Pieces of forest tree seed units… …Pieces of unscientific nonsense

Introduction and Objectives

As all forest tree, shrub and other seed analysts know, when performing a purity test in accordance with the International Rules for Seed Testing (ISTA Rules), in compliance with Rule, if

Pieces of seed units larger than one half their original size with a portion of the testa attached are found in the working sample, they must be considered to be pure seeds.

Objective 1

The primary objective of this review is to demonstrate that such pieces of forest tree seeds/seed units are not merely inert, they are actually dead, particles. Whereas they may be inert by the time they find their way to the analyst’s purity table, by the time they are removed from storage for use in a tree seedling nursery they are dead. Classifying them as pure seeds has no scientific basis.

The context for the discussion that follows is found

  1. in the second sentence of the fourth paragraph of the Introduction to the ISTA Rules (Anon. 2009) which claims that:

    The test methods used must be based on scientific knowledge and the accumulated experience of those working in seed testing and quality control.

    In this the word “must” is vital. And,

  2. in the fifth sentence of the first paragraph which claims that:

    The primary aim of the ISTA Rules is to provide testing methods for seeds designated for growing of crops or production of plants.

    In this the words “growing” and “plants” are vital.

Objective 2

Abundant evidence will be documented to demonstrate that as far as Purity Test Rule and forest tree seeds are concerned, these claims are patently untrue.

They are, in fact, scientifically bankrupt.

General Overview of the Purity Test

Before going into too much detail, a brief overview of the basics of the Purity Test is in order.

Rule 3.1 Object states (in part):

The object of the purity analysis is to determine:

  1. the percentage composition by weight… and by inference the composition of the seed lot, and

  2. the identity of the various species of seeds and inert particles constituting the sample.

Item (b) is the critical factor in this review: that is, the identity of inert particles in the sample.

To understand this correctly the ISTA definitions that follow Rule 3.1 must be considered. It is essential that the details be restated here to form the basis of the discussion that follows. Thus, Rule 3.2.1 Pure seed begins with:

The pure seed must refer to the species stated by the applicant, or found to predominate in the test, and must include all botanical varieties and cultivars of that species including:

  1. The following structures (even if immature, undersized, shriveled, diseased or germinated, provided they can be identified as of that species) unless transformed into visible fungal sclerotia… smut balls or nematode galls:

    1. Intact seed units (= commonly found dispersal units i.e. achenes and similar fruits, schizocarps, florets. etc.) as defined for each genus and species in the Pure Seed Definitions (PSDs) in Table 3B Part 2.

      In Poaceae (Gramineae):

      1. florets with an obvious caryopsis containing endosperm,

      2. free caryopses.

    2. Pieces of seed units larger than one-half their original size.

      (Sub-item 2 above is referred to throughout the following text as Rule

Rule, as was pointed out by Ashton (2000), does not define how to measure “one-half.” Does it mean that only the length of the seed is to be considered, or all of its dimensions, i.e. seed mass (see also Gorian et al. 2006). It was contended that for symmetrically-shaped seeds an estimate based on a linear measurement should be considered. For asymmetrical seeds it may be necessary to consider more than one dimension. However, actual measurements do not necessarily give a more accurate result because the appearance of the missing fragment can only be estimated. By way of confirmation, Ashton (2000) reminded everyone that seeds that are “exactly one half their original size are classified as inert matter” (Rule

In this review only the length of the seed will be considered, even though—as will be illustrated—forest tree seeds are usually asymmetrical.

The Rules continue with:

From the above main principles, exceptions are made for certain genera of Poaceae (Gramineae) (Table 3B, Part 2): … (details omitted as they are unimportant herein).

Rule 3.2.2 Other seeds (mentioned here for the sake of completeness, to be considered later)… shall include seed units of any plant species other than that of pure seed…

Then comes the really important part:

Rule 3.2.3 Inert Matter

Inert matter shall include seed units and all other matter and structures not defined as pure seed or other seed as follows:

  1. Seed units in which it is readily apparent that no true seed is present. Does this not apply to shriveled, diseased, pre-germinated seeds, or broken seeds of any size?

  2. Florets of those species listed in with a caryopsis less than the minimum size. (Remaining details not important for forest tree seeds.)

