized and depended upon these productive marine environments through time.

Diagnosis: OL/OH = 1.00–1.15, OL/SuL = 1.15–1.25, OsL/CaL = 0.70–0.90. Tall, sole-shaped otoliths with thick profile. Dorsal rim dome-shaped, evidently elevated posterior to midline; ventral rim either horizontally straight or smoothly curved, occasionally with large undulations. Sulcus very broad, shallow, median, well-differentiated into ostium and cauda. Ostium subtriangular, opening widely antero-dorsally. Cauda broad, rectangular.

Description: Otoliths tall, sole-shaped, thick; outer face strongly convex, inner face nearly flat. Dorsal rim curved, markedly elevated posterior to midline. Anterior rim with obtuse, upward-directed rostrum. Posterior rim straight, strongly inclined between postero-dorsal and postero-ventral angles, especially pronounced in larger specimens. Ventral rim horizontally straight to smoothly curved, occasionally with large undulations. Sulcus shallow, very broad, median; ostium subtriangular, opening widely antero-dorsally with shallow colliculum; cauda rectangular, broad, shallow. Dorsal depression fan-shaped, moderately deep.

Remarks: Among the six extant Hoplostethus species inhabiting the pan-Caribbean and East Pacific (H. atlanticus, H. fragilis, H. mediterraneus, H. mento, H. occidentalis, and H. pacificus), H. boyae most closely resembles juvenile otoliths of H. occidentalis (Haimovici et al., 2024: p. 139, but see Conversani et al., 2017: pl. 8). Other extant species typically exhibit a more variable dorsal rim, including flattened or undulating forms, often with digitiform projections, which may be a reflection of ontogenetic variation (Kotlyar, 1996). The new species differs by its consistently gently curved, dome-shaped dorsal rim, observed across both juvenile and adult stages. Compared to fossil congeners, such as the European Miocene species Hoplostethus praemediterraneus Schubert, 1905, and the Pliocene Hoplostethus pisanus Koken, 1891, H. boyae exhibits a more elevated dorsal profile and a shorter posterior rim, making the overall outline more compact and erect.

Occurrence: Currently known only from the Piña Norte locality, Panama (Upper Miocene, Chagres Formation).

Remarks: All Carapus otoliths in the Piña assemblage are small and represented by juvenile specimens. They closely resemble an undescribed fossil Carapus otolith illustrated by Schwarzhans & Aguilera (2016: fig. 12), although preservation quality in our material is poorer. Given the incomplete preservation and small size, the specimens are conservatively assigned to the genus level. Schwarzhans & Aguilera (2016) further referred to a larger specimen from the Pliocene of Jamaica, depicted by Stringer (1998), but in our view, confirming such a connection requires additional material.

Remarks: A comprehensive review on the otolith taxonomy of fossil Ophidiidae from the region, and particularly the genus Lepophidium, has been provided by Schwarzhans & Aguilera (2016). The juvenile otoliths assigned to Lepophidium limulum closely match the type material illustrated by Schwarzhans & Aguilera (2016: fig. 56). The species is characterized by a gently declining dorsal rim, a moderately proportioned sulcus with a slight ventral notch at the posterior of cauda.

Remarks: A peculiar, thickset otolith is assigned to Opistognathus based on its distinctive sulcus morphology. The ostium curves sharply upward anteriorly, while the cauda initially bends slightly upward before flexing ventrally in its posterior portion. This sulcus configuration matches the pattern seen in extant Opistognathus otoliths (see Nolf & Stringer, 1992: pl. 15, fig. 10). Due to limited material, identification is restricted to the genus level.

Remarks: Two thin otoliths are assigned to the family Carangidae based on their pronounced concavity of the outer face and the presence of a typical percomorph-type sulcus. However, due to incomplete preservation, particularly of the anterior regions, further identification to genus or species level was not attempted.

Holotype: ASIZF 0101015 (Fig. 12A), Piña Norte, Panama. Upper Miocene, Chagres Formation. OL = 2.78 mm, OH = 2.29 mm.

