Drive uphill on a Hawaiian island and you can move through several growing environments before the distance on the odometer seems significant. That makes elevation a sensible question when planning a turmeric crop. It does not make elevation a complete explanation of yield.
For Hawaiian Red, we have something much more useful than a generic claim about tropical plants: replicated Hawaiʻi trials that identify the cultivar and report fresh yield. They support a practical starting budget for warm sites. They also show why a neat “pounds per acre at each elevation” chart would imply more certainty than the experiments provide.
Our working estimate: for a warm, well-managed site reasonably comparable with the lower-elevation trials, we would begin planning around 20–30 metric tonnes per hectare of fresh fingers, approximately 17,800–26,800 lb per acre. This is our conditional planning range, not a published prediction interval. At cooler upland sites, the evidence is too confounded to offer a reliable healthy-crop forecast; a small local trial should establish that number.
Start with the yield the researchers actually measured
The anchor is Tavares and colleagues’ 2022 Hawaiʻi study. The table below selects Hawaiian Red from its seven site-year trials. Yields are fresh marketable fingers, excluding mother corms, estimated using 22,222 plants per hectare. Each trial used four experimental blocks; five harvested plants per block were subsamples, not twenty independent field replications.
| Site and harvest year | Elevation | Metric t/ha | lb/acre | Disease context |
|---|---|---|---|---|
| Waimānalo, 2019 | 23 m / 75 ft | 31.35 | 27,970 | No major disease outbreak reported |
| Waimānalo, 2020 | 23 m / 75 ft | 21.64 | 19,307 | No major disease outbreak reported |
| Waimānalo, 2021 | 23 m / 75 ft | 25.33 | 22,599 | No major disease outbreak reported |
| Kauaʻi, 2020 | 160 m / 525 ft | 24.78 | 22,108 | No major disease outbreak reported |
| Kauaʻi, 2022 | 160 m / 525 ft | 23.72 | 21,162 | No major disease outbreak reported |
| Kula Agricultural Park, 2019 | 428 m / 1,404 ft | 36.59 | 32,645 | No major disease outbreak reported |
| Maui ARC, 2020 | 990 m / 3,248 ft | 5.69 | 5,076 | Severe rhizome disease |
The low-yield Maui ARC trial had severe rhizome disease. That observation cannot identify a temperature-only penalty. For the broader study designs and additional regions, see our growing-trial catalog.
What we would propose at different elevations
The bands below are practical reading aids, not biological thresholds. Crossing 600 feet does not suddenly change a turmeric plant. Use the observed values as reference points, then decide whether your site resembles the trial in growing conditions and management.
| Elevation band | Evidence and proposed use |
|---|---|
| 0–600 ft (0–183 m) | Five lower-site observations provide the strongest local reference. If the site is comparably warm, well drained, adequately watered, and planted for a full season, use our 20–30 t/ha planning range initially. Budget below that when establishment, irrigation, or disease status is uncertain. |
| 600–1,600 ft (183–488 m) | One favorable trial near 1,404 ft exceeded the proposed warm-site range. We would still start a comparable warm site at 20–30 t/ha, using the higher observation as an upside scenario rather than the default expectation. One productive field does not establish a mid-elevation optimum. |
| 1,600–3,000 ft (488–914 m) | No Hawaiian Red observation in this seven-trial set directly fills the interval. We would not interpolate a central yield from altitude alone. Use a measured local pilot; until then, test several downside cases instead of treating a single number as likely. |
| 3,000–3,500 ft (914–1,067 m) | The approximately 3,248 ft trial is a combined cool-site/disease stress case. Its 5.69 t/ha result is useful for asking whether a project could tolerate poor performance. It is neither a healthy-crop ceiling nor a reliable expected yield for every upland farm. |
| Above 3,500 ft (above 1,067 m) | These data do not support a cultivar-specific Hawaii estimate. Establish performance in a small, measured planting before scaling. A greenhouse or a different cultivar would be a different production system, requiring its own evidence. |
The 20–30 t/ha range is an editorial planning judgment, a conservative rounded interpretation of the warm-site results, below their best outcome. It is not a confidence interval, a profit projection, or a claim that most future farms will land inside it. The upper and lower limits do not include every possible weather, planting-stock, or disease failure.
For an untested intermediate site, a useful sensitivity exercise is to calculate what happens at 100%, 75%, 50%, and 25% of a chosen reference yield. Those fractions are assumptions that you control. We are not claiming that a particular elevation causes a particular fraction.
