Limitations to U/Th dating of reef-platform carbonate boulders produced by high-energy marine inundations in the Tuamotu Archipelago (French Polynesia)

Abstract

On intertropical coastlines, coral boulders located on the reef platform are often used to characterize high-energy marine inundations of the past. Precise U/Th coral dating provides opportunities to assess the timing and recurrence of these events. However, this method of dating is subject to several limitations, as revealed by studying live coral cover at five islands in French Polynesia, since the early 1990s. We show that live coral assemblages experience very different changes within the islands of the same archipelago. Moreover, our observations show that the passage of a cyclone causes almost complete disappearance of the most recent coral in the coral boulder formation zone. This analysis shows that the probability of taking coral samples that were alive on the day of the high-energy marine inundation is extremely low. We have clarified the limitations of this method based on the dating of coral assumed to be alive on the day of the well-documented high-energy marine inundation of 1903 AD. The dating of 24 samples taken from 22 boulders from 3 islands of the Tuamotu Archipelago shows that important parts of the coral were dead when displaced by the marine inundations, thereby overestimating the timing of the disturbance event. As such, U/Th coral dating has limited application for establishing the precise timing for high-energy marine inundations with a return period of less than a century, but may be useful for exploring infrequent and extreme events such as the cyclone that ravaged the Tuamotu atolls in 1903. Indeed, we observed that this cyclone is the one that produced the world’s largest reef-platform coral boulders (more than 100 m3). We also provide evidence of an earlier event in the center of the Tuamotu Archipelago, during the middle of the first century AD.