Paratypes: Five specimens: ASIZF 0101016–1020 (Figs. 12B–12F), same data as holotype. OL = 1.89–4.55 mm, OH = 1.53–4.14 mm.

Additional material: Nine specimens, unfigured, same data as holotype.

Etymology: Named in honor of Werner Schwarzhans (Natural History Museum of Denmark) for his outstanding contributions to the study of fossil and extant fish otoliths, particularly in tropical America.

Diagnosis: OL/OH = 1.10–1.30 (mean = 1.20, n = 6), OL/SuL = 1.05–1.15 (mean = 1.10, n = 4), OsL/CaL = 0.60–0.90 (mean = 0.61, n = 6). Pentagonal otoliths with thick profile. Dorsal rim gently angled, highest anterior to midline; ventral rim gently angled or curved, deepest slightly anterior to midline; posterior rim nearly vertically straight. Sulcus broad, median, shallow, clearly divided into ostium and cauda. Ostium oblong, filled with colliculum, opening widely antero-dorsally. Cauda horizontally straight, narrow, slightly flexed at tip, nearly reaching posterior rim.

Description: Otoliths pentagonal, thick; thickness mostly from convex outer face umbo, inner face slightly convex. Dorsal rim gently angled, highest point anterior to midline, ending in postero-dorsal angle (most manifest in larger specimens). Ventral rim curved or subtly angled, deepest point slightly anterior to midline. Posterior rim nearly vertically straight. Sulcus broad, median, bounded by well-developed cristae. Ostium oblong, filled with colliculum, opening widely antero-dorsally; ostial crista superior markedly bent antero-dorsally, crista inferior gently curving upward. Cauda horizontally straight, narrow, nearly reaching posterior rim, slightly flexed at tip. Dorsal depression shallow, wide.

Remarks: The pentagonal outline of M. schwarzhansi is superficially similar to otoliths of several other families, such as Lactariidae and Epigonidae. However, none of these possess the markedly upward-directed ostium and sharply bent ostial crista superior observed in Malakichthys (Lin et al., 2023b; Ng et al., 2024b) (Fig. 13). Small specimens of M. schwarzhansi also resemble otoliths of Ambassis (Ambassidae), but Ambassis otoliths differ by having a more pointed posterior rim and a slightly widened caudal tip, features not observed in Malakichthys (see Fig. 14).

A large otolith previously illustrated by Aguilera & Rodrigues de Aguilera (1999: pl. 1) may belong to this species, although it shows a more strongly elevated dorsal area, possibly reflecting ontogenetic variation. Additional specimens are needed to confirm this assignment.

The genus Malakichthys comprises eight extant species distributed in the Indo-Pacific (Yamanoue & Matsuura, 2004; Ng, Liu & Joung, 2023). Other members of the order Acropomatiformes, such as Parascombrops, display much wider geographic and stratigraphic distributions (Schwarzhans & Prokofiev, 2017). The occurrence of M. schwarzhansi in the Late Miocene of Panama suggests that the genus had a broader Neogene distribution than it does today.

Occurrence: Panama: Upper Miocene, Chagres Formation in Piña Norte locality, Colon. ?Venezuela: Lower Pliocene, Cubagua Formation, northwestern Venezuela.

Otolith density, sample coverage, and diversity indices

A total of 6,211 otoliths were collected from 34 bulk sediment samples, yielding an average otolith density of 278.80 ± 135.59 otoliths/kg (Table S1). Our otolith collection was represented by 31 taxa belonging to 12 families, plus nine additional specimens remaining indeterminate (Table 1; Fig. 15). Rank abundance of otolith families remains stable with or without the unweighted sample CH18-1-1 (Fig. 15). Sample coverage, based on specimen counts, reached 99.87%, indicating a high level of sampling completeness. Rarefaction curves based on species richness (⁰D) suggested that the estimated diversity could increase to approximately 35 taxa with additional sampling effort, with or without the unweighted sample (Fig. 16). However, rarefaction curves for Shannon (¹D) and Simpson (²D) diversity indices approached asymptotes, indicating that the most abundant and dominant taxa were successfully captured (Fig. 16). This pattern suggests that any additional taxa would likely be rare and of low-abundance.

