Taphonomy of fossils from the hominin-bearing deposits at Dikika, Ethiopia.

Two fossil specimens from the DIK-55 locality in the Hadar Formation at Dikika, Ethiopia, are contemporaneous with the earliest documented stone tools, and they collectively bear twelve marks interpreted to be characteristic of stone tool butchery damage. An alternative interpretation of the marks has been that they were caused by trampling animals and do not provide evidence of stone tool use or large ungulate exploitation by Australopithecus-grade hominins. Thus, resolving which agents created marks on fossils in deposits from Dikika is an essential step in understanding the ecological and taphonomic contexts of the hominin-bearing deposits in this region and establishing their relevance for investigations of the earliest stone tool use. This paper presents results of microscopic scrutiny of all non-hominin fossils collected from the Hadar Formation at Dikika, including additional fossils from DIK-55, and describes in detail seven assemblages from sieved surface sediment samples. The study is the first taphonomic description of Pliocene fossil assemblages from open-air deposits in Africa that were collected without using only methods that emphasize the selective retention of taxonomically-informative specimens. The sieved assemblages show distinctive differences in faunal representation and taphonomic modifications that suggest they sample a range of depositional environments in the Pliocene Hadar Lake Basin, and have implications for how landscape-based taphonomy can be used to infer past microhabitats. The surface modification data show that no marks on any other fossils resemble in size or shape those on the two specimens from DIK-55 that were interpreted to bear stone tool inflicted damage. A large sample of marks from the sieved collections has characteristics that match modern trampling damage, but these marks are significantly smaller than those on the DIK-55 specimens and have different suites of characteristics. Most are not visible without magnification. The data show that the DIK-55 marks are outliers amongst bone surface damage in the Dikika area, and that trampling is not the most parsimonious interpretation of their origin.