From Hawaiʻi to the Mainland: Hawaiian Hiwa to Kali Hiwa

About this discussion

The companion article reports what 13 certificates of analysis show about a micropropagated Hiwa line and the Hawaiian stock it came from. It carries the methods, the tables and the figures, and it states its limits. This piece is the discussion around that article: why the question was worth asking, what the numbers mean for someone growing or buying kava, and which parts of the answer are still missing. Every figure quoted here is taken from the article without change; where a statement is my own reading rather than a reported result, it is written as such.

Why the question is worth asking

Kava sets no seed. A named cultivar is a clone, kept alive only by cuttings, and its chemistry is expected to travel with it. That expectation is what allows a six-digit chemotype code to stand in for a cultivar name in the trade and in the literature.

Material entering the mainland United States does not arrive as a cutting handed from one grower to the next. It is routinely put through tissue culture, both to multiply a very small amount of founder stock and to clear it of the pathogens that travel with field-collected material. The Kali Hiwa line went through that process, and has since been grown in California and Florida under conditions that share almost nothing with a Hawaiian hillside. Growers and buyers reasonably ask whether the plant is still what its name claims. Until now there was no published record for an American kava line that answered the question in either direction.

Where the Hawaiian stock came from

The Pacific kava story begins with one plant. Lebot et al. (1992) proposed that every kava cultivar in the region descends from a single ancestral plant in northern Vanuatu, most probably Piper methysticum var. wichmannii, cloned and carried eastward by stem cutting over roughly three millennia. Hawai‘i was the last stop on that route, and the plant arrived as living cuttings in the hands of voyagers rather than as seed (Johnston & Rogers, 2006).

What happened after landfall is the part that bears on Hiwa. Hawaiian planters could not breed new varieties, because the plant cannot reproduce sexually. What they could do was notice the somatic mutations an existing clone threw off and propagate those separately — the conscious selection of sports that Handy et al. (1991) describe. Lebot et al. (1999) tested the consequence, examining DNA in 22 plants and composition and morphology in 44, and concluded that the thirteen cultivars then recognised in Hawai‘i are most likely somatic mutations descended from a single introduced variety, and that the Hawaiian set as a whole runs highest in the kavalactones valued for their especially pleasant effects.

The older record implies a much larger population of such selections than survives. Winter (2004) recovered 49 varietal names and descriptions and estimated that at least 35 distinct varieties once existed. Thirteen were still identifiable by the time systematic collecting resumed. The stock in cultivation today is not a continuous farm lineage: ‘awa had largely fallen out of general cultivation, and what is grown now was reassembled from surviving plants, principally through the collecting Joel Lau carried out across the islands in the 1980s, the distribution of that material to botanical gardens and private collections, and the revival work of the Association for Hawaiian ‘Awa from 1998 onward (Johnston & Rogers, 2006).

Why Hiwa, of the thirteen

Hiwa was not picked at random. The Hawaiian descriptions agree on the trait that names it: hiwa denotes black or dark, and the plant is black-stemmed with long internodes, few lenticels and an erect habit reaching eight feet or more in full sun (Johnston & Rogers, 2006). The nineteenth- and early twentieth-century accounts Winter (2004) compiled describe it as very black, with long stem sections and dark leaves.

The planting Kali Hiwa came from, filmed in November 2012 at Puʻuʻala Farm on the Hāmākua Coast. Tyler Blythe is recording while Zach Gibson, the farm's owner and his mentor, walks him through the varieties growing there — the black-purple stems and long internodes that name hiwa. This is not a stand-in for the type: it is the planting the Kali Hiwa line was taken from.

Its standing in old Hawai‘i is recorded in more than one independent source. Handy (1940) called it the darkest of all varieties and the ritualistic ‘awa used by the kahuna, possibly reserved for that use, all the other varieties being common ‘awa which anyone might use. Kekahuna named ‘awa hiwa and ‘awa mō‘ī the most sacred of the Hawaiian varieties, offered to Pele and other deities, and recorded that the great chiefs also drank the sacred ‘awa while permitting the commoner only the non-sacred kinds (Kekahuna, 1963, as cited in Winter, 2004). Beckwith recorded the same restriction: the commoner could use only the most common variety, the rarer kinds were reserved for the chiefs, and hiwa was among those used for the gods and on ceremonial occasions (Beckwith, 1970, as cited in Johnston & Rogers, 2006). It is the variety offered to hula deities in return for knowledge and inspiration, in the line he ‘ike pū ‘awa hiwa (Winter, 2004).

A note on wording, since the point gets flattened in retelling. The sources place hiwa with the gods, the priesthood and the chiefs; they do not describe a single royal cultivar, and ‘awa mō‘ī is a separate variety with its own name. That is the framing the companion article uses, and it is the framing I would ask people to repeat.

