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

Part II

A Partial History of the Purity Test

Seeking the origin of what is now Rule 3.2.1.1.2, a crucial 1965 report concerning the purity test (Proc. ISTA vol. 30) (see literature cited for full reference) by Prof. Dr. Harald Esbo (Statens Centrala Frökontrollenstalt, Solna, Sweden), who became President of ISTA 1965–1968, is quoted at some length as follows:

Concerning ‘Pure seed’, two different basic principles have been under discussion for many years. Up to 1950 the so called ‘Stronger method’, abbreviated S.M., was prevalent and prescribed by the ISTA rules, but in that year the so-called ‘Quicker method’ (Q.M.) was agreed upon by the ISTA Congress in Washington. According to S.M., only seeds which could possibly give rise to normal seedlings were considered pure seed; whereas Q.M. is including all questionable, damaged or badly developed seeds into the pure seed fraction leaving the evaluation of live or dead to the germination test. As the new method was saving time it was named Quicker Method. Nowadays this method is quite predominating. According to ISTA rules, pure seed includes all varieties of each kind under consideration as stated by the sender or found by the laboratory test, also,

  1. undersized, shriveled or immature seeds

  2. pieces of seeds more than half their original size

  3. fruits whether they contain a true seed or not unless it is readily apparent that no true seed is present

  4. diseased seeds

  5. and free caryopses.

There cannot be any doubt that Dr. Esbo (1965) was declaring that anything in the working sample that resembles a seed, even a clearly diseased or shriveled seed, except for a piece of seed that is less than one-half its original size is to be called a “pure seed.”

Where is the science claimed by ISTA to support this? Is this concept not scientifically bankrupt?

This concept is scientifically bankrupt.

Dr. Esbo then went on to state:

No doubt Q.M. has diminished the influence of personal judgement and led to more uniform results and is really time-saving.” (Esbo, 1965).

The matter of “time-saving” is repeated in item ii). This will be discussed below, together with the claim of diminished “personal judgement.”

Also note the inconsistent use of “seed” and “seeds” throughout.

In the same 1965 publication, Dr. O.L. Justice (Market Quality Research Division, Agricultural Research Service, United States Department of Agriculture, Beltsville, Maryland, U.S.A.) wrote that:

The story of the ‘Stronger Method’ and the ‘Quicker Method’ of testing for purity may well become a legend in seed testing. Briefly, Europeans contended that seed structures of the kind under consideration which were incapable of germination should be removed from the pure seed fraction and included with inert matter. The Americans and Canadians, on the other hand, believed that such structures generally should be left with the pure seed and the germination test allowed to determine their planting value. As a consequence, both the Stronger and Quicker methods were included in the original rules adopted in 1931, and retained until 1953 (ISTA, 1931). Dr. W.J. Franck of the Netherlands made a strong plea for adoption of a single purity procedure at the Washington Congress in 1950. This and similar representations by the late Dr. H.A. Lafferty of Ireland in 1963 paved the way for general acceptance of a single method of determining purity.” (Justice, 1965.)

The above quotes are from two pivotal papers. The key here is the 1950 Congress being held in Washington (D.C.) at which many members of the AOSA would have been present. A large proportion of these individuals probably enjoyed membership in both AOSA and ISTA as many do at present. It is worth noting that the AOSA Rules were first promulgated in 1908, almost a quarter of a century before the ISTA Rules were first adopted in Wageningen in 1931, so the Q.M. is likely to have been well established within the AOSA Rules long before the 1950 Congress. So, as Drs. Esbo and Justice reported, the North American contingent held sway over the Q.M.

But why did the Americans and Canadians believe that seed structures which were “incapable of germination” should be left with the pure seed, leaving it to the germination test to determine their planting value? Also, exactly what did Dr. Justice mean by “the story… may well become a legend in seed testing?” Perhaps the current review is part of that legend.

Dr. Esbo stated that it was not until 1950 that the ‘Quicker Method’ only was agreed upon at the Washington Congress. Previous documents show that the “Piece of seed” issue had been in the ISTA Rules since at least 1938 after adoption by the General Assembly at ZĂĽrich in 1937. Dr. Justice noted that both the Stronger and Quicker methods were included in the original rules adopted in 1931, and retained until 1953.

