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

Part I

The Structure of Forest Tree Seeds

Except for Sequoiadendron, for which there are no illustrations, the remaining 16 of the gymnospermous genera included in the Rules, viz. Abies, Calocedrus (formerly Libocedrus), Cedrus, Chamaecyparis, Cryptomeria, Cupressus, Juniperus, Larix, Picea, Pinus, Pseudotsuga, Sequoia, Taxodium, Taxus, Thuja, and Tsuga will be discussed first, followed by 25 angiospermous genera, viz. Acacia, Acer, Aesculus, Ailanthus, Alnus, Betula, Carpinus, Castanea, Cornus, Corylus, Eucalyptus, Fagus, Fraxinus, Ginkgo, Gleditsia, Ilex, Liriodendron, Platanus, Populus, Quercus, Robinia, Salix, Sophora, Tilia, and Ulmus.

A tropical forest tree seed genus (Tectona–teak) will be included also.

The Illustrations

Except for two photographs (Figures 1 and 2) in Part I, all seed structures will be illustrated by the excellent line drawings published in “Seeds of Woody Plants in the United States” (Schopmeyer 1974), used with permission. Many of these drawings were published in the first “Woody-Plant Seed Manual,” USDA Misc. Publ. 654 (Anon. 1948), and have been reused—and added to—in the recent “The Woody Plant Seed Manual” (Bonner et al. 2008) for which this reviewer was privileged to revise the Abies (fir) chapter. The early original drawings (Anon. 1948) were made by several artists, including Leta Hughey, W.H. Lindemann, A. Kuban, N.T. Mirov, and Helen M. Dille. The added genera in Schopmeyer (1974) were drawn mainly by Suzanne Foster Manley. Such drawings were also used in Young and Young (1992), which was published as a revision of Schopmeyer (1974).

For purposes of this review, the line drawings have been modified, using Photoshop, by adding check marks to the vertical scale to show proportional seed lengths of 25%, 50%, and 75% of the overall size from either the chalazal (cotyledonary) or micropylar end. To indicate what happens when seeds are broken, portions of the line drawings have been erased and a broken (dashed) line added to indicate the position of each inferred break at the relative positions along the length of the seed. In all other aspects the drawings remain unaltered. Part II will include a few additional photographs taken by the reviewer, plus others provided by the Royal BC (British Columbia) Museum.

A. Gymnospermous species

As is demonstrated in all the illustrations (Figs. 3–28 in the text and Appendix I), with the exception of Taxus (yew) (Figs. 23–28), all gymnospermous seeds possess the same basic internal structures: an almost linear embryo (2n-diploid) that occupies, at seed maturity, 75% or more of the length of the cavity within the female megagametophyte (1n-haploid) tissue, surrounded completely by an intact seedcoat/testa (1n). Thus, (except for Taxus) the mature embryo is almost the full external-length of the seed.

I. Cone-bearing genera

Pinaceae

Internal structures are shown to full advantage in Figure 1, a photograph of the sagittal half of a mature Douglas-fir (Pseudotsuga menziesii) seed. The embryonic cotyledons can be seen; at their base is the apical meristem from which the new shoot will form. The root meristem, from which the seedling radicle develops, is surrounded by its protective root cap of parenchymatous tissue.

Figure 2 is a similar photograph, again of a Douglas-fir seed, but with an immature, developing embryo, although the megagametophyte tissue is fully developed. It can be seen clearly that the cavity within the megagametophyte tissue is already fully formed, that is, the embryo does not have to “force” its way through this tissue as it elongates. The megagametophyte tissue contains the energy for the growing embryo; this energy is absorbed by the embryo via the cotyledons (Edwards 1969).

Because they possess the same basic internal structures, what happens to these structures when the seeds are damaged will be discussed in full detail for two genera, viz., Larix (Pinaceae) and Thuja (Cupressaceae). These two genera have been selected because the available illustrations provide not only an internal impression, as if by x-ray, but also an exterior, adaxial view of the complete seed, which represents the view that the purity analyst would have.

The internal morphology of Taxus (yew) is different and a separate discussion will be required. All other Gymnospermous species are illustrated in Appendix I.

Larix (larch)

(or, as Monty Python’s Flying Circus described it, “Number 13: the Larch”!)

Although the larch seed illustrated in Figures 3–11 is a North American species (Larix laricina), analysts familiar with this genus will know that all larch seeds are easily recognizable as such. A small seed at 3.5 mm in length, L. laricina retains a remnant of the seed wing that, as in Douglas-fir (Pseudotsuga), is fused to the seed coat (PSD51) (Fig. 3).

This means that when these seeds are processed after extraction from the cones in which they developed the wing has to be broken off, a likely source of seed damage. As already mentioned, on the vertical axis to the left in all the figures are check marks showing 25%, 50% and 75% of the seed length, from either end of the seed. Thus, when a piece of a broken seed is considered (as in Fig. 4), the size of the piece can be readily assessed.

The piece of seed shown in Fig. 4 is approximately 55% of the original length as measured from the chalazal (cotyledon) end of the seed. Thus, it is larger than one-half the original size, so an analyst must classify it as a “pure seed.” The view of the internal structures (on the left) shows it has lost its embryonic root meristem from which the radicle would develop. It should be clear that this piece of seed more than one-half the original size, with a full compliment of seed coat (or testa) attached, represents the same situation—separated cotyledons—for which seeds of the Fabaceae (Leguminosae) would be classified as “inert matter.” But only seeds of the Fabaceae (Leguminosae), not seeds of Cupressaceae, Pinaceae, Taxaceae or Taxodiaceae. Why? Something appears to be amiss here.

Even if this piece of seed was sown, or included in a germination test, it would be impossible to form a primary root. Because all gymnospermous seeds exhibit epigeal germination, this seed has zero possibility of producing a new plant. Yet, even though it is easily recognizable to the purity analyst for what it is, as is clear in the drawing of the external (adaxial) view of the seed, it must be classified as a “pure seed” according to Rule 3.2.1.1.2. If a tetrazolium test (Rule 6) was performed on such a seed, how would the staining pattern (if staining occurred) be interpreted for an incomplete embryo?

In similar fashion, Fig. 5 shows a piece of a larch seed approximately 55% of the original as measured from the micropylar end. That is, it is larger than one-half the original size so the analyst must classify it as a “pure seed.” The internal view on the left shows that it lacks the embryonic cotyledons and the apical meristem. Were it to be sown, or included in a germination test, the remnant of the embryo probably would swell and the broken surface would extrude beyond the megagametophyte surface. Because there is no chalazal end to the seed the elongating embryo remnant would meet no resistance and would not emerge via the micropyle (epigeal germination).

Even if the radicle were to penetrate the micropyle, the lack of cotyledons means that this seed is doomed. The cotyledons are the organs that absorb the energy (mainly sugars) for growth from the megagametophyte (Edwards 1969), and their absence means the seed will be unable to grow. Yet it must be classified as a pure seed according to Rule 3.2.1.1.2. If a tetrazolium test (Rule 6) was performed on such a piece of seed, how would the staining pattern—if any staining occurred—be interpreted for an incomplete embryo?

