Full moon and produce

Sowing and harvesting by the moon — the "agricultural calendar" and biodynamic farming — is folklore the world over. The waxing days before the full moon, the "fruit days," are said to be best for fruit. We test it against how much produce reaches the market: the daily wholesale volume of fruit and vegetables at Tokyo's central wholesale market.

Total wholesale volume (all produce)

Vegetables (mostly leaf and root crops)

Fruit (what the "fruit days" should favor)

Data: 2004–2026, Tokyo Metropolitan Central Wholesale Market daily report (produce, all markets), 5,919 trading days

Farming by the moon

Deciding when to sow and harvest by the moon's phase is folklore the world over — Japan's old lunisolar farming calendar, and Europe's biodynamic agriculture (begun by Rudolf Steiner, popularized by Maria Thun's lunar planting calendar). The waxing days before the full moon, the "fruit days," are said to be best for fruit: the moon is imagined to draw water and life upward, swelling seeds and fruit.

If the moon really does ripen crops, those vegetables and fruit are harvested, carried to market, and sold at auction. So — around the full moon, the volume of produce reaching market ought to rise. We test that against some 22 years of daily wholesale volume at the Tokyo Metropolitan Central Wholesale Market, one of the largest produce markets in the world.

How this verdict is computed

The data is the daily wholesale volume of produce (all markets) from the Tokyo market's "daily report" (2004–, Shift-JIS CSVs aggregated over every trading day). Besides the total, we split out vegetables (mostly leaf and root crops) and fruit — the crop the "fruit days" are supposed to favor.
A produce market has a strong weekly rhythm (closed days) and seasonality (what's in season). So we take the average for the same weekday-and-month as the expected value, and compare actual ÷ expected as an index — the same adjustment used in the births, accidents and fish topics.
Using the moon's age at noon Japan time, we mark "full-moon days" (the instant ±24h) and "new-moon days," and compare each group's mean index against the baseline (100).

See the methodology for the full criteria.

A caveat on the data

Market arrivals are a proxy for supply, not the harvest itself. There is a short lag from harvest to shipment and auction, and storable root crops and imported fruit flow decoupled from any single harvest date. Still, if the moon shaped the harvest, its shadow should fall across 22 years and thousands of trading days of arrivals — and if the "fruit days" worked, it should show in fruit first.

What clearly moved the volume

What clearly moved market volume was the season — the round of temperature and sunlight — and the market's own opening days. Fruit gathers in summer, leafy greens firm up in winter; that rhythm is stamped strongly into the numbers. If the lunar calendar plays a part, it is still lost inside that big rhythm and can't be told apart yet.

What we can't see yet

This result doesn't mean "the moon has nothing to do with it." Here is what the current method can't see:

Growth and flavor: the lunar calendar talks about how well things ripen, how they taste and keep. Market volume can't see that
Time lag: an effect of sowing day would show up months later at harvest, not in the arrivals on the full-moon day itself
Individual crops: if some crops responded more than others, the all-produce total would dilute it

What we want to check next

Split arrivals by crop
Link sowing dates to harvest dates

Perhaps the wisdom of farming by the moon holds something we haven't yet turned into numbers. We'd like to find it in the data one day.

Data sources

Tokyo Metropolitan Central Wholesale Market daily report (daily wholesale volume, produce, all markets; source: Tokyo Metropolitan Central Wholesale Market)
Lunar age and the instants of the phases are computed in-house from the algorithms in Jean Meeus, "Astronomical Algorithms" (UTC basis)

No difference