The Potentiality for Long Range Oceanic Dispersal of Cotton Seeds

The presence of wild forms of the New World amphidiploid cottons on remote islands in the Pacific raises the question whether they could have reached their present habitats by oceanic drift. The same question may be raised concerning the origin of the amphidiploid cottons as a whole, since their putative parents (or the diploid species most nearly related to them) are confined to Africa and America, respectively. Tests of seed buoyancy and seed viability in experimental tanks of salt water indicate that the upper time limit for seed buoyancy is a little over two months. This is sufficient to effect the transport of seeds over relatively short distances (e.g., throughout the Caribbean islands and from mainland South America to the Galapagos Islands) but totally inadequate for trans-Atlantic or trans-Pacific dispersal. If one accepts the possibility of seeds being "rafted" over longer distances, the experiments indicate that the seeds of the Hawaiian wild species, G. sandvicense, have sufficient tolerance to salt water immersion to have been carried from the Central American mainland to their present location (a required period of nine months). Seeds of the Wake Island form of hirsutum have salt water tolerance comparable to that of sandvicense, but it seems doubtful that this would be sufficient to maintain them for the necessary period of time (19 months). The other Pacific forms tested lack the degree of salt water tolerance required to have reached their present locations as a result of oceanic drift. Reasons are given for believing that these forms arrived from Central America by some route other than the South Pacific Equatorial Current. The seeds of neither G. herbaceum var africanum nor G. anomalum (relatives of the diploid parents of New World cottons) appear to have sufficient salt water tolerance to survive a trans-Atlantic crossing.