4    Breeding methods

In chapter 3, we drew up principles for organic breeding in the form of propositions. In this chapter we assess the suitability of breeding methods on the basis of these principles. Breeding methods never stand on their own; here, we have placed them in their agricultural and socio-economic context. Also in this chapter, we will refute the claim that the further development of organic agriculture depends on gene technology.

After a general overview of current plant breeding practices, we will discuss the breeding methods at each level of incorporation. An extensive description of the various breeding techniques is included in appendix 4.         

         4.1    A critique of developments in agriculture and plant breeding

        4.1.1    Developments in agriculture

The years after World War II were the reconstruction years. Industrial and agricultural production had to be restored as quickly as possible. At first, food and currency were scarce and agricultural policy had to address these restrictions. The obvious choice was a policy strongly focused on increasing production. Food prices were kept low to enable a low wages policy and the export of agricultural products provided hard cash which was pumped into the budget deficit (Dutch Industrial Board for Agriculture (Landbuwschap), 1961).

Post-war agricultural policy was not merely a national affair; it was also one of the first areas in which the member states of the European Economic Community (founded in 1958) cooperated. The EEC's Common Agricultural Policy was aimed at the realisation of five specific goals. These were (EEC Treaty, Article 39, section 1)(Rood, 1993):

1.    increased agricultural productivity;
2.    an adequate standard of living for the agrarian population;
3.    market stability;
4.    securing food supplies;
5.    fair prices.

EEC policy was implemented with the major aim to increase the production potential. The Dutch government's vision on an optimal agricultural production encompassed intensive farming, economies of scale and specialisation (Wiskerke, 1997; Jongerden & Ruivenkamp, 1996). The result of this vision was that by 1990, for example, mixed farming systems had become a rarity.

Agricultural policy transformed agriculture from a sector in the true sense to a mere link in the food-production chain. This chain has four links: 1) primary supplying industries; 2) farms; 3) processing and manufacturing industries, and 4) distribution to consumers. Over the years, production-based agriculture has become fine-tuned to the other links in the chain: breeders supply high-yield varieties, the agro-chemical industry supplies the means to optimise growing conditions for these varieties, farmers create optimal farm conditions for the implementation of these resources, the processing industry shapes farm products into food which is shipped to the shops to be sold to consumers.

         4.1.2    Developments in plant breeding

The same trend discussed in section 4.1.1 can be found in the development of plant breeding. For centuries, farmers had been responsible for their own seed. In the sixteenth century, domestic and international trade grew and prompted some farmers to propagate seed professionally.

Seed trading companies did not start breeding varieties until the 1950s, when the agricultural sector began its transformation into a mosaic of specialisation. The role of science in agriculture grew and the new knowledge supplied by science ultimately led to a greater control of farm circumstances through soil cultivation, land consolidation, chemical fertilisers and pesticides. All this required varieties to be highly uniform. In the race to maximise yield and uniformity, the large variations in landraces were dismissed as suboptimal and undesirable. Subpopulations were extracted from the broad genetic variety of landraces and a limited number of types were selected. In this race, the limitations of breeding based solely on selection soon became evident and crossing techniques were subsequently developed.

Crossing techniques launched the breeding sector into a new, second, phase. From now on, farming and breeding were two distinct activities. Breeding was carried out by specialised firms that pursued their own economic goals. The bond between breeding and farm conditions was finally broken. New varieties were characterised by uniformity and yield. Many of these new varieties were hybrids, which are typically highly uniform, highly productive plants because of heterosis. This degree of uniformity had never been achieved in cross-pollinators. With hybrids, the selected characteristics are not passed on to progeny. Breeders therefore also had economic motives for developing hybrid varieties. However, breeders' focus on hybrids has led to a stagnation in the breeding of seed-forming varieties, which therefore have a lower yield and quality than hybrids.

