Printer Friendly

Balancing investment incentives and social benefits when protecting plant varieties: implementing initial variety systems.

THE APPLICATION of intellectual property rights (IPR) to plant varieties affects a number of groups with different priorities: researchers, growers, seed companies, and consumers, making it a complex matter to strike the optimal balance of interests. Further complexity is added by the variability possible across plant species, time, and the sequential nature of most breeding programs, whereby the best current varieties are sought as the base material for breeding subsequent improvement. Changes in legislation regarding the application of IPR to plant varieties have also affected the interactions among public and private breeders. As a consequence of these multiple factors, several distinct forms of IPR for protecting plant varieties have developed over time. In this paper, we consider the functionality of a recent (1991-1994) change in IPR legislation intended to provide an additional incentive for one form of plant breeding. For background, one must consider the development and effects of past and current forms of protection for plant varieties

Alternative Forms of Protection for Plant Varieties

Protection for plant varieties has been slow to develop. Worldwide, application of all forms of IPR were implicitly or explicitly excluded from agricultural plants throughout the 19th century. In the USA, a specific form of protection for plant varieties, the Plant Patent Act, was adopted only in 1930, and that protection was and is limited to asexually propagated plants. In Europe, early versions of protection for all plant varieties began to be adopted in the 1940s, with the International Convention for the Protection of New Varieties of Plants (UPOV) established in 1961. That system of protection, known as Plant Variety Protection (PVP), was first adopted in the USA as the Plant Variety Protection Act of 1970, with the most recent amendments made in 1994 when the U.S implemented the UPOV 1991 Act. The use of utility patents for protecting plants was facilitated when, in 1980, the U.S. Supreme Court in Chakrabatry declared that "anything under the sun that is made by man" is patentable subject matter. That case was in specific reference to microorganisms. However, in 1985 in the internal decision ex parte Hibberd the U.S. Patent and Trademark office (USPTO) extended patent protection to plant varieties, a decision recently upheld by the Supreme Court in Pioneer vs. J.E.M. Hence, in the USA, there are currently three means of protecting plant varieties: (i) Plant Patents (PP, which covers only asexually propagated varieties), (ii) Plant Variety Protection (PVP, which covers all sexually or asexually propagated plant varieties, pure lines and the hybrids they produce), and (iii) Utility Patents (UP, which covers all plant varieties, including pure lines and their hybrids, if included in the specific claims).

Plant Patents are limited to the USA, but the 54 members of UPOV all operate similar PVP systems (membership list at; verified 2 March 2004). The use of UP for plant varieties is more complex to establish but is generally limited to the USA, Japan, and Australia. Of these methods, PVP and UP are the most general and are currently extensively used. Since these PVP and UP are not limited to the special case of asexually propagated varieties, as are plant patents, and since most major crops are seed propagated, this analysis will focus on PVP and UP.

Limitations of the Existing Forms of Protections

PVP was developed in concert with UP, but with several fundamental differences to accommodate the particular characteristics of plants. A key distinction between PVP and UP protection of plants is the express allowance under PVP for using PVP-protected varieties as breeding material without requiring the permission of the rights holder (Article 15.1(iii) UPOV 1991 Act). With patents, the so-called research exemption is limited and depends on the specific claims in the patent itself. The claims in some recently granted U.S. utility patents on plant varieties appear to have been drafted expressly to prevent breeding use (see Lesser and Mutschler, 2002). The interest of a seed company in protecting its germplasm base, as represented by its current varieties, from use in breeding by other programs is a logical one. Breeding progress often proceeds by serial improvements on the current optimal materials so that companies' germplasm can be their major asset. However, in the view of many plant breeders, the progression of plant breeding is restricted if breeding access to varieties is prohibited, or even questionable. The International Seed Federation (ISF) (2002) declared itself "strongly attach[ed] to the [research] exemption provided for in the UPOV Convention." For many then the use of patents for protecting plant varieties is problematic. The breeding exemption under PVP is therefore a public benefit compared with UP.