Size Matters I

A critical comparison must be made between the larger piece (more than one-half the original size, so “pure seed”) that is shown in Fig. 4 with the smaller piece (less than one half of the original, so “inert matter”) that was broken off when the piece of seed shown in Fig. 5 was formed. These two pieces are shown side by side in Fig. 6.

Likewise, Fig. 7 shows side by side the larger piece (more than one-half the original size, so “pure seed”) shown earlier in Fig. 5 with the smaller piece (less than one half of the original, so “inert matter”) that was broken off when the piece of seed shown in Fig. 4 was formed.

What is the difference between these pairs of pieces of seeds? The only difference is an arbitrary, small variation in size. In each case, one piece is 55% of the original size, and must be classified as a “pure seed,” while the other is 45% of the original size and is to be considered “inert matter.” Yet neither have any potential for producing a new plant.

What scientific principle is being invoked to distinguish them from one another? In other words, why are both pieces (larger and smaller) not considered being “pure seed” or, conversely, not “inert matter?”

No scientific principle, merely an arbitrary difference in size.

From the discussion presented so far, it should be clear that no scientific principle is at work. That the distinction is pure unadulterated rubbish.

Size Matters II

Rule 3.2.1.1.2 assumes that the analyst knows the original size of the seed. This, in itself, appears to be a contradiction, since the seed has been broken and a piece is missing. As already noted, this Rule was introduced well before forest tree seeds were included in the Rules, being designed primarily with crop seeds in mind. Because they are highly bred, crop seeds are likely to be quite uniform in all aspects, including size. Thus, it is probably quite easy for the crop seed purity analyst to judge that this piece of seed is larger or smaller than one-half of the original size, because all seeds in that working sample look alike. However, see Ashton (2000).

For the vast majority of tree seeds, even those produced in forest tree seed orchards, seed size is far from uniform. Experience with noble fir (Abies procera in North America, a.k.a. A. nobilis in Europe) showed that seed size varies from mother tree to mother tree, from one aspect of the crown to another (in the northern hemisphere, seeds developing on the south aspect of the crown tend to be larger), from cone to cone on the same branch, within a single cone (seeds near the cone extremities are usually much smaller than those in the cone’s centre), and even between the two seeds borne on a single ovuliferous scale (Edwards 1969). In Sitka spruce (Picea sitchensis) seed size is known to affect seedling attributes (Chaisurisri et al. 1994).

As all tree seed producers/processors are aware, this variation in seed size must be considered when the seedlot is cleaned, especially if an aspirator cleaning system (air is drawn through the seed mass) is used to lift lighter particles such as needles, cone scale fragment, empty seeds, etc. from the heavier ones. This is a modern, mechanized method of winnowing. For such a system to be efficient, that is, to avoid small filled seeds from being removed along with large empty seeds, the seedlot first must be sized.

For gymnospermous seeds this is usually done by passing the seedlot (after wing removal) over a series of vibrating, inclined screens. Three, sometimes four, different mesh sizes are used; the largest seeds are retained on the upper screen, the medium-sized seeds retained on the screen below, and the smallest seeds falling through to the lowest screen through which any dust or other very fine debris falls onto a conveyor belt for disposal. Each size class is then cleaned separately in the aspirator cleaner. Air speed is adjusted so that light, empty seeds for that size component can be drawn off for disposal leaving the heavier filled seeds. Seeds containing an insect larva may also be removed at this time. If there were any broken seeds to begin with most, if not all, may have been removed as debris. But there is no guarantee that the seedlot is free of all pieces of seeds.

When each size component has been satisfactorily cleaned, all the filled seeds from each size component are then thoroughly remixed. This is a vital step in retaining the genetic integrity of the seedlot. However, because such sizing is an integral part of tree seed processing, application of Rule 3.2.1.1.2 to tree seeds makes no sense whatsoever.

While other methods of cleaning may be used, it must be recognized that seed size varies within any and all seedlots. The degree of variation differs among species, and from seedlot to seedlot within a species, but the purity analyst must apply the Rules, particularly Rule 3.2.1.1.2 uniformly. For a working sample from a seedlot with even a modest degree of variation in seed size, it is difficult if not impossible for the analyst to correctly judge whether a “piece of seed” is “larger than one half the original size” because the “original” seed is not available to make a correct decision. There are other reasons, already alluded to, for not classifying “pieces of seeds” as pure seed and these will be discussed more fully below.

As an example of actual variation in seed size, or more correctly seed mass (compare with Ashton 2000), Fig. 8 is included here courtesy of Dr. F. Gorian (Italy). In his presentation at the 2006 ISTA Forest Tree and Shrub Seed Committee (FTSSC) Seminar in Verona, Dr. Gorian showed that, across its natural range, seed mass in Larix decidua (European larch) may vary by up to 250%. Suppose the purity analyst is given a working sample from a range-wide collection of European larch (Mother Nature forbid that seeds from across the range should be mixed into one seedlot. This is highly unlikely given today’s knowledge concerning matching seed collection with planting site). How would the analyst know if a piece of seed similar to that shown in Fig. 4 or Fig. 5 is a small piece (smaller than one-half the original size) from a large seed (P4 in Fig. 8), and therefore “inert matter,” or a large piece (larger than one-half the original size from a small seed (P1 in Fig. 8), and therefore a “pure seed?” As will be discussed in Part II, the analyst is not supposed to stop and consider such matters. And if one analyst does so, the next analyst might not agree (again, see Ashton 2000). This is but one example to illustrate how nonsensical Rule 3.2.1.1.2 is for forest tree seeds.

Critics might regard a piece of seed that is 55% of the length of the original is borderline for Rule 3.2.1.1.2. (but see Ashton 2000) and, even though it meets the “larger than one-half the original size” criterion, larger pieces should be considered. So what happens to the internal structures if the larch seed shown in Fig. 3 was broken so that a piece close to 75% of the original size measured from the chalazal end is encountered? This is illustrated in Fig. 9 showing a piece of seed that contains almost a complete embryo, only the tip of the embryonic radicle, perhaps only part of the root cap is missing; typically, the root meristem does not extend to the tip of the root cap. So if sown, or included in a germination test, a germinant might be seen, even a normal-looking one. However, the megagametophyte tissue has been exposed so unless this seed is used—sown or germinated—fairly soon it will perish in storage. More on tissue exposure later.

Likewise, Fig. 10 shows a piece of seed measuring 75% of the original size from the micropylar end; this piece also contains an almost complete embryo, only the tips of the cotyledons have been lost. However, the megagametophyte tissue at the chalazal end is exposed, so unless this seed is used—sown or germinated—fairly soon it will perish in storage. Even if it does germinate the embryo will elongate via the chalazal end, not via the micropyle, because there is no resistance to encourage the radicle to emerge via the micropyle (epigeal germination). This means that the cotyledons will lose contact with the megagametophyte tissue, the embryo’s source of energy, and it will cease to grow. Its usefulness for producing a new plant is completely compromised therefore.

What if the seed is merely chipped, at either end, so that the piece measures 90% or more of the original? Fig. 11 shows just such a seed with just enough of the seedcoat missing to expose the megagametophyte tissue at the micropylar end. To the uninitiated this might appear to be a piece of seed that could give rise to a new plant, so it is worthy of the connotation “pure seed.” However, since the megagametophyte tissue is exposed, this piece of seed has little to no chance of producing a new plant should it undergo dry, cold storage.