The first tissue culture methods were developed in the 1930s to increase the possibilities for crossing wild varieties with cultivars in order to transfer disease resistance qualities from the wild plant. In vitro techniques facilitated crosses between increasingly dissimilar species, thus broadening a cultivar's gene pool. Micro-propagation and in vitro selection cut down on breeding time. Currently, tissue culture techniques are used in breeding to increase variation, to simplify selection and to speed up propagation so that seeds can be marketed sooner. Tissue culture techniques have become a standard element in breeding programmes (see appendix 5), illustrating just how far removed breeding is from farm practice.

The cleft between breeding and farming is deepened by gene technology, which enables characteristics to be incorporated at an even lower level. Gene technology marks phase 3 in the development of the breeding process. Genetic engineering has stimulated innovations, from breakthroughs in crossing to the introduction of foreign genes in a plant or organism. Biotechnology does not stand on its own, rather it is a new and supplementary option in breeding (table 4 and appendix 5).

         4.1.3    A critique of current plant breeding objectives

The breeder's role in the agricultural chain is to supply farmers with the best varieties. In this view, top quality is equated with high productivity. However, in practice these varieties will only prove the 'best' choice and give the promised yield in the right conditions. These conditions pertain to the standardisation of the agro-ecological system through the use of chemical fertilisers, pesticides, soil disinfection and anti-lodging substances. Chemical fertilisers in particular helped standardise the production regime by distributing nutrients uniformly in the field (Almekinders et al., 1995). Soils with different production characteristics are standardised to facilitate production maximisation in the long term. However, the generous application of fertilisers and manures makes crops susceptible to disease, which in turn stimulates pesticide use and focuses breeding efforts on resistance (Visser et al., 1997).

In the post-war period, efforts were concentrated on increasing agricultural production. Plant breeding, too, focused on increasing the ratio of harvestable yield to the total biomass of a cultivar as, for example, with short-straw varieties in cereals. Since the 1950s, the total production volume per farm has increased tenfold and production volume per hectare almost fivefold (Van der Linden, 1996). This is partly attributed to plant breeding programmes, and partly to intensification and standardisation in farm management. One could call it a synergy between breeding and farm management aimed at boosting production.

In conventional farming and breeding circles, the harvestable part of the plant reflects the plant's worth. But seeing a plant only as the sum of its edible parts is not a complete picture. In organic farming, crop residues are not useless waste products. This biomass may play an important role in the plant's ability to reduce weeds (plant structure, tillering speed and range, foliage, stress tolerance and hardiness), in contributing to soil condition or it may be used as straw for livestock. In other words, organic farming values non-harvestable parts of the plant more than does conventional farming, which has had to resort to chemicals to fill in for the plant where certain qualities have been selected out (for example, chemical weedkillers are used in crops where dense foliage has been selected out).

Resistance breeding is also aimed at increasing yield. Commercial breeders focus their efforts on absolute, monogenetic resistance since this can be selected fairly quickly and easily and gives the plant maximum protection against a specific pathogen. Absolute resistance does not alter the crop, thereby securing yield (and the appearance of edible parts of the crop). At the same time, however, the selection process in pathogens will be directly solely at overcoming that monogenetic resistance. Absolute monogenetic resistances are invariably overcome within a few years and new resistances must be incorporated.

For this reason, organic agriculture prefers increasing polygenetic resistance to disease and maintaining variation in resistance within the crop population. Plants remain subject to a marginal, acceptable level of disease, but the resistance tends to be more durable. Crop yield may be lower and the appearance of the edible parts of the plant may be imperfect. However, the main disadvantage is that this approach can be a difficult and lengthy process. In short, striving for speed and efficiency in breeding may conflict with finding durable, organic solutions to resistance problems in agriculture.

As stated, conventional, commercial breeders prefer absolute monogenetic resistances because of 'breeding economics'. Breeders try to keep the costs of developing new varieties down by reducing the breeding cycle (see inset 'a critique of the breeding race'). Compared to monogenetic resistance breeding, breeding for polygenetic and/or partial resistance considerably lengthens the breeding cycle. These resistances are therefore less interesting from an economic point of view. However, fundamental research programmes on polygenetic resistance are being carried out (Bonnier, 1991).

