Otoliths are tiny calcium carbonate structures located in the inner ears of bony fishes, and they are among the most abundant and durable microfossils in marine sediments. Because each species produces otoliths with a distinctive shape, these structures can be identified much like teeth or bones, allowing paleontologists to reconstruct ancient fish communities even when skeletons are absent. The research team collected 151 bulk sediment samples weighing more than 747 kilograms in total from three localities along the Dahan River: Tachi-1, Tachi-2, and Sanxia. The sediments were soaked, sieved through a 500-micrometer mesh, and then painstakingly sorted under a stereomicroscope, with every diagnostic otolith hand-picked and cataloged into the collections of the Biodiversity Research Museum at Academia Sinica.

The three sampling sites tap different slices of geological time within a thick, roughly 2,500-meter succession of foreland basin deposits associated with the ongoing Taiwan orogeny. Tachi-2 and Sanxia expose the Early Pliocene Erchiu Formation, deposited in wave- and tide-influenced open-marine settings, while Tachi-1 samples the lower, coastal portion of the Early Pleistocene Cholan Formation, dated to approximately 2 million years ago. The Sanxia outcrop had never before yielded formally published paleontological material, making its rich shell beds a genuine surprise. Combined with previously documented Late Miocene otoliths from the same river section, the new collections complete a temporal archive that few marine sections anywhere in the northwestern Pacific can match.

What sets this study apart from traditional otolith paleontology is its explicitly quantitative community framework. Rather than simply listing species, the researchers applied Hill numbers, a family of diversity metrics that capture taxonomic richness, the diversity of common species, and the diversity of dominant species, all standardized through specimen-based rarefaction with 1,000 bootstrap replicates. The pooled assemblage yielded 80 observed taxa, extrapolating to roughly 93 at doubled sampling effort. While total richness was broadly similar across the three sites, ranging from 48 to 54 taxa, the Pleistocene Tachi-1 site showed markedly higher community evenness, consistent with a heterogeneous, high-energy estuarine to deltaic environment, whereas the Pliocene sites were dominated by a few abundant taxa, notably small gobies and juvenile flatfishes that likely reflect ancient nursery grounds.

The heart of the analysis lies in how the fish communities changed through time. Using permutational multivariate analysis of variance, or PERMANOVA, based on Jaccard dissimilarity, the team demonstrated that assemblage compositions differed significantly between all pairs of epochs. A principal coordinates analysis resolved the samples into three clean, non-overlapping clusters corresponding to the Miocene, Pliocene, and Pleistocene. Crucially, the researchers then partitioned total beta diversity into its two fundamental components: turnover, which measures true species replacement, and nestedness, which measures the loss or gain of species without replacement. The result was striking. Across the Miocene-to-Pliocene transition, median turnover dissimilarity reached 0.97 while nestedness was a negligible 0.01; across the Pliocene-to-Pleistocene transition, turnover was 0.81 against nestedness of 0.10. In plain terms, the fish communities were not simply losing species or shrinking into impoverished subsets of their former selves. They were being rebuilt, species by species, through wholesale replacement.

Indicator species analysis gave that replacement a face. The Miocene assemblage was defined by extinct sciaenids, members of the croaker and drum family, including Nibea chaoi, Taosciaena hui, and Taosciaena jiangi, the last of which showed the strongest possible association with its epoch. The Pliocene was marked by the codlet Bregmaceros nectabanus and the goby Myersina filifer, while the Pleistocene was characterized by the anchovy-like Coilia cf. mystus and three sciaenids that are still alive today: Collichthys lucidus, Larimichthys polyactis, and Pennahia pawak. The most dramatic example of within-family replacement occurs in the Sciaenidae itself, where the extinct Miocene genus Taosciaena vanishes entirely from younger strata and is succeeded by modern genera such as Johnius and Pennahia. A parallel pattern appears in the smelt-whitings, with Sillago japonica emerging as a key Plio-Pleistocene indicator.

To place these local patterns in a regional context, the team compiled 5,083 fossil and archaeological occurrences of sciaenid otoliths from across the western Pacific, spanning Japan, Taiwan, Thailand, and Brunei Darussalam. The synthesis reveals a stepwise evolutionary drama. During the Late Miocene, 13 of 19 recognized sciaenid genera were present, including four extinct genera restricted to that epoch, such as Paleosciaena in Japan, Taosciaena in Taiwan, and Advenasciaena and Bruneisciaena in Brunei. Then, beginning in the Pliocene, the modern cast entered the stage: Collichthys, Johnius, Otolithes, and Pennahia made their first appearances, followed by Chrysochir in the Late Pleistocene. Because these occurrences are both temporally structured and geographically coherent from the East Asian margin to Southeast Asia, sciaenid otoliths now offer a practical biochronological tool for correlating Neogene and Quaternary coastal deposits throughout the region.

Perhaps the most provocative implication concerns molecular clocks. Molecular phylogenetic studies have estimated that the family Sciaenidae originated in Tropical America during the Late Oligocene to Early Miocene, around 25 to 30 million years ago, and that several extant western Pacific lineages diverged deep in the Early Miocene, roughly 18 to 20 million years ago. Yet the empirical otolith record tells a different story: rich and diverse sciaenid assemblages already existed in Eocene shelf environments between roughly 33 and 50 million years ago, with the earliest known record coming from the early Eocene of the U.S. Gulf and Atlantic coastal plains, which is difficult to reconcile with a young American origin. Meanwhile, extant western Pacific genera are absent from Early and Middle Miocene regional records, which are instead dominated by extinct endemic forms. If the molecular dates are correct, they would imply unsampled ghost lineages of approximately 10 to 15 million years for multiple modern genera, a gap the authors caution could also reflect incomplete sampling or earlier origination elsewhere followed by dispersal.

The drivers behind the gradual modernization likely include the tectonic development of the Taiwan orogeny, late Neogene climate transitions, sea-level oscillations, and possibly the establishment of the modern-style Kuroshio Current during the Pliocene, though the authors emphasize that these paleoceanographic mechanisms remain working hypotheses until further facies-standardized sampling can separate local environmental tracking from genuine regional turnover. What the data do establish firmly is that the Plio-Pleistocene transition represents a critical interval of faunal replacement and the establishment of modern shelf lineages in Taiwan, achieved through progressive succession rather than abrupt extinction and recovery. By linking local paleoecological signals with regional biostratigraphy, the study demonstrates how something as small as a fish’s ear stone can anchor a continent-scale narrative of evolutionary change, while providing empirical calibration points that molecular phylogenies urgently need. For an underexplored corner of the Pacific, the humble otolith has proven to be an extraordinarily eloquent witness to eight million years of oceanic history.

Subject of Research: Neogene–Quaternary teleost community turnover and sciaenid biostratigraphy in the western Pacific based on fossil otoliths

Article Title: Gradual teleost modernization in the western Pacific: Neogene-Quaternary otolith evidence

Article References: Gradual teleost modernization in the western Pacific: Neogene-Quaternary otolith evidence. (n.d.). https://doi.org/10.3897/sjp.145.211471

Image Credits: AI Generated

DOI: 10.3897/sjp.145.211471

Keywords: otoliths, teleost fishes, Neogene, Quaternary, Taiwan, western Pacific, Sciaenidae, beta diversity, paleoecology, biostratigraphy, molecular clock, Pliocene