As noted, the PVP research exemption specifically allows the use of protected varieties in breeding programs. The resultant new variety, if distinct, would be noninfringing and may qualify for protection in its own right. A trait valuable in multiple varieties, for example a virus resistance in soybeans, can therefore legally be utilized in a competitor's breeding program. This prevents the originating company from reserving the trait for its sole use, but also limits the opportunity of the originating company to recover the investment in the initial incorporation of the trait into soybeans from a wild species.

The process of the initial transfer of a trait from a wild or agronomically inferior donor source is called "prebreeding" or "germplasm enhancement." Prebreeding entails considerable cost and risk. The transfer of the desired trait might not produce usable materials at all. Even if the prebreeding program does succeed, the incorporation of the trait into a useful variety could require 10 or more years, depending on the complexity of the trait, the sexual barriers between the donor and the crop species, and the possible need to break linkage between the gene(s) controlling the desired trait and linked gene(s) controlling undesirable traits. However, once the resistance has been incorporated into a commercially acceptable variety, a competitor could readily transfer it to a new variety in very few breeding seasons, since the trait already resides in an agronomically superior line. Therefore, the initial breeder who invested in the prebreeding program to transfer the novel trait has limited time and opportunity to recover the benefits of this investment, while the competitor uses the trait with considerably less risk or investment of time and resources. The ISF (2003) describes the situation this way, "Plant breeders are particularly in need of such protection [from close copies]:

* because their investment is necessarily expensive ...;

* because of the ease with which many varieties [...] can be multiplied, once they exist. [...] It is the nature of living things to reproduce themselves.

Without appropriate IPR protection, as soon as a promising new variety appears on the market, it will be bought up and multiplied by seed traders to sell on, or by farmers for their own use. In consequence, the original breeder will never be able to sell enough of the new variety at a price that will enable him to cover the cost of the years spent breeding it."

As a result of the lack of financial incentives, prebreeding efforts are largely limited to the public sector, with the exception of the most valuable crops, and benefits to society from the transfer of novel crop traits to the public occur more slowly. This then is a limitation of PVP. [For more detail on the application of IPR to plant varieties, see e.g., Rories (2001) or Sechley and Schroeder (2002)].


Overall then, there are issues or limitations in the use of either PVP or UP as has been applied to plant breeding. Research access is restricted by UP, as they have recently been used, while the use of PVP results in a system that may not provide sufficient incentives for higher risk breeding or prebreeding programs. Recognizing that situation, the parties to UPOV adopted a hybrid version of protection in the 1991 Act, which retains the research exemption while introducing the concept of improvements used in patent practice. An improvement patent is one which adds to the technology of a prior patent. In many instances, the patented improvement cannot be made, used, or sold without infringing the basic patent so that the inventor of the improvement is "dependent" on the owner of the prior patent to grant permission for its use (Silberman, 1995). Clearly, once the prior patent has expired or is otherwise no longer in effect, use permission is no longer required. Under UPOV, improvement similarly involves adding to prior art (in the form of existing varieties) but, unlike UP, physical derivation and a proportion of common genetic material in the prior and improved varieties are required as well.

The purpose of this article is to explain why the UPOV approach to improvement is ultimately unworkable. We then consider several alternative forms of implementation that overcome the identified limitations of the UPOV approach and recommend one specific alternative approach. The proposed approach applies specifically to PVP and does not involve UP except to the extent, under an identified (but not recommended) approach, a research exemption might be added under UP.

Attention here is not focused on the use (or misuse) of protected varieties as a vehicle for delivering bioengineered traits through the introduction of particular gene constructs. If that were the sole basis for the modification of PVP, then a direct approach of disallowing genetic transformation as a means of creating distinctness would suffice. Indeed, that is what Article 14.5 (c) does (see below).

