Development of an Efficient Cleaning Technique for Small Vertebrate Fossils from the Bayanshiree Formation, Mongolia

Uchida, M., Okoshi, T., Takahashi, A., Chiba, K., Buyantegsh, B., Mainbayar, B., Badamkhatan, Z., Saneyoshi, M., Tsogtbaatar, K.

The 6th International Symposium on Asian Dinosaurs in Japan 2025

Asian Dinoaur Association

Fukui, Japan

For over a century, research and excavation efforts have been undertaken in the Upper Cretaceous strata of the Gobi Desert in Mongolia, driven by the discovery of exceptionally well-preserved vertebrate fossils, particularly dinosaurs. Over the last ten years, the Institute of Paleontology, Mongolian Academy of Science, and Okayama University of Science have conducted excavation surveys in this region. In 2019, a densely fossiliferous layer containing small vertebrate remains was discovered at Bayan Shire, the type locality of the Upper Cretaceous Bayanshiree Formation, located in the southeastern Gobi Desert. The recovery of small vertebrate fossils from this formation is exceedingly rare, and the specimens recovered are critical for reconstructing terrestrial ecosystems during the early Late Cretaceous. However, the fossils are microscopic and often embedded within indurated matricies, necessitating the application of efficient preparation techniques. Traditionally, screen-washing has been employed for preparing small vertebrate fossils with various refinements—both physical (e.g., modifications to sieve design and water flow) and chemical (e.g., the use of heavy liquids and hydrogen peroxide)— to improve extraction efficiency. This study seeks to evaluate screen-washing methods that are suitable for field conditions in Mongolia and capable of enhancing fossil recovery efficiency. Our primary focus is on optimizing the washing process and advancing both mechanical and chemical preparation protocols. The material used in this study was collected from the Bayanshiree Formation at Bayan Shire locality. Samples were collected from a fossil-bearing layer composed of poorly sorted medium- to fine-grained sand. The samples were first divided into gravel-sized and sand-sized fractions. Two experimental treatments were employed for the sand-sized fraction. In Experiment A, the samples were directly screen-washed with water, while in Experiment B, they were washed after preliminary dry sieving. For the gravel fraction, Experiment C involved ultrasonic cleaning, and Experiment D utilized hydrogen peroxide immersion and rinsing with water. All samples were subsequently air-dried in trays, and fossils were picked under a microscope. For sand-sized fractions, there was no significant difference in the substrate reduction rate (i.e., the amount of removed matrix) between Exp. A and B, and no apparent damage to fossils was observed in both experiments. However, the application of dry sieving in Exp. B substantially increases the amount of sediment samples that can be processed in one cycle. For gravel-sized fractions, ultrasonic cleaning in Exp. C enhanced the efficiency of the subsequent preparation regardless of the variation in the matrix hardness due to the gravel composition. This effect is potentially due to ultrasonication-induced pressure changes, which promote air release from and water infiltration into the matrix, thereby softening the matrix to a disaggregatable state. In contrast, despite five days of hydrogen peroxide immersion, Exp. D yielded limited improvements in matrix disaggregation. In summary, pre-dry sieving notably increased processing throughput for fossil-bearing sand samples and reduced overall preparation time by approximately 70%. Conversely, significant challenges remain in the efficient processing of gravel samples, requiring further methodological reassessment in future studies.