Use elevation to investigate climate, not replace it
The National Weather Service’s Hawaiʻi climate summary describes temperature generally declining with elevation and strong differences in rainfall over short distances. The University of Hawaiʻi Rainfall Atlas lets growers compare rainfall patterns and their uncertainty, while the Climate Atlas provides additional variables such as air temperature.
Use those resources to screen a location, then measure the planting itself. A map’s annual average does not tell you whether water stands in the root zone after an afternoon downpour. Nor does it tell you the temperature of a raised bed under mulch, the timing of a dry spell, or how long the crop remains actively growing.
For comparing sites, keep a short record of planting date, soil temperature, rainfall and irrigation, drainage after storms, and the date growth begins to slow. A warm site without adequate water can lose its apparent advantage. A cooler site with good drainage and healthy planting stock may behave differently from a diseased trial at a similar elevation.
Our soil-temperature guide separates pre-sprouting experiments from field-growing conditions. A temperature that produces sprouts quickly in a chamber is not, by itself, a demonstrated optimum for season-long rhizome yield.
Work through the units before making the budget
A metric tonne is 1,000 kilograms; a hectare is approximately 2.471 acres. Therefore:
lb/acre = metric t/ha ×892.18
For example, 25 t/ha is about 22,304 lb/acre. A half-acre at that reference productivity produces about 11,152 lb of fresh fingers. If you deliberately stress-test the plan at 75% of the reference, the result is about 8,364 lb. The 75% factor is your scenario, not a measured effect of elevation.
Keep the crop area consistent with the study’s area basis. Do not multiply a whole-farm acreage by a dense-bed yield while ignoring roads, work areas, or unplanted land. Conversely, do not subtract pathways twice if the reference plant density already includes its access spacing. Record how you define a production acre.
Do not mix fresh fingers, total fresh rhizomes including mothers, and dried powder. Adding mother corms changes the product category; drying removes water and processing introduces further losses. A fresh-yield figure cannot establish powder output without measured recovery data. The same caution applies to saleable grades: a buyer’s size specification may differ from a research definition of marketable material.
Explore a fresh-finger yield scenario
This calculator scales a reference yield by area and your chosen productivity assumption. It does not calculate a yield penalty from elevation. Reference values already concern marketable fingers; the percentage below is an additional scenario adjustment.
Why we have not fitted an elevation-yield equation
Software can draw a line through seven points very easily. That line would not resolve the design problem. Elevation travels together with site, soil, year, weather, planting date, management, and disease. The trial was valuable for comparing cultivars across environments; it was not an experiment that changed only elevation while holding everything else constant.
Giving the line a slope or an impressive-looking coefficient would make a weak planning tool appear precise. A new site halfway up the mountain is not guaranteed to sit halfway between two observed yields. Nor can the three seasons at one location be treated as three independent pieces of evidence about different elevations.
Our chart therefore displays observations without a fitted trend. Its most useful message is the contrast between information we have and information still missing. The model worth building next is a local one that includes measurements relevant to the crop, with new harvests reserved to test how well it predicts.
Turn an unknown site into useful evidence
A manageable pilot should answer the question you need for your next planting. Keep the cultivar and planting-stock source documented. Use several separated beds or blocks that represent the field’s variation, rather than selecting only the best corner. Mark the planted area, count the original planting positions, and record losses instead of quietly excluding failed plants.
At harvest, weigh fresh fingers and mothers separately using the same cleaning and draining procedure. Record healthy saleable material, disease losses, and other rejects. Note the harvest date and crop duration. Report both yield per planted area and yield per surviving plant: a high surviving-plant yield can conceal a poor stand.
Repeat the observation through another season before treating a strong first result as normal performance. If testing two practices, such as mulches, spread both treatments across comparable blocks and keep other management as consistent as possible. For a larger trial, an extension specialist can help arrange the comparison so field variation does not masquerade as a treatment effect.
Finally, separate harvest weight from business capacity. Labor, washing, storage, customer grading, seed retention, and dispatch can each limit what is sold. This article models an ingredient yield; it does not claim that our farm’s announced annual capacity has already been harvested, sold, or verified by these experiments.
Continue with the trial catalog, pest and disease guide, and documented history of Hawaiian Red. Together, those records give a more useful picture than altitude alone.
Sources & further reading
Put the evidence in context.
Growing conditions and cultivar identity matter. Use the trials as reference points, record your own site conditions, and keep forecasts separate from measured harvests.
Browse the trial catalogRead our farm’s growing practices