Return to METRIA 9

To Clone or Not to Clone: A Debate

Daniel K. Struve and Michael Knee

The Department of Horticulture and Crop Science
The Ohio State University
2001 Fyffe Ct.
Columbus, OH 43210

A paper from the Proceedings of the 9th Metropolitan Tree Improvement Alliance (METRIA) Conference held in Columbus, Ohio, August 8-10, 1996.

Knee: I am not sure I want to take sides in this debate; am I supposed to be the fall guy or the straight guy or the ignorant fool who's uninformed statements reveal the logic of your arguments?

Struve: No, perish the thought - I really am confused about the rights and wrongs of clonal vs seed propagation. I think it would be useful to discuss the arguments - we can each play devil's advocate without getting personal. By the way, nice tie. (Yeah right!).

Knee: OK, let's see whether a dialectical approach - thesis and antithesis - really does get us to a synthesis. First I think we should agree on some ground rules - what should we be concerned about when we try to decide whether to clone or not to clone, that is the question? Nice shirt. (NOT!)

Struve: Well my preference for clones is based on a singular issue: the members of a clone are genetically identical to the mother plant whereas members of a seed generated population are genetically variable, since they contain random assortments of the genes from their parents. The importance of genetic uniformity is that the individuals within a clone are phenotypically identical.

Knee: I guess that is generally true, although random mutations or sports can arise in clonally propagated material and we can expect progressive divergence over time if plants are propagated from different mother plants. The apple industry has made great use of sports and some of them are quite distinct from the original selection. For instance, there are over 150 strains of 'Red Delicious' apple. Furthermore seed-propagated material can be very close or identical to a parent plant if it is an inbred line or the result of strange, but natural, reproductive processes like apomixis. And don't forget about phase change characteristics, mature/juvenile traits (Figure 1).

Struve: (to the audience: "These are but wild and whirling words, my lord.") OK but we can agree that clones are usually more uniform genetically than seed propagated plants?

Knee: Well yes but is that all that is different? Is a tree grown from a seed going to be just like one from a cutting? What about their root systems?

Struve: I see where you are going and I don't want to go there, but I'll answer your questions as if rooting of stem cuttings were the only asexual propagation system (Figure 2). A seedling will usually have a main (tap) root with many lateral roots uniformly distributed about the main root. On the other hand a cutting on its own roots will have a shallow, less branched root system. But, despite the apparently inferior root system, plants originating from rooted cuttings have grown as well as seed-propagated material. However, the best of both worlds can often be obtained by grafting buds from a superior clone onto seedling rootstocks.


"Your bate of falsehood takes this carp of truth;
and thus do we of wisdom and of reach,
With windlasses and with assays of bias,
By directions find direction out."

Struve: Say what! ?

Knee: I think we are starting to compare apples and oranges; the cost of producing grafted trees will be greater than that of seedlings or own-rooted cuttings. Each of these propagation methods is likely to have its place in particular markets.

Struve: I'm not letting you get off that easily, I still want an argument; I mean discussion. Ask me about phase changes and juvenility. Wait, why am I setting my self up? You are supposed to be the straight man.

Knee: OK, a twenty year-old ash tree is going to have different foliage, bark type, limb structure, vigor, disease and insect resistance, and general appearance than a one year-old seedling. When asexually propagated, the mature characteristics of the adult phase will be maintained in buds taken from the twenty year-old tree; it may not be as fit (stress-resistant) as the seedling.

Struve: Good point, but I think we can deal with that. For one thing we can be careful to take buds from the lower part of the tree where the juvenile phase persists; if adult phase characteristics are undesirable. But many adult phase characteristics are desirable: lack of thorns, flowering, etc. Besides, when budded, many adult phase characteristics revert to a more juvenile form. And there is tissue culture as a method of asexual propagation.

Knee: Tissue culture induced phase change is well documented in the industry. Weeping birch is an example. Tissue cultured weeping birch don't.

Struve: The percentage of variants from tissue culture depends on the method of multiplication: axillary bud proliferation, adventitious shoot regeneration, or protoplasm fusion. Besides, most variants "out grow" the juvenile phase.

Knee: Yes, but there's nothing like reproduction through seeds to reset the maturation clock and restore the full vigor of the juvenile phase. I think that an organism propagated clonally may well get old; if nothing else it can accumulate viruses that depress growth. Most viruses are eliminated in the process of seed formation.

