CHOOSING SEED SOURCES

What is meant by biodiversity?

The great range of variation found in living things, and in the ecosystems they form (D 10 in Manual 4)

Do you mean all of it?

Yes, in the widest sense of the word; but
No, in the way this relatively new term is commonly used, which usually implies that a particular plant or animal, or a specific habitat, is becoming depleted in its variability, and that this needs to be corrected.

But why should I bother about variation anyway?

Because it is a fundamental feature of living organisms, which:

1. provides the basis for their evolution as well as for the functioning of ecosystems;
2. allows us to go on choosing forms that are more suited to a particular purpose; and
3. will permit future generations of human beings to have access to sufficient diversity.

Are we talking about genetic variation?

The variability we see is the expression of inherent tendencies in a particular set of environmental conditions. For some characteristics of trees, the genetic contribution seems to be relatively small, whereas other features appear to be strongly inherited. It is these latter tendencies which are especially implied in the term biodiversity.

Surely people want to reduce genetic variation in crops, trees and animals!

Yes, that's right; though not too much! A balance needs to be struck between:
domestication, which is the process of finding, selecting, testing and using superior forms for particular sites and purposes; and also breeding new ones; and
conservation, which is concerned with retaining enough biodiversity for the future.

What are the chief ways of domesticating trees?

1. Direct vegetative propagation of selected individuals (A 10 in Manual 1, C 5 in Manual 3), followed by testing the clones in practice (Manual 5);
2. Choosing particular seed sources, and testing them under local conditions, for example:
   1. which provenances or local races appear to be most suitable (B 21);
   2. the parents that give seedlings with above-average performance (B 22);
3. Cross-pollinations to produce hybrids with new combinations of characters (B 23); and
4. New techniques of biotechnology, some of which can allow genetic material from unrelated organisms to be added.

But if you propagate from fine trees that just grew in a favourable environment, it won't do any good, will it?

You've put your finger on the chief problem that hampers tree improvement. There is so much variation there, and we usually do not know to what extent it reflects:

1. genetic tendencies, passed on in the pollen as well as by the seed parent (B 11);
2. environmental influences of many kinds, during the whole time the trees were growing;
3. chance occurrences, that may have damaged or favoured them; and
4. internal factors, such as the effects of competition for sugars and nutrients on seed size, and then on seedling growth; or the differences between juvenile and mature shoots.

Doesn't that make it a hopeless task?

No, just one that is not all straightforward! However, the prospects of tree improvement are actually good, because:

1. its potential in ‘wild’ or undomesticated species (the great majority of trees) is much greater than in crop plants which have been selected and bred over long periods;
2. many features that are important to growers are strongly inherited; and
3. several worthwhile steps forward can be taken ahead of the results of longer-term research results.

Surely it's too difficult for an individual grower?

No, because most people raising young trees in a nursery can learn how to:

1. recognise characteristics of trees that would be more suitable for the site and the purpose of planting;
2. appreciate the main pitfalls to avoid; and
3. choose seed sources that are more likely to give superior planting stock.

Supposing that much of the replanting of trees needed in the tropics was achieved - but with unimproved or inferior planting stock. What an opportunity would have been missed!

Yes, it certainly would. But mightn't genetic engineering allow quicker progress?

The recent very rapid developments in this field mean that there are now several other potential methods of domestication. For tropical trees, one needs to take into account:

1. the possibilities of bypassing some of the barriers that have hampered work;
2. whether any of the new techniques would be applicable to the ways in which trees are propagated and planted; and
3. what risks might accompany their widespread use, as some people fear with GM (genetically modified) agricultural crops.

What sort of changes might apply to tropical trees?

For example, inserting a gene that makes a tree flower early and heavily has been shown to be possible (see Weigel and Nillson, 1995 in B 51). One could ask:

• Would this be desirable in a fruit tree, but not in one grown for vegetative products?
• Might bigger fruit yields be achieved and sustained by sufficient leaf production?
• How much would it cost, compared with other methods?
• What problems could occur, and what might happen in subsequent generations? and
• How about the transfer of other desirable characters, as is now possible with mango?

