The Grade on the Bag Predicts Less Than the Season Does

Yongchun Foshou's character is set mostly by its season. The grade letter on the bag governs a narrower trait. A study published on December 15, 2025 in npj Science of Food¹ measured 40 samples of this oolong across two harvests and four commercial grades, and found that the spring and autumn pickings differ so widely in their aroma and polyphenol chemistry that they behave as two teas rather than two batches. Grade's authority is concentrated almost entirely on one thing: how much of the amino acid pool, and therefore how much mellowness and umami, survives into the finished leaf.

The same paper built the first dedicated flavor wheel ever made for this cultivar, drawn from those forty samples and nothing that came before them.

The cultivar nobody had measured

Yongchun Foshou is a geographically protected oolong from Yongchun County, Fujian Province, made from the large-leaf Foshou cultivar. Buddha's Hand. The name describes the leaf, which is unusually large and shaped like an open palm. Locally the tea is held in high regard for a rich aroma and a mellow body, and that reputation has done the work that measurement should have been doing.

Consider what was actually on the record before this study. Earlier research on Foshou covered roasting, storage, and potential health benefits. Its core flavor chemistry was undocumented, no comprehensive profile of its aroma existed, and the question of how season and grade interact to shape the cup had never been asked. A tea can be prized in its own county for generations and still be, in the literature, a blank.

The blank matters more here than it would for a better studied leaf, because Foshou does not behave like the oolongs the reader has probably calibrated on. Premium Tieguanyin is prized for fruity and roasted character, and its grade is chemically signaled by compounds such as nerolidol and the indoles. Foshou runs herbal and woody instead, and its own grade marker is a different compound entirely: 2-nitroethylbenzene, a signature that belongs to this cultivar alone. Judging Foshou by Tieguanyin's markers would rank it wrongly on axes it was never running on.

Forty samples, 214 compounds

The design was straightforward and unusually well replicated. Two harvest seasons, spring (YCFS-S) and autumn (YCFS-A). Four commercial quality grades within each season, A through D, highest to lowest. Five biological replicates per grade per season, and three technical replicates per sample. Forty samples in total.

Volatiles were captured by headspace solid-phase microextraction (HS-SPME) and read on a gas chromatograph coupled to a mass spectrometer (GC-MS). Nonvolatiles went through liquid chromatography (HPLC): amino acids, tea polyphenols, caffeine, soluble sugars, the individual catechins (C, GC, EC, EGC, ECG, EGCG), flavonoids, and the pigments theaflavins, thearubigins, and theabrownins. A trained panel ran quantitative descriptive sensory analysis alongside. The statistics were chemometric: two-way ANOVA with effect-size analysis, and a discriminant model (OPLS-DA) to pick out which compounds actually drove the split.

The aroma inventory came to 214 distinct compounds, sorted into ten chemical categories: 39 esters, 33 terpenes, 31 alcohols, 30 aldehydes, 27 ketones, 27 hydrocarbons, 10 acids, 9 aromatics, 7 heterocyclic compounds, and 2 ethers.

The instrument that produced the surprise is the effect-size analysis. An ANOVA that only asks whether a factor matters will report that both season and grade matter, which any drinker could have guessed. Partial eta-squared asks a harder question: of all the variation observed in a given compound, how much of it does this factor own? Ask that, and the two factors separate cleanly.

What the season decides

Season was the dominant factor for most metabolites measured, with partial eta-squared above 0.84 for tea polyphenols, caffeine, and the catechin EC. A value that high means season explains the overwhelming majority of the variation in those compounds. Grade, against it, is noise.

Amino acids ranged from 0.62 to 2.61 percent across all samples, running higher in spring; flavonoids ranged from 7.57 to 15.03 percent, also higher in spring. Caffeine was higher in spring too. The aroma compounds linalool and phenylacetaldehyde reached their peak concentrations there. On the panel, spring read as subtly fruity and herbal.

Autumn inverted much of it. Tea polyphenols overall were higher, and so were the catechins EC, EGC, ECG, and EGCG. On the aroma side, autumn carried more (E)-nerolidol, geraniol, and alpha-ionone, plus 3-octen-2-one and dihydroactinidiolide. The panel read it as floral, sweet, and woody.

Thirty aroma compounds were the ones statistically distinguishing the two seasons from each other, and a further twelve phytochemicals returned VIP scores above 1 for seasonal discrimination, marking them as the strongest individual drivers of the spring and autumn split. Two bags of the same grade from the two harvests are not two examples of one tea.

What the grade decides

This narrower finding is the one worth carrying to a merchant.

For amino acids specifically, grade was the primary driver, with a partial eta-squared of 0.834 against season's 0.42 on the same compound class. That is a clean reversal of the general pattern, and it is confined to that pool. Amino acids are what give a cup its mellow, savory weight, and the sensory panel followed the chemistry exactly: higher grades scored better on mellowness and on umami.

Grades A and B retained more amino acids and more aroma compounds overall. Twelve aroma compounds served as the critical grade markers: linalool, 2-nitroethylbenzene, 3,5-octadien-2-one, alpha-ionone, hexanal, heptaldehyde, 1-octanol, 1-octen-3-ol, nonanal, benzaldehyde, dihydroactinidiolide, and phenylacetaldehyde. A further 31 aroma compounds distinguished grade variations by VIP score, and 20 of those overlap with the 30 that marked the seasonal split. The compound lists are not clean or separate; the effect-size analysis is what actually tells the two apart.

So grade earns its place on the label. It simply governs one trait with real force rather than governing quality in general.

A lower grade is not less of everything

The C and D samples showed elevated hexanal and heptanal, two green, grassy aldehydes that count as off-notes in a finished oolong and work directly against a smooth cup. That much is expected.

What is not expected is the rest of the profile. Lower grades carried higher concentrations of the pigments theaflavins and thearubigins, and higher levels of the catechins C and EGCG, than the higher grades did. More pigment, more of two catechins, less amino acid.

A lower grade of Yongchun Foshou is therefore a different chemical object, not a thinner one: astringent-leaning in its pigment and catechin content, short on the amino acid smoothness that makes the higher grades read as mellow, and carrying green aldehydes that the better lots have shed. Whether that trade is a loss depends on what the drinker is after, though the panel's verdict on mellowness was not ambiguous.

The limits, held honestly

This is one cultivar. An understudied, geographically specific one, whose grade marker is a compound that does not mark grade in the oolongs most readers know. Four commercial grades of Foshou, assessed by whatever standard Yongchun applies, are not a general theory of oolong grading, and the Order does not extend a finding past the samples that produced it. Forty samples, two seasons, one county.

Nor does the study speak to what happens across successive infusions. It measured leaf and panel, not the arc of a session, so the Rule of this Order stands unaffected and untested by it.

The paper changes the question a novice should ask before buying. Grade tells you how mellow the cup will be. Season tells you what tea it is. Ask for the harvest first, and expect the merchant to know. Until December 2025 there was no wheel to hang any of that on. There is one now, drawn from 40 samples of a leaf shaped like a hand.