River mussel shells are commonly found during our excavations at Cooper’s Ferry. Mussels were eaten by the early foragers who lived in the canyon, providing a widely available source of protein. In addition to revealing aspects of forager diets, we can submit mussel shell samples for radiocarbon dating, which reveals when the mussels were eaten at the site. Over the past nine months, we have been collecting shell carbonate samples from mussel shells recovered from Cooper’s Ferry and other sites in the lower Salmon River canyon as part of a larger study of past climate. Because we radiocarbon date the shells we are using for our paleoclimate study, we also learn more about the ages of the canyon’s archaeological sites. This year, we’ve received the results of radiocarbon dating from 77 shell samples. These shells were recovered from LU6, which is the dark brown layer we’ve been working in for the past few summers. Normally, we wouldn’t submit so many samples for radiocarbon dating; however, our paleoclimate project relies on large numbers of directly dated shells that span the last 12,000 years and we’re fortunate that Cooper’s Ferry holds so many shells to study. The 77 shell samples returned radiocarbon ages spanning the time period between 8030 ±37 radiocarbon years before present (RYBP; “present” being the year AD 1950), which, in calendar years, would be between 8971 cal BP to 9011 cal BP and 9138 ±38 RYBP (10234 cal BP to 10297 cal BP). Because these 77 shells are found throughout the upper 40 cm of LU6, we’ve not yet determined the lower chronological limit of LU6; however,we expect that the lower part of LU6 will date older than 9138 RYBP. Because the shell samples come from different excavation units that were dug into LU6, and because LU6 has some topographic variation that makes correlation between the vertical positions of radiocarbon dates more complicated than if the dates were found in a series of perfectly horizontal deposits, more work is needed to know just how finely we can divide up stratigraphic time at Cooper’s Ferry. Suffice to say that we are confident that the upper half of LU6 dates between 8030 to 9138 RYBP (8971-10,297 cal BP). The presence of these radiocarbon dates in LU6 makes perfect sense since this layer contains both Windust and Cascade projectile point forms. Throughout the southern Columbia River Plateau region of the interior Pacific Northwest, the transition from the Windust archaeological phase to the Cascade phase is seen to occur between 9,000-8,000 RYBP. In the lower Snake River of eastern Washington, radiocarbon dates on river mussel shell have been viewed as unreliable because living river mussels had incorporated ancient carbon introduced into Snake River water from deep aquifers. To check whether river mussel shells from Cooper’s Ferry would provide reliable radiocarbon ages, I submitted five shells collected from a beach along the Salmon River for radiocarbon dating. These five shells returned ages of “modern”, showing an absence of an old carbon effect in Salmon River water.
Apart from defining the age of LU6, these new dates help us to better understand the larger chronological sequence of the Cooper’s Ferry. In 1997, I directed the excavation of a 2 x 2 m test unit (Unit A), which contained four other sediment layers below LU6. In a lower layer of wind-blown dust (LU3 in the figure below), we discovered a large pit that contained four stemmed projectile points. This large pit, designated Pit Feature 2, also contained three charcoal fragments, which were submitted for dating and returned ages of 7300 ±70 RYBP, 8710 ±120 RYBP, and 11,370 ±40 RYBP; a fragment of animal bone returned an age of 12,020 ±170 RYBP. Two radiocarbon dates came from the surface of LU3, where we also saw the upper limits of Pit Feature 2. These two dates are: 10,050 ±180 RYBP (on bone) and 11,410 ±130 RYBP (on wood charcoal). Because the two bone samples were not subjected to the most rigorous pretreatments, they might not be very accurate. These presence of the youngest two ages (7300 and 8710 RYBP) can be interpreted from different hypothetical viewpoints. First, the younger dates might accurately clock the age of Pit Feature 2 and the older two ages (11,370 and 11,410 RYBP) could predate the creation of Pit Feature 2 and the occupation of LU3. Second, the younger two charcoal dates might have been introduced to Pit Feature 3 (i.e., moved downward by burrowing rodents) thousands of years after the pit was created and buried at the site and the feature dates to 11,370-11,410 RYBP. Third, it is possible that none of the available radiocarbon ages accurately date the earliest occupation and pit feature at Cooper’s Ferry and that their true age has yet to be revealed.
In my view, the 77 new radiocarbon ages from LU6 clearly show that the 7300 and 8710 RYBP dates do not accurately clock the age of Pit Feature 2. Photographs and drawings clearly show the presence of a rock cairn on top of Pit Feature 2 at the upper limits of LU3; thus, the feature does not begin from LU6. In consideration of the available data, we can reject the first scenario as a plausible hypothesis. Given the 77 radiocarbon dates from LU6, the deeper 7300 and 8710 RYBP ages cannot represent the age of Pit Feature 2 nor LU3 and give support to the second hypothesis that features the role of bioturbation. The question is, do LU3 and Pit Feature 2 date to 11,370-11,410? This question cannot be fully answered right now; however, it is important to point out that the available data do not require us to reject the hypothesis that these late Pleistocene-aged radiocarbon dates could date the early occupation of Cooper’s Ferry.