Why does conservation matter as well as domestication?

Because a diversity of seed sources is always important to maintain a wide range of genetic variation into the future; especially because:

1. the natural range of many tree species is now much less extensive than it used to be, and some parts may have been completely cleared;
2. some species have been wiped out entirely or are in danger of disappearing;
3. the best timber trees (D 36 in Manual 4) are likely to have been preferentially felled, probably causing negative selection (B 22), instead of some of the best being retained as seed trees (D 24); and
4. products like medicines, forage, firewood and poles (D 33–35, D 38–39) may have been removed by careless exploitation where they could have been harvested in perpetuity (D 5).

So conservation has to do with everything from future breeding programmes (B 23) to maintaining the tree cover that is important for so many reasons (B 1).

But people often talk just about preserving rare animals!

Yes they do, partly because they make attractive subjects for television documentaries. Others speak as though recording the distribution of plant species in woodlands was the only important thing. Actually, conservation has many sides to it, including not wiping out:

1. the animals that pollinate flowers (B 11) and distribute seeds (B 12);
2. the kinds of micro-organisms that form important close associations with tree roots (C 30–32 in Manual 3; D 32 in Manual 4); nor
3. particular tree species that can play key roles in making nutrients (D 13) available in the ecosystem.

Conservation also needs to be incorporated into the local patterns of land-use, in order to secure the future livelihoods of the local people.

Which landscape patterns do you think can combine people's immediate needs with conservation for the future?

I see this connection being achieved in several traditional and new ways, including:

1. keeping some portions of natural and semi-natural stands, rather than removing the woodland completely;
2. leaving some trees as individuals or small groups scattered across an area of farmland, instead of creating a landscape bare of trees;
3. avoiding large blockes of pure, even-aged plantations (D 53–54) and
4. planting a number of suitable species as shade trees, boundary lines and small plots.

So one could have uniformity in some places and diversity in others!

Exactly! The two aims of domestication and conservation are best seen as complementary, rather than in conflict.

Then why don't we see that put into practice?

Why indeed? Perhaps some of the many reasons include:

1. inappropriate assumptions brought to the tropics from the temperate zone (D 1 in Manual 4);
2. exploitation of natural resources in place of managing them sustainably (D 60);
3. replacing mixed woodland with single-species plantations (D 30) or grazing land;
4. lack of detailed knowledge about the complex interactions that maintain wooded ecosystems, and neglect of known ways of improving degraded land (D 22, D 32);
5. loss of the local understanding of trees that used to be transmitted from one generation to the next;
6. the idea that the bigger an enterprise, the more money one can make; and
7. increase in human populations, and urbanisation.

Couldn't the approaches be altered?

Well, the way things are viewed has been changing quite rapidly, so that:

1. more farmers are seeing that they need trees (D 3, D 21 in Manual 4);
2. timber companies are beginning to realise that mere exploitation means that a natural resource comes to an end, when it could have gone on indefinitely;
3. economists are taking sustainability into account, instead of simply costs, expected returns and the interval of time between them;
4. foresters are appreciating the importance of minimising soil disturbance (D 60), using local as well as exotic species (D 31), and planting mixtures of trees (D 30, D 53);
5. a number of governments are taking tree-planting seriously (D 5); and
6. international agencies are using up-to-date language, and have changed their emphasis considerably (C 53 in Manual 3; D 70 in Manual 4).

What about conserving by making large Reserves?

Many countries have designated sizeable areas as Nature Reserves (D 28 in Manual 4) and Wildlife Parks, though this has not always been done using the knowledge and with the collaboration of local people. Other kinds of conservation work include:

1. growing small groups of a species in Botanical Gardens, rather than only a single specimen;
2. long-term storage of seeds (B 34) and pollen; and
3. an on-going programme of education about the importance of conservation.

Can I do anything about it?

Yes! Everyone who is growing trees can help to combine domestication and conserving biodiversity, for example by:

1. growing a number of different kinds of trees, rather than just a few;
2. thinking about which seed sources are likely to perform best, rather than being content with any seed of that species;
3. not using a single seed origin year after year, or taking seed from a very few parent trees, or covering whole landscapes with one clone; and
4. exchanging information, experience and seeds with others (B 24; C 53 in Manual 3).