To minimise the duration of the breeding cycle, the basic structure of a variety is fixed. In effect, high yield varieties and elite lines are intercrossed so that a new variety can be selected in the shortest possible time. When the breeding materials do not possess the desired quality themselves, genetic variation is introduced 'from outside', for example through parent lines developed from wild plants or gene constructs.

Diversity in the breeding cycle is usually restricted to elite material. In other words: previously selected breeding material is used as much as possible. This reduces the gene pool for breeding. Landraces and wild plants are not incorporated in the breeding cycle as much as they could be because they extend the cycle, thus increasing the costs of the programme. Cost considerations also play an important role in the implementation of biotechnology. It is generally assumed that incorporating a single characteristic into an existing variety by means of a gene construct would shorten the 'improvement cycle', compared to crossing and generative selection, while at the same time the gene pool is expanded. Using gene constructs, resistances could be incorporated that would not be possible through classical methods.

As gene constructs can be used to produce new varieties, biotechnology as a whole can be said to complement traditional breeding methods. A major disadvantage with the use of gene constructs, however, is that they are based particularly on a monogenetic approach to plant characteristics.

Breeding economics also limits diversity. Profit maximisation demands a maximum distribution range. Investments need to be paid back faster because varieties have shorter life cycles (see inset). These aspects conflict with the development of regionally adapted varieties. The greater the geographical range in which varieties are marketed, the more generic, or average, the variety's qualities must be. These varieties can be grown in many different regions in standardised conditions, however standardisation in turn reduces the specific characteristics for a region. The bottom line is that regional ecological characteristics are adapted to 'generic' seed.

To some extent, genetic modification opens up new possibilities for region-specific varieties by adapting certain characteristics of the basic variety to the farming circumstances in different regions. The basic variety is developed first for the greatest possible geographical range. This variety is the starting point for differentiated varieties, which may be adapted to specific regional conditions such as cold or (semi-)arid climates or disease. The result would be some local diversification of the basic variety based on one or very few characters.

However, regional specificity has a different meaning in organic farming. According to the principles of organic farming, seed stock is selected for the whole range of ecological characteristics of a region. We suspect that the real motive for transformation techniques is to market varieties across a greater geographical range, thus raising the potential returns for the variety. Consider cold-resistant soya, which can be grown over a greater geographical range than can 'traditional' soya, thus raising potential revenues of this variety. In this case, the distance between breeding and organic farming is even greater than we thought.

Clearly, the current breeding system is part of the agricultural and socio-economic fabric of conventional farming, but this is a different fabric from that of organic agriculture. It is important to assess the suitability of conventional breeding methods for organic agriculture according to ecological and socio-economic principles.

         4.2    The suitability of current breeding methods for organic agriculture

         4.2.1    Selection methods at the crop level

The two basic principles of breeding are: 1) to create and collect genetic variation, followed by 2) selection of the most promising genotypes. Genotype preferences are determined by agricultural need and by the socio-economic motives of the breeding firm.

A plant's phenotype is the result of genotype - environment interaction, therefore selection at the crop level is the same as selecting for phenotype. The course of the selection process can be influenced by carrying out selections in specific growing circumstances.

Various methods exist to select individual plants from plant populations which possess the desired combination of characteristics. In mass selection, what appear to be the best plants are selected and maintained in bulk without testing the progeny separately. Plants with undesirable characteristics are eliminated. Mass selection is commonly applied in the early stages of a breeding programme when there are too few potentially desirable genotypes for repeated testing. In pedigree selection, seed-propagating plants are retained or eliminated on the basis of the performance of their progeny. This selection method works well for self-pollinators. In cross-pollinated populations, some cross- pollination in the selection field is inevitable, making pedigree selection less useful for these types of plants. In indirect selection, selection is not based directly on a desired characteristic, for example because this characteristic cannot be observed until the plant has flowered; rather, plants are selected for a strongly correlated characteristic. A lot of research goes into finding such correlations.