Taphonomy and preservation

Otoliths are exceptionally abundant at the Piña site and are readily visible on the surface of the exposed sediments. Closer examination reveals that the otoliths are not randomly or evenly distributed but instead exhibit distinct clustering patterns within the sediment layers (Figs. 2C–2D). This clustered distribution suggests that otolith burial was not continuous, but occurred episodically.

Piscivorous predation, digestion, and subsequent excretion are important processes in the formation of otolith assemblages in marine sediments (Schäfer, 1972; Nolf, 1985; Welton, 2015; Lin et al., 2019; Agiadi et al., 2022). Predator feeding events can result in the accumulation of thousands of otoliths, especially by large predators such as whales, dolphins, and tunas (Fitch & Brownell Jr, 1968; Lin et al., 2020). At Piña, fossils of marine predatory mammals (dolphins and predatory whales) as well as piscivorous billfishes and sharks are common (Fierstine, 1978; Vigil & Laurito, 2014; Carrillo-Briceño et al., 2015; Pyenson et al., 2015; De Gracia et al., 2022), supporting a predation-mediated accumulation model. Therefore, the clustered distribution of otoliths in the sediments is consistent with deposition from predator excretions rather than from background mortality or mass mortality events.

Moreover, otoliths appear closely associated with ichnofossils attributed to Ophiomorpha (Stiles et al., 2022; Figs. 2C–2D). This suggests that burrowing organisms may have contributed to the local redistribution and concentration of otoliths within their burrow systems, either by incorporating otoliths during their activities or perhaps by selectively concentrating organic-rich material containing otoliths (Fig. 2D). While this burrowing activity does not increase the overall abundance of otoliths in the sediment, it may create localized zones of higher otolith density within and around burrow structures.

Surface-exposed otoliths are, on the whole, heavily weathered, often cleaving in half and exposing their whitish internal structure. Better-preserved otoliths were obtained from excavated blue-grey sediments found around 1–10 cm deep into the exposed sediments, which is where we focused sampling. In cases where lower taxonomic assignment was not possible, this was usually due to specimens being juveniles rather than from poor preservation. Nonetheless, a substantial proportion of specimens are moderately eroded, resulting in loss of outline details, resulting in taphonomic scores of 2 or 3 following Agiadi et al. (2022). This, combined with the prevalence of juvenile specimens, contributed to a relatively high number of otoliths being assigned only to the family level (Table 1).

Paleoenvironmental and paleoecological implications

The otolith assemblage at Piña is extraordinary in both abundance and composition, providing compelling evidence for a unique paleoenvironmental setting. Otolith densities in the Chagres Sandstone at Piña are exceptionally high, with individual sediment samples frequently exceeding 300 otoliths/kg, and one sample exceeding 775 otoliths/kg (Table S1). For context, Stringer et al. (2020) reported that a clay interbed of the Oligocene Glendon Limestone in Mississippi, USA, yielded 811 otoliths/kg, which they suggest may reflect enrichment driven by piscivorous predators. Thus, the densities observed at Piña rank among the highest otolith densities ever recorded from fossil assemblages for which sediment weight was systematically measured (cf. Leonhard & Agiadi, 2023). This unprecedented abundance coincides with a unique taxonomic dominance, where mesopelagic lanternfishes (Myctophidae) constitute over 96% of otoliths and represent more than 50% of the taxa (18 out of 31). Other pelagic fishes, such as hatchetfishes (Polyipnus) and codlets (Bregmaceros), were also present, albeit at much lower frequencies (each approximately 1.5% of the total otolith counts). Shallow-water taxa are rare, represented only by Carapus and Opistognathus. Nevertheless, these three pelagic families (Myctophidae, Sternoptychidae, and Bregmacerotidae) were the top three most abundant families (Fig. 15) at the site, demonstrating the importance of mesopelagic fauna.