The Improvement Concept under UPOV

Relevant Articles

To consider the concept of improvement adopted by UPOV, and the implication of applying this concept in varietal protection, it is necessary to consider the exact language in the relevant articles of the act. Dependency is set out in Article 14.5 of the 1991 Act, which reads as follows in its entirety (available at

(5) [Essentially derived and certain other varieties] (a) The provisions of paragraphs (1) to (4) shall also apply in relation to

(i) Varieties which are essentially derived from the protected variety, where the protected variety is not itself an essentially derived variety,

(ii) Varieties which are not clearly distinguishable in accordance with Article 7 from the protected variety and

(iii) Varieties whose production requires the repeated use of the protected variety.

(b) For the purposes of subparagraph (a)(i), a variety shall be deemed to be essentially derived from another variety ("the initial variety") when

(i) it is predominantly derived from the initial variety, or from a variety that is itself predominantly derived from the initial variety, while retaining the expression of the essential characteristics that result from the genotype or combination of genotypes of the initial variety,

(ii) it is clearly distinguishable from the initial variety and

(iii) except for the differences which result from the act of derivation, it conforms to the initial variety in the expression of the essential characteristics that result from the genotype or combination of genotypes of the initial variety.

(c) Essentially derived varieties may be obtained for example by the selection of a natural or induced mutant, or of a somaclonal variant, the selection of a variant individual from plants of the initial variety, backcrossing, or transformation by genetic engineering.

Interpretations of Article 14.5

Interpretation of Article 14.5 must consider the foundation laid by the prior sections. Subsections 14.1-.4 set out conditions for infringing PVP, that is, those acts such as selling which require the permission of the breeder to perform legally. Hence, 14.5(a) states that essentially derived varieties also require the permission of the breeder for infringing-type acts, such as selling, exporting, etc. Under 14.5(b)(i), an essentially derived variety can have but one initial variety. Moreover, if B is derived from A and C from B, then both B and C are essentially derived from A if both B and C express the essential characteristics of A. With protection typically lasting 20yr, it can be very lucrative for a line to be identified as the initial variety (A).

A variety, once essentially derived, is always essentially derived--that is a physical characteristic. However, an essentially derived variety is infringing only if its initial variety is protected. No infringement exists if the initial variety was never protected, or if its protection lapses. The essential derivation literature uses the term "dependency" to refer to those essentially derived varieties which require permission for commercialization, a nomenclature we follow here. For terminological convenience here, an infringing improvement patent will likewise be referred to as being "dependent."

Beyond these key articles, the 1991 Act provides no direction on how, in practice, to identify essentially derived varieties. Other UPOV documents have made the following related points:

"All related parties agree that the dependency relations should be handled by the breeders themselves, without the involvement of the authorities responsible for administering the plant variety protection system." (UPOV, 1996, Annex).

That is, the determination of essential derivation is to be treated differently from the granting of PVP certificates. Most countries have crop committees that establish the characteristics in which distinctness (one of the requirements for protection) will be recognized in deciding if a PVP is to be granted, and may even adopt statistical standards compared to a 'reference' variety (for a description of test practices, see UPOV, 2002). For example, onions may be (among other characteristics) distinguishable by resistance to sprouting in storage, with for instance 1% fewer sprouts after 4 mo in storage required to establish distinctness.

According to the UPOV interpretation, initial varieties are not to be established in a similar manner. Rather, the holder of a variety he/she believes to be an initial variety will notify the owner of the allegedly dependent variety, with a solution to be worked out through negotiations or, if need be, the courts.

"... essential derivation would not be based only on characteristics used for distinctness. In addition, genetic conformity was not the only criterion for the judgment of distinctness." (UPOV, 2000).

"... a variety should only be essentially derived from another variety when it retained virtually the whole genotype of the other variety." (UPOV, 1992, par. 8).

"The essential characteristics are those which are indispensable or fundamental to the variety. 'Characteristics' would seem to embrace all features of a variety including, for example, morphological, physiological, agronomic, industrial and biochemical characteristics. 'while retaining' requires that the expression of the essential characteristics be derived from the initial variety." (UPOV, 1992, par. 9).