  3. Pieces of broken or damaged seed units half or less than half of the original size.

  4. Those appendages not classed as being part of the pure seed in the pure seed definitions for the species (Table 3B Part 2). Appendages not mentioned in the pure seed definitions must be removed and included in the inert matter. (This would apply to seed wings of certain gymnospermous species—see item 7 below.)

  5. Seeds of… Cupressaceae, Fabaceae (Leguminosae), Pinaceae, Taxaceae, and Taxodiaceae with the seed coat entirely removed. In Fabaceae (Leguminosae), separated cotyledons are regarded as inert matter, irrespective of whether or not the radicle-plumule axis and/or more than half of the testa may be attached.

  6. Seeds of Cuscuta spp. … (other details not important herein.)

  7. Unattached sterile florets, empty glumes, lemmas, paleas, chaff, stems, leaves, cone scales, wings, bark, flowers, nematode galls, fungus bodies… smut balls, soil, sand, stones, and all other non seed matter.

  8. All material left in the light fraction… (other details not important herein).

Attention is drawn to Rule (item 5 above) which states that “in Fabaceae (Leguminosae), separated cotyledons are regarded as inert matter, irrespective of whether or not the radicle-plumule axis and/or more than half the testa may be attached.”

No mention is made of separated cotyledons in the Cupressaceae, Pinaceae, Taxaceae or Taxodiaceae. Clearly, the Fabaceae (Leguminosae) have been given special treatment.

Perhaps this is because the Cupressaceae, Pinaceae, Taxaceae and Taxodiaceae (and the Berberidaceae and Brassicaceae [Cruciferae]) were “shoe-horned” into Rule well after this Rule had been written, and no-one gave the slightest consideration to their requirements. This will become a major issue in discussing gymnospermous and angiospermous tree seeds.

Why is this? Why do only some of the Fabaceae (Leguminosae), which includes the forest trees Acacia, Gleditsia, Robinia, Sophora (see Table 1), receive special attention, but not other forest tree seeds?

Probably because forest tree seeds were not recognized, i.e. were not included in the Rules until the late 1950s/early 1960s.

Also, there is no further reference to the radicle-plumule axis with more (or less) than half the testa still attached, so it has to be assumed that this part—the radicle-plumule axis—of the seed is also to be classified as “inert matter.”

Should this not be made completely clear?

This is obviously a lack-a-daisical approach to this issue, and it has never been thought through scientifically.

Based upon the preceding General Overview, what actually comprises a Purity Test in non-ISTA language forms the basis of this review. As all purity analysts know, it is a visual examination (a rapid visual examination, speed is so important—to be discussed in PART II) of the external appearance of the particles in the sample. As will be demonstrated, pieces of broken tree seeds of any size are easily recognizable and, because they have zero potential for producing a new plant, they must be classified accordingly: not “pure seed.”

However, does this mean that the analyst is being asked to perform a growth assessment? No, the analyst is being asked to recognize that pieces of broken tree seeds of any size have no more potential for producing new plants than do unattached sterile florets, empty glumes, chaff, smut balls, etc. (see item 7 above) and so should classify such pieces as “inert matter,” if not as “dead matter” (see following).

So what is really meant by “pure seed?”

The word “pure” can be defined in several different contexts (see Dictionary quotes). However, the use of “pure” by ISTA has never been defined in relation to seeds, despite the many Pure Seed Definitions.

In the Glossary at the end of this review are several definitions of the word “pure” quoted from well-known English Language Dictionaries. Prominent among these are the definitions “uncorrupted; faultless; genuine article; free from alteration; free from anything debasing or deteriorating; unadulterated.” (See especially Grove 1961, and Murray 1971).

“Unadulterated” is a significant word in this context. According to the Concise Oxford Dictionary of Current English (Fowler and Fowler 1964), it means the opposite of “adulterated” which, in turn, means “spurious,” or “counterfeit.”