This is documented in Proc. ISTA Volume 10 (undated), Fourth Part (International Rules for Seed Testing), section II Purity (beginning on page 412) which provides a “Definition of pure seed according to the a) Stronger Method (S.M.)” and “b) Quicker Method (Q.M.).”

These original definitions read as follows:

  1. Stronger Method (S.M.)

    All seeds of the kind under consideration (in so far as it is possible to ascertain from their appearance alone) both fully developed and uninjured, as well as such injured or not fully developed seeds as may possibly develop normal sprouts, should be considered ‘pure seed.’ However, certain cases are defined below where, in the interest of uniformity and accuracy, it is desirable to deviate from this general rule.

These “certain cases” included Clovers, Grass seeds, Beet seed, and any insect-eaten seed… if the damage is confined to the endosperm, but if the radicle is injured the seed must be considered as “inert matter.”

It is important to note that even in 1938 the Rules spoke of injured (or not fully developed) seeds as pure seeds if they might yield a normal germinant.

As an addendum, “For grass seeds, those with injured germs so as to preclude any possibility of germination shall be considered as inert matter.” (There is no mention of forest tree seeds—because they came to the Rules much later.)

However, this appears to be a clear indication that at this time (1938) the S.M. purity test did consider the question of viability. But compare the Quicker Method as quoted below.

Continuing with the 1938 definitions, we read:

  1. Quicker Method (Q.M.)

    All seeds of the kind under consideration (in so far as it is possible to ascertain from their appearance alone) should be considered as ‘pure seed,’ regardless of whether they are shriveled, sprouted, cracked or otherwise injured, provided that in the case of broken seeds any fragment larger than one-half shall be considered as ‘pure seed’, while pieces that are one-half or less should be considered as ‘inert matter.’ In making the purity analysis the question of viability of the seed must not be considered.

(Note this last instruction regarding viability, a clear break from the S.M.)

Another contradiction regarding the purity analysis and viability, we find:

  1. Valid for both methods (S.M. and Q.M.)

    If a sample contains a great many severely injured, poorly developed or discoloured seeds, this fact should be reported on the international analysis certificate, and in such cases it is advisable to make a supplementary germination test in soil.

This special definition c) then speaks to Legumes, Grasses, mangel, beet and sugar-beet clusters, and goes on to allow that:

If the sample received for analysis contains seeds (intentionally or unintentionally) which closely resemble those seeds with which they are mixed, the separation of the pure seed and the admixed ingredient in the normal working sample may be slow and laborious, and in such cases the following quicker method is permitted.

What followed is an even “quicker method” using as few as 200 seeds to sort out the pure and admixed seeds, determine their percentages, and report them (admixed seeds) on the certificate.

As already mentioned, there was no mention of forest trees seeds in the original Rules. The actual date (possibly 1956) of their introduction to the Rules in uncertain as this review is written, but it must have been before 1961 when a Forest Tree Seed Number was published (ISTA 1961). Dr. Esbo mentioned the coniferae in his 1965 report. Tree seeds were introduced to the AOSA Rules in 1965. Unfortunately, for current review purposes, no access to ISTA (or AOSA) publications prior to 1965 is available.

These rather lengthy quotations have been included here to place the “Piece of seed…” issue firmly it in its historical context. Clearly, broken seeds have been a headache since day 1 (1938) of the ISTA (and probably 1908 AOSA) Rules, and especially in 1950. They continue to cause some problems even today, perhaps for different reasons. This review of the “Piece of Seed” issue is not the first; others have trodden the same path previously (see Ashton 2000), and the issue “transcends forest tree seeds,” to quote recent correspondence with the ISTA Purity Committee. As Dr. Justice (1965) indicated, the 1950 Washington Congress was aimed at cleaning up the confusion created by using both methods.

Dr. Esbo (1965) pointed out that the so-called Q.M. became the preferred one after 1950 but, as Dr. Justice noted, against European desires. While the “European desires” are not discussed further by either Esbo or Justice, it is obvious that the Europeans did not wish to adopt the Q.M.—and this must be perfectly clear—because the so-called Quicker Method simply buries, or hides, or disguises broken seeds in the pure seed fraction, so that they are no longer accounted for on the certificate as they were when the S.M. was used. Yet if seeds had been “admixed” with other seeds, this was not only to be determined, but reported on the certificate as called for in definition c) above for both methods.

