Floating the sweet potato to polynesia: Considering the feasibility of oceanic drift for the prehistoric introduction of the sweet potato (Ipomoea batatas) to Pacific Islands

Abstract

The domesticated sweet potato (Ipomoea batatas) traces its evolutionary past to the Americas, yet, this tuberous vegetable has been a cultural and dietary staple for Pacific Islanders for over a millennium. Potential introduction of the crop to the Pacific by human action is supported by archaeological evidence, but phylogenetic analysis of historically collected plant tissues suggests that the sweet potato may have pre-dated human presence on Pacific Islands, pointing to a non-human-mediated introduction. Here we use laboratory and numerical experiments to evaluate the feasibility of oceanic drift as the introduction mechanism for sweet potato seed capsules from South America to the Pacific Islands. First, present-day and ocean paleocurrents and winds are used as input into numerical experiments that simulate drift trajectories starting in South America. These simulations are used to determine likely departure and arrival sites and to estimate trip duration. Drift voyage duration values are used to inform laboratory experiments designed to determine the effects of exposure to seawater on seed and seed capsule buoyancy and viability. Minimum simulated trip duration for arrival to the easternmost islands of Polynesia range from about 80 to 120 days. Viability of seeds decreases but is not eliminated after 121 days of exposure to sea water. None of the tested seed capsules remained buoyant after 40 days, indicating that while colonization by drift is possible, it would require some transport mechanism such as floating vegetation mats. While modeled landfalls occurred over a large area, most were concentrated along a linear distribution extending from Rapa Nui to islands in Kiribati, including the islands of the Marquesas and Tuamotu archipelagoes. Drifts departing from the segment of the South American coast between approximately 18S to 8S have a greater probability of making landfall. The main conclusions from the drift experiments are not impacted by changes in ocean and atmosphere circulation associated with glacial to interglacial climate variability.