Agriculture Reference
In-Depth Information
SECTIon TWo - SEEd LoT HETEroGEnEITy
Seed lots are by deinition homogenous from bag to bag or container to container. While it is recognized that
some variability will occur, excessive variability (heterogeneity) in content or quality becomes unaccept-
able. When this occurs, the seed lot should be divided into two or more smaller lots of acceptable homogene-
ity or bulked or reblended for greater uniformity.
The consequences of excessive heterogeneity are, in turn, excessive variability in test results on repeated
samples from such lots. This further results in invalid quality labeling and problems in application of toler-
ances. Equally important, it results in a variable product and potential loss of consumer conidence in the
seed and its performance.
Seed Lot Size restrictions in International Shipment
The International Seed Testing Association (ISTA) rules require that seed lots in which differences in qual-
ity between containers or primary samples drawn from them become visible to the sampler be declared
unacceptable and be refused. In cases where variation is not obvious, heterogeneity tests are used to deter-
mine acceptability. Research has shown that larger seed lots, because of their size alone, tend to be more
heterogeneous than smaller lots. Thus, ISTA has set limits on lot size, beyond which excessive heterogene-
ity is thought to be a problem. Maximum ISTA lot sizes for agricultural and vegetable species are found in
Table 2A (Part 1) of the ISTA rules, 2010 and range from 10,000 kg (22,000 lb) for small-seeded grasses to
30,000 kg (45,000 lb) for large-seeded types such as grains and pulses. The lot size for tree and shrub seeds
ranges from 500 to 1000 kg for most species, with a few exceptions (Table 2A [Part 2]). Seed lots moving in
international trade must meet these size limits to be eligible for Organization for Economic Cooperation and
Development (OECD) certiication. These regulations have been established to prevent excessive heteroge-
neity in OECD certiied seed in international trade. The following information on heterogeneity, including
the examples used, is largely based on information in the ISTA rules and related studies.
Types of Heterogeneity
There are two types of heterogeneity that relate to seed lots.
In-range heterogeneity
compares observed
and theoretical (expected), variance caused by random sampling variation and additional variation in a seed
lot. Such heterogeneity can cause variability in any quality test, including purity, germination, or other seed
count. Heterogeneity within each container is not involved since samples are independently drawn from
each container.
Off-range heterogeneity
involves comparing differences (variation) between samples of
similar size drawn from a lot which may be caused by mixing two lots of widely different quality without
thorough blending. This can result in widely different estimates of various quality tests.
determining Heterogeneity
Two different tests are used by ISTA in determining heterogeneity. The H-test, which was irst developed
by Leggatt in Canada in 1933 and modiied by Miles from Purdue University in 1963, has been used since
that time. The R-test was proposed by Miles et al. (1960a). It has been revised by Banyai and was introduced
into the ISTA rules in 1985.
The basic principle of the heterogeneity test is to measure heterogeneity of a seed lot by comparing
the theoretical variance with the actual, or measured variance. Variation in results of purity and germination
follows the Binomial distribution, while those of other seed counts (e.g., noxious weed seeds) follow the
Poisson distribution.