Fig. 12 illustrates a piece of seed with just enough of the seedcoat missing to expose the megagametophyte tissue at the chalazal/cotyledon end. If used promptly after it was discovered, it might germinate, with the embryo exiting via the chalazal end. But the cotyledons will soon lose contact with the megagametophyte tissue after which growth will cease. In a germination test this would produce an abnormal germinant—cotyledons exiting first. Because of this, such a piece of seed has no usefulness in producing a new plant. If placed in storage, this piece of seed will perish.

Why must either of these pieces be classified as “pure seed?”

Because ISTA is not applying the scientific principles involved.

The Rules state that even if a larch seed is broken at 75%, or even 90%, of its original size it must be recognized as “pure seed.” This is because the Rules are not conversant with tree seed biology.

Incidentally, chipping Abies procera seeds at the micropylar end and then soaking them for 24 hours prior to sowing in a nursery flat has the same effect as 3–4 weeks of pre-chilling (Edwards 1969).

While seed breakage is more likely across the length of any gymnospermous seed, other damage may occur. For example, the seed may be broken along its length, as shown in Fig. 13. In this example the embryo remains completely intact, but the megagametophyte tissue is exposed along the damaged edge. Should this seed be used—sown or germinated—relatively soon after it is discovered, a normal germinant may be forthcoming, although it is possible that the elongating embryo will bend and force its way sideways from the surrounding tissues, with the radicle not exiting via the micropyle. Its usefulness for producing a new plant is highly questionable therefore. If stored, as in other examples above, exposure of the megagametophyte tissue will cause this piece of seed to perish.

Sometimes the seed coat may be merely cracked, as shown in Fig. 14. Although having the appearance of an otherwise intact seed to the purity analyst, exposure of the megagametophyte tissue means that this seed, essentially 100% of the original, is likely to perish if placed in cold, dry storage.

All of the issues presented above apply equally well to seeds of all other members of the Pinaceae (Abies, Cedrus, Picea, Pinus, Pseudotsuga and Tsuga), illustrations for which are found in Appendix I.

One more seed type, a member of the Cupressaceae, (Thuja occidentalis, northern white cedar), will be discussed in detail for a few additional reasons. As with the larch seed, the line drawings available for this species offer both an external adaxial view of an intact seed and the x-ray-type of view showing the internal structures. Some additional notes will be included for other genera.

Cupressaceae

While seeds of all genera in the Cupressaceae presented here are similar to the Pinaceae in their internal structures, in that their embryos are linear and occupy the bulk of the seed length, Calocedrus, Chamaecyparis and Thuja add a new twist to the discussion, viz., seeds of these three genera bear two wings that are not removed in routine processing because doing so is known to cause damage. For this reason damaged seeds may be less frequent. However, they need to be discussed in some detail in regard to Rule 3.2.1.1.2.

Thuja (arborvitae)

Similar to the illustrations for Larix, line drawings are available (Schopmeyer 1974) for Thuja occidentalis that offer both an external adaxial view that the analyst would have of an intact seed, plus the x-ray-type of view showing the internal structures. Illustrations for pieces of seed 51–55% of the original size, as well as for larger pieces, 75% and 90% are shown.

For Thuja occidentalis, northern white-cedar, the intact seed as viewed from its adaxial surface is shown at left in Fig. 15, with the “x-ray” view at right. Because the seed coat is fairly soft and easily damaged, the two wings are not routinely removed in seed processing. By now it should be clear what occurs if seeds of this species are broken, either into pieces that are more than one-half their original size, and to be regarded as “pure seeds” (Rule 3.2.1.1.2), whether 51–55% of the original from the chalazal end (Fig. 16) or from the micropylar end (Fig. 17). It should be obvious also what happens to the internal structures if the Thuja seed is broken at approximately 75% (Figs. 18 and 19) or even 90% (Figs. 20 and 21), from either end. A Thuja seed broken lengthwise down one side is shown in Fig. 22. Once again, because of the need to store surplus seeds the lifespan of broken specimens can be expected to be very short.

The following are some brief additional notes on Calocedrus, Chamaecyparis, Cupressus, Ginkgo and Juniperus. Illustrations for these genera can be found in Appendix I.

Calocedrus (incense cedar)

Although appearing to have only one wing, each Calocedrus decurrens seed actually has two: a longer, wide wing that extends lengthwise beyond the seed on one side, and a narrower, much shorter one just emerging alongside the first from the opposite side (Appendix I). Both wings are persistent and project past the narrow micropylar (radicle) end of the seed rather than the chalazal (cotyledon) end as in most other conifers. During germination the radicle emerges from the narrow, winged end of the seed. How the internal morphology of an intact seed is impacted if it is broken into a piece 55% or more of the original size (as measured from either end) is also shown in Appendix I. The same concerns apply as for the larch and arborvitae seeds even if the seeds are broken into larger pieces.

Chamaecyparis (white cedar)

Like those of Thuja, seeds of Chamaecyparis lawsoniana, Port-Orford cedar, bear two wings, one on either side of the seed as seen from an adaxial view (Appendix I). Again, these are not routinely removed in seed processing because the seed coat is fairly soft and the seed easily damaged. Broken seeds may be infrequent, therefore, but should damage occur the same concerns as for Larix and Thuja apply.

Cupressus (cypress)

Cypress seeds vary widely in shape and size (2–8mm long), those of Cupressus arizonica var. arizonica shown in Appendix I being among the smallest. In general, they bear either a minute wing (a tegumentary extension of the seed coat) or none at all, making dewinging unnecessary. For these reasons cypress seeds are unlikely to be subject to the same damage as those of the Pinaceae. However, their embryos are linear and occupy the bulk of the seed length, so if damage does occur why should pieces of Arizona cypress seeds of any size be classed as pure seeds?

Juniperus (juniper)

The fleshy cones of Juniperus scopulorum, usually referred to as berries, produce only 1–2 wingless seeds each, and breakage is usually minimal during seed processing. Should it occur, the linear embryo occupying the majority of the seed length (Appendix I) will suffer the same fate as those of the other seeds discussed so far.

Taxodiaceae

Cryptomeria (cryptomeria, sugi)

A monotypic genus, the embryo of a Cryptomeria japonica seed is linear and occupies the bulk of the seed length (Appendix I), so it is almost identical to seeds of the Pinaceae. If a sugi seed is broken, the same reservations as all the previous seeds must be applied.

Sequoia (redwood) and Sequoiadendron (giant sequoia)

Although seeds of Sequoiadendron giganteum (no drawing available) tend to be larger (3–6 mm) than those of Sequoia sempervirens (Appendix I) both contain linear embryos that occupy the bulk of the seed length. When mature, the brown wing (a tegumentary extension of the darker seed coat) of S. sempervirens is about equal in width to the seed. Even though redwood and giant sequoia seeds are not normally dewinged, should damage occur the same reservations must be applied to seeds of both genera, whether they are broken from either the chalazal or micropylar ends, and no matter what size the pieces of seeds represent.