Together, these interpretations indicate that the criteria for the establishment of initial variety status are four-dimensional. An essentially derived variety must (a) express the characteristic(s) used in establishing distinctness of the initial variety, (b) express the other major phenotypic features of the initial variety, and (c) retain almost all the genotype of the initial variety. And finally, (d) the essentially derived variety must be physically derived from the initial variety (Hunter, 1992).

Comparing essential derivation in PVP with dependency in patents, only requirement (a) above applies for establishment of dependency under patent statutes. That is, a subsequent improvement invention that displays the nonobvious characteristic of a patented product would be dependent; physical derivation and the expression of the other characteristics of the initial patented invention are irrelevant for determining dependency under patent statutes. The underlying reason for this difference is that PVP protects the whole plant, or more formally the particular unique combinations of traits embodied in the initial plant variety, while patents protect individual unique characteristics themselves. This distinction has major consequences for the operation of initial and essentially derived variety systems.

As an example of the consequences of the UPOV-suggested approach, Troyer and Rocheford (2002) explored selecting essential derivation among U.S. maize varieties on the basis of a threshold of 75% relatedness, the tentative American Seed Trade Association threshold guideline. They determined that many closely related inbred lines exist because the major firms all began with the same materials from public breeding programs and subsequently used similar breeding methods. However, when considering performance of closely related varieties, "Evidently introgression of a small amount of DNA can have significant effects on agronomic performance. The quality (positive effect) of introgressed DNA improving useful agronomic traits is more important than quantity of DNA similar to the IV [initial variety] parent." They conclude that, for hybrid field corn, a genetic similarity threshold of 90% is a more appropriate standard for essential derivation than the proposed 75% threshold.

Assessment of Forms of Protections

Functioning of Current System

It is now possible to consider how the initial variety system could function in the cases of two different types of plant breeding innovation. Both are common means of producing improved varieties and so are relevant to assess in the context of initial-dependent varieties. Consider first the case of breeding a resistant variety of soybean by transferring a simply inherited single-gene based virus resistance from a wild relative into a commercial variety. The originating company invests the resources needed to transfer the resistance from a wild species to an agronomically acceptable variety. The resistance would be the distinctness characteristic of this new variety. If the resistance is commercially valuable. that is, significantly reduces risk of crop loss and the cost of disease controls and/or increases yields, then competing breeders will wish to make the resistance available in their new varieties as well. Indeed, depending on the need for the new trait, the other companies may have to include the disease resistance for their new varieties to be competitive so they can retain or increase their market share. This secondary transfer of the resistance gene is easily accomplished by use of the initial resistant variety as a germplasm source in the second breeding program. If the resultant varieties qualify as dependent, the outcome is that the initial breeder will receive royalties from numerous competitors using the viral resistance over the life of the PVP on the initial variety. This is exactly the case intended by UPOV--the rewarding of prebreeding. However, whether the second variety is dependent depends not only on the presence of the resistance gene, but also on the determination of the relatedness of the new variety to the initial variety, and the relatedness standard used.

In the following discussion of relatedness, we are not concerned with the procedure by which relatedness is actually measured. The means of measurement is critical to the actual operation of the system, but for the present, without loss of generality, it is possible to abstract from that level of detail.

If the relatedness standard is set high--say 90%--it would be relatively easy to use the initial resistant variety in a breeding program and select for virus resistant varieties that do not meet that threshold of relatedness, and thereby avoid dependency. As the threshold for relatedness is increased, the ease with which dependency is avoided increases. Because the owner of a nondependent variety is not subject to royalty payments and indeed can commercialize the variety without restriction, there is a definite incentive to avoid dependency. That avoidance might be done through making additional crosses, or selecting against common molecular markers. This approach would lead to the selection of genetic material of no practical value other than reduction of relatedness. Such cosmetic breeding rewards the company for accomplishing a breeding goal that does not contribute anything of practical agronomic value, so is socially inefficient. It also reduces royalties to the initial innovating breeder, thereby reducing the potential benefits to society that intellectual property protection is intended to generate. Therefore, for the use of the concept of dependency in PVP to function as intended, the threshold for relatedness level would have to be set relatively low. This is counter to the interpretation by UPOV (1992, Par 8--see above).