Dr. Justice commented that when the Q.M. is used, it is left to the germination test to determine the “planting value” of the seeds. Since it is still the method used today, it is necessary to ask if the Q.M. is really doing its job of determining the purity value of the seeds for planting/sowing, or is it abrogating its responsibility to the germination test? There appears to be some contradiction here that deserves more attention.

Some Rhetorical Questions

For what purpose are seeds of any plants tested? Does not the above represent a major contradiction of terms for testing? Why conduct a purity test if it offers no (or very little) assistance in determining the value of the seeds for growing new plants? Is such a test not a waste of time?

It may be inferred that by 1950 broken seeds had become an even more significant issue than previously (1930s) for producers of crop seeds, and ISTA was persuaded (by the North Americans) to do something about it. There is a suggestion here that pressure of a non-scientific nature was brought to bear.

Most importantly, we must ask: Is this practice honest? In other words: is it not deceitful to hide broken seeds in the pure seed fraction? How happy is the buyer of ISTA-certified forest tree seeds to find that, despite a high purity percentage, the poor germination is due to bits of seeds that are useless?

In any case, there was another option in 1950 that would have been completely supported scientifically, would have satisfied the need for speed, and which could have been applied to forest tree seeds when they were introduced to the Rules. It is tempting to call this the Really Quicker Method (R.Q.M.) although it does involve a little bit of paper work. A definition (similar to those for the S.M. and the Q.M.) might read as follows:

The advantages of the R.Q.M. include the gain in speed because the analyst does not have to judge (even though this judgement may take only a few seconds) if pieces of seeds are larger or smaller than one-half the original size, so another analyst working on a sample from the same seedlot is much more likely to agree with the result. Because of this, Dr. Esbo’s claim that the Q.M. has really “diminished the influence of personal judgement,” is highly dubious. Again, see Ashton (2000).

Consider the following hypothetical example using the same working sample in each case:

Case 1: Using S.M.

Pure seeds

87%

Broken seeds (all sizes)

(These are noted on the certificate)

12%

Other crop seeds

0%

Weed seeds

0%

Inert matter

(Pieces of seeds of all sizes already accounted for)

1%

The subsequent Germination test showed 99%.
 

Case 2: Using Q.M.

Pure seeds

98%

Other seeds

0%

Weed seeds

0%

Inert matter

(Broken seeds are not mentioned on the certificate)

2%

The subsequent Germination test showed 90%.

So what are the differences between the S.M. and the Q.M. (and R.Q.M.)? Using the Q.M. (or R.Q.M.), we record:

  1. 12% increase in purity (over the S.M.).

  2. 1% increase in inert matter (due to pieces less than one half of the original).

  3. 9% decrease in germination.

The results of the Q.M. indicate that the lot appears to be very high in purity (99%) and not too poor in germination (90%). What seed purchasers do not know until the seeds are sown is that there is a substantial number of broken seeds—even though inert matter appears to be quite good at 2% because broken seeds are not recorded on the certificate—broken seeds do not germinate (perhaps they did not contain the germ), and which may clog up their modern seed sowing machinery. Alternatively, had the R.Q.M. option been used the same increase in purity, the same increase in inert matter, with identical germination results would have been obtained—at the expense of recording the fact that the sample contained broken seeds on the analysis certificate. So the seed purchasers would have been aware in advance that there were a number of broken seeds in this particular seedlot. But perhaps ISTA (and AOSA) did not (still does not) wish seed purchasers to know this.

As has already been explained, most modern forest nurseries use vacuum devices to sow (not plant) seeds in small containers (see Figs. 80 and 81). Having bits and pieces of broken seeds in the mix to be sown plays havoc with the production line, costs skyrocket because the line has to be shut down for cleaning. In a nursery producing millions of seedlings each growing season the problem can be immense.

How can undersized, shriveled or immature seeds, pieces of seeds more than one half their original size that do not contain the germ, fruits that may not contain a true seed, and diseased seeds be considered to be “pure seeds?”

See Esbo 1965

It would seem that perhaps some other pressure was brought to bear that was not scientific. That is, of a commercial nature. Whatever, it has probably never been documented.

The main emphasis placed by Dr. Esbo was on speed, hence the so-called “Quicker Method.” As has been seen, this is because in the Q.M. there are only 3 components (potentially 4 if weed seeds occur in the sample) to be sorted, whereas there were 5 in the S.M. Dr. Esbo claimed that:

No doubt Q.M. has diminished the influence of personal judgement and led to more uniform results and is really time-saving.