Taxodium (bold cypress)

An important timber species in the United States, but introduced into Europe in the mid-17th century both as an ornamental and for wildlife values, Taxodium distichum seeds are larger than all the seeds discussed so far (Appendix I). They are irregularly shaped with thick, horny, warty coats and projecting flanges. Just as in other gymnospermous seeds, baldcypress embryos are almost linear—slightly curved—and occupy the bulk of the seed length. If such seeds are broken either from the chalazal or micropylar ends the same reservations as with all the previous seeds must be applied.

II. Non-cone-bearing genera

The discussion so far has been on tree seeds that develop in, and are extracted from, cones. Special consideration is now given to two non-cone bearing gymnosperms, the yew and ginkgo (or maiden-hair tree) the latter being linked botanically to vascular cryptograms.

Taxaceae

Taxus (yew)

Although a major tree species, if only as an ornamental, yew trees do not produce cones, even though they are generally regarded as a member of the Coniferae (Dallimore and Jackson 1954). Taxus seeds form singly in scarlet, fleshy arils (outgrowths of the ovuliferous scales) that are similar in general appearance to berries, but they are not berries in the strict botanical sense. A line drawing of Taxus brevifolia, the Pacific West Coast yew is used for illustration purposes (Fig. 23). It is very similar to the European yew, Taxus baccata, except that its needles are shorter (hence the specific name brevifolia). Both species grow very well in graveyards when their roots reach decaying bodies.

The main difference in the internal morphology of a yew seed is that the embryo is proportionately much smaller that that of other species discussed so far. It is linear, but occupies no more than 50–60% of the seed length (Fig. 23). Thus even though the seeds are very hard (and difficult to germinate), and there is no wing to remove, should a seed be broken into a piece approximately 51–55% as measured from the chalazal end (Fig. 24) it will no longer have an intact radicle; as in other gymnospermous genera, yew also exhibits epigeal germination. Thus, such a piece of seed slightly larger than one-half the original size, while classified as a pure seed according to Rule 3.2.1.1.2 has zero chance of producing a new plant.

Conversely, if the yew seed is broken at 51–55% from the micropylar end (Fig. 25) there is a possibility that the embryo might remain relatively undamaged—only the tips of the cotyledons will have been lost—and germination might occur. As foresters working with yew seeds are well aware, germination of Taxus seeds can be very difficult to induce, requiring many months, up to a year or more, of moist pre-chilling. It is doubtful that broken seeds would survive such prolonged stimulation. If a yew seed is broken at around 75% of its original length (Fig. 26), and even more so at 90% (not illustrated), as measured from the chalazal end it will no longer have an intact radicle. Should the seed become broken at 75% (Fig. 27) or 90% (not illustrated) of its original length as measured from the micropylar end, the embryo may remain intact. If broken lengthwise the embryo may remain intact (Fig. 28), but the exposed megagametophyte and embryo may not withstand extended pre-chilling, and the seed will likely perish. Yew seeds, like those of other trees, must be collected when there is a sizeable crop, and then placed in dry, cold storage until they are prepared (pre-chilled) for sowing. Even if they do survive in storage, what chance does a broken yew seed have of producing a new plant? Especially if it does not contain the germ.

Ginkgoaceae

Ginkgo (maidenhair-tree)

Differing from the Coniferales or Coniferae, especially in that its method of fertilization is effected by motile male sperms, Ginkgo is linked to the Vascular Cryptograms (ferns and their allies) (Dallimore and Jackson, 1954). Other workers regard it as a deciduous gymnosperm (Shepperd 2008). Single naked ovules ripen into large, drupe-like seeds with fleshy, smelly outer layers, and a thin, smooth, cream-coloured, horny inner layer. Embryo ripening continues in a large percentage of the “fruits” over winter, after they have fallen to the ground.

Internally, these fruits comprise a linear embryo (Appendix I) that occupies only 25–30% of the length of the seed, situated at the opposite end from the attachment scar, and surrounded by endosperm (2n). Like the yew, if a ginkgo seed is broken around 60% of its original length from the scar end, the entire embryo will be lost in the smaller piece (smaller than one-half the original length). The piece larger than one-half its original size has zero potential for producing a new plant because the embryo is not present, but Rule 3.2.1.1.2 says it must be regarded as a “pure seed.” If the seed is broken at around 60% from the opposite end, so that the piece contains a complete embryo, it is unlikely to produce a new plant not only because the exposed surface of the endosperm will cause the piece of seed to perish in storage, but also because the elongating embryo (if that occurs) will push the cotyledons out of contact with the endosperm and growth will cease. Normal germination is epigeal, but since there is no resistance from the chalazal end, the radicle will not emerge via the micropyle. Thus, this piece of seed, despite the intact embryo, has zero potential for producing a new plant. Even if larger pieces—75% or even 90%—of the original size containing the entire embryo are formed, the same limitations will apply.

It must be clear by now that there is nothing to commend broken gymnospermous seeds of any size, even up to 90% of the original size, as qualifying as “pure seeds,” even if they are sown soon after collection. For yew and ginkgo, unless the correct piece of seed is sown, that is, the one containing the embryo/germ, no germinant will be forthcoming. For all seeds discussed so far, the application of Rule 3.2.1.1.2 is nonsensical.

Fungi and Bacteria

Even with the seed coat merely cracked, the internal tissues of all damaged seeds are subject to fungal and bacterial infections. The more tissue that is exposed the greater the likelihood of some form of infection. Intact seeds bear such spores in their seed coats, but the internal tissues may not be affected until the seeds begin to germinate. In germination tests of many coniferous species, especially Abies, the presence of Mucor and Penicillium can be substantial, and germinants may be abnormal as a result. With this knowledge, when a crack in the seed coat/testa is visible to the analyst, it must not be classified as a “pure seed.”

B. Angiospermous Species

In most instances, the effect of Rule 3.2.1.1.2 for angiospermous tree seeds is not the same as it is for gymnosperm genera. For discussion purposes, seeds of angiospermous genera have been organized into two groups based on information contained in the “Woody Plant Seed Manual” (Bonner et al., 2008), as follows:

  1. endospermic, which includes Cornus, Fraxinus, Gleditsia, Ilex, Liriodendron, Platanus, Tilia.

  2. non-endospermic, which includes Acacia, Acer, Aesculus, Ailanthus, Alnus, Betula, Carpinus, Castanea, Corylus, Eucalyptus, Fagus, Populus (rudimentary endosperm only), Quercus, Robinia, Salix, Sophora, Tectona, and Ulmus.

Representative illustrations for each group, rather than all genera, will be discussed. Illustrations for other genera can be found in Appendix II.

(a) Endospermic seeds

The internal structures of seeds of Cornus, Fraxinus and Platanus closely resemble those of the gymnosperms in that there is a linear embryo that occupies the bulk of the seed length.

Cornaceae

Cornus (dogwood)

In Cornus seeds the embryonic cotyledons are thick and wide on an embryo that occupies 80% or more of the length of the “bony stone” surrounded by endosperm (Fig. 29). Occasionally a stone may contain two embryos as shown in the transverse section (B), or just a single embryo (C).

If such a seed containing a single embryo (Fig. 30) is broken so that a piece of seed 51–55% of the original length as measured from the chalazal (cotyledon) end of the seed, essentially all of the hypocotyl and the primary radicle tissue will be lost (Fig. 31). The endosperm will also be exposed, so the potential for a new seedling will be zero.