Moreover, the timing of the determination of dependency is out of phase with the timing of establishing and executing a breeding plan. The planning occurs very early in the breeding program as parents are selected, while the determination of dependency is made on the release of the new variety. Since it will not always be possible to know beforehand which variety owner will claim initial variety status, breeders will have the incentive to use only their own or unprotected varieties in breeding programs (recall, dependency exists only when the initial variety has active protection). The result is either a break in the process of building new varieties from the best of existing varieties or increased investment in cosmetic breeding to circumvent the threshold of relatedness. Neither is socially desirable.

A further complication arises when considering the potential effects which the determination of initial varieties and dependency have on the next generation of varieties--the consequences of the system on "pyramiding" or breeding for varieties that combine two or more new traits. Consider the case above where a breeder has added resistance to virus 'X', producing a variety that achieves initial variety status. Subsequently, another breeder transfers resistance to equally devastating virus 'Y' into the same crop. The variety produced by the second breeder can become an initial variety, but only if it does not use the initial variety to include resistance to X in the breeding program, rather transferring Y resistance to a typical X-susceptible background. If the second breeder wishes to use the resistance to 'X' as well, to combine the benefit of resistances to both viral pathogens, the new variety would likely be dependent because of its inclusion of resistance to 'X.' Therefore, if only a single initial variety is permitted, the breeder accomplishing the transfer of 'Y' resistance will receive no additional reward for this achievement. Hence, there is either no incentive for competitors to engage in further pre-breeding, or there is no incentive to produce varieties that combine more than one novel desirable trait. As a result, the optimal commercialization of the results (resistance to both virus X and Y in a single variety) may be withheld until protection of the initial variety lapses (which can be 20 yr or longer).

Note further, the owner of the initial X-resistant variety may have incentive for additional background breeding so long as X and Y resistance can be linked. Then, the royalty fee for X can be raised to compensate for the added value of Y resistance. That is, initial variety status for a single improvement can be leveraged into a dominant position in new variety production. Indeed, the initial variety owner also may license Y resistance from a second breeder, but the terms would not be expected to be favorable to the latter firm for it would have little alternative opportunity to benefit at least until the X-resistant variety protection expired. If, however, through breeding, Y resistance can be separated, no breeder has an incentive for additional prebreeding. None of these cases is desirable from a societal perspective.

Much of the effect of initial variety legislation depends on the threshold of relatedness imposed. Consider the second case where the distinctness trait is quantitative in nature, and probably due to multiple genes--say a modest but measurable decrease in onion sprouting in storage. If the relatedness threshold is set moderately low (75%), then the new varieties produced by breeders who use the initial variety in their program will likely be dependent. Breeding around a low degree of relatedness is impractical in this case, so the competitor will have an incentive to exclude the new variety from the breeding program.

Conversely, if the relatedness standard is set high (90%), then competitors have a ready option to breed around it, leading to the same inefficiency noted above. Under this scenario though, cosmetic breeding probably imposes less inefficiency than the avoidance of existing protected varieties, leading to a preference for higher relatedness levels. [This seems to be the case Troyer and Rocheford (2002) are describing.] The inherent complexity is that the degree of relatedness required for dependence must change according to the importance of the contribution of the distinctness trait; the two should vary inversely. That though would be a hugely complex system to manage in the public interest.

Finally, the relatedness level can be set very low, say 20%. However, that would mean an essentially derived variety could be dependent on up to five initial varieties, in clear contradiction to the current intent under Article 14.5(b), as interpreted by UPOV (1992, par. 8).