The moment and focus of selection are shifting. Field selection is shifting to later generations because specific characteristics are first selected in the laboratory. Secondly, the focus of selection itself is shifting towards increasingly specific, distinctly quantitative characteristics (size, colour, blemishes etc.) which can be entered as data into the computer. This simplifies the selection process and cuts costs. However, in this process selectors lose their feeling for the plant as an interconnected whole, and reliance on breeders' special, artisan skills, the unique ability to pick out an unusual or harmonious plant, is reduced. Conventional breeders admit that this is the other side of progress and a reduction of their art. Organic breeders should be given ample opportunity to develop that earlier expertise (Kunz et al., 1991b).

All selection methods in the (experimental) field at crop level pertain to a genotype -environment interaction. Therefore these techniques are well-suited for organic farming, which is aimed at optimising the plant-environment interaction. By selecting at different locations in organic conditions, breeding can boost a plant's adaptability to regional, natural circumstances. Organic field conditions further aid the selection process. This applies to the pathogenic stress, root formation and adaptation to organic fertilisation and soil quality. These selection methods are also suitable for participatory plant breeding (see section 3.2.1), involving farmers in regional selection processes by setting up experimental and selection fields on their farms (which are representative for the whole region).

Research on genotype - environment interactions plays an increasingly important role in population genetics. New ecological insights obtained through fundamental research will facilitate the further development of variation and selection methods at the crop level. This is likely to provide a new impulse to the breeding of seed-forming varieties, thus reducing or even eliminating the gap with hybrid varieties.

                        Proposition:    Variation and selection techniques at the crop level are highly appropriate for an organic breeding system.

         4.2.2    Crossing methods at the whole plant level

In plant breeding, crossing is the oldest and most common method of obtaining a new variety when selection within existing varieties has ceased to yield the desired result. To obtain plants with new combinations of characteristics, pollen from selected paternal plants is applied to the pistil of maternal plants. In combination breeding, these plants will be existing varieties within a species and crossing rarely causes a problem. In species-crossing, cultivars are crossed with their wild equivalent, which can sometimes only be achieved after bridge-crossing. Generally, next to the desirable traits, many undesirable wild characteristics are also transferred which are subsequently eliminated by repeated backcrossing with the cultivar. During the backcrossing process, selection occurs for the desired characteristic and undesirable characteristics are rogued out.

                        Proposition:    In principle, crossing methods which follow the steps of natural reproduction - pollination, fertilisation and seed formation - are appropriate for organic breeding.

Incompatibility problems often arise during species crossing. These pertain to pollination and fertilisation as well as to the fertility of progeny. Some techniques, such as temperature treatment, grafting and cutting (part of) the style, improve the chances of fertilisation. These techniques are applied in breeding bulbs such as tulips and lilies.

As long as seed formation occurs on the plant, these techniques are considered acceptable for organic breeding, provided that progeny are fertile. Some species crosses may only succeed if certain techniques at a lower level of incorporation are carried out first. However, organic agriculture may reject those techniques, thus effectively ruling out a number of species crosses.

                        Proposition:    Species crossing techniques which depend on an artificial intervention are acceptable for organic agriculture provided that pollination and seed formation occurs on the plant.

         4.2.3    Hybrid varieties

The techniques used in hybridisation have been debated for some years now in organic circles in general and in biodynamic circles in particular. A special section is therefore set aside for the discussion of hybrids. Discussions generally focus on two aspects of hybrids: a) how natural is hybridisation and b) what are the socio-economic consequences.

Many discussions do not differentiate between the various forms of hybrids. However, this is essential if we are to make a sound assessment of hybrids in the light of organic principles. Our differentiation is presented below.

             How natural are hybridisation techniques?
In cross-pollinating crops it is particularly tricky to select a new variety; variation among plants of one variety is inevitable. Just how much of a problem this variation poses for farmers or breeders is debatable. Hybridisation is a way of achieving highly uniform varieties, if that is an absolute requirement. Parent lines have to be artificially inbred for a number of generations to achieve homozygous lines. The F1 (the hybrid variety) is the result of a cross between two homozygous inbred lines. F1 hybrids have a uniformly high vigour (heterosis). Artificially inbred lines are often weak and have little vigour. Breeders select for vigour, so that the strongest lines are used to produce the most vigorous F1. The F1 line represents the new variety, therefore large numbers of seed need to be produced. In plants such as maize, where male and female inflorescence are physically separated, the male inflorescence may be removed mechanically. In other crops, natural male sterility is introduced in the maternal line through crossing or protoplast fusion. This technique is applied in the development of parent lines in headed cabbage. Male-fertile restorer lines (so-called B lines), whose genetic make-up is identical to that of the maternal line, are used to maintain the maternal line. In seed-forming plants, the restorer genes are also present in the paternal line, so that the F1 also has fertile pollen, in other words, it retains its ability to form seed.

