Description: Publication date: 15 September 2018 Source: Palaeogeography, Palaeoclimatology, Palaeoecology, Volume 505 Author(s): Elver Luiz Mayer, Leonardo Kerber, Ana Maria Ribeiro, Alex Hubbe A recurrent aspect of the fossil record is the observation of a disproportionate number of specimens or individuals of a single taxon in some deposits, which is stated as dominance. Commonly, the dominance is explained as the result of catastrophic or short-term mass death events or are proxies for palaeoecological inferences regarding gregariousness. However, taphonomic, stratigraphic and chronologic analyses of fossiliferous deposits have shown that this is not always true. To contribute to the study of dominance in fossil assemblages, we describe a probable Quaternary skeletal accumulation dominated by the extant gregarious rock-dwelling rodent Kerodon rupestris recovered from Sumidouro do Sansão, a 65 m deep pitfall cave in northeast Brazil, and discuss the palaeoecological implications of our findings. We provide taxonomic identification, taphonomic analyses, and chronological assessment. Besides K. rupestris (minimum number of individuals, MNI = 35), we recorded three taxonomic groups, that are the ground sloth Catonyx cuvieri (MNI = 1), the anteater Tamandua tetradactyla (MNI = 1), and the cougar Puma concolor (MNI = 1). The taphonomic analysis of the K. rupestris remains supports the idea that entrapment of individuals was the main process of bone accumulation and that the death of cave inhabitants followed by short transport to the main hall possibly occurred. Kerodon rupestris remains persisted on the surface of the cave deposit for different time spans and were exposed to fragmentation, weathering, invertebrate boring and encrustation inside the cave environment. Direct dating suggests differences in the ages obtained and intermittent deposition of the individuals inside the cave. It highlights the influence of taphonomic controls on gregarious taxa, such as K. rupestris , towards dominating attritional time-averaged assemblages. Furthermore, it demonstrates that such assemblages do not necessarily support inferences about single event mortality and gregariousness of the dominant taxon. Graphical abstract