Rapa Nui Journal: Journal of the Easter Island Foundation Volume 10 Article 1 Issue 2 June 1996 1996 Rat Colonization and Polynesian Voyaging: another hypothesis Atholl Anderson Australian National University Follow this and additional works at: https://kahualike.manoa.hawaii.edu/rnj Part of the History of the Pacific slI ands Commons, and the Pacific slI ands Languages and Societies Commons Recommended Citation Anderson, Atholl (1996) "Rat Colonization and Polynesian Voyaging: another hypothesis," Rapa Nui Journal: Journal of the Easter Island Foundation: Vol. 10 : Iss. 2 , Article 1. Available at: https://kahualike.manoa.hawaii.edu/rnj/vol10/iss2/1 This Research Paper is brought to you for free and open access by the University of Hawai`i Press at Kahualike. It has been accepted for inclusion in Rapa Nui Journal: Journal of the Easter Island Foundation by an authorized editor of Kahualike. For more information, please contact [email protected]. Anderson: Rat Colonization and Polynesian Voyaging: another hypothesis Rat Colonization and Polynesian Voyaging: another hypothesis Atholl Ander on Division ofArchaeology and Na/ural His/OIY, Australian National University Robert Langdon (1995:77) disputes the long-standing Humboldt connecting these islands with New Zealand, and proposition that Rattus exulans was dispersed by Polynesian few large rafts of vegetation are debauched by their rivers voyaging and suggests tbat over hundreds of thousands or (there could bave been logs, but tbese are bigbly unstable in millions of years it "succeeded in getting from one island to a seaway), the probability of a successful natural drift event another witbout any human aid at aiL." Between this and the occurring during a maximum timespan of 2000 years before conventional view lies the possibility, not yet explored in demonstrated buman settlement of New Zealand, cannot be detail, that some rats were transported on canoes that had lost bigh. Nevertheless, the disposition of wind and current is their human crew. I discuss this is relation to New Zealand, quite favorable to a drift crossing by canoe. but the principles are the same for Easter Island. Oceanic circulation in tbe south Pacific forms a vast The New Zealand case is especially pertinent because anticyclonic vortex or gyre wbich extends from the Coral sea Rattus exulans remains from non-cultural sites have been to the vicinity of Easter Island. "The southern border of the dated to as early as about 2000 b.p. (Holdaway, Worthy, gyre consists of an easterly current whjch flows in the Anderson and McGlone n.d.), some 1200 years earlier than subtropical zone (25-35°S), favored by the dominant westerly any reliable radiocarbon dates on archaeological sites winds" (Rougerie and Rancher 1994:15). Thus one potential (Anderson 1991). The status of these results is still uncertain drift route clips the northern extremity of New Zealand. and, in addition, there are several dates of similar age on During El Niiio conditions, the surface current flow slackens Ratllls exulans remains from tbe Shag River Mouth but there is a marked increase in winds from tbe westerly archaeological site where the cultural remains are otberwise quarter and therefore in the speed and breadth of the eastward securely dated to 620 b.p. (Anderson, Allingham and Smith swface flow. J996). Nevertbeless, tbe initial results are sufficiently For most of New Zealand, tbe more important waters are provocative to encourage speculation about the plausibility of those originating in the South Equatorial Current. These flow rat colonization in the effective absence of people. through Polynesia and East Melanesia and then swing south at about 20° latitude, as they reach the continental shelf, to Drift voyaging to distant lands become the East Australian Current. Off the coast of New Unaided dispersal of rodents on mats of vegetation that South Wales, slow-moving gyres are spun off towards originated in Ecuadorian rivers is argued for the Galapagos Tasmania and New Zealand, while the remainder of the water Islands. Steadman and Ray (1982:19), say this happened on is turned eastward to New Zealand along the Tasman Front at at least three occasions, with two genera arriving long before the Subtropical Convergence Zone, usually located 37-42°S. human colonization of the Americas, and Oryzomys more So, there are several drift routes whjch could bring material recently (possibly with prehistoric people, although that is southeast towards New Zealand. The landmass stretches tbought unlikely by Terrell 1986:94-106). The absence of across an arc of 33 degrees from New Caledonia, and 50 Megaoryzomys from the more isolated Galapagos islands degrees from either Norfolk Island or a point 100 kID off (Steadman 1986), also suggests a non-cultural mechanism, if Sydney and is, therefore, a very large target in Oceanic terms. it is assumed that watercraft