What is a provenance?

One part of the natural range of a tree species.

How does that matter to someone planting trees?

Because trees which originate from particular provenances often thrive much better on a given site than those from other seed sources of the same species.

Why does this happen?

Over very long periods of time, the trees in one area tend to become a bit unlike those of the same species in different places. This is because they:

1. become adapted to the particular conditions that prevail locally, such as:
   1. commonly occurring soil conditions (D 12 in Manual 4);
   2. average climate (D 11);
   3. rare extremes of temperature, rainfall or wind; and
2. do not usually interbreed freely across long distances, because pollen and seeds seldom travel that far in any quantity.

So they become separate genetic origins within that species?

Yes, that's right; and these have the potential to perform better, or less well, at a given site and for a particular purpose.

How big can such differences be?

Occasionally they can involve success compared with complete failure;
Often they are large enough to be easily observed; but
Sometimes it is difficult to know if they are present, because:

1. a lot of variation within each provenance obscures the smaller differences between one seed source and the next;
2. features are involved that have not shown up in these conditions, or not yet;
3. the site is so variable that any genetic tendencies are obscured; or
4. no appreciable differences actually exist between the provenances.

What features of trees can be affected?

Depending also on the environment, any characteristic of the growth and development of the trees may show differences, such as the:

1. proportion of trees surviving, and their root development (C 11 in Manual 3);
2. rate of height and diameter growth, and the timing and duration of these (C 12);
3. various features of the leaves;
4. branching habit and stem straightness;
5. onset, timing and heaviness of flowering (B 10–11);
6. size, flavour and ripening of fruits (B 12); and perhaps
7. susceptibility to diseases or pests; or to rare climatic events.

How big an area does one provenance cover?

This varies a lot, depending on:

1. how variable the species is; and
2. the size of its natural range; but also
3. how much interest has been shown in it.

Sometimes a provenance may be designated as a:

1. whole country;
2. particular part of a country, or a
3. more specific area, like a particular river valley or small island.

Supposing the natural range is very small?

Then it may be better to concentrate on:

1. trying to make sure the original seed sources are not lost altogether (B 20);
2. selecting from the variation between different parent trees (B 22); and later on
3. crossing promising parent trees with each other, maybe including attempts to make hybrids with a related species (B 23).

Would using one provenance reduce the variation amongst my trees a lot?

No, it is a relatively small step in the overall process of domestication (B 20).
However, it is usually best to grow more than one provenance, in case disadvantages are found at a later stage.

Which kinds of trees have been studied most?

A lot of work on seed sources has for instance been done with teak and with several species of pines and eucalypts.

Is that because they are widely planted exotics?

Yes; and because their provenance differences have often been found to be large (B 51).

However, indigenous species can also be expected to show considerable variation, particularly when they have a wide natural range. If they have not yet been tested, concentrate on collecting your seed locally, from above-average parent trees (B 22; C 5 in Manual 3).

How are provenances tested?

1. Seed is collected from a sizeable number of trees in selected places covering the natural range, and the seeds within each seed-lot are well-mixed;
2. Seedlings are raised in tree nurseries under reasonably uniform conditions (C 7 in Manual 3), with similar sowing dates;
3. Field trials are planted (D 29 in Manual 4) in which plots of those provenances which have produced enough healthy seedlings are randomised (D 55). The trials are usually sited in areas similar to the proposed planting sites, with several sites in each region and sometimes many sites scattered across the tropics; and then
4. Observations and measurements are done in a standardised way, the results are analysed (C 67–69; and Manual 5), and they may be published (B 51).

That sounds too difficult for me to undertake!

It may not be necessary, if seed is available of already tested provenances; but
It can be relatively simple to collect and test a small set of local seed sources.

What is the most important thing to remember in doing a provenance trial?

Careful labelling and good records! Most of its usefulness will be lost (C 54) if one cannot distinguish later on which provenance is which, or tell where the seed was collected (C 64–65).

But supposing nobody knows what seed source was used for a thriving stand?