The heterosis effect is sometimes seen in F1 hybrids resulting from self-pollinators (eg lettuce, tomato, wheat), but for practical reasons few such hybrids have yet been produced. In these cases, the production of F1 plants poses a problem. It is difficult to eliminate the pollen of the female parents. In addition, the form of the flower rarely allows efficient pollination with pollen from male parents carried by the wind and/or insects. For some crops, such as tomatoes, mechanical emasculation and pollination of the female line is worth the effort because of the sheer numbers of seed produced. This seed often has a fair economic worth as well.

In general, inbred lines in cross-pollinators have less vigour. Sometimes tissue culture techniques must be used to maintain them. If we state that organic breeding methods may not affect the reproductive ability of the plant, then hybrids based on such inbred lines, such as leek hybrids, cannot be used in an organic farming system. This also applies to cytoplasmic male-sterile female lines for which no restorer line is available. These practices impede a sustainable use of the variety in cultivation.

If cytoplasmic male sterility is applied, the female line, and in seed-propagating plants the F1 generation, can be maintained with a restorer line. This illustrates a degree of specialisation at crop level, with respect to reproduction, which may be considered acceptable for organic agriculture.

             Socio-economic consequences
Our concern about hybrids also pertains to the socio-economic benefits for seed companies. (Commercially) propagating progeny of hybrids has little use, since the characteristics of the two parent lines will be split up when the hybrid (F1) is reproduced. The use of hybrids only makes sense when new F1 seed is bought each year. This is one of the reasons why the breeding sector is so keen to produce hybrids, especially in countries without breeders' rights. Next to uniformity and heterosis, hybrids naturally prevent propagation. Breeders are so keen to rule out propagation by farmers that hybrids are also marketed when the heterosis effect is almost nil, for example with carrots and onions.

Organic agriculture has largely gone with the flow of leaving propagation to specialised companies, mostly by choice. Hybrid varieties can be suitable for one-time propagation in organic conditions, but do not allow further adaptation in subsequent generations according to a cycle of propagation and selection. However, there is no question of choice when only hybrid varieties are available for a certain crop. This restriction is felt most by farmers who could feasibly fit the propagation-selection cycle into their farm management system, such as with cereals. Thus hybrids have no place in a participatory breeding system.

Hybrids prevent further propagation not only by farmers but also by breeders from rival firms. Unlike with seed-forming varieties, breeders will not part with their hybrid parent lines, thus preventing the use of these lines in other breeders' programmes. The result is erosion of the genetic base of a cultivar.

For many crops, breeders' interest in seed-forming varieties has waned and the lack of improvement is reflected in the price of these varieties. In many crops, such as maize, cabbage and tomatoes, seed-forming varieties have been wholly replaced by (more expensive) hybrids.

It is generally believed in organic circles, that many seed-forming varieties would perform just as well as today's hybrids if the same amount of capital had been pumped into their development. For the reasons mentioned above, hybrids are not really compatible with organic farming principles. However, since not all hybrids are alike, they should not be judged as a group. Hybrids therefore fall in the grey category of what is and what is not acceptable in organic agricultural practice.

The agricultural advantages of hybrids are already so embedded in organic agriculture that we propose allowing some hybrids, which are more or less in keeping with organic principles (the 'yes, provided that' approach). At the same time, the organic sector calls for an initiative to develop seed-forming varieties to prevent genetic erosion. The following propositions can be formulated:

                 Propositions:    Hybrids are appropriate for use in organic farming, provided that:

                         Suppliers of organic products should be consulted in order to determine whether or not the uniformity requirement has been (over) accentuated, comparable to developments in conventional quality criteria, such that varieties that score high on yield and uniformity, but less on for example taste, should be recommended more strongly.