Struve: "He waxes desperate with imagination." I think you are raising specters to frighten us - I have yet to see evidence of aging of clones and there are some famous examples of seemingly immortal plants. I am assuming that propagation is done by competent nurseries who do not willingly or knowingly propagate diseases with their nursery stock.

Knee: In your opening argument, I sensed that you wanted to make a big thing out of the genetic uniformity. Why is that?

Struve: The great variation that exists in most woody plant species requires that clones be developed for use in urban sites. The resulting phenotypic uniformity is desirable and essential from an aesthetic/design perspective. Further, asexual propagation captures all of the genetic variation (additive and non-additive) thus maximizing the efforts from breeding and selection programs. Clones developed by the green industries have significantly improved the quality of the urban forest; instead of ill-adapted seedlings, the urban forest is composed of aesthetically pleasing, genetically uniform superior individuals.

Knee: Au contraire, that "great variation" is just what I want to preserve. We all know about the problems of monoculture in crop production. In agriculture it is possible to plant a different crop next year if this year's crop is wiped out by disease or cold. The urban forest is not an annual crop, unless you are growing Paulownia tomentosa in hardiness zone 5. Therefore greater precautions must be taken to assure genetic diversity. That is why we have the 30-20-10 rule (no more than 30% of the urban forest should be composed of a single family, no more that 20 from a single genus and no more that 10% from a species). I would like to add, that no more than 5% of any one clone. This rule is consistently violated, so I would not like to see a further expansion of clonal production.

Struve: Clonal propagation does not necessarily lead to monoculture. A monoculture is an establishment of trees with restricted genotypes over large contiguous areas. A large planting of a single species is not necessarily a monoculture. If we plant several clones of a species that originate from genetically diverse parents selected for uniform phenotype but diverse genetic background and wide adaptability and pest resistance, then the dangers of clonal propagation are greatly reduced, while maintaining phenotypic uniformity.

Knee: But the dangers are not eliminated.

Struve: "The professor protest too much, methinks." You always have to weigh the risks against the benefits. There are no guarantees. But let me finish my thought. Selecting a parent from the eastern range of a native North American species for rapid growth rate and crossing it with a drought resistant parent from the western portion of the specie's range has the potential for creating a population from which an individual with the best traits of both parents can be selected and introduced as a named cultivar. If this process is repeated with unrelated parents, thus generating many unrelated clones, then the diversity of the urban forest can be increased.

Knee: It also has the potential for creating a population with the undesirable traits from each parent. Do we really know which are the best trees for every situation? Isn't the urban forest itself the place where we find out the genotypes that survive best? Long ago I was told that "to get a tree you should plant three". Shouldn't the surviving trees form the breeding population? Besides, you have described a one generation breeding program; a terminal breeding program. How will further improved cultivars be introduced?

Struve: Recurrent selection from a continuously improved genetic base can result in a constant supply of improved clones (Figure 3). In this breeding and selection scheme, a broad genetic base is maintained in the breeding population. Only the few truly superior individuals are selected from all the progeny of the crosses for introduction into the trade as named cultivars. Then the best individuals from the most recently created population are selected as parents for the next round of crosses.

Knee: Good theory, but what happens in practice? Aren't most clonal introductions selected from nursery environments or isolated individuals along fence rows? Remember, phenotypic superiority does not ensure genetic superiority; especially in hardwoods which are extremely site sensitive.

Struve: Save for a few government sponsored breeding programs, you have described the traditional method of new cultivar introduction.

Knee: Something is rotten in the state of cultivar introduction. Well, maybe but I think you will find that selection breeding programs all have one inherent flaw; they eventually run out of variation. We need to keep that reservoir of genetic variation as wide as possible through seed propagation.

Struve: True there are some problems with recurrent selection. Inbreeding, the loss of genetic variation, is unavoidable within a recurrent selection breeding program. However, inbreeding can be managed by adopting breeding strategies such as sub-line breeding, introducing new genetic variation from unselected populations and genetic engineering. I am not saying that the nursery industry uses breeding programs, just that the appropriate technology exists and could be adopted. An additional benefit of clonal propagation is that it can exploit the special advantages of root stocks, such as size control and resistance to biotic and abiotic stresses.

Knee: Dwarfing root stocks are already being used, although inadvertently; I'll refer you to our later discussion on "Wolf Trees" . The point is, we are planting half-clones; phenotypically superior scions grafted on root stocks of unknown genetic potential.