Then it is best to regard it as a local race, whether it:

1. was originally introduced (and has perhaps now become naturalised); or
2. is a local species, but all the natural stands have disappeared.

Note: watch out in such cases for any signs of inbreeding depression (B 22).

Doesn't importing seed cost a lot?

Not as much as using unsuitable sources for planting young trees, which are likely to:

1. grow poorly and so not fulfil their desired functions; and perhaps
2. remain to ‘contaminate’ better sources of seed with their pollen in the future.

Collaborating with other growers, and exchanging seeds, are ways of reducing the expense (B 24; C 54 in Manual 3).

How would you summarise the aims of choosing provenances?

As a first step in the genetic improvement of trees, which:

1. is often easy to take;
2. may make an appreciable difference; and
3. can lead on to other kinds of selection (B 22–23; A 12–13 in Manual 1).

Can't I just collect seed from any tree of an appropriate provenance?

Although this is quite often done, it is generally unwise, because you:

1. will be missing an opportunity to choose better sources; and
2. may actually be making an unconscious negative selection.

This is even more important than avoiding non-genetic problems like collecting fruits and seeds that contain a lot of the resting stages of pests or diseases (B 30–31).

How would I be missing out on better sources?

Selecting suitable provenances is only the first stage in domestication by seed (B 20–21; and C 5 in Manual 3). The next step is to collect seed from chosen parent trees rather than just from any of them. This is because:

1. there will still be a great deal of genetic variability left to be tapped; and so
2. one can expect to make quite a lot more positive selection.

Unselected seed collections quite often mean that the average performance is kept down by all the poorer trees in the provenance.

But choosing the parent trees isn't bound to give me better plants!

No, that's right. This is because:

1. the parent trees are being selected by their appearance, rather than by progeny tests;
2. the inherited characteristics of parents are not all passed on to the next generation;
3. the seedlings will be grown under different environments than the parent trees.

So why is it worthwhile, then?

Because choosing the parent trees carefully is a straightforward step that makes it more likely that the young trees you raise in your nursery and plant out will establish and grow well.
Your aim is to move towards planting stands and individuals with genetic potentials that are probably more appropriate to the particular site and purpose.

Could I really be going backwards rather than forwards by not bothering?

Yes, you definitely could; for instance because of:

1. unconscious selection towards unfavourable characteristics; or
2. inbreeding depression because the parent trees are too closely related to each other.

What sort of unfavourable features might I be encouraging?

Here's one example. Some trees:

1. start flowering earlier in life (B 10); though what you might be wanting is rapid production of a tall, straight stem;
2. can dominate the pollen produced, but perhaps be broad-crowned and branchy; or
3. produce many more seeds than others in the same stand, yet you may be looking for vegetative growth (C 10–12 in Manual 3) rather than reproductive activity (B 10–12) in your planted trees.

By unconsciously encouraging such unfavourable features in the seedlings you grow, it is likely that a lot of energy, sugars and minerals may be ‘wasted’ in the planted trees through unnecessary reproduction.

Can you give me an example?

For instance, many of the teak plantations made in Nigeria in the 1970s from tested provenances came into very heavy flowering soon after closing canopy. Since the large inflorescences are formed in the terminal bud of the leading shoot and strong upwardly growing branches, frequent forking occurs, so that there would hardly be any straight main stem to harvest for timber. It is very probable that this situation resulted from unconscious selection for early and heavy flowering, since:

1. it is clearly much easier to collect fruits underneath prolifically bearing trees;
2. such flowering characteristics are often strongly inherited; although
3. they were not marked features of the original introductions from Asia, nor of present-day stands there.

In West Africa, some young teak trees even started to flower in the nursery!

But what could be done about that?

One might for instance:

1. give preference to any available seeds from under sparsely fruiting trees of good stem form and branching habit;
2. import new seed of the best provenances (B 21, B 24);
3. create some seed stands out of thriving parts of an old provenance trial;
4. try out flower-inducing techniques (B 14) on good trees that are largely vegetative;
5. cross promising trees together (B 23); or
6. fell some good trees and use the coppice shoots as a sources of rooted cuttings (A 11 in Manual 1; C 5 in Manual 3) for clonal trials.

