Germination Capacity: The Share of Normal Seedlings in a Defined Test
Ten seeds sit on damp paper, and after three days seven show a white root tip. That is not 70 % germination capacity. Germination capacity is the share of seeds that produce a normal seedling under defined conditions, meaning a young plant with a sound root and a sound shoot. What gets counted is what developed properly, on a fixed day and at a fixed temperature. A seed lot here is one packaged batch from one harvest, and the figure always describes the batch rather than any single seed. Which is why every percentage needs the conditions that produced it.
What a germination test counts
At the end of the test every seed goes into one of three groups: normal seedling, abnormal seedling, dead seed. For hemp a seedling counts as normal when its main root is neither stunted nor deformed, or when it has instead grown numerous side roots, and when its stem and leaves stand undeformed. Everything else is abnormal, even where the hull has split and something is visibly protruding.
The reason for the split is practical: an abnormal seedling will not become a plant in a bed. A seed that pushes out a root and then stops has germinated and still delivered nothing. The international testing rules therefore define the abnormal seedling by what it lacks: any prospect of developing into a serviceable plant under good conditions.
The conditions are part of the figure
For hemp the procedure is documented. Four replicates of one hundred seeds each go onto rolled germination papers moistened with a weak potassium nitrate solution and rest in sealed bags so the moisture cannot swing. Temperature alternates across the day: sixteen hours at 20 °C in darkness, eight hours at 30 °C in light, a night-and-day regime rather than steady warmth. The count is made seven days after setting up.
That these details are not a formality shows in a comparison of six arrangements applied to the same five lots. Every lot gave different figures; in one they ran from 42.75 % to 59.75 %, a spread of roughly a quarter of the value reached there. Sand as a base gave the highest figure in three of the five lots and, in another, the lowest of all six arrangements. A germination figure without its conditions cannot be compared with any other.
Two counting dates, two statements
Counting happens twice. After three days the test records germination energy: only those seedlings whose essential structures are already fully formed. It describes speed and serves as a first indication of how a lot will perform in the field.
The second date, at seven days, gives germination capacity itself, split into normal seedlings, abnormal seedlings and ungerminated seeds. The two figures can sit far apart: a lot that starts slowly and still comes up in full has low germination energy and high germination capacity. Read only the earlier date and it looks like a poor lot.
Why freshly harvested seed tests badly
A seed can be alive and still not germinate. That state is called dormancy, and in hemp it lasts only briefly after harvest. In work on two varieties, germination twenty-one days after harvest stood at 79 % and 91 %, while a staining test, which turns living tissue in a cut seed red, found 97 % and 96 % living seeds in the same lots. The gap came from resting seeds, not dead ones.
Two routes close it. Prechilling the moistened seed at 10 °C for five days before the test itself raised germination to the level of the staining test. And simply leaving the lot alone does the same: from twenty-eight days after harvest onward, prechilled and untreated samples no longer differed, the dormancy having broken down by itself. Test a fresh lot and you test it too early.
What the figure does not tell you
Germination capacity says how many seeds give a sound young plant under good conditions. It says nothing about how many do so under poor ones. There is a separate term for that, seed vigour, meaning performance under adverse conditions. The rulebooks of the two major testing bodies carried no procedure for it in hemp in 2019; one proposal stresses the seed for forty-eight hours at 43 °C and near-total humidity, then germinates it under the usual regime. The figures dropped sharply: for one of the varieties tested, to 64 % and 70 %, against 97 % to 98 % living seeds in the staining test on the same lots.
Nor does the figure say anything about its own reference base when nobody states it. The studies read here set up one hundred seeds per replicate without addressing how pure the material was beforehand, and neither of them gives the spread still permitted between the four replicates before a test has to be repeated.