The expansion of humans from the closely spaced islands of the western Pacific into the vast eastern Pacific via long distance oceanic voyaging is the last great expansion of humankind into previously uninhabited areas.  The arrival of humans and associated plants and animals into pristine island ecosystems, and subsequent modifications to varied island environments through hundreds of years of cultural development and rising populations provide unique records of human-environment interactions.

Accurate and precise dating is needed to decipher this history. For six decades, Polynesian archaeologists have depended upon ¹⁴C dating to provide the temporal framework for their investigations, however, in spite of significant improvements over the years, ¹⁴C dating has severe inherent limitations in this context.  Foremost among these is the need to calibrate¹⁴C ages in order to correct for the effects of past variation in ¹⁴C levels in Earth's atmosphere.  The ¹⁴C calibration curve is well known but its complex shape can induce large errors in calibrated ¹⁴C ages.  See Figure 1.  Accordingly, alternative dating methods are needed to resolve current debates and problems in Polynesian cultural history. ²³⁰Th dating of corals from archaeological contexts, based on the decay of trace amounts of uranium incorporated into corals as they grow, has already proven effective in providing high-precision chronologies for temple construction and occupation sites in Hawai'i and Mo'orea, and for dating the first settlement of Western Polynesia. Our project has extended the application of ²³⁰Th dating to one of the most common artifact types in central Eastern Polynesian archaeological sites: abraders of Acropora and Porites coral. Small pieces of fresh coral were used for a variety of shaping and smoothing tasks by Polynesians, such as fashioning fishhooks from mollusk shells.  As a result, such coral fragments are common in Polynesian archaeological sites and sequences.  Dating the last increment of coral growth prior to use can provide a close constraint on the timing of associated human activities.  The ²³⁰Th dating technique requires relatively few assumptions, does not require external calibration, and is capable of providing highly precise ages in the period of interest; e.g., ²³⁰Th ages for 1000-year-old corals have typical uncertainties of five to ten years.

The project has four primary goals: (1) Develop precise chronologies for key archaeological sites in the Mangareva, Marquesas, Cook Islands and Austral archipelagoes, thereby refining our understanding of the archipelagoes' cultural sequences. (2) Test the fit between the ²³⁰Th chronologies and existing ¹⁴C chronologies for these key sites. (3) Determine whether abrading tools made from both Acropora and Porites genera are equally appropriate as materials for precise dating. (4) Determine the Δ-R marine reservoir values for ¹⁴C dating through dual dating of coral abraders with ¹⁴C and ²³⁰Th, thereby aiding future radiocarbon dating and calibration of materials from marine environments (e.g., fishbone, shell, pelagic seabird bones).

Our first journal article from the project, "²³⁰Th dating of coral abraders from stratified deposits at Tangatatau Rockshelter, Mangaia, Cook Islands: Implications for building precise chronologies in Polynesia", appeared in Journal of Archaeological Science in early 2019.  In it, we report twenty-five ²³⁰Th dates for Acropora and Porites coral abraders from a well-studied, stratified archaeological sequence at Tangatatau Rockshelter, Mangaia (site MAN-44, Cook Islands), develop novel screening criteria to identify reliable ²³⁰Th dates from buried contexts, and compare our ²³⁰Th dates on corals with a recent age model for the site based on ¹⁴C dates for short-lived plant materials.  We find that ²³⁰Th dates deemed reliable based on objective screening criteria cluster within individual layers, preserve stratigraphic order among layers, and generally agree with the previous ¹⁴C chronology but provide more precise dates (median ± 7.4 yr, 95% confidence level). These results indicate that ²³⁰Th dating of corals from stratified sequences has great potential for developing refined chronologies for Polynesian archaeological sites. 

Synthesis of ²³⁰Th dates for corals from other sites (>100 in all) in the Cook, Austral, Mangareva, and Marquesas Islands of central Polynesia are in progress, as is ¹⁴C dating of selected ²³⁰Th-dated corals to determine Δ-R marine reservoir values.

The project has provided in-depth training in state-of-the art techniques of isotope geochemistry and mass spectrometry to Dr. Elizabeth Niespolo while she was a graduate student at UC Berkeley.  Early career scientists trained at the Berkeley Geochronology Center, as Dr. Elizabeth Niespolo has been as part of this project, have gone on to career positions working in the U.S. National Labs and at the U.S. Geological Survey, where they apply their skills to problems in national security, energy research, and mitigation of natural hazards.  Dr. Niespolo has accepted a tenure-track position at Princeton University where she plans to establish teaching and research programs in geochemistry and geochronology that will further increase US capability and capacity in these important areas.

Last Modified: 09/23/2019
Modified by: Warren D Sharp