Being the international agency for testing and certifying seed quality, ISTA can define a “pure seed” in any manner it pleases. Since the Rules were first promulgated, ISTA has drawn up (concocted might be a better term) 63 Pure Seed Definitions (PSDs) (Rules, Table 3B Part 2). To the non-analyst lay-person, such as the forester, even the farmer, anyone who is a seed user, the term “pure seed” connotates, i.e. suggests, that the seed(s) in question might produce new plant(s). Of course, the purity analyst cannot determine if the seed(s) will germinate because (as already noted) the purity test is not a growth test. Nevertheless, some unbroken, i.e., intact, seeds may be devoid of an embryo and associated tissues (empty seeds), or may contain an insect larva, but this cannot be determined by the purity analyst and so they may appear superficially to be sound, that is “pure seeds.” Conversely, as already claimed in the foregoing, the analyst should recognize that broken tree seeds are (so-called) “inert matter.”

The heart of this review is aimed at showing that if some forest tree seeds are broken into pieces of any size, including those “more than one-half their original size,” these cannot be regarded as “pure seeds” according to the dictionary definitions in the Glossary because they are not “free from alteration” (Grove 1961), they have been adulterated. That is, they are “not unadulterated,” they are “spurious.” At the expense of repetition, such pieces of forest tree seeds do not require any growth tests to be visually and rapidly recognized for what they are, viz. not “pure seeds” but so-called “inert matter.” Actually, as will be reasoned below, they are dead matter.

Indeed, the analyst is required to recognize them and to judge if they meet the “more than one-half their original size” criterion (Rule The scientific knowledge referred to in the Introduction to the Rules, and upon which test methods must be based, will be documented in the following pages to establish, beyond any doubt, that for forest tree seeds such “pieces of seeds,” larger (or smaller) than one-half their original size, have absolutely no potential for producing new plants. They have no more value to the forester, the silviculturist, and any other seed user, than the “chaff, stems, cone scales, wings, bark, flowers, nematode galls, fungus bodies… smut balls, soil, sand, stones” listed in Rule, and should not be classified as “pure seed,” but as “inert matter.” Or, as mentioned above, even Dead Matter.

So what is meant by “inert matter?”

Inert matter applies to anything that has no life functions, such as broken seeds.

Also in the Glossary are several definitions of the word “inert” quoted from well-known English Language dictionaries. Most speak of the “lack of power to move,” but do not state that this “lack of power” is permanent. Two definitions (Grove 1961, Stein and Urdang 1967) do include the word dead, or lifeless, so it becomes a matter of context in which the word is applied. In a human context for example, a person may fall down and lie “inert,” that is, “without power to move or act” until medical help arrives. The person is revived and recovers the power to move.

Even intact, fully-developed forest tree seeds are “without power to move or act,” that is, they are inert when they are dormant, which is all of the time they are in storage until they are appropriately prepared for sowing in a nursery to produce reforestation stock: that is, until they are revived for germination. In other words, dormant seeds do not move, they do not grow because the power to move or act is not available in their dormant state, and they remain in this state for months, to several years (Bonner 2008).

Thus, based on official English-language definitions, an intact, fully developed, mature, dormant Douglas-fir (Fig. 1) (or Scots/Swedish pine, or Sitka spruce, etc.) seed would be classifiable as bothpure seed” and “inert matter.” That is, an “inert,” pure seed.

Seeds of a Protoaceae (Leucospermum spp.) and two legumes (Liparia spp. and Acacia spp.) collected in 1802–03 were germinated 150 years later at the Royal Botanic Gardens, Kew, United Kingdom. Supported by carbon-dating, other seeds have survived over 200 years (Daws et al. 2007), likely because they remained intact.

That is to say, these seeds remained “inert” this length of time (200 years), but they did not perish. Until they were ready to germinate, by having been subjected to the correct pre-germination treatment, they remained “inert.” That is, they did not move, they did not grow. But they were not dead as a tetrazolium test (Rule 6) would have quickly determined.

To press the issue of dormant-inert-not dead (likely intact) further, it is of value to note that seeds of a non-tree member of the Fabaceae (Leguminosae), the genus Lupinus (included in the Rules), were unearthed in 1954, kept in a bottle for 12 years before being sent to the National Museum of Canada where they were carbon-dated as being between 10,000 and 15,000 years old. Several were placed on damp filter paper and germinated within 48 hours. These seeds were some 8,000 years older than the then oldest-known seeds, those of the sacred lotus, that germinated after lying dormant (inert-not dead) in a far-Eastern peat bog (Porsild 1967).