However, the proposed R.Q.M. would have been both quicker (than the S.M.) and would have completely resolved the “influence of personal judgement” (in the Q.M.). (See Ashton 2000).

Dr. Esbo noted (in passing as it were) that whereas the S.M. only permitted seeds that could possibly give rise to normal seedlings were to be classified as pure seeds, the Q.M. totally disregards this—in the interests of speed—leaving the evaluation of live seeds to the germination test. As already questioned, does this not mean that the purity test is abrogating its responsibility to the germination test, and is not determining the planting/sowing value of the seeds?

So why was speed so important?

A very slow weighing procedure.

An equipment bottleneck

In the early days of seed testing, some 80+ years ago, purity analysts were limited to weighing the submitted samples, the working samples and the components thereof, using an analytical beam balance (Fig. 82). For today’s analysts who may not be familiar with this equipment, a brief description is in order.

It was a very sensitive instrument yielding precise weights, but it had to be used in a draught-free environment. As shown here, it came with its own glass-sided case; the front panel would be slid upwards (Fig. 83) to provide access to the weighing pans, then lowered to make the weighing and to avoid ambient air movements. Before using the balance the operator had to make certain that it was precisely leveled on the work bench using the adjustable screw feet in each corner underneath the case.

When not in use, the beam was lowered by rotating a knob (visible in the center of the lower edge of the case) anti-clockwise until the beam made contact with two supports (near its ends) to take the load off, and therefore protect, the knife edge. For weighing, the knob was rotated clockwise to lift the beam off its supports. With the beam balancing nicely on the knife edge, the analyst then had to make certain that the pointer (just visible in front of the white scale between to the two pans) was reading zero. To access the balance within its case, the front panel had to be lifted (Fig. 83) so that the small screws (counterbalances) visible at the ends of the beam could be turned clockwise or anti-clockwise as required until the pointer registered zero. Done with the beam lowered.

All these preparations had to be made before weighing could begin. The sample would be placed in the left pan (usually), then the beam raised by rotating the front knob, and weights added to the right pan. Weights were kept in a small box designed specifically for their use (Fig. 84).

These weights, usually made of brass, had to be used with considerable care, and were not to be handled except by the use of forceps, especially the smaller weights that consisted of cut pieces of metal (brass or other non-corrosive material). Such a precaution was necessary because handling them by hand, without forceps, could easily affect their weight, especially the very small ones. All weights, even the larger ones seen in the rear row of the box, had to be handled very carefully.

Before a seed sample was weighed it would be placed in a small container, usually a metal dish that would fit onto the balance pan. The “tare weight” of this container had to be determined in advance, its weight recorded so that it could be subtracted from the seed sample weight. In most instances, the tare weight had already been inscribed on the pan by the manufacturer, making life a little easier. Using such a dish also made removing the samples from the balance pan relatively easy. Again, the dishes had to be handled with care, usually using the same forceps.

Raising the front panel of the case, the sample to be weighed (in its dish) would be placed on the left-hand pan of the balance. With experience, the analyst would be able to guesstimate which of the larger weights to add (to the right-hand pan) to begin the weighing procedure. This was done with the beam lowered. The front panel of the case would then be lowered, gently, and the beam raised off its anchors. If the pointer did not register centrally on the scale, the beam would then be lowered, the front panel raised again, and the amount of weight adjusted. This process was repeated until the weight(s) in the right-hand pan exactly balanced the sample as shown by the pointer registering zero on the scale. The operator then had to lower the beam onto its anchors, open the case again, and remove the weights from the pan and record their total. Although not shown in Fig. 84, weights to fractions of a milligram could be achieved by this process.

Having weighed the submitted sample, and recorded its weight, the analyst would divide it to obtain the working sample on which the purity analysis was to be made. The working sample for that species (Rule 3.5) was weighed using the same process, and its weight recorded. The purity test proper would begin by separating the components of the working sample—pure seeds, crop seeds, other seeds, damaged seeds (using S.M.) and inert matter—and then weighing and recording each in turn. All this had to be performed in a draught-free room because the beam balance was especially sensitive to air currents. Hence the sliding, closeable front panel. When completed, the beam would be lowered one last time, and the weights (if they had not been dropped on the floor!) returned to their allotted spaces in the box, the box closed and put away for future use.

