Geng Detection Report: "Early Devonian pterygotid eurypterids from Yunnan Province, China" (Wang et al., 2025, J. Paleontol.)

Academic fraud report · Geng Detector

Summary

This report assesses "Early Devonian pterygotid eurypterids from Yunnan Province, China" by Maxwell Wang, Simon Braddy, Victoria Tang, and Zhiheng Ma, published in the Journal of Paleontology (DOI: 10.1017/jpa.2025.10135). The paper is primarily a morphological/taxonomic description, so no image-based duplication or advanced statistical fabrication was tested. The overall verdict is HIGHLY SUSPICIOUS, driven not by misconduct per se but by multiple low-level errors indicating inadequate proofreading. Most critically, the occurrence locality for Pterygotus wanggaii lists coordinates as 103.698351°N, 103.698351°E, with identical latitude/longitude values placing the site well north of the Arctic Circle rather than in Yunnan (~25°N). Additional issues include a duplicated Agassiz 1839 reference with formatting drift, misspelling of "University College London" as "Universiy College London," and inconsistent species spelling (wanggaii vs wanggai) and authorship-year citations (2025 vs 2023). These errors raise concerns about core locality data reliability and editorial oversight, but do not constitute direct evidence of fabrication.

Verdict

🟠 Highly Suspicious — No direct evidence of systematic fraud was identified (consistent with the paper's morphological focus), but multiple low-level errors in core data and references suggest poor proofreading and raise concerns about the reliability of key locality information.

Key findings

🔴 Core locality data error (Finding 1): Coordinates for *Pterygotus wanggaii* given as 103.698351°N, 103.698351°E. Latitude and longitude values are identical, and a latitude of ~103.7° N lies north of the Arctic Circle, contradicting the stated Yunnan Province setting (~25° N) and other sites mentioned (e.g., Kudangtang reservoir).
🟡 Reference list duplication (Finding 2): The Agassiz, J.L.R., 1839 reference appears twice in the References, with minor capitalization/punctuation differences ("Fishes of the upper Ludlow rock... Counties of Salop... and Stafford, with Descriptions..." vs. "Fishes of the Upper Ludlow rock... Counties of Salop... and Stafford: with Descriptions..."), suggesting duplicate imports.
🟡 Affiliation misspelling (Finding 3): "University College London" rendered as "Universiy College London."
🟡 Taxonomic/citation inconsistency (Finding 3): Species cited as *Pterygotus wanggaii* in the text but as *Pterygotus wanggai* in Table 1; the species source is attributed to "Ma et al. (2025)" in Table 1 but "Ma et al., 2023" in the abstract.
✅ Measurements (Finding 4): Morphometric values (e.g., lunules and coxal denticles for DS-t0120230907a-b) appear within normal ranges for fossil measurements; no suspicious digit-preference patterns were observed.
⚠️ Image analysis (Finding 5): Pixel-level image reuse checks were not possible because original high-resolution image files were not available; only figure captions were reviewed, with no obvious internal contradictions.

Evidence highlights

Coordinate string exactly as reported: 103.698351°N, 103.698351°E — identical N and E values; latitude inconsistent with Yunnan geography.
Duplicate Agassiz 1839 entries distinguished only by capitalization ("upper" vs. "Upper") and comma vs. colon before "with Descriptions..."
Author affiliation typo: "Universiy College London" on Page 1.
Nomenclature mismatch between main text (*wanggaii*) and Table 1 (*wangi*/*wanggai*), and year discrepancy (2025 vs. 2023) for the naming authority.

Notes

DOI 10.1017/jpa.2025.10135 preserved verbatim.
Confidence in the verdict is moderate: the coordinate error is self-evident, but whether it reflects a typo, copy-paste artifact, or more serious locality-data manipulation cannot be determined from the manuscript alone.
Statistical and image-based fraud detection tools are largely inapplicable to morphological paleontology papers; conclusions are limited accordingly.
Suggested follow-up: contact authors for raw specimen data and locality verification, monitor PubPeer for author response, and check for an erratum before escalation to the journal editor.
Report is AI-assisted and intended for academic discussion; formal determinations require institutional investigation.