Cornus seed, 55% size, cotyledon end
Line drawing showing a Cornus sericea seed broken at 51–55% of its length from the chalazal end.
Cornus seed, 55% size, radicle end
Line drawing showing a Cornus sericea seed broken at 51–55% of its length from the micropylar end.

Conversely, should the break occur at 51–55% of the original size measured from the micropylar (radicle) end of the seed (Fig. 32), most of the embryo will be lost, and the endosperm will be exposed. Should the embryo remnant elongate it will emerge toward what remains of the chalazal (cotyledon) end of the seed because there will be no resistance to cause the radicle to exit via the micropyle; like gymnospermous seeds, germination is epigeal. Growth will be limited due to the lack of an energy source created by the missing cotyledons. Thus, even the short-term potential for a new plant will be zero. But Rule 3.2.1.1.2 tells the analyst this must be declared a “pure seed.” Once more, the inconsistent application of Rule 3.2.3.5 must be noted regarding separated cotyledons.

Cornus seed, 75% size, cotyledon end
Line drawing showing a Cornus sericea stone broken at 75% of its length from the chalazal end.
Cornus seed, 75% size, radicle end
Line drawing showing a Cornus sericea stone broken at 75% of its length from the micropylar end.
Cornus seed, 90% size, cotyledon end
Line drawing showing a Cornus sericea stone broken at 90% of its length from the chalazal end.
Cornus seed, 90% size, radicle end
Line drawing showing a Cornus sericea stone broken at 90% of its length from the micropylar end.

Even with pieces of seeds measuring 75% of the original size from either end (Figs. 33 and 34) there is no possibility for a new plant to be produced for exactly the same reasons. Although the embryo may not be damaged if the seed is broken at 90% from either end (Figs. 35 and 36) the endosperm will be exposed and viability likely lost. So the chances of producing a new plant are slim. Similar concerns must be expressed for a seed broken along its length (Fig. 37).

Again, the analyst will easily recognize these pieces for what they are. But, in all cases, Rule 3.2.1.1.2 requires the analyst to classify these pieces as “pure seeds.”

Oleaceae & Platanaceae

Fraxinus (ash) & Platanus (sycamore)

Ash fruits are elongated, winged, single-seeded samaras borne in clusters. Individual seeds vary in size, the one shown here of green ash (F. pennsylvanica) is approximately 15 mm long and the embryo occupies 90% or more of this length (Fig. 38). Germination is again epigeal, so if the seed is broken at 51–55% of its original length measured from either end, a major portion of the embryo will be lost: either the entire hypocotyl and primary radicle (Fig. 39), or the bulk of the cotyledons (Fig. 40) that are needed to provide energy for growth. In both cases, the internal tissues will be exposed. The effect of breakage at 75% from both the cotyledon and micropylar ends is shown in Figs. 41 and 42, respectively.

Fraxinus seed, 55% size, cotyledon end
Line drawing of a Fraxinus pennsylvanica seed broken at 51–55% from the chalazal end.
Fraxinus seed, 55% size, radicle end
Line drawing of a Fraxinus pennsylvanica seed broken at 51–55% from the micropylar end.
Fraxinus seed, 75% size, cotyledon end
Line drawing of a Fraxinus pennsylvanica seed broken at 75% from the chalazal end.
Fraxinus seed, 75% size, radicle end
Line drawing of a Fraxinus pennsylvanica seed broken at 75% from the micropylar end.

American sycamore seeds (Platanus occidentalis) (Fig. 43) closely resemble ash seeds in their internal structure—an elongated linear embryo from which significant portions will go missing if the seed is broken at 51–55% from either end (Figs. 44 and 45).

Platanus seed, 55% size, cotyledon end
Line drawing of a Platanus occidentalis seed broken at 51–55% from the chalazal end.
Platanus seed, 55% size, radicle end
Line drawing of a Platanus occidentalis seed broken at 51–55% from the micropylar end.
Platanus seed, 75% size, cotyledon end
Line drawing of a Platanus occidentalis seed broken at 75% from the chalazal end.
Platanus seed, 75% size, radicle end
Line drawing of a Platanus occidentalis seed broken at 75% from the micropylar end.

In both genera, even if the pieces of seed, from either end, measure 75% of the original size (Figs. 41 and 42 (Fraxinus), 46 and 47 (Platanus) there is no possibility for a new plant to be produced for exactly the same reasons. Internal tissues will be exposed so broken seeds will perish in storage. Again, these “pieces of seeds” are easily recognized for what they are but, in all cases, Rule 3.2.1.1.2 requires the analyst to classify them as “pure seeds.”

Tiliaceae

Tilia (basswood, linden)

Linden embryos also occupy the bulk of the seeds, but they are far from linear (Fig. 48). It is clear that should a seed be broken from either end, at 51–55% (Figs. 49 ands 50), at 75% (Figs. 51 and 52) or even 90% (not illustrated), there is no possibility for a new plant to be produced for exactly the same reasons given above for Cornus, Fraxinus and Platanus. Exposure of the internal tissues will cause such broken seeds to perish during storage. Despite such “pieces of seeds” being easily recognizable, Rule 3.2.1.1.2 requires the analyst to classify them as “pure seeds.”

Tilia seed, 55% size, cotyledon end
Line drawing of a Tilia americana seed broken at 51–55% from the chalazal end.
Tilia seed, 55% size, radicle end
Line drawing of a Tilia americana seed broken at 51–55% from the micropylar end.
Tilia seed, 75% size, cotyledon end
Line drawing of a Tilia americana seed broken at 75% from the chalazal end.
Tilia seed, 75% size, radicle end
Line drawing of a Tilia americana seed broken at 75% from the micropylar end.

Fabaceae, Aquifoliaceae, Magnoliaceae

Gleditsia (honeylocust), Ilex (holly), Liriodendron (tuliptree)

Although they depend on endosperm (2n) tissue to provide the energy for germination, seeds of the other three genera in this group differ in that their embryos are small: relative to overall external length of each seed, they occupy approximately 15% in Gleditsia (Fig. 53), less than 10% in Ilex (Fig. 54), and 20–25% in Liriodendron (Fig. 55). Thus, broken seeds present somewhat different problems.

After any amount of breakage from the chalazal end of the seed (the opposite end from where the radicle would normally emerge—epigeal germination), even if the piece is 75–90% of the original length, the embryo structure is likely to be lost within the piece “smaller than one-half of the original size” which must be classified as “inert matter.” In each case, despite the absence of the embryo, the “pieces larger than one-half the original size,” Rule 3.2.1.1.2 requires the analyst to classify them as “pure seed.”

If the seeds are broken from the micropylar end, so that the embryo is contained within the pieces, any “piece of seed larger than one-half the original,” i.e. a “pure seed,” has a chance—but not a very strong one—of germinating. Whether a normal germinant would result is unknown. And this would only apply if the seeds were sown or germinated before being placed in storage.

Both the impact of seed storage, already referred to frequently, and the designation of which piece of seed is greater than one-half the original, will be discussed more fully at the end of Part I.

In all cases, the identity of which piece of seed is being classified should be important. However, as will be discussed in Part II, not only has this detail not been taken into account, it has actually been denied to the analyst.