It will be readily apparent that this was a slow, laborious process; the purity analysis of a single submitted sample might take 30 minutes or more, even for an experienced analyst, to complete. Anything that could be done to speed up the process was most welcome. This is where the Q.M. really came into its own. By weighing only three components instead of five, the time spent per working sample was reduced by 40%. This is what Dr. Esbo meant when he claimed that the Quicker Method was “really time-saving.”

However, since the early 1960s electronic balances have come to the fore (Fig. 85). Although these need to be protected from air currents also—they also have their own cases—the time required for weighing the components of a purity analysis has been reduced to a fraction of that using the beam balance. Instead of minutes per component, each can be weighed in seconds. More modern electronic balances now print out the sample weight at the push of a button. Some may even be connected directly to a computer. In either case, the analyst does not need to record the weights, a time-saving factor in its own right. With electronic balances being readily available, there is no scientific or operational reason for not weighing broken seeds of any size, and reporting them. Therefore, Rule 3.2.1.1.2 has become completely redundant for forest tree seeds, perhaps for all types of seeds.

Having gained the superior weighing speed provided by electronic balances, why does ISTA continue to insist on using the Q.M. and not the S.M.?

See Esbo 1965.

As has been discussed in PART I, there is no scientific basis for calling broken forest tree seeds of any size “pure seeds.” That Rule 3.2.1.1.2 still has to be applied is merely a throw back to the status of the purity test when forest tree seeds were introduced to the Rules, when extremely slow sample weighing was carried out using beam balances. Modern balances have cut the work time required for a purity analysis to a fraction of what it used to be, and there is no excuse for not recognizing this and changing the Rules accordingly, at least for forest tree seeds.

On page 34 of his 1965 paper Dr. Esbo wrote:

As far as crop plants are concerned, all seed-like structures which according to the definitions already given can’t be classified as pure seeds, are included in inert matter.

Then follows a list of “such structures” including “seeds of coniferae with the seed coat entirely removed.” So forest tree seeds, at least coniferous seeds, had come up over the horizon, indicating that they had been included in the Rules prior to writing his report.

Further on, Dr. Esbo states that:

Common to all these categories of seed-like structures which are classified as inert matter is the fact that they are deemed as not to be able to give rise to a seedling.

Note that they are merely “deemed” to be useless for producing new plants, not that scientific knowledge says that they are useless.

Here is the purity test being used to assess growth potential, something that the Rules insist must not be taken into account—the purity test is not a growth test.

Again, in a subsequent paragraph when writing about weed seeds, Dr. Esbo states that seed-like structures which:

…by visual examination based on long experience can be evaluated with certainty as having no chance of giving rise to a plant should be deemed as inert matter.

Why is he writing this? The purity test is not a growth test. But Dr. Esbo seems to ignore this. But seed like structures… evaluated with certainty as having no chance of giving rise to a plant… certainly includes forest tree seeds of all sizes and all kinds.

Therefore, it must be said that Dr. Esbo lacked the scientific knowledge, and experience, that the current Rules insist upon (Rules, Introduction, page 1), scientific knowledge and experience that has firmly established that pieces of forest tree seeds of any size should be classified as inert (if not dead) matter, or simply as Impurities. This is not a growth test: it is scientific knowledge being applied.

Again, here in the Q.M. Dr. Esbo states that the analyst must make an evaluation of no chance of giving rise to a plant, that is, a viability judgement. While such “evaluation” or judgement is not a test per se, it does contradict the notion that the purity test is not a viability test.

The fact that the purity analyst is expected to distinguish between seed-like structures that are not able to give rise to a seedling, and those that may give rise to a seedling, Dr. Esbo’s claim that the “Q.M has diminished the influence of personal judgement” appears to be highly suspicious, contradictory, false and based entirely on innuendo.

As was illustrated in detail for Larix, in Part I, one analyst may classify a piece of seed measuring 51–55% of the original size, that is, more than one-half the original size based on a guess at the original size (see Fig. 8), (with a portion of the testa still attached, of course) as pure seed, while the next analyst, possibly in the same laboratory, may regard it as inert matter. Ashton (2000) reported that at a Workshop in Hungary in 1997, most participants estimated several fragments of Avena fatua caryopses as being greater than one half, but these fragments had been deliberately made to be exactly one half of the original seeds, and therefore should have been classified as inert matter. If this happens when analyzing the purity of wild oat seeds, not just at a Workshop, but also in the laboratory (again Ashton 2000), how inconsistent can matters become when analyzing the purity of forest tree seeds that may vary in size by more than 250% (see Fig. 8)?