Identifying Different Forms of Genetic Improvement

If the quest for a means of applying essential derivation seems particularly convoluted, there is an underlying reason. The two examples of enhancements given above--virus resistance and reduced sprouting--are quite distinct in source and heritability of the respective trait. The virus resistance example involves the introduction of a wholly new trait, for this particular species, and a trait that is likely to be simply inherited (largely controlled by one or very few genes). Here, we refer to this type of trait as a discrete trait. When one sexually introgresses a single gene trait into new background, the size of an introgression carrying the gene, and so the amount of DNA the introgression contains, can vary widely, as demonstrated by Young and Tanksley (1989). However, even the largest of these introgressions represents a small proportion of the total genome of the plant. Therefore, since the transfer of such discrete traits does not require the transfer of a significant amount of genetic material, there is no need for a standard of the proportion of relatedness. At one extreme, this situation is the closest to that involved in the consideration of GMO plants. However, we do not consider the case of genetic transformation here, since for GMOs, there is a more direct way to define essential derivation than that in Article 14.5.

The second example, sprout resistance in onions, is an incremental improvement in a trait already possessed by varieties of the crop species, similar to other important traits such as yield, days to maturity, and fruit size. Although there have been instances in which transfer of a single gene or QTL can have a substantial incremental effect on such traits, in most cases, these traits are controlled by multiple genes, or can be the result of multiple combinations of genes, and hence inherited in a more complex manner. We refer here to those cases as complex traits. Documenting derivation for such complex traits could involve a measurement of the proportion of genetic material, which should be set sufficiently low to discourage cosmetic breeding (i.e., 75 rather than 90%).

Attempting to identify a single approach (as the UPOV system does) that accommodates such disparate forms of variety improvement is clearly problematic. We turn now to an analysis of alternative approaches to those proposed by UPOV.


In summary, the proposed UPOV essential derivation system provides perverse incentives--positive financial incentives for breeders which lead to socially inefficient outcomes. Those perverse incentives include the following.

1. Incentives for nonproductive cosmetic breeding if the relatedness level is set high (as would typically be needed for discrete traits) because the genetic composition of the base and improved varieties is so similar.

2. Incentive to use one's own or unprotected varieties in a breeding program rather than the best available one if protected as it could subsequently be established as an initial variety.

3. Lack of incentive to add (pyramid) additional useful traits onto a dependent variety, reducing the amount or introduction rate of useful traits.

4. Practical considerations dictate that the relatedness level be fixed by species, which is nonoptimal as (i) more important contributions should have a lower level of relatedness to restrict cosmetic breeding and (ii) complex traits would necessarily require a lower level of relatedness which means a single standard must be either too low or too high, depending on whether the trait is single or multigene based.

Alternative Approaches

Alternative 1: Establish Distinctness Standards for Initial Varieties

UPOV (1996, Annex) proposes essentially derived status should be established privately among the affected parties. One alternative is for an agency (probably a plant variety protection office currently responsible for establishing distinctness standards) to set minimum standards for qualification as an initial variety. For example, if a 1% lower sprouting rate were set as a critical threshold for distinctness in onions, then perhaps 5% would qualify for initial status. That is, via initial variety status, a broader scope of protection is granted for a larger contribution to production value. Australia, in a variant of the proposed approach, presently requires the secretary of the variety rights office to declare a variety essentially derived; the burden of proof for establishing initial variety status lies with the variety owner (Plant Breeders' Rights Act of Australia Articles 12(c) and 40.5, available at; verified 2 March 2004).

This approach would be especially applicable to complex traits. Minimum distinctness standards would resolve the problem of granting initial variety status on a variety that makes a limited contribution. However, unresolved remains a means to provide an incentive for subsequent major advances within the same initial variety.

Alternative 2: Drop the Relatedness Requirement

This approach would make the distinctness requirement the only one for essential derived status. That is, if a variety expressed the distinctness trait of an initial variety, it would automatically be essentially derived, closely paralleling improvement inventions under patent law. The requirement for direct descent (use of the initial variety in the breeding program) could be retained, so that independent discovery would be allowed, unlike patents.

This approach is particularly applicable to discrete traits, those that by definition did not previously exist and hence any detectable expression would constitute a substantial change.

Alternative 3: Allow for More Than One Initial Variety

This approach would provide an incentive for the inclusion of more than one significant new trait in a dependent variety, and help preserve the important process of cumulative breeding. However, this approach would effectively necessitate the removal of the relatedness requirement and hence is applicable for discrete traits only.