(b) Non-endospermic seeds

Acacia, Acer, Aesculus, Ailanthus, Alnus, Betula, Carpinus, Castanea, Corylus, Eucalyptus, Fagus, Populus (rudimentary endosperm), Quercus, Robinia, Salix, Sophora, Tectona, and Ulmus

The embryo fills the entire seed coat in seeds of all these genera, the bulk of the tissue being the essential cotyledons (hypogeal germination); in comparison, the radicle and hypocotyl are diminutive, in many cases no more than 10–20% of the overall seed length, even smaller in Salix and Tectona. To illustrate the damage that will occur if these seeds are broken, focus is placed on Corylus and Quercus—hypogeal germination—and Alnus and Fagus—epigeal germination. To complete the picture, a special mention will be made of a tropical tree seed, Tectona (teak). Illustrations of the other genera in this group can be found in Appendix II.

Betulaceae

Corylus (hazel)

The hard-shelled nuts of the hazel are quite large (Fig. 56), and are almost completely filled by the cotyledons (hypogeal germination). However, the hypocotyl and radicle occupy no more than 20% of the length of the seed at the micropylar end. As in the honeylocust, holly and tuliptree seeds discussed earlier, even a piece of a hazel nut that measure 75–80% of the original length of the seed measured from the chalazal end will not contain the hypocotyl and radicle (Fig. 59), let alone a piece only 51–55% of the original (Fig. 57). Despite the complete absence of embryonic structures, Rule 3.2.1.1.2 dictates that they must be classified as “pure seeds.”

Corylus seed, 55% size, cotyledon end
Line drawing of a Corylus cornuta seed broken at 51–55% from the chalazal end.
Corylus seed, 55% size, radicle end
Line drawing of a Corylus cornuta seed broken at 51–55% from the micropylar end.
Corylus seed, 75% size, cotyledon end
Line drawing of a Corylus cornuta seed broken at 75% from the chalazal end.
Corylus seed, 75% size, radicle end
Line drawing of a Corylus cornuta seed broken at 75% from the micropylar end.

In contrast, seeds broken in the same proportions measured from the micropylar end (Figs. 58 and 60) might produce a seedling, but it may not be a vigorous one since its cotyledonary-source of energy has been reduced and, because the internal tissues have been exposed, it may not survive storage. Once more, such pieces must be classified as “pure seed” according to Rule 3.2.1.1.2.

Alnus (alder)

Alder seeds, represented here by the quite small-seeded species, Alnus rubra (red alder) (Fig. 61), are produced in cone-like strobiles and are easily extracted. Air dried, they can be stored for 2 years in sealed containers, but if dried to less than 10% moisture content (m.c.) they hold their viability for 10–20 years when stored at below freezing in sealed containers. They do not appear to be readily broken during extraction and processing, but Figs. 62 and 63 illustrate the damage that could happen to the internal structures if the seeds are broken.

Alnus seed, 55% size, cotyledon end
Line drawing of an Alnus rubra seed broken at 51–55% from the chalazal end.
Alnus seed, 55% size, radicle end
Line drawing of an Alnus rubra seed broken at 51–55% from the micropylar end.
Alnus seed, 75% size, cotyledon end
Line drawing of an Alnus rubra seed broken at 75% from the chalazal end.
Alnus seed, 90 size, radicle end
Line drawing of an Alnus rubra seed broken around 90% from the chalazal end.

Because the germ is so miniscule, even seeds broken at 75% of their original length measured from the chalazal end will not produce a new plant (Fig. 64). Yet such “pieces of seeds” must be regarded as “pure seeds” according to Rule 3.2.1.1.2. It is irrelevant to consider the effects of storage on such “pieces,” including seeds that have suffered longitudinal side breaks (not illustrated). Even at around 90% from the chalazal end (Fig. 65), germination is doubtful because the radicle/hypocotyl probably has been damaged. As in other species, seeds broken 51–55% (or more) of their original length as measured from the micropylar end might germinate if they are sown soon after collection and processing. However, it is doubtful they would survive below freezing storage at less than 10% m.c.

Fagaceae

Quercus (oak)

Acorns vary in size from 6 to 37 mm in length, those of Quercus rubra, the northern red oak (Fig. 66), being quite large; like the fruit of the black oaks, those of the red oaks require two years to mature. There are two fleshy cotyledons—no endosperm—that occupy the bulk of the acorn with the germ developing very close to the micropyle. So, if broken more than 50% of the length from the micropyle (Fig. 67), that is, a “piece of seed larger than one-half the original size,” there is a good chance of obtaining a new plant. In fact, the official germination test method involves soaking acorns in water for 48 hours, then cutting off roughly one third of the fruit from the cotyledon (cup scar) end, and planting the cut surface in wet sand. When the radicle appears and has grown down the side of the acorn to reach the surface of the sand, it is deemed to have germinated.

Quercus seed, 55% size, radicle end
Line drawing of a Quercus rubra seed broken at 51–55% from the micropylar end.
Quercus seed, 75% size, cotyledon end
Line drawing of a Quercus rubra seed broken at 75% from the chalazal end.

However, if the piece of acorn comprised 75–80% of the length of the nut from the cotyledonary end (Fig. 68), it would lack the germ and would have no potential for producing a new plant. But it must be classified as “pure seed” according to Rule 3.2.1.1.2. Even intact acorns are “recalcitrant” seeds (the term “less orthodox” preferred now) when it comes to storing them for any time. That is, they cannot tolerate desiccation below a high minimum moisture content—up to 45–50% in white oaks, somewhat lower in red and black oaks—making them very difficult to store. A side break (not illustrated) would exacerbate this condition. White oak acorns need to be collected soon after they have fallen from the tree to retard early germination. For these two primary reasons broken acorns have no chance of surviving any duration of storage and have no value for producing new plants. Pieces of acorns of any size are not “pure seeds.”

Fagus (beech)

It should be clear by now that if seeds of Fagus grandifolia (American beech) (Fig. 69) are broken at approximately 75–80% from the cotyledon end (Fig. 70), there is no chance of obtaining a new plant because the very small germ that occupies only the narrowest part of the nut at the micropylar end, has been lost. Conversely, there may be a faint chance of a germinant if the nut is broken more than half way from the micropylar end (Fig. 71), an even larger piece increasing this possibility. Research on the European beech, Fagus sylvatica, has shown that its intact nuts can be stored for at least 6 years if dried properly and placed in sealed containers at temperatures from -5 to -15°C, which is long enough for reforestation operational needs. Similar studies on Fagus grandifolia are not known (Bonner 2008) or for broken seeds of either species, but broken seeds should be considered a write-off for producing new plants. Based on this, Rule 3.2.1.1.2 must not be applied. That is to say, pieces of beech seeds of any size must not be classified as “pure seeds.”

Fagus seed, 75% size, cotyledon end
Line drawing of a Fagus grandifolia seed broken at 75% from the chalazal end.
Fagus seed, 60% size, radicle end
Line drawing of a Fagus grandifolia seed broken around 60% from the micropylar end.