Yet Dr. Esbo claimed that the Q.M. removes such inconsistencies.

These are just some of many contradictions.

But there are more. As quoted earlier, the so-called Q.M. includes “all questionable, damaged or badly developed seeds” (more than one-half their original size) in the pure seed fraction.

Expressed in other terms, the analyst must include any damaged seeds with the pure seeds so that they become hidden. Because nothing is written on the certificate (Rule 3.7) regarding broken seeds the seed buyer is hood-winked into thinking he/she is purchasing a crop with a high degree of purity. As shown in the numerical example earlier, the germination test will show this is not a top quality seedlot.

But is this not a highly questionable practice, and basically dishonest?

Besides, if ISTA was really interested in speeding up the purity test (as Dr. Esbo claimed) why not put all questionable, damaged or badly developed seeds of any size into the inert matter (or, preferably, Dead matter = Impurities). This would lower the resulting purity of the seeds, say from 98% to 90%, but the germination test result would be higher. Not only would this really remove any judgement on the part of the analyst (a serious concern of Dr. Esbo), but it would make the certificate more honest.

Elsewhere in his 1965 report, Dr. Esbo wrote that:

A reliable evaluation of the purity of seed as well as other properties can however be done only by people having a special education and a certain knowledge in botany and agriculture and by persons who are quite impartial and totally independent of consequences of the evaluation.

Here the operative words are people having a special education and a certain knowledge in botany and agriculture. Since tree seeds had begun to be introduced to the Rules when he wrote this, should not and forestry (and horticulture) have been added? Dr. Esbo had to be aware of some of these additions to the Rules because he makes reference to the coniferae. But it seems clear that, in 1965, this special education and knowledge in forestry was completely lacking in agricultural officiandos who were then controlling the Rules, and it appears to have been lacking over the past 70+ years (50 years since Dr. Esbo wrote his piece).

As mentioned earlier, by 1950 (the Washington Congress) broken seeds clearly had become a significant issue, especially for American and Canadian seed producers and ISTA was persuaded to do something about it. That is, some 63 years ago ISTA opted for speed over accuracy, precision and honesty in the purity test.

Was there a hidden agenda? Was there a commercial agenda? There certainly was no scientific agenda where forest tree seeds were concerned.

One More Look Back

Exactly how was the “Stronger Method” of the purity test applied before the change in 1950? According to the version published in ISTA Proceedings Vol. 10, number 1, (undated), originally there were five groupings, i.e., “pure seed,” “crop seeds,” “weed seeds,” and “inert matter,” with broken seeds being set aside, weighed separately and reported on the certificate. In the “Quicker Method” there were only three groupings, i.e., “pure seeds,” “other seeds,” and “inert matter.” The original grouping “weed seeds” was deleted, “crop seeds” became “other seeds,” “pieces of seed larger than one-half their original size” were included (hidden) within “pure seeds,” while pieces smaller than one-half were included within the “inert matter.” Neither larger nor smaller pieces of seeds were accounted for on the certificate. By not doing so this saved a few minutes.

Basically, in the Q.M. anything and everything that was not classified as “other seeds” or “inert matter” was (and still is) to be considered as “pure seeds. Exceptions were to be made for certain cases of special types of damage to “clovers, grass seeds, beet seed clusters,” and even “insect-eaten seed”… provided said “damage is confined to the endosperm.” Seeds with damaged endosperm were to be included as “inert matter.”

In the very early years when both methods were employed, the main difference between the S.M. and the Q.M. appears to be that the S.M. paid no regard to the size of any broken seeds and all pieces were placed in a separate pile, and – most importantly – the weight of broken seeds was recorded on the certificate. In contrast, the Q.M. all along has required that broken seeds more than one-half their original size should be considered as “pure seeds” (Rule 3.2.1.1.2). This is a major difference, one that needs to be questioned now that the “requirement for speed” riddle has been exposed—and solved.

