Technical Data on the Origin of Kava
An extract from "Legend and history: Did the Vanuatu-Tonga kava trade cease in A.D. 1447?" by David Luders, Journal of the Polynesian Society - V. 105, no. 3 (1996): 287-310

    It is to the agronomist Lebot and his various collaborators that we owe the elucidation of the origin of domesticated kava.  Lebot's work in the early 1980s in Vanuatu produced strong circumstantial evidence for supposing that domestication of kava occurred in Vanuatu and subsequent publications, in particular Lebot, Merlin and Lindstrom (1992), bring the work to a conclusion.  In the latter publication the evidence is more direct and powerful than that presented by Brunton (1989), who argues that kava might have been domesticated elsewhere in Melanesia and seems to favour Papua new Guinea in this respect.
    From the beginning, Lebot showed that there was a much greater range in Vanuatu than anywhere else of kava varieties and that these varieties were usually more potent chemically than those from elsewhere (e.g., total kavalactone analyses of Vanuatu kavas were two to five times that of common Fijian varieties).  The chemical analyses were also quite diverse.  Equally significantly, the name for kava in local languages was as diverse as nigui (Hiw, Torres Is), maloku (Marino, Maewo), mele (Sa, South Pentecost), bir (Tur, Santo), hae (Malo), nimvulum (South West Bay, Malekula), nikawa (Kwamera, Tanna), kava (Aneityum) to select some (Lebot and Cabalion 1986:83-93).  This suggests an origin more ancient than in Fiji or Polynesia.  Only the last two of these names are cognate with the Polynesian kava.  These occur in the south and it is quite likely that they were introduced there from Polynesia (Crowley 1994:95; Lebot, Merlin and Lindstrom 1992:52).
    Lebot, Merlin and Lindstrom present the results of extensive analysis of many kavas using three means: morphotypes, chemotypes and zymotypes.   Morphotypes are essentially varieties, distinguished by physical characteristics.   Chemotypes are groupings made on the basis of the chemical analysis of the kavalactones, the active substances in kava.  Zymotypes are distinctions made on fundamental genetic characteristics; this kind of analysis can be loosely termed genetic finger-printing. 
    Although it can be said that these studies were not exhaustive of every kava extant, they range so widely and so much further than any other that their results are compelling.  They reach a conclusion almost impossible to deny.
    Some two-thirds of the morphotypes studied are found in Vanuatu and nowhere else.  Of the eight chemotypes, six are present in Vanuatu and in no other place are there more than three: in Samoa alone were three recorded; in the rest of Polynesia one or two , and in Fiji one.  The distribution of zymotypes is still more persuasive.  There are ten of them, of which seven are Piper wichmanii ("wild" kava) and three are Piper methysticum ("cultivated" kava).  Five of the "wild" kavas were found only in Papua New Guinea.   All but one of these are genetically remote from domesticated kava.   The two "wild" kavas not found in Papua new Guinea are found in Vanuatu (and one is found in the Solomon Islands) and these are genetically the closest to "cultivated" kava.  No "wild" kavas were found in Polynesia or Micronesia.
    The distribution of the three "cultivated" zymotypes also points to domestication's having occurred in Vanuatu.  One occurs only in northern Papua New Guinea and is of minor significance.  The other two occur in Vanuatu.  One of these appears also in southern Papua New Guinea.  The other is the sole genetic type occurring in Fiji, Polynesia and Micronesia.
    It is difficult to find any grounds for doubting the conclusion of Lebot, Merlin and Lindstrom that kava was domesticated in Vanuatu.  The kava to be found in Fiji and Polynesia is the result of a thin trickle of varieties out of Vanuatu.   This reflects the testimony of Firth's record.  Just where Vanuatu domestication occurred is still conjectural but the evidence points to northern Vanuatu, possibly Maewo island.

An extract from "Legend and history: Did the Vanuatu-Tonga kava trade cease in A.D. 1447?" by David Luders, Journal of the Polynesian Society - V. 105, no. 3 (1996): 287-310

.