Verbenaceae

Tectona (teak)

To illustrate that the questions surrounding pieces of tree seeds are not limited to temperate zone species, a tropical forest tree, Tectona grandis (teak), is included here. As Fig. 72 illustrates, the germ occupies less than 10% of the overall seed length. Thus, if the seed is broken so that a piece approximately 90% of the original size measured from the chalazal end is created (Fig. 73), the germ will be lost and there will be no possibility for obtaining a new plant. But which piece of seed is a decision denied the analyst (see Part II). Conversely, only a piece broken from the micropylar end (Fig. 74) has any chance of producing a new plant.

Tectona seed, 90% size, cotyledon end
Line drawing of a Tectona grandis seed broken around 90% from the chalazal end.
Tectona seed, 90% size, radicle end
Line drawing of a Tectona grandis seed broken around 90% from the micropylar end.

Because teak seeds are truly orthodox in storage behavior, viability of intact seeds may be maintained for up to 7 years at low temperature and relatively low moisture content (12%). Seeds stored in sacks in a dry warehouse maintained their viability for 2 years (Schubert and Francis, 2008). One method of extracting the 1–3 seeds from each fruit is by working a bag half-filled with dry fruits against the ground with a foot, then winnowing to separate the chaff. This would appear to be a very good opportunity for some seeds, small as they are, to be broken. There is no evidence that pieces of broken teak seeds maintain their viability as well as intact seeds, so why are they to be considered as pure seeds, especially if they are broken from the cotyledon end? Detailed information on many other tropical tree seeds can be found in Vozzo (2002).

Myrtaceae

Eucalyptus (eucalyptus) a special consideration

The genus Eucalyptus requires special attention where Rule 3.2.1.1.2 is concerned. There are several reasons for this:

  1. There are more than 523 known species of eucalyptus and 138 varieties, and new species and varieties are still being described (Krugman and Whitesell 2008). Among these species and varieties, as well as within a species, the length of fertile seeds varies from as small as 0.75 mm for E. camaldulensis, to 4.25 mm for E. citriodora. Fig. 75 illustrates the internal organs of an intact E. rudis seed, a small member of the genus.

  2. For any given seed collection within a species (or variety) seed length may vary considerably. For example, in E. camaldulensis, one of the very smallest eucalyptus seeds, seed length may vary from 0.75 mm to 1.75 mm (Krugman and Whitesell 2008), i.e. by a factor of more than 230%.

  3. PSD 60, as in all other definitions, requires that the pure seed fraction shall include pieces of seeds more than one-half the original size (with or without testa, of course). This means that the analyst is required to recognize broken seeds of E. camaldulensis possibly as small as 0.38–0.41 mm (380–410µ) in length, that is, approximately 51–55% of the original; at 65% of the original, pieces will be only 0.45–1 mm (450–1000µ) in length; and even at 75% of the original, pieces will be no larger than 0.56–1.3 mm (560–1300µ) long.

  4. PSD 60 goes on to state that “in many” (of the more than 500) “species it is impossible to differentiate with certainty between seeds and ovulodes…” Does this not also mean that it is impossible to differentiate “with certainty” “pieces of seeds” possibly as small as 380–870µ (microns) long, even if they are distinguishable from the ovulodes? In other words, how can a piece of seed be recognized “with certainty?” What power microscope must the analyst use to attempt this recognition?

    Eucalyptus seed, 90% size, cotyledon end
    Line drawing of a Eucalyptus rudis seed broken around 90% from the chalazal end.
    Eucalyptus seed, 90% size, radicle end
    Line drawing of a Eucalyptus rudis seed broken around 90% from the micropylar end.
  5. Even if pieces of seeds can be distinguished from the ovulodes by the analyst, a piece of seed broken more than 90% of the original length as measured from the chalazal end will be useless for propagating a new plant, because the minute hypocotyl and radicle will be missing (Fig. 76). Although “orthodox” in storage behaviour, a eucalyptus seed broken at 95% from the micropylar end (Fig. 77), even if it can be identified properly by the analyst, stands little chance of surviving prolonged storage.

So why does PSD 60 include Rule 3.2.1.1.2?

Because the scientific principles that apply to Eucalyptus have not been understood by ISTA.

It is clear that the biological knowledge of eucalyptus seeds has been completely ignored, that there is no legitimate scientific reason for inclusion of Rule 3.2.1.1.2. The only reason that Rule 3.2.1.1.2 is included in PSD 60 is to make it consistent with the other 62 PSDs. In any event, the procedure for the germination test (by weight rather than by number) makes the purity test for this genus appear to be redundant.

Other angiospermous seeds are illustrated in Appendix II.

Some additional biological considerations for forest tree seeds

  1. Seed crop periodicity

    Unlike the farmer’s seed crop plants, most of which produce annually, tree seed crops are periodic, and usually unpredictable. Whereas some individual trees may produce some cones or fruits every year, there may not be enough to warrant the expense of their collection on an annual basis. Most trees, most stands of trees, produce sizeable seed crops erratically, sometimes every second, third, or fourth year, or even less frequently, depending on the species and the geographic area in which the trees are growing. The terms “bumper crop” years, or “heavy mast” years, are commonly used to indicate, based on pre-collection inspections, that the crop is collectible, meaning it is worth the expense and effort of bringing it into the processing plant. In British Columbia it has been documented that only 8 “collectible” crops of Douglas-fir seeds occurred over 40 years (Lowry 1966). Meteorological conditions during the year of reproductive primordia formation, and in the following year when pollination, fertilization and seed maturation occur have major impacts on the amounts of seeds produced on forest trees (Eis 1973; Owens and Molder 1985). In contrast, teak trees (Tectona grandis) produce good seed crops more or less every year (Schubert and Francis 2008).

    In the hemlocks, (Tsuga spp.), there are physical as well as biological constraints to producing cones of any quantity less than 2 years apart. This is because the female reproductive primordia are formed at the tips of branches and branchlets. Having produced a good crop on most if not all branch/branchlet tips this year, the mother tree must grow vegetatively for a year to produce fresh tips. Thus, at least a year must pass before this particular mother tree can form fresh reproductive primordia. Development of these are at the mercy of meteorological conditions over the following 2 years.

  2. Collecting cones, fruits and seeds

    Compared to the farmer who may drive his combine harvester around his more-or-less level fields adjoining his farm house, the forester must make forays into the hills, perhaps to the summits of local mountains. Before doing so, the forester must make certain that gathering together the required collecting crew, equipment (including safety equipment) and supplies is worth the cost of travelling, often many kilometres, to the collection site. How to plan and organize cone collections, as well as guidelines for collecting, have been published (Dobbs et al. 1974; 1976; Edwards 1981a, 1985; Eremko et al. 1989; Portlock 1996).

    Cone collecting of Douglas-fir and other regional conifer species by helicopter was pioneered in British Columbia (Camenzind 1990), and is now also used in other regions of the Pacific Coast (and probably elsewhere). This method (Fig. 78) permits seeds to be harvested from regions that are inaccessible overland by road. Clearly, the cone crop, and the seed crop contained therein, must be large enough to warrant the use of this method, not only because of the cost, but also because of the danger involved. Whatever the collection method, for any given seed source as many cone and seeds as possible have to be collected, because there may not be another crop for several years and annual reforestation targets have to be met.