The cultural record has a chemical correlate, and it is the reason a kava drinker notices Hiwa. The published reference assay for the cultivar — air-dried lateral roots of a 5.5-year-old plant from the Hāmākua coast, analysed for Lebot et al. (1999) — returned the chemotype 462351 at 11.23 per cent total kavalactones (Johnston & Rogers, 2006, table 5.3): kavain first, methysticin second, dihydrokavain third. Kavain-led profiles carrying a low proportion of dihydromethysticin are the ones the chemotype literature associates with the most desirable effects (Lebot et al., 1992), and cultivars rich in kavain, methysticin and dihydrokavain are those used in Vanuatu for a pleasant, fast-acting daily drink rather than ceremonially or medicinally (Lebot & Siméoni, 2004). A kavain-led Hiwa sits at the drinkable end of the range Pacific growers were selecting for.

Why a clone is not a seed crop

A sterile cultigen has no seed bank. Each named cultivar exists only as living tissue, and a clone that is lost cannot be bred back from anything. Of the 35 or more varieties the Hawaiian record implies, thirteen were still identifiable at the end of the twentieth century. Multiplying Hiwa through tissue culture and establishing it on the mainland is therefore a conservation measure as much as an agricultural one — but it is only a conservation measure if what comes out of tissue culture is still Hiwa. That is the whole of why the article was written.

How the line has been grown

Kali Hiwa has been carried through several production systems since micropropagation, and that history is part of what the continuity question is being asked about. In 2016 the line was grown hydroponically in deep water culture at the American Kava Research greenhouse in Sacramento, California, by Tyler Blythe. Lateral roots from that system assayed 6.5 per cent total kavalactones, and the culture allowed them to be harvested sustainably: roots were cut from the suspended mass while the plant stayed in production, rather than the stump being lifted and the plant ended at harvest. The same line has since been grown in greenhouse soil in California and in the field in Florida, which is the range of conditions the assays discussed below sit across.

That 6.5 per cent figure comes from the Association's own growing records rather than from the thirteen certificates analysed in the companion article, and it is quoted here as cultivation history rather than as part of the evidence base. It is worth stating because it bears on a practical question the article does not address: whether a soilless system that permits repeated lateral-root harvest from a living plant can hold a commercially meaningful kavalactone concentration. On this line, in this greenhouse, it did.

Figures 2 and 3. The Kali Hiwa line in deep water culture at the American Kava Research greenhouse, Sacramento, 2016, with the lateral root mass suspended clear of the vessel and available for harvest without lifting the plant; and mature stems of the same line, carrying the dark pigment that names hiwa alongside the green stripe variation.

The cultivar's morphology is documented in the Association's botanical illustration of Piper methysticum var. Kali-Hiwa (Figure 4), which records the inflorescence, the leaf, the basal stump, the lateral root mass and the dark stems and long internodes that name hiwa. The illustration also records the green stripe variation — a stem colour form carried within the line, in which the dark pigment is broken by green along the internode — so that the variation is documented as part of the cultivar rather than read as a different plant.

Figure 4. American Kava, Piper methysticum var. Kali-Hiwa. Botanical illustration by S. McWilliams for the American Kava Association, 2025, recording the inflorescence, leaf, basal stump and lateral roots of the cultivar, with the green stripe variation of the stem shown at upper right.

What the record shows

The short version is that the part of the chemotype that matters held, and the part that moved moved for reasons that have nothing to do with the plant changing.

Kavain ranked first and methysticin second in all 10 root, powder and lateral-root samples of both lineages, across two islands, four assay dates spanning seven years and three material states, by margins of 3.23 to 11.12 relative percentage points — far outside anything analytical variation could produce. That two-position head is the conserved feature of the cultivar (Figure 1).

Figure 1. Kavalactone rank order for every Hiwa assay in the study, each row showing that sample's six kavalactones in descending order of concentration. Kavain leads and methysticin follows in every root, powder and lateral-root sample across both lineages and all four assay dates. Solid cells mark the kavain > methysticin > dihydrokavain reading; dashed cells mark the four that read otherwise — crown root and peeled stump, which reorder along the plant-part gradient, and both 09/25 materials, in which yangonin and dihydrokavain are separated by less than the method resolves. Chemotype codes beneath each sample name show that the lower three positions differ from one another within a single cultivar assayed on a single plate, while the head does not.

Compared like for like, the root and chip powder of the 09/25 Kali Hiwa harvest sat 1.20 percentage points from the mean of the two Hawaiian root samples across all six relative fractions. That number only means something against ordinary variation measured the same way, and the article supplies three such yardsticks from within the same dataset: 1.84 points average spread among five Hawaiian powders milled and assayed together in one job, 2.01 points between Hawaiian root and Hawaiian powder, and 2.77 points between the tissue-cultured line's own lateral-root harvests of 2019 and 2025. The line is closer to its Hawaiian source stock than it is to itself six years earlier.

The external check agrees. Re-expressed on the same basis, the published Lebot et al. (1999) Hiwa assay sits 1.26 points from the Hawaiian root mean and 2.29 points from the 09/25 powder — a different plant, island, decade and laboratory, inside the same envelope.