Alternative 4: Adopt a Special Research Exemption for Patented Plants

Under this approach, the initial-derived variety system would be dropped from PVP and incorporated into the patent statutes. What would exist would be two distinct types of protection for different magnitudes of nonobviousness. PVP legislation (shorn of essential derivation) would apply to incremental improvements (1% reduced sprouting), while greater enhancements (5% reduction in sprouting) would be protectable under UP.

Technically, precedent exists for different forms of protection for different degrees of inventiveness. Some countries (Germany and Japan among others but not the USA) have both utility patents and petty patents (also called utility models). Petty patents provide briefer terms of protection with lower nonobviousness standards. In the current context of PVP, existing PVP systems shorn of initial varieties would serve proximately as petty patents while UP would apply to initial varieties.

To allow for access to protected varieties for breeding purposes, a statutory research exemption would need to be added. At present, the research exemption under patent law is based on case law, and hence uncertain and open to different interpretations. A specific research exemption clause for plants would resolve that matter. There is precedence in patent statutes for clauses specific to individual classes of products. For example, the U.S. Patent Act (Articles 162 and 112) for plant patents requires a description "as complete as is reasonably possible" while that for other products must be "full, clear, concise, and exact."

If such amendments to patent law were legislatively possible, what would the effects be on the incentive for plant breeding? Patent law has no relatedness requirement so that this approach would follow Alternative 2 above. There is also no limitation on a single dependency under patent law so that Alternative 3 would be incorporated as well. As patents are examined, among other steps, the claimed nonobviousness is compared with the standard for the product class. Since the Patent Office has discretion over the minimal degree of nonobviousness, a relatively high level could be adopted, which resolves the consequences identified above under the UPOV-type system of potentially granting initial variety status for modest distinctness contributions.

Several additional caveats need to be considered for the protection of plant varieties under utility patent law. One is that farmer seed saving is not expressly permitted under patent law, whereas it is allowed under UPOV systems. The so called "farmers' privilege" is absolute under UPOV 1978 for it is not included in forms of infringement, but a national option under the 1991 Act (Article 15.2). However, the USA and most countries do permit unlimited seed saving.

The economic consequences of forbidding seed saving are somewhat complex in advanced agricultural systems, and would vary depending on the crop considered. For small developing-country farmers, allowing seed saving is considered essential, but then most of the seed used in those environments would not be patented. There is no requirement under patent law to show direct derivation (the use of the protected variety in the breeding program). If direct derivation were considered a vital part of determining infringement, that condition would have to be added. Finally, it should be noted with some concern that in current U.S. UP for plant varieties, in which corn and soybeans dominate, the non-obviousness standard appears to be low (see Lesser and Mutschler, 2002). This, also, would have to be adjusted as well for the proposed system.

Using UP to protect germplasm enhancement would be operational for discrete traits, which can be conveyed with a limited amount of genetic material. That, however, is not the case with complex traits, so the solution is only partial. Moreover, UP are relatively costly to prepare, file, and manage compared with PVP, which would place an economic burden on the breeders of minor crops. Overall, the application of UP for essential derivation of plant varieties is perhaps too complex to be warranted.


The principal point made here is that the initial-essentially derived variety system, as proposed by UPOV, where the status of a variety is worked out solely between the parties, is unlikely to work. The complex interaction between distinctness and relatedness of plant varieties suggests that, at a minimum, some oversight body must be involved to establish consistent standards. But the larger problem is that the UPOV approach attempts to provide protection for two very different forms of genetic enhancement, discrete and complex, with a single system. As the alternatives above show, no single approach is optimal for the two forms. The fundamental contradictions between the two forms of variety improvement and a single system are (i) more significant improvements should have a lower level of relatedness requirement to limit further cosmetic breeding, and (ii) complex, multigene improvements by definition involve more genetic differences than simple, single gene ones.

As a consequence, no system applied to both types of genetic enhancement can be optimized for both.