Supposing I am growing a tree species mainly for its fruits?

In that case you might welcome a tendency for:

1. earlier flowering, meaning that fruits could be produced nearer the ground;
2. heavier flowering, giving fruiting a higher status in relation to vegetative growth.

You could also select parent trees on the basis of fruit quality, size or season of ripening.

And what is inbreeding?

The fertilisation of female parts with pollen coming from trees that are closely related.

Why does that matter?

Most tree species are outbreeding, with successful progeny coming chiefly from the inter-breeding of distantly related or unrelated parents.

NOTE: seedlings from closely-related trees are generally poorer, and may be stunted.

Examples where inbreeding depression is particularly pronounced are Eucalyptus ?grandis and the temperate zone conifer Douglas fir.

Isn't that rather like human beings?

Yes - our societies have developed stringent rules against close sexual relationships, presumably because the children of near relatives often get strongly inherited disabilities and illnesses. For instance, the royal family in 19th century Britain had become inbred, and tended to suffer from haemophilia, a highly inherited condition affecting particularly the boys, in which the blood fails to clot.

How close is bad, for trees?

With most species, it is generally best to avoid collecting seed from:

1. single isolated trees, which is likely to be self-pollinated (B 11);
2. groups of 2–4 trees that are far away from any others of the same species; and
3. stands which themselves originated from:
   1. one or a very few parent trees;
   2. a very small seed-lot; or
   3. one or two clones.

When would inbreeding depression show up?

It might be already detectable in the nursery, particularly if a pot-plant trial is done (C 15 in Manual 3), using a well-known seed-lot as a standard; but
It may not show up until a field trial has grown for some years, involving more expense and a longer delay before the unsuitable seed sources are eliminated.

And trees can sometimes pollinate themselves!

Yes, this is not uncommon, if they produce both viable pollen and receptive female parts at the same time (B 11). However, there are often internal barriers to fertilisation (B 10), and ‘selfed’ seed is frequently empty. Any self-pollinated seedlings that occur naturally or from a breeding programme are generally inferior to those from cross-pollinations (B 23, B 50). For instance robusta coffee farms made from rooted cuttings need a minimum of two clones to set (B 12) any coffee berries at all from the flowers.

Are some kinds of trees different from this?

Yes, a few are inbreeders, in which seed from closely related trees does not show any depression, for this is the normal situation.

How can seed stands help?

Making a seed stand is the next stage in domestication by seed. It is a good way of avoiding both negative selection and inbreeding depression, and is not hard to do.

This how to set about it:

1. Choose a good stand of the species which has started flowering, if possible one whose seeds have already been found to give promising planted trees;
2. Carry out a thinning in which parent trees that show poorer characteristics for your purposes are removed;
3. Protect the seed stand (Manual 5), look out for signs of flowering (which may be enhanced by the thinning), and collect fruits or seeds (B 31) from as many of the parent trees as possible; and
4. Test the seed stand by comparing seedlings from these open-pollinated progeny and an ordinary seed-lot in a simple field trial (C 7 in Manual 3; D 29 in Manual 4).

How big would the seed stand need to be?

This will vary according to species and site, but for adequate pollination usually needs to be at least 0.5 ha. You could use past nursery and seed collection records (C 54, C 63–66), together with the expected demand for seedlings, to give you an rough idea of how many parent trees might be needed, after thinning. Then double the number to be on the safe side, since surplus seed might be stored (B 33), exchanged or sold (B 24).

Could I choose part of a larger stand?

It is usually better to find a place where the whole stand can be converted, or considerable amounts of pollen may be coming in from other trees nearby.

Where should it be established?

If possible a seed stand should be:

1. close at hand, rather than in a remote spot; and
2. not amongst large numbers of inferior trees of the same or a closely related species.

Are seed stands unsuitable for some species?

Yes, those in which flowering is very irregular.

Could I plant trees for a seed stand?

This could be sensible if:

1. there is no suitable existing stand available for converting; and
2. the species can be expected to come into flowering reasonably early in life.