If the green industry took to heart the principle of genetic diversity and adhered to the 30:20:10 rule, the genetic quality/diversity of the urban forest would be improved. Monocultures are to be avoided.

Struve: Yes they are, but their existence is not from design or ignorance of biological principles. Rather is results from biological limitations such as low root regeneration potential in nursery stock, absence of asexual propagation methods or complex seed dormancy. In short the species in the urban forest are the ones that transplant readily and are easy to propagate. An urban forest of drought resistant species (which are difficult-to-transplant) will remain a dream, if they can not be successfully transplanted.

Knee: Some of the biological limitations are being overcome by new production methods.

Struve: Now, I would like you to admit some of the problems with sexual propagation like yearly variation in seed production. Asexual propagation is more likely to yield a consistent yearly supply of propagules, unlike seed propagation which is subject to "alternate bearing" habit common in unselected genotypes.

Knee: "Therefore, since brevity is the soil of wit,and tediousness the limbs and outward flourishes,I will be brief."

Before we leave asexual propagation, I want to ask some summary questions about actual industry practice for clonal introductions

Struve: My silence is your answer. But before we leave asexual propagation, the prevalence of clonal planting stock has occurred because of economic incentives. There were few clones available before 1949 when the Plant Protection Act was used to patent 'Moraine' honeylocust. Licensing fees are a significant driving force in new cultivar development.

Knee: The same economic gains can be realized with sexually propagated crops: vegetable and flower seed producers have made a business of it.

Struve: Whether you'll admit it or not, the overall quality of nursery stock now being grown is higher than 20 or 30 years ago. I say that the increased quality resulted from the economic incentive of patenting plants. Now, about that yearly variation in seed production?

Knee: "O, my offence is rank, it smell to heaven."

The key to preventing boom and bust seed years is, of course, unselected and unmanaged seed sources. There are management practices that reduce alternate bearing, the apple industry uses them all the time. We don't have to accept the boom and bust cycle in seed supplies. Besides, extra seeds from good seed years can be stored for use when seed production is low. If clonal propagation is so much better we should start a research program on vegetative propagation of oaks.

Struve: Wait we are taking seeds. Storage protocols have not been developed for all species, especially recalcitrant species like oak which can not be stored without reducing seed quality.

Knee: For recalcitrant species, extra seedlings can be produced in good years and used in the poor years. The seed can be "stored" as seedlings. There can also be great yearly variation in clonal stem cutting rooting response. And what about grading of rooted cuttings according to root system quality; only cuttings with many well distributed roots should be grown on. Just because a cutting roots does not make it commercially acceptable.

Struve: Can we stick with the disadvantages of seed production? Realistically, how are seed supplies managed? Are there seed orchards? Are seed lots identified by source? In my younger years I picked seed from river birch trees in North Carolina, sold them to a seed dealer in the Northeast, who sold them to a seedling grower in Oregon (Figure 4). I can only guess where the seedlings were lined out and in what landscape they were planted. Also, and I'm not proud of this, I have collected seed from wolf-trees.

Knee: There may be practices that need adjustment. However, progressive seed suppliers and nursery managers are offering source identified seed and planting stock. In most forestry nurseries, source identified planting stock is standard practice.

Struve: So why are seeds collected from putative undesirable genotypes, wolf trees? Also, what is a "wolf tree".

Knee: Wolf trees are short lived, typically of poor form but have abundant and frequent seed crops. Unfortunately, seedlings produced from these seeds tend to inherit the undesirable characteristics of their parents (Figure 5).

Struve: So why are seed collected from wolf trees?

Knee: It's an economic issue. Seed collectors are paid by the number of pounds of seed that they collect; the faster the rate of collection, the greater the rate of pay. For instance, why climb to the top of a large phenotypically superior tree to collect seed, when you can stand on the ground and collect seed from a wolf tree? Unless there is a price differential between seed lots reflecting their genetic superiority there is no economic incentive to risk life and limb. Until nursery managers or landscape architects recognize the value of parentage and its effect on nursery profitability and the design aesthetics, there will be no change in seed collection practices. Bottom line: genetically superior seed pays for itself although the production cycle and into the landscape.

Struve: Genetically superior seed is a nice concept, but is it practical? Can a nursery manager profit from genetically superior stock?

Knee: In a study with birch you showed that the average price per tree ranged from $6.00 to $36.60 per tree depending on the seed source.

Struve: I also demonstrated the value of asexual propagation in capturing all the genetic variation. Clonally propagating the best individual within the best family returned $56.60 per tree as opposed the growing seedlings of the best family which returned $36.60 per tree.