Before pursuing this further, consider the following analogies:

The diamond, pine needle, and seed potato analogy

A jeweler cuts a rough diamond into two pieces—one is larger than one-half the original size (to be used as a pendant), while the other piece is smaller than one-half the original size (to be used in a ring). Both pieces remain pure diamond. The original size is in no doubt.

A forester plucks a needle from a pine tree, breaks it into two pieces—one piece being larger than one-half the original size, while the other piece is smaller than one-half the original size. Both pieces remain pure pine needle. Again, the original size is in no doubt.

A farmer randomly removes a seed potato from a storage bin to check its condition. Satisfied after hacking it into two unequal pieces he throws them back into the bin. The larger piece was larger than one-half the original size, while the other piece is smaller than one-half the original size. Both pieces remain pure seed potato. The original size is in no doubt.

Why, then, does ISTA insist that when a seed is broken into two pieces, only the piece larger than one-half the original size is to be regarded as “pure seed,” while the other piece (smaller than one-half the original size) is not to be regarded as “pure seed?” Are not both pieces of the seed organically still “pure seed” using the diamond, pine needle and oak leaf analogy? In any case, unless all the seeds in the seedlot are very uniform in size, how can the analyst be sure of the original size?

There is a difference, of course, between the diamond, pine needle, seed potato and the seed. The diamond has always been inanimate (see Stein and Urdang 1967; Cayne and Lechner 1987 in the Glossary), whereas the pine needle was animate until plucked from its “mother trees,” at which time it became inanimate, that is it is en route to becoming dead. The seed potato remains animate unless it goes mouldy. Likewise, a forest tree seed (or any other type) containing a germ is a living, animate, breathing (=respiring) entity. Its outer cover, the seed coat or testa, protects the tissues that will give rise to a new plant under the proper conditions. Damaging this cover, that is adulterating—even cracking—it, exposes the internal tissues to (a) disease, and/or (b) excess drying when the seeds are prepared for storage (more on this below). The diamond is not subjected to these stresses; while the pine needle eventually crumbles and dies, as does the seed potato if not planted soon. Likewise, pieces of broken seeds.

In the previous analogy, the diamond—piece of or complete—is not merely “inert,” nor is it dead since it never had any living properties. It always rested in peace, it had no mortal coil to shake off because it was always inanimate. But it remains “pure” diamond.

Therefore, it must be asked if ISTA (and other seed testing agencies such as AOSA—the Association of Official Seed Analysts of North America) is using the correct term to describe forest tree seed units from which the seed coats have been entirely removed as “inert matter” (Rule As will be discussed in detail below, these seed units are not merely “inert” as defined above. They are dead.

Again, any doubts on this can be quickly assuaged with a tetrazolium test (Rule 6). They are as DEAD as any chaff, stems, leaves, cone scales, wings bark, soil particles, sand, stones (as stipulated in Rule But soil particles, sand, stones are not dead. They never had any life, so they cannot be dead or, by definition “inert.” They should be referred to more correctly as inanimate.

Why not simply call such seeds “impurities?”

Because ISTA has chosen, inadvertently over the past 80 years, to play games with basic definitions.

It is probable that it will be argued (“an exchange of ignorance”—see below) that the words inert, dead and inanimate are one and the same. However, the main issue remains:

How can dead seeds be “pure seeds?” That is, seeds known to the analyst to be dead based on scientific knowledge and experience?

Because ISTA has failed to recognize them as such, choosing to confuse the result of the Purity Analysis.

Table 1 lists the 19 PSDs that apply to the 21 families (44 genera) of forest trees to which the Rules apply. All PSDs (11, 47, 49, 50, 51) for the 17 gymnospermous genera state that pieces of seed units larger than one-half of their original size (Rule are to be considered as “pure seed,” provided a portion of the testa remains attached. No explanation is given as to whether this must be a small or large portion of the testa.