Another reason would be if you had evidence that a special seed stand would give better progeny. Crossing chosen trees with each other (B 23) is the stage that comes after selecting and testing for better existing sets of characters.

How can domestication by seed be taken further than selecting provenances and parent trees?

By making specific crosses between chosen parent trees, and then growing on and testing their progeny.

Is this worth bothering with?

Yes it is; for the research has quite a large practical potential; but
No it may not be, for the individual grower to undertake.

So it's no use to me?

Yes, it could indeed be valuable, for instance if it had already been found that:

1. crossing two provenances produced considerably better young trees for your district;
2. specific pairs of parent trees combined to give superior progeny than others; or
3. a hybrid between two species combined many of their useful qualities.

But how would I be able to utilise those research results?

1. Though reading or hearing about new developments (B 51);
2. By visiting a research area or demonstration plots (D 29 in Manual 4);
3. From being able to obtain improved seed of these kinds (B 24); or
4. By planting a special seed stand yourself.

Special seed stand.

How would that differ from an ordinary seed stand?

In an ordinary seed stand (B 22), you are simply removing the trees from an existing stand that you think may be inferior, so that they contribute neither as males nor as females (B 11) to the seeds you collect; whereas
In a special seed stand, you could for example be planting alternate rows of those provenances that had already been shown to produce superior progeny.

The likelihood and the extent of genetic improvement are both greater in the special type, provided that the flowering times of the provenances are similar.

What would happen if they did not overlap much?

Most of the seeds will be progeny of the two provenances, rather than crosses between them.

Could I plant a seed stand with parent trees that were known to combine well?

Usually no, because this would involve difficult procedures like grafting mature scions on to seedling rootstocks (A 3 in Manual 1), progeny testing and making a seed orchard; but Occasionally yes, if mature cuttings (B 10) of tested parental clones can be rooted easily, and they can be stimulated to flower fairly soon after planting (B 14).

And what about hybrids between species?

These could have the biggest potential, but are hardest to achieve, because:

1. there are often genetical barriers which mean that few or no viable seeds (B 13) are produced;
2. the usual pollinators (B 11) may not operate efficiently between the two species;
3. flowering times are even more likely to differ; and so
4. seed supplies are usually very limited.

But couldn't one collect seed from just a few outstanding hybrids?

Generally no; because:

1. using promising first generation hybrid trees as parents commonly produces inferior and very variable progeny in the next generation;
2. the few hybrid trees may well be closely related, giving problems of inbreeding depression (B 22); and
3. some hybrids may fail to produce any viable seed at all.

Can pollinations be done by the grower?

Sometimes this may be possible on a small scale. For example, in wind-pollinated trees that produce a lot of pollen fairly early in life, it is possible to do ‘assisted’ pollination, where additional pollen, collected from desirable trees, is ‘puffed’ towards the receptive female parts of the other partner, in addition to natural open pollination; but

Usually crossing needs a research station with a long-term breeding programme. Then controlled pollination can be done, in which all male parts removed before they shed any pollen, and the desired female parts are enclosed in bags before they are receptive. This excludes unwanted pollen, so that when pollen from a specific tree or trees is ‘puffed’ on to them, and the bags resealed, all seeds will be of known parentage.

Supposing the pollen is shed after the females are receptive?

That's a good point. Then cross-pollinations depend on keeping pollen from a previous flowering. This involves:

1. drying the pollen, so that it does not go mouldy;
2. storing it, usually at lower temperatures; and
3. developing simple pollen germination tests (B 48), often in ‘hanging drops’ of a strong sugar solution, in order to check whether the procedures have kept the pollen viable.

Long-term storage of pollen is sometimes used as a method of conserving biodiversity (B 20).

What about flowers that are pollinated by insects or bats?

The quantity of pollen is typically much less, and it is generally sticky rather than powdery. However, on reasonably small trees it may still be possible to do assisted and controlled pollinations with a very small brush.

How does a research station set up a breeding programme?