would tend to spread taxa 10 addition, there is no doubt that material is deposited on throughout the archipelago. the western New Zealand coasts from tropical sources. There The eastern islands of the Galapagos lie about 1000 kID have been no recent studies of the quantity or distribution of from the mainland, but currents off the Ecuadorian coast are such exotic flotsam in New Zealand generally but local relati ely weak and variable. The more powerful Humboldt observations show that it occurred frequently on Northland current, wbich runs through the Galapagos, sweeps away beaches (Smith et aI., 1990:16) and both natural debris of from the mainland coast of Peru at about 3000 kID distant tropical origin and several artifacts of apparent Melanesian from the islands, the maximum likely distance involved. origin have been recovered from one short stretch of coastline There are no other cases of natural over-water dispersal of west of Wellington (Keyes and Hall 1967). cursorial mammals to match that of the Galapagos although For the transportation of rats, bowever, much more is some placental rodents were rafted across narrow water-gaps required than drifting debris. Anything floating very low in to Australia and the Philippines (Diamond 1987). tbe water will move at the speed of the currents and tbat is For New Zealand the over-water distance is within the probably too slow for survival of rats between New Zealand parameters of the Galapagos example (1000-3000 kID), but and tbe tropical Pacific. Material drifting under the influence Rattus exulans had almost certainly not reached the nearest of the South Equatorial Current, for instance, takes two to six potential source islands, Norfolk Island at 800 kID distant and months to cross from Fiji to Australia, about 2600 kID (Smith New Caledonia at 1400 kID distant, prior to about 3000 b.p.­ eta!., 1990:7, Smith 1992:59),atanav rage speed of 0.6-1.8 distribution in the western Pacific is largely covariant with kID per hour. As a "western boundary current", comparable to Lapita expansion. Since there is no current comparable to the the Gulf Stream, the East Australian Current moves at up to Rapa Nlli Journal 31 Vol 10 (2) June 1996 Published by Kahualike, 1996 1 Rapa Nui Journal: Journal of the Easter Island Foundation, Vol. 10 [1996], Iss. 2, Art. 1 9 kID per hour (CressweJl 1987), but the speed drops scenario that seems quite plausible is that vessels sailing in substantially as it dissipates eastward in the Tasman Sea light northerlies, common conditions in the Coral Sea during where the main influence on surface drift becomes the summer (Smith 1992:58) strayed too far south into the strong, persistent westerlies which are predominant south of 300 S. persistent northwesterly conditions of the northern Tasman Material from the Coral Sea would probably drift in a broad Sea, became disabled and were driven on to New Zealand. arc towards Australia, down the east coast and then across the There is, of course, no need to propose that a vessel was Tasman, a distance of about 4000 kID. At a current speed of overtaken by disaster only at a point close to the origin of a 20-30 kID per day it wouLd take 133-200 days to reach New voyage. Much of a passage might have been sailed quite Zealand. rapidly, leaving only a brief period of drift. A more crucial factor in the drift rate and direction of In any event, drift times from the nearest archipelagoes immobilized objects is the influence of the wind. Sea surface­ northwest of New Zealand-Norfolk Islands, New Caledonia layer drift speeds can be predicted at 1-6% of wind speed at and southern Vanuatu-are well within the range (mean of81 10m (Brown 1991 :Table 1). So, for example, in a Fresh days, 2000 kID, in Table 1) of those that have been Breeze (17-21 knots), which is a common range for trade successfully accomplished historically. This is much less the winds in the open ocean, the surface layer will move case to the north and northeast. The disposition ofcurrent and downwind (deflected slightly to the left in the southern weather systems changes unfavorably for drifting to New hemisphere) at a speed of 7.4­ Zealand and th distances 9.4 kmJday at 1% to 44.6-56.2 lengthen substantially eastward. kmJday at 6%. Material which Rarotonga, for example is 2500 sticks above the water is km away, or 73 days drift at the subject to additional wind mean speed of the historical data pressure and will "sail" faster in Table I, and it is unlikely that than the surface-layer drift a successful drift from that speed. How much faster is direction could attain that rate. dependent on a range offactors (height, surface area to wind, Vulnerability of Watercraft draft, weight and so on), of Unballasted wooden vessels which the proportion of an are relatively immune to sinking object's height above water and canoe hulls continue to float level is the most crucial.