Table 1. PSDs in numerical order for forest tree seeds

A. Gymnosperms




Piece of Seed (POS)


Cupressaceae, Taxodiaceae

Juniperus, Taxodium

POS with testa



Picea, Pinus II

POS with testa


Cupressaceae, Taxodiaceae

Calocedrus, Chamaecyparis, Cupressus, Thuja Cryptomeria, Sequoia, Sequoiadendron

POS with testa




POS with testa



Abies, Cedrus, Larix, Pinus I, Pseudotsuga, Tsuga

POS with testa

B. Angiosperms (and a Vascular Cryptogram – see Ginkgo)




Piece of Seed (POS)


Ginkgoaceae, Hippocastanaceae

Ginkgo, Aesculus

POS with or without testa


Fabaceae (Leguminosae)

Gleditsia, Robinia

POS with testa



Populus, Salix

POS with/without testa (part of or none)




POS with testa




POS with/without pericarp/testa (part of or none)




POS with testa


Aceraceae, Magnoliaceae, Oleaceae, Simaroubaceae, Ulmaceae

Acer, Liriodendron, Fraxinus, Ailanthus, Ulmus

POS with/without pericarp/testa (part of or none)



Alnus, Betula

POS with/without pericarp/testa (part of or none)




POS with/without testa




POS with/without pericarp/testa (part of or none)




POS with/without pericarp/testa (part of or none)


Betulaceae, Fagaceae, Tiliaceae

Carpinus, Corylus, Castanea, Fagus, Quercus, Tilia

POS with/without pericarp/testa (part of or none)




POS with/without pericarp/testa (part of or none)




POS with/without testa

Eleven of the 14 PSDs (10, 11, 12, 20, 48, 50, 52, 53, 54, 55, 56, 57, 58, 60) to which the 28 angiospermous genera are assigned, allow that pieces of seed units more than one-half their original size are to be recognized as “pure seedeven if no pericarp or testa remains attached. That is, completely naked seeds are to be recognized as “pure seed” provided they are larger than one-half of their original size. This is hair-splitting at its finest.

It should be clear from this that the ISTA idea of “pure seed” differs entirely from that of seed users—the farmer, the forester, the horticulturist.

A Brief Historical Note

A more complete history of the Rules will be presented in Part II, but it is useful to mention here that the early Rules (roughly 80 years ago) were formulated to deal with the quality of agricultural crop seeds. Forest tree seeds were not introduced into the Rules until some 50 years ago (in 1956?). The precise date of this introduction is uncertain as this review is written, but it is known that tree seeds were included in the Rules by 1961 when this reviewer underwent basic training at the ISTA-Accredited Seed Testing Laboratory of the United Kingdom’s Forestry Commission, Alice Holt Lodge Forest Research Station (see Reviewer’s Credentials at the end of this review). It is probably safe to say that the provision for Pieces (or Piece) of seed (or seed units) of non-tree species that were larger than one-half the original size as pure seeds was already contained in the 10 PSDs to which some tree seeds were assigned. For the bulk of other forest tree seed genera 9 new PSDs had to be written. Each and every one of these PSDs contains Rule That is to say, although certain tree seed genera were shoe-horned into existing definitions, all tree seeds were expected to comply with the existing piece/pieces provision without exception. That is, no consideration was given to the scientific knowledge of forest tree seeds.

In contrast, it must be noted that special provisions had already been made for certain Poaceae (Gramineae) genera and these provisions exist today as Rule So why do some genera in the Poaceae receive special attention, but not tree genera in the Cupressaceae, some Fabaceae, Pinaceae, Taxaceae and Taxodiaceae? Why has ISTA persevered in its studious avoidance of forest tree seed biology, the science of which is abundantly documented and which clearly points to any piece of tree seed as being no more valuable to the forester for producing a new plant than a similarly-sized piece of rock, needle, cone scale, or twig?

It is probably true that other tree seed scientists have asked these or similar questions in the past.

Method: Organization of the review

The Discussion that follows is divided into to two main parts, followed by a Literature Cited list, the Reviewer’s qualifications, and three Appendices, two of which include the bulk of the illustrations, and a Glossary.

Part I addresses the structure of forest tree seeds, and how these structures are impacted when the seeds are broken into pieces. Aspects of tree seed biology that demonstrate that broken forest tree seeds have zero use for the production of new plants will be included to further demonstrate that they must not be given the connotation of “pure seeds.”

Part II comprises a more inclusive historical overview of the purity test to show, among other things, how it existed in the first published Rules, how and when it was modified, and who promoted the changes. This will provide an opportunity to raise some ethical questions.