Additional attention needs to be drawn to the 1965 words of Dr. Esbo (President of ISTA 1965–1968) especially regarding item (c) above:

In order to achieve a high degree of uniformity in seed testing, which is one of the major aims of ISTA, it is essential that clear definitions of the different words and terms be settled and made available to all parties involved.

So why “seed” and “seeds”?

One of the aims of this review has been to make different words and terms clear. In English “seed” may be used as a collective singular, thus meaning many seeds. In doing so, the verb particles around it often become confused. For example, “this seed is to be sown next week” (only one seed?) or “these seeds are to be sown next week.” The use of “seeds” when referring to more than one seed is highly preferable, because the writer’s syntax becomes more consistently correct. Other languages may have similar, or lesser, difficulties. This may appear to be a very picky point, but if the Rules are to be precise, as Dr Esbo believed, the wording needs to be studied carefully.

All of this reflects the seed processing technology that existed 60+ years ago. It is assumed, perhaps incorrectly, that even for crop seeds major advances have been made and broken seeds are much less of a problem, or have been completely eliminated. If the latter is true then Rule 3.2.1.1.2 must be amended, if not extinguished.

From the reviewer’s personal experience, improvements in processing of coniferous tree seeds have advanced greatly during a 28-year career as a seed biologist. There were reports from a few seed producers that damaged tree seeds could still be found occasionally, but their frequency has greatly diminished. During the reviewer’s decade-plus tenure as manager of the ISTA-accredited laboratory CAN07, no broken seeds were found from local seed producers who sold their products internationally. Thus, the “Piece of seed larger than one-half the original size, with or without integument, provided a portion of the testa is attached” (Rule 3.2.1.1.2) issue basically became a non-issue, despite what the Purity Committee expressed in 2006 in an e-mail relating to proposed PSD revisions for a number of gymnosperm species (see below).

A Step Further

In reality, the issues regarding forest tree seeds and the Purity Test go beyond “pieces of seeds.” In PSDs 10, 12, 52, 53, 54, 55, 56, 57, 58 and 60, even intact seeds that are devoid of (“without,” or “entirely removed”) testa are regarded as pure seeds. In PSDs 11, 47, 49, and 50 intact seeds with only a portion of the testa attached are regarded as pure seeds. As was noted earlier, no definition of “a portion” is provided. Perhaps a “portion larger than one-half of the original?”

In PSD 51 “portion of the testa” becomes “part of the testa,” another example of inconsistency in wording.

What published scientific evidence is there that allows naked, or near naked tree seeds, to be regarded as “pure,” that is, good/sound seeds capable of producing new plants, especially if such seeds have been in dry, cold storage for any length of time, or have been infected with fungal or bacterial organisms? If there is such scientific evidence, it should be brought to everyone’s attention, especially that of this reviewer. However, this is grist for another mill.

Two Final Questions

What is the purpose of this long-winded review?

That the Rules continue to perpetuate the unscientific myth (or legend according to Dr. Justice, 1965) of Rule 3.2.1.1.2, was brought to the reviewer’s attention in 2006 when the Chair of the Purity Committee insisted on its inclusion in a proposal submitted to revise PSD 51 (Abies, Cedrus, Larix, Pseudotsuga and Tsuga). Whereas this demonstrated the Purity Chair’s attention to detail and correctness in drafting PSDs (new or otherwise), it served as a wake-up call regarding the ambiguity and scientifically bankrupt nature of Rule 3.2.1.1.2 as it is applied to forest tree seeds.

Is it not time for ISTA to move the purity test into the 21st century?

All forest tree seed analysts have the botanical and forestry education, as well as expertise and experience, so collectively they should be able to make a strong case for revision of the ISTA Rules as they are applied to forest tree seeds. Failure to do so in effect means that ISTA will be satisfied with the early 20th century status quo, that is, the Dark Ages for purity testing.

Based on the foregoing it is strongly suggested that the following statement be included in the Introduction to the Rules:

“Seed quality is a concept made up of different attributes. Seed is a living biological product and the methods used for testing must be based on scientific knowledge of the seed in question, as well as on the accumulated experience of seed analysts. The information gained from these tests must be aimed at providing information to the producer, the processor, the warehouseman, the merchant, the farmer, the horticulturalist, the forester, the certification authority and to the government or agency responsible for seed control, to determine if the seeds are of high enough quality to produce an abundant crop of the species under examination.