The depositional environment of the Chagres Formation has been debated, with interpretations ranging from bathyal depths based on benthic foraminifera and fish assemblages (Collins et al., 1999) to shallow nearshore settings based on ichnofossils and sedimentological evidence (Stiles et al., 2022). Body fossil assemblages from the Chagres Formation include several indicators traditionally associated with deeper marine environments. Similarly, the occurrence of taxa such as Coelorinchus, Hoplostethus, and Malakichthys in our otolith assemblage could be consistent with deeper depositional environments, as these demersal genera are primarily associated with bathyal depths in modern oceans (Schwarzhans, 2013a; Lin, Girone & Nolf, 2016; Lin et al., 2017b; Lin et al., 2018). The dominance of myctophids themselves, being oceanic mesopelagic fishes that undertake diel migrations between surface waters at night and deeper waters of 300 m or more during the day, has often been taken to support interpretations of deeper water deposition.

However, sedimentological and ichnofossil evidence presents a contrasting picture. Stiles et al. (2022) provide compelling trace fossil and sedimentological evidence for shallow-water deposition, identifying archetypal Cruziana ichnofacies assemblages and sedimentary structures indicative of storm wave base environments (<50 m depth). At the same time, the total absence of typical shallow-water indicators such as ariid lapilli in our collection, which are near-ubiquitous in many Neogene shallow-marine otolith assemblages in the Caribbean (Aguilera et al., 2020), strongly suggests a depositional setting distal from the coast. Furthermore, some elements of the ichnological and sedimentary record could also be consistent with energetic upper-bathyal gateway systems where persistent bottom currents (contourites) and intermittent downslope processes redistribute bioclasts. Miguez-Salas, Rodríguez-Tovar & De Weger (2021) demonstrated that similar trace fossil assemblages and contourites can develop in deeper marine environments within oceanic gateways (the Rifian corridor), where strong bottom currents create dynamic conditions even at bathyal depths. Such a mechanism may be certainly applicable to the Isthmus of Panama in the Late Miocene, where tidal differences between the Caribbean and Panama could have created a highly energetic system even at depth.

A critical spatial consideration is the current geographic position of the Chagres Formation outcrops. Located approximately 20 km from the modern shelf edge at modern sea level, a bathyal depositional setting (>300 m depth) would require exceptional tectonic displacement to account for the current coastal position while maintaining stratigraphic coherence. Even accounting for the active tectonics during isthmus formation (O’Dea et al., 2016), known processes cannot reasonably explain how bathyal deposits could be displaced >20 km landward over the ∼6–7 Ma timespan. Typical rates of crustal shortening and arc migration in active margins (1–5 mm/yr; Coates & Obando, 1996) would require implausibly high displacement rates to account for such repositioning from a beyond-shelf-edge setting to the current coastal position, irrespective of eustatic sea level changes. The depositional depth of the Chagres Formation clearly requires further investigation through integrated geotectonic, sedimentary, paleoecological, and relative and global eustatic sea level analyses. The extraordinary dominance of myctophids suggests environmental conditions that differed significantly from typical modern shelf or bathyal settings. While myctophids can occur in relatively shallow waters starting at about 50 m over continental shelves (Lin et al., 2019; Heard et al., 2021), they are typically rare at such depths (<20%) and represented by only a few species that tolerate shelf conditions (Schwarzhans, 2013a; Lin, Girone & Nolf, 2016; Lin et al., 2017b). However, the Chagres assemblage may represent a paleoenvironmental setting with no exact modern analogue. Stiles et al. (2022) proposed that the orientation of the Caribbean coastline during Chagres Formation deposition could have facilitated Ekman-style seasonal upwelling. During the Late Miocene, the coastline orientation may have been positioned such that trade winds flowing parallel to the coast generated northward transport of surface waters, creating favorable conditions for localized upwelling in the Chagres region. Such a configuration would be consistent with the extensive evidence for seasonal upwelling systems documented throughout the Caribbean during the Late Neogene (O’Dea et al., 2007; Grossman et al., 2019; Jackson & O’Dea, 2023).