    Once the seeds have been extracted from the cones or fruits, they must be cleaned. This is the stage at which seed damage is most likely to occur. The cleaned seeds must be conditioned and placed in storage to meet future reforestation requirements. The sample submitted to the testing laboratory and which arrives on the analyst’s purity table, has been withdrawn from storage. This brings us to a crucial issue for pieces of forest tree seeds.

  3. Seed storage

    Literally thousands of research papers, entire libraries, have been written on seed storage, especially tree seed storage. The principles and practices of forest tree seed storage in general have been recently reviewed by Bonner (2008). This includes the most up-to-date information for all the gymnospermous and angiospermous genera included in the Rules. An earlier, excellent source is Wang (1974).

    “Orthodox” versus “Non-orthodox” a.k.a “Recalcitrant” seeds.

    Thirty or so years ago forest tree seeds were classified either as “orthodox” or “recalcitrant” for storage considerations. Currently, “orthodox” and “recalcitrant” are viewed as extremes of a continuum, and terminology is moving away from “recalcitrant” to “non-orthodox.”

    Coniferous seeds are regarded as “orthodox,” meaning that intact seeds can withstand being dried to low moisture contents in the range of 5–10% of their fresh weight and held for several years at -17°C without losing their viability. Below 5% m.c. autoxidation is likely. In contrast, many angiospermous seeds are “non-orthodox,” meaning that they do not take kindly either to being dried or to being refrigerated, and their viability suffers drastically and rapidly. Oak acorns (Quercus) are borderline “non-orthodox,” while chestnuts (Castanea) are very “non-orthodox.” Seeds of these genera, and Aesculus, are highly perishable even when intact.

  4. Role of the seed coat (or testa)

    Even for seeds of species regarded as being orthodox for storage, the role of the seed coat, or testa as ISTA prefers, is crucial. The longevity of coniferous seeds that survive being dried to low moisture contents and refrigerated depends upon their seed coats remaining intact. This is the key.

    Structurally the seed coat is an inactive (non-living) covering that imbibes moisture rapidly when the seeds are prepared for germination, but which more importantly protects the internal tissues. In gymnospermous seeds these tissues are the female megagametophyte from which the embryo derives it source of energy for germination, and the embryo from which a new plant may be (but not always) derived. The seed coat acts to prevent germination until climatological conditions are advantageous. That is, in large part the seed coat is the cause of seed dormancy. It has been demonstrated in noble fir (Abies procera in North America, Abies nobilis in Europe) that chipping the seed coat at the micropylar end of the seeds, then allowing the seeds to imbibe water, brings on rapid germination without need for pre-chilling (Edwards 1969). However, if the seed coat is damaged in any way, even cracked, it loses its protective-envelope abilities and seed viability suffers if and when the seeds are returned to storage. Cracks or other breaches of the seed coat also allow microorganisms to enter. Respiration increases (Leadem 1993) until moisture levels have decreased to prevent further gas exchange. If this happens when the seeds are merely cracked, what chance is there for seeds that have been broken into pieces, no matter what proportion the piece is relative to the original size of the seed?

    As already discussed, the seed coat acts to prevent fungal and bacterial infection. Damage it and the seed likely loses its viability to disease. Even with the seed coat merely cracked, the internal tissues of all damaged seeds are subject to fungal and bacterial infection. The more tissue that is exposed the greater the likelihood of some form of infection. Even intact seeds bear such spores on/in their seed coats, but the internal tissues may not be affected until the seeds begin to germinate, and the germinants may be abnormal as a result. With this scientific knowledge, as claimed in the Introduction to the Rules, why must the analyst classify damaged seeds as “pure seeds?”

    “Non-orthodox” (recalcitrant) seeds, with their high moisture contents, are potentially very easily damaged during seed handling. According to Bonner (2008), the most important non-orthodox species in North America are the oaks (Quercus spp.) that have single-seeded fruits (acorns) with rather well-protected embryonic axes because of their strong outer covering structures. However, these coverings may become damaged, and the acorns broken into pieces if care in handling is absent. In sharp contrast, the seed of the silver maple (Acer saccharinum) has a large embryo protected by a soft pliable pericarp that is very susceptible to bruising during seed handling (Bonner 2008). Rough handling may break this seed into pieces.

    The principle objective of storage is to reduce seed metabolism as much as possible without damaging their viability, and to prevent attack by microorganisms. The ideal metabolic rate during storage will conserve much of the stored energy reserves (needed for germination), yet allows the seeds to maintain the viability of the embryos. If seeds are broken into pieces, even if the outer coverings are merely cracked, let alone totally removed, internal changes include a loss in moisture content and an increase in respiratory processes that rapidly lowers the integrity, i.e., the viability, of the embryos.

    Thus, when the ISTA Rules state that a “pieces of seed units larger than one-half their original size” must be regarded as a “pure seed” for 41 genera of Forest Tree Seeds (Table 1), the question must be asked: who wrote this Rule, and, on what scientific knowledge is it based?

    Throughout this review, it has been emphasized that the Rules advertize that a “pure seed” means a “good seed,” i.e., one that, at least visually, has potential for producing a new plant (see Reminders). Churchgoers will be familiar with the 18th century German hymn “Wir Pl├╝gen” (words by the German poet Matthias Claudius, 1782, music by the German composer Johann Abraham Peter Schulz, died 1800) that in English begins:

    We plough the fields and scatter the good seed on the land…

    The emphasis here is on scattering (i.e. sowing, not planting) the “good seed” to produce a new crop of plants. Thus, more than 200 years ago farmers were anxious to sow “good,” i.e., potentially viable, seeds. What would be the point of sowing pieces of seeds that were not viable? Two hundred plus years ago, or even now?

  5. Modern forest nursery practice

    When production of nursery seedlings for reforestation peaked at 265,000,000 plants, as it did in British Columbia (BC) in 2015, nursery managers would not be very happy to find they were saddled with pieces of seeds that are useless. However, the BC Ministry of Forests collects, processes and tests its seeds in-house, the latter following the ISTA Rules in principle, but not applying Rule 3.2.1.1.2 since seeds are provided to provincial nurseries. Seedlots that showed unusual amounts of impurities, including broken seeds, would be, and still are, recleaned. This is the practice at most if not all modern forest seed processing plants and nurseries. Particularly at nurseries that grow upwards of 14 million seedlings in mini-containers, such as is shown in Fig. 80. Fig. 81 shows the containers, or cavities as they are known, with spruce seedlings.

    To be cost-effective, this type of seedling production requires a plant in each and every cavity. Even with highly pure seedlots germination may not be sufficient to guarantee a seedling from every seed sown, so to avoid empty cavities two or three seeds maybe sown in each cavity. Research has developed guidelines for the numbers of seeds to be sown based on laboratory germination performance. If more than one seedling appears in the cavities they are culled leaving the most vigorous plant to grow. Other research studying the effect such culling has on the genetic makeup of the final seedling crop, that is, the genetic base of the new forest, has also been conducted (El-Kassaby et al., 1993.)

    The main issue is that broken pieces of forest tree seeds are not only useless in this system, they add to the costs (Edwards, 1981b).

    The previous discussion illustrates a situation that has existed in the International Rules for most of the 20th century. Based upon this review (with more to follow) it is hoped that the ISTA purity test can be brought into the 21st century.