This is not as straightforward as with an annual crop plant, where there are well-tried methods. Here are the traditional stages for trees, with some points to consider:

1. Choice of species - because fashions change, it may be that by the time the improvement has been achieved, a species may have become less popular. So choose several species where improvement is clearly needed, perhaps including multipurpose trees (D 40 in Manual 4).
2. Selection of candidate ‘plus’ trees - well-above average individuals chosen throughout the region by methods similar to those described in A 12 in Manual 1, bearing in mind the chief purposes for the tree planting.
3. Collecting mature scions high in the crowns, and grafting them on to seedling rootstocks - this usually involves climbing, as well as finding the best grafting method.
4. Growing the grafts in a research nursery, and planting them in a ‘clone bank’.
5. Doing a controlled crossing programme between parents when they flower - this often runs into difficulties with clones that remain largely or entirely vegetative (B 11).
6. Raising each of the labelled progenies in standard conditions in the research nursery (C 7 in Manual 3).
7. Planting replicated progeny trials in relevant planting areas (Manual 5).
8. Assessing these progeny trials at least twice, calculating each progeny's average performance in the desired characters, and finding the contributions made by different parents. Bear in mind the problems of reliability of early assessments.
9. Re-grafting the superior parent clones, and planting mixtures of them that combine well in seed-orchards that should produce improved seeds in bulk (D 24).

Can't this very long process be cut short?

Well, the seed-orchards can be planted at stage (D), and used first for making crosses, and later for seed production. This can work out:

1. well, if many of the crosses are found to give superior progeny; but
2. a costly waste of time and effort, if they are average, inferior or show inbreeding depression.

I still don't see quite why it's so much harder with trees than with bananas or cereals!

Both are short-lived crops in which amounts and quality of the yield can be directly selected for each year. So:

With bananas, clones which have been selected many hundreds of times are grown vegetatively from offsets (A 3 in Manual 1) that are nearly always genetically identical to the best of the plants that have just been harvested; and

With many cereals, because they are inbreeding species (B 22), the seed for the next crop can be taken from each years' harvest, again allowing positive selection repeated each year for centuries.

While for trees?

With most of the thousands of kinds of trees, since they are long-lived, outbreeding perennials:

1. no seeds are produced during the juvenile period, so repeated selection would take many decades;
2. progeny do not necessarily carry the desired characteristics of the parent trees;
3. there are many important features of trees, rather than a single clear-cut ‘yield’;
4. a number of these may not show up until the progeny have grown into sizeable trees, so choosing the best often has to be delayed for several years;
5. by that time it may be difficult to produce large supplies of improved seed; and
6. markets may anyway have changed.

Well, I can see why trees have mostly stayed undomesticated!

Yes, that's right - but there are other approaches (B 2)! The best way forward for most species is probably by combining seed and vegetative propagation in a different way.

How would that be done?

1. Starting with:
   1. provenance selection (B 21) and seed stands (B 22); and at the same time also with
   2. rooting juvenile cuttings taken from the best sources then available (A 10–12 in Manual 1);
2. Choosing promising parent trees and doing some ‘open-pollinated’ crosses between them;
3. Doing early tests (A 13) of various kinds on the progeny;
4. Selecting the most promising of these seedlings as stockplants (A 20) to produce clones of juvenile rooted cuttings;
5. Testing these in clonal trials that also contain the best current seed-lot as a comparison.

Are there any other possibilities of more rapid domestication?

Yes, there are several (B 20). For example:

1. hybrids between species could be multiplied vegetatively to produce planting stock, even if there were too few of them to test as seedlings;
2. meristem culture (A 5) of superior trees, perhaps including the stimulation of somatic embryos (sometimes called ‘artificial seeds’. These could be multiplied as micropropagated material, and then carefully ‘weaned’ and ‘hardened’ (A 54);
3. close-planted, miniaturised seed orchards might be planted for tree species where flowering can be reliably stimulated (B 14), and these could:
   1. come into seed production much earlier;
   2. produce large quantities of fruits or cones at lower heights above the ground; and
   3. be managed as short ‘seed-bushes’ rather than being let grow tall.

Large-scale improved seed production has occasionally been achieved in crop plants by using a male-sterile form for one parent. Then all the seeds collected from this parent will be crosses, not selfs.