Many Caribbean coastal systems experienced strong seasonal upwelling during the Late Neogene, as observed in isotopic fluctuations within fossil shells and intracolony variations in module size in bryozoans from Florida and the Dominican Republic to Costa Rica and the Isthmus of Panama (O’Dea et al., 2007; Grossman et al., 2019; Jones & Allmon, 1995; Anderson et al., 2017). Additionally, the ecological composition of other nearshore fossil assemblages around the Caribbean during this period strongly support high levels of upwelling and productive coastal ecosystems, many of which contain abundant otoliths from fishes indicative of high productivity (Allmon, 1992; Jackson & O’Dea, 2023) and taxa shared with the Chagres Formation (Aguilera & Rodrigues de Aguilera, 2001). These productive ecosystems subsequently collapsed at the end of the Pliocene, giving rise to the modern, aseasonal and oligotrophic Caribbean (Jackson & O’Dea, 2023).

To explore this question further, we analyzed the proportion of mesopelagic planktivorous (principally myctophids) otoliths relative to all other otoliths in 187 modern Caribbean shelf sediment dredge samples (data from O’Dea et al., 2007; Jackson & O’Dea, 2023). When plotted against depth (Fig. 17), these data reveal that modern myctophid otolith assemblages increase from nearshore environments, peaking at around 120–150 m, before declining towards the shelf edge at 200 m. While intriguing, however, this shelf-focused analysis is somewhat inconclusive as it does not include bathyal depth samples and therefore cannot address the full depth range where myctophids are most abundant in modern Caribbean shelfs.

Regardless of depth, the tropical upwelling system observed at Piña differs substantially from modern temperate upwelling systems like the colder Humboldt and California Currents, where anchovies, sardines, and other small epipelagic fishes typically dominate (Chavez et al., 2003)—taxa almost entirely missing from the Piña assemblages. Instead, we argue that the Piña ecosystem was shaped by three key factors: warm tropical temperatures, high productivity from seasonal upwelling, and intense predation pressure. High tropical temperatures would have increased metabolic rates, which when combined with elevated primary productivity from upwelling, would have created conditions where predation rates can be extremely high (cf. Kordas et al., 2022). This would, in turn, favor the selective survival of predator-avoiding planktivorous fishes like myctophids, whose diel vertical migrations serve as a principal mechanism of predator avoidance. Despite these adaptations, a substantial proportion of myctophids still fell prey to predators. However, the extraordinary productivity, coupled with rapid demographic turnover of these small fishes, would have sustained large prey biomass, resulting in the high abundances of myctophid otoliths preserved in the sedimentary record.

Evidence for high predation pressure at Piña is substantial, including numerous predatory shark taxa such as Otodus megalodon (Carrillo-Briceño et al., 2015), and the frequent occurrence of large predatory vertebrates (e.g., dolphins, billfishes) and abundant elasmobranch teeth at the Piña locality (Fierstine, 1978; Vigil & Laurito, 2014; Carrillo-Briceño et al., 2015; Pyenson et al., 2015; De Gracia et al., 2022). Particularly striking, but as yet unremarked, is the extraordinary abundance of cookiecutter shark teeth (Isistius), which while observed in other Neogene tropical American sediments, reach exceptional frequency at the Piña Chagres site (see Table S1). As Isistius is a facultative ectoparasite that often feeds on large marine mammals, fishes, and sharks (Papastamatiou et al., 2010), their abundance testifies to a remarkably high density of large-bodied animals. The Piña scenario parallels aspects of tropical upwelling systems in the modern Arabian Sea, where primary production predominantly channels relatively directly into mesopelagic fish communities (Gjøsaeter, 1984; García-Seoane et al., 2025). In these systems, high planktonic productivity supports dense aggregations of myctophids, bypassing some of the longer and more complex food chains, and this energy is efficiently transferred to higher trophic levels, particularly to apex predators.

The combination of warm temperatures, strong coastal upwelling, extremely high planktivore abundance, and intense predation pressure provides a plausible explanation for the ecological observations at Piña. The Piña assemblage therefore represents a remarkable fossil example of a Late Miocene upwelling-driven, mesopelagic fish-dominated ecosystem, providing valuable insights into trophic dynamics and ecosystem structure during a critical period in the formation of the Isthmus of Panama (Fig. 18). However, we explicitly acknowledge that depositional environment requires further work to resolve the contrasting evidence.