The tooth sits in wet plumbing
A partial left third molar, specimen MPC 3000, assigned to Cardiatherium and dated at about 8.8 million years, comes from the Bahía Inglesa Formation on the Chilean coast in a peer-reviewed paper.[1]
Living capybaras are semiaquatic herbivores of perennial wetlands east of the Andes. The same paper pairs the tooth with pedogenic-carbonate isotopes that indicate C3-dominated vegetation and with phytoliths of palms and dicots, and reads those together as persistent freshwater and riparian ecosystems in the southern Atacama in the late Miocene.[1]
Two roads, one wet interval
The authors say the occurrence requires ecological connectivity between foreland wetlands and the Pacific margin. They leave two routes open: a trans-Andes corridor, or a northern path from the Amazon basin to the Pacific through westward-draining systems. Either reading is a hydrologic map, not a proof that one mountain pass was used.[1]
That is the useful lever. A textbook that treats the Atacama as hyperarid without interruption for tens of millions of years is being asked to make room for a wetter late Miocene coast. The paper leaves later drying undated and offers no monsoon forecast. It does move the baseline: desert reconstructions that skip this wet interval misread how sensitive that coast was to water.[1]
The next wet proxy
The next check is another independently dated Hydrochoerinae fossil, or another wet-climate proxy of the same late Miocene slice, on the Pacific Atacama margin. Until that arrives, the honest unit remains one tooth, C3 soils and palm phytoliths in one formation, not a closed history of the whole desert.[1]