What is the most important thing to remember when crossing trees?

To label and keep good records (C 54, C 64–66 in Manual 3)! Otherwise considerable progress could be made, but then the advantages lost.

What are the different ways that I could get the seeds I need?

1. By collecting them yourself;
2. Through exchanging seeds with another grower; or
3. By buying them locally or from abroad.

Occasionally there might be an opportunity to take part in local research or an international provenance trial, when seeds might be provided in return for growing them carefully and keeping simple records.

Which is the best way of ensuring regular seed supplies?

By exploring all three methods, because:

1. the same approach may not suit all the species you want to grow;
2. one method can act as a back-up in case another fails; and
3. you will be more likely to discover seed-lots, local races, provenances or species that could supplement or replace those you have been using.

Isn't it generally best to collect one's own seed?

Yes, it often may be, since you can have better control over:

1. which provenances and seed parents are used (B 21–22);
2. how and when the work is done; and
3. the way the seeds are handled after collection.

What kind of problems might I have?

Particularly if seed collection has not been done before, you would need to learn about and train staff and workers in:

1. looking out for flowers and ripening fruit (B 11–12);
2. making timely collections of good fruits or seeds (B 31) from sufficient numbers of trees (B 22), perhaps involving climbing;
3. separating and cleaning seeds (B 32) before sowing them at once (B 13) or drying and storing them (B 33); and
4. labelling the different seed-lots (C 54 in Manual 3), and keeping them separate.

Another kind of difficulty comes from lack of suitable fruiting trees, or of access to them.

What advantages are there in exchanging seeds?

1. You each receive some seeds without having to do the collecting and handling;
2. Seeds that are surplus to requirements need not be wasted; and
3. Both parties can obtain a greater diversity of seeds without the need for payment.

For example, a village nursery in northern Ghana, set up with training and modest funding from a non-governmental organisation, is now distributing seeds to neighbouring villages where the advantages have been noted.

Couldn't there be problems with seeds exchanged with someone else?

Yes; you would need to look out for:

1. low germination levels (B 30, B 48), because of restricted pollination, previous handling, transport conditions or because the seeds are old;
2. inferior genetic quality (B 20–23), because they were taken from few or unsuitable parents;
3. poor condition, because the seeds are mouldy, or contain pests or troublesome weed seeds; or
4. a lack of records about where they actually came from.

How could I get over such difficulties?

By building up a good relationship and agreed standards with other growers, so that you can all trust each other to:

1. provide good, viable seed of known origin; and
2. declare, and if necessary replace, any seed-lots that are below standard.

It might for example be possible to form a seeds co-operative, which might also handle seed storage and testing.

Both farmers have planted a boundary line.

Could I sell seeds if I had collected more than enough?

Yes, if there is a demand for them, and you can provide quality and information.

Is it really worth spending money on buying seeds?

If it can be done, it may often be very useful, for instance if:

1. there are no local seed sources of a particular species;
2. local collections have been unsatisfactory, or can be expected to be so;
3. you want to broaden the genetic base by including different provenances (B 21); or
4. improved seed from known parents could be obtained, or a hybrid tested (B 23).

Bought seed should generally be of high standard, recently tested for its germination percentage, and indicating any pre-sowing treatments (B 34) or special propagation conditions needed.

Wouldn't there be difficulties in getting seed from abroad?

Yes, there could be several problems, including:

1. knowing who has which kinds of seeds, though see Kindt et al. (1977) in B 52;
2. unavailability of the sort of seed ordered;
3. poor storage conditions in transit, such as very high or low temperatures (B 30);
4. bureaucratic delays at the port or airport, including import licences, phytosanitary certificates and extra handling charges; and
5. delayed arrival of the seed until after the best time for sowing (B 43).

If someone is travelling, bringing the seeds in their hand luggage is often the best option.

Why is keeping records of each seed-lot necessary?

Because:

1. then later on you can identify which of them germinated well and thrived, and which had problems;
2. plan your next seed collections on this basis; and so
3. lay good foundations for improving the genetic quality and health of your young trees, and the reliability and efficiency of your nursery operations (C 54, C 64–66 in Manual 3).