Knee: Yeah, well. But don't forget the extra cost of asexual propagation. That difference is not all profit.

Struve: So what is the proper way to manage seed resources?

Knee: The best method is to establish a seed orchard, a group of trees managed exclusively for frequent and abundant crops of genetically superior seeds.

Struve: So, you establish a seed orchard (Figure 6). That is no guarantee of genetically superior seed. Doesn't it depend on the selection criteria applied to the seed orchard trees?

Knee: True, genetic superiority is not assured by selecting phenotypically superior individuals as seed orchard parents. Genetic superiority is assured only after progeny tests. In a progeny test, the genetic worth of an individual tree is determined by the performance of its progeny. Based on the progeny test results, the inferior individuals are removed from the seed orchard. Genetically superior seed is assured only if the mother trees are genetically tested.

Struve: How many mature trees of a given species are needed to supply the annual needs of the U.S. landscape industry. I have read estimates that one mature sweetgum tree can produce 20,000 plantable seedlings per tree per year. Just a few trees could supply the needs of the entire U.S. landscape industry. What about genetic diversity? Is it advisable to have the sweetgum genetic base for the U.S. to be represented by six to 10 trees? How narrow of a genetic base will you accept for a species? And are different seed orchards needed for different regions of the country?

Knee: True, just a few mature individuals of many species would supply the planting needs for the U.S. But the genetic base can be easily broadened by planting more trees. And yes, different seed orchards probably are needed for different regions of the country.

Struve: Inclusion of greater number of parent trees is desirable but, what about the increased cost of managing additional seed orchard trees needed to assure a broad genetic base?

Knee: I offer dwarf apple trees as an example of how to increase the number of individuals per acre (and potentially the genetic diversity) without significantly increasing seed orchard acreage. With dwarfing root stocks, many trees per acre can be planted, thus increasing the number of individuals per acre without increasing the size of the orchard.

Struve: But dwarfing root stocks are not available for commonly used landscape species.

Knee: Dwarfing root stocks are widely used in the nursery industry, but inadvertently so. You remember those clones you liked so much? Well, we are really planting half clones when seedling root stocks are budded. Seedling root stocks derived from wolf trees.

Knee: "Let me see. Alas, poor Budgraft. I made h'm, Dan, a graft of infinite life, of most excellent fancy. He hath promised me profits a thousand times. And now how abhorred in my imagination is it? It failed, a thousand times failed. Prithee, Dan, tell me one thing."

Struve: "What is that my friend?"

Knee: "Dost thou think 'Bradford on Bartlett' looked o' this fashion i'th' earth?"

Struve: "E'en so."

Knee: "And smelt so? Pah?"

Struve: So, the genetic quality of the root stock is as important as the genetic quality of the scion.

Knee: I'm glad you learned something today.

Struve: Are there other ways to draft seed orchard trees?

Knee: Seed orchard trees can also be dwarfed by chemical and mechanical means. For instance, an apple tree study found that application of chemicals significantly reduced tree size, without decreasing seed yield, size or quality. Seedling size and quality was also unaffected by growth regulator application to the mother tree. Restricting root volume is a mechanical dwarfing method. Container production could be incorporated into a mechanical dwarfing system. There are other advantages to dwarf seed orchards; dwarfed trees are easier to manage: pesticide application and seed harvesting are more easily done on small trees than on standard sized tree.

Struve: So you are back to picking seed from wolf trees.

Knee: Trees with wolf-like phenotypes, but not genotypes.

Struve: But how do you control pollination? 400' wide pollen isolation zones are required to dilute wild pollen. What about adjacent seed orchards of naturally hybridizing species, such as oaks? These isolation zones require large acreage and are expensive to maintain.

Knee: Dwarf seed orchard trees create many options. Dwarf trees could be grown in Pot-N-Pot system within a polyhouse. In spring, passive solar heating would accelerate development, bringing seed orchard trees to anthesis before native plants, accomplishing pollen isolation by time, rather than distance. The plastic skin and house orientation would serve as physical pollen barriers allowing hybridizing species to be grown adjacent to each other.

Struve: Well, I hope I have convinced you that genetically superior and diverse sexually propagated plant material can be economically produced and can be an essential component of the urban forest.

Knee: And I hope I have convinced you that genetically diverse, asexually propagated plant material can be an essential component of the urban forest.

Return to METRIA 9

Format updated July 22, 2008