Here, we therefore propose a dual system that accommodates both possibilities. For discrete contributions, the existence of the trait in varieties for which the initial variety was included in the breeding ancestry would be prime face evidence of essential derivation, with no need to establish a proportion of genetic material. Clearly, more than a single initial variety would be possible for discrete contributions.

For complex traits, the dual standard of the proportion of genetic material and existence of the distinctiveness trait (in other words, the current UPOV system) would be used. The degree of relatedness could be set relatively low (say, 75%) to avoid the incentive for cosmetic breeding. Indeed, for complexly inherited traits, there would be a high presumption of derivation if the purported initial and derived varieties contained that degree of common genetic material. Clearly, for complex traits, multiple initial varieties are not possible so that no provision need be made.

Because of the complexity of the issues identified here, it is hardly surprising there are virtually no known examples of the application of initial variety status. Some in the industry suggest that is due to forbearance--the voluntary avoidance by breeders of source materials that could be claimed as initial varieties. If true, then there is no benefit to undertake prebreeding efforts, which could result in a possible loss in variety enhancement. Perhaps then, this is the time to reconsider the use of UPOV's initial and essentially derived variety system.


This research was supported in part by USDA IFAFS project, "Innovation and Dynamic Efficiency in Agricultural Biotechnology". Dr. Nuria Urquia FAO/AGPS provided helpful comments on an earlier version.


Hunter, R.B. 1992. Essentially derived and dependency, some examples. Mimeo, UPOV, Variety Identification Steering Committee, Geneva.

International Seed Federation. 2002. No. 02.088. Nyon, Switzerland.

International Seed Federation. 2003. Fostering plant innovation: ASSINSEL brief on review of TRIPS 27.3b. Nyon, Switzerland. Available at (verified 9 February 2004).

Lesser. W., and M. Mutschler. 2002. Lessons from the patenting of plants, p. 103-118. In M. Rothschild and S. Newman (ed.) Intellectual property rights in animal breeding and genetics. CABI Publishing, Wallingford, UK.

Rories, C.C.P. 2001. Does the U.S.P.T.O. have authority to grant patents for novel varieties of sexually reproducing plants? J. Patent and Trademark Office Soc. 83:737-758.

Sechley, K.A., and H. Schroeder. 2002. Intellectual property protection of plant biotechnology innovations. Trends Biotechnol. 20: 456-461.

Silberman, A.B. 1995. The relationship between basic and improvement patents. JOM 47:50.

Troyer, A.F., and T.R. Rocheford. 2002. Germplasm ownership: Related corn inbreds, Crop Sci. 42:3-11.

UPOV. 2002, General introduction to the examination of distinctness, uniformity and stability and the development of harmonized descriptions of new varieties of plants. TG/1/3. Geneva.

UPOV. 2000. Progress Report on the Work of the Technical Committee the Technical Working Parties and the Working Group on Biochemical and Molecular Techniques and DNA-Profilingin Particular. C/M/10. UPOV, Geneva.

UPOV. 1996. Model law on the protection of new varieties of plants. UPOV, Geneva.

UPOV. 1992. Essentially derived varieties. IOM/6/2, Sixth Meeting with International Organizations. Geneva.

Young, N.D, and S.D. Tanksley. 1989. RFLP analysis of the size of chromosomal segments retained around the Tm-2 locus of tomato during backcross breeding. Theor. Appl. Genet. 77:353-359.

W. Lesser and M. A. Mutschler *

W. Lesser, Department of Applied Economies and Management, Cornell University, Ithaca, NY 14853; M.A. Mutschler, Department of Plant Breeding, Cornell University, Ithaca, NY 14853. Received 30 June 2003. * Corresponding author (
COPYRIGHT 2004 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Perspectives
Author:Lesser, W.; Mutschler, M.A.
Publication:Crop Science
Article Type:Author Abstract
Date:Jul 1, 2004
Previous Article:CSSA officers, boards, and committees, 2004.
Next Article:A decade of QTL mapping for cyst nematode resistance in soybean.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters