Agriculture Reference
In-Depth Information
mately the same number of seeds per gram. Regular AOSA purity tolerances are given in tables in the rules.
The following example of a chaffy seed lot illustrates each of the four components of a labeled sample com-
pared to the complete test results conducted on a second check sample. Original and check results are aver-
aged to give both equal credibility. Then the tolerance for each component of the purity test is developed
separately based on the mean of both test results. When this is done, the percentages of pure seed, other crop
seed and inert matter are shown to be outside of tolerance while the percent weed seed is within tolerance.
In the example below, a chaffy grass seed lot was labeled as having 98.00% pure seed, 1.01% other
crop seed, 0.11% weed seed, and 0.88% inert matter. A subsequent check test from different submitted
samples estimated the lot as having 94.12% pure seed, 2.21% other crop seed, 0.17% weed seed, and 3.50%
inert matter. The procedure for determining whether these results are within tolerance is given below.
Purity components
Label or
irst test
Second test
Mean of
two tests
Tolerance
values
Difference
between tests
Pure Seed
98.00
94.12
96.06
1.36
3.88*
Other Crop Seed
1.01
2.21
1.61
0.93
1.20*
Weed Seed
0.11
0.17
0.14
0.28
0.06
Inert Matter
0.88
3.50
2.19
1.05
2.62*
*Outside of tolerance according to Table 13B, AOSA rules.
Special Tolerances (AoSA)
Special AOSA tolerances are used for seed lots composed of more than 5% each of two kinds (or cultivars)
of pure seed with different weights per seed. Additional kinds (or cultivars) of pure seed may be present in
any amount. Special tolerances are also applied to each component of the purity test in the same way, but
are
much more complicated
to calculate. The AOSA special tolerances are given in Tables 13C, D, and E
of the AOSA rules (2010). Tolerances for the chaffy types are wider than those for nonchaffy types because
of greater amount of experimental error in chaffy species over and above that due to random sampling
variation.
Two methods can be used for assessing the difference between two test results, the “short method”
and the “long method.” Either method may be used, however, the short method should be tried irst. The
example of the short method given below was developed by Miles in 1963 and is used in the current version
of the AOSA rules (2010).
The average of the irst and second tests for bluegrass in a chaffy mixture is 16.25. Opposite 16.25 in
column A of Table 13E, the smallest and the largest tolerances are 3.22 and 11.23, respectively. Opposite
16.25 in column B, the smallest and the largest tolerances are 2.07 and 2.79, respectively. The smallest of
the four tabulated values is 2.07. If the difference between the two tests does not exceed 2.07, then the irst
and second tests for bluegrass are within tolerance. The largest of the four values is 11.23. If the apparent
difference exceeds 11.23, it is outside the range of tolerance and the difference is declared real (signiicant
difference is referred as “deiciency” in AOSA rules). If the apparent difference is between 2.07 and 11.23,
the correct tolerance must be computed as shown in Table 13C, AOSA rules (2010). Also, if the extent to
which a difference exceeds the tolerance is desired, then the correct tolerance must be computed by the long
method, using Table 13E if the seed is chaffy or 13D for non-chaffy seeds.
Table 13C of the AOSA rules shows the procedure for computing the correct tolerance for the blue-
grass and red fescue components of a chaffy seed mixture, when the short method is not appropriate.
Table 15E of the AOSA rules is used because the seeds are chaffy. The average percent of two blue-
grass tests is 46.2. Opposite 46.2 in columns A and B, the smallest and the largest tolerances are 3.60 and
10.35, respectively. The apparent difference between the two tests of 7.6 is between these values; therefore,