Two further axes point the same way. Flavokavain ratios placed every assay carrying them well within noble character, at FK/KL 0.018 to 0.037 against a noble group mean of 0.13 and K/FKB 20.7 to 40.7 against a noble mean of 7.31 (Lebot et al., 2019). HPTLC under the British Herbal Pharmacopoeia (1996) identification returned a profile consistent with authenticated P. methysticum root references. These three axes fail on different things, which is what makes their agreement worth something: botanical identification would fail on a different plant, the rank head on a different cultivar, the flavokavain ratios on a two-day or wichmannii accession.

And the material performs in the cup. Under two standardized aqueous preparations the line delivered 90 to 334 mg of kavalactone per 125 mL serving at roughly four fifths single-cycle recovery, with 91.2 to 97.2 per cent of the delivered kavalactone travelling on suspended particulate rather than in solution. The spread in per-serving strength is a function of dose, water volume and plant part, not of the material.

What the third digit is worth

This is the practical finding, and the one I expect to be argued about. Hiwa is spoken of as a 462 cultivar. Both certificates of the 2025 lot read 463. In that lot dihydrokavain and yangonin were separated by 1.4 per cent of the reported concentration, a fraction of the 7.5 per cent repeatability that AOAC SMPR 2018.005 permits for individual kavalactones. Neither ordering is established by the method. The difference between a 462 and a 463 reading of Hiwa carries no analytical weight, and a change in that digit is not evidence that the plant changed.

Plant part matters far more than the third digit. Three parts of one plant, taken on one day and assayed in one job, returned three different chemotypes: lateral roots reproduced the Hawaiian head, crown roots exchanged the first two positions, and peeled stump reordered all three. The movement is a gradient, not a scramble — kavain falls and the dihydro- forms rise from lateral root to stump, with total kavalactone content falling alongside, which is the distribution kava shows generally. A chemotype reported without its material class is not interpretable, and reporting only the lateral-root figure would be picking the best number from a set the laboratory issued together.

For anyone reading a certificate with cultivar identity in mind:

• Read the first two positions, and treat the rest as informative rather than diagnostic.
• Check the material class before comparing anything to anything — lateral root, whole root, crown, stump and powder are not interchangeable.
• Divide each analyte by the sample's total kavalactone content before comparing samples, so that potency differences from age, drying and site drop out.
• Ask what the method's repeatability is before treating a narrow gap between adjacent analytes as real.
• Treat flavokavain ratios and a botanical identity test as separate evidence, because they fail on different things than the rank order does.

Why no identity threshold is proposed

An earlier pass at these data tried to build one: a standardised-distance criterion calibrated on the five October 2018 Hawaiian powders. The article reports it as a negative result, and I think that is the honest place for it. The criterion rejects the Moloka‘i Hiwa as outside the envelope, and the Moloka‘i sample is Hiwa. Variation among samples milled, stored and injected together in one job is not a usable estimate of the variation a cultivar shows across sites, seasons and material states, and a tolerance fitted to it will throw out genuine material. Anyone building an authentication threshold for kava cultivars needs replication at the level of the variation the threshold is meant to tolerate.

What this does not settle

Identity here is inferred from composition alone. No genotyping was performed. Microsatellite or SNP work of the kind Lebot et al. (1999) applied alongside chemistry would test clonal identity directly, and its absence is why the article claims chemotype continuity rather than verified clonal identity. That is the first thing a reviewer will press on, and it should be.

The certificates were also commissioned for commercial quality control by the party that grows and sells the material, at a single laboratory, with duplicate injections rather than independent replicate assays. The design is retrospective. Nothing in the analysis repairs that, and the article says so in its own limitations.

The comparison rests on two Hawaiian root samples against one Kali harvest, both mainland harvests came from the same California greenhouse, and no assay exists from shortly after micropropagation — so the record establishes that the profile matches now, not that it never departed. The study is also a continuity analysis within one cultivar: it does not test whether these features would tell Hiwa apart from another kava variety.

What would close the gaps

The work I would like to see done next, in rough order of how much it would add:

• Genotyping of both lineages, source stock and micropropagated line, alongside the chemistry.
• A whole-root or dried-powder assay of the micropropagated line matched to the Hawaiian material class, so the like-for-like comparison rests on more than one pair.
• Replicate harvests from both growing sites, including the Florida field material, which nothing here tests.
• A comparison panel of other named cultivars run against the same three axes, which is what a discrimination claim would require.
• An assay taken shortly after micropropagation on any future line, so continuity is documented from the start rather than reconstructed.

Where to take the discussion

The article and its supplement carry every certificate value, so any reader can recompute the distances and disagree with them on the arithmetic rather than on the framing. The two questions I would most like comment on are whether the two-position rank head is the right unit of cultivar identity for kava reporting generally, and what replication a distance based authentication criterion would need before it could be published as one. Comments and corrections are welcome through the Association.

Acknowledgements

All analyses discussed here were performed by Flora Research Laboratories, LLC, Grants Pass, Oregon. Thanks to James Neal-Kababick, Laboratory Director, and analysts Reine Bravo, Pierce Prozy and Stacy Wise, whose certificates over seven years are the entire evidence base for both the article and this discussion.