         4.2.4    Breeding at the cell level

Breeding at the cell level comprises tissue culture. Tissue culture techniques fall under the general term 'biotechnology' and in many cases have been used in breeding for many years. These techniques have paved the way for genetic engineering. Until now, organic agriculture has refrained from judging breeding techniques at the cell level. However, if we are to draw up preconditions for a breeding system according to organic principles, these techniques, too, should be assessed for their suitability for such a system.

Tissue culture methods are based on the premise that each cell in a plant can potentially grow on to form a whole new plant. This process is called regeneration. The entire operation is carried out in sterile conditions on a nutritive medium of water, minerals, sugars and vitamins. The addition of synthetic plant hormones (auxin and cytokinin), which balance cell division and differentiation, helps direct the regeneration process along the desired lines. The plants are cultured in a climate-controlled room. Light intensity, length of day and daytime/nighttime temperature can be set and controlled.

Tissue culture methods can be divided into three general groups:

        a)    methods to induce genetic variation. For example in vitro pollination and embryo culture play an important role in species crosses, it increases the gene pool of the cultivated crop;

        b)    methods to accelerate the analysis and selection of large numbers of seedlings by preliminary selection on the culture medium;

        c)    methods to obtain large quantities of plant material in a short space of time.

Up to a point, organic agriculture has no objections to these techniques because, from an organic viewpoint, the cell is the most basic form of life. This life-form remains intact during the tissue culture period. However, if organic breeding takes the road of developing varieties that are optimally adapted to organic farming conditions, organic breeders will be taking a detour with tissue culture methods, since the necessary adaptation to organic conditions still has to take place although at a later stage. The plants are temporarily removed from their natural (organic) environment and after culture, need time to adapt again to organic growing conditions. Moreover, lines which may have possessed favourable characteristics for organic farming may be lost during the mutation induction and selection steps of the culture process. When crossing takes place on the plant and selection occurs in the field, breeders can select plants that perform well in organic conditions right from the start. When tissue culture methods are used, this selection must occur at a later stage as the cultured plants adapt to their new environment.

                        Proposition:    Plant - environment interactions do not take place during breeding at the cell level. These techniques are an ecological detour with respect to the goal of organic breeding and are thus not suitable for use within an organic breeding system.

However, the use and acceptance of tissue culture techniques is common, even in organic agriculture. In effect, the organic sector has missed its chance to set out on a different course in this respect. The sector may state that it is against the application of these techniques in an organic breeding system, but then it must accept the mammoth task of developing andoptimising other techniques which are acceptable from an organic point of view and which have the same benefits as tissue culture.

However, it should not be ruled out that certain tissue culture methods may be accepted in organic breeding in the long term, when technical know-how and scientific knowledge is able to provide us with a better understanding of ecological laws. Further research in this field is a must.

A transition period will have to be instated until varieties are available which have been developed with acceptable techniques. The lack of such a transition period would be an unacceptable blow to the development of organic farming. A 'no, unless' guideline could be used, with the 'unless' pertaining to suitable alternatives.

                        Proposition:    To avoid putting the clock back, a transition period will apply for cell breeding methods in organic agriculture; a 'no, unless' guideline will be used until suitable alternative breeds become available. Research must be carried out to develop techniques at the cell level which do not take an 'ecological detour'.

         4.2.5    Breeding at DNA level

Breeding techniques at the DNA level can be divided into two categories: genetic modification techniques and diagnostic techniques. They will be described and assessed separately.

         4.2.5.1    Genetic modification techniques

It is common knowledge that organic farming organisations have declared themselves against the use of genetic modification in organic agriculture. Below, we explain the sector's objections. These objections are both ecological and socio-economic.

             A brief description of the technique
Molecular techniques are a logical consequence of the blurring of the species concept by 'artificial' crosses and the desire for a larger gene pool (that is, more characteristics). The techniques are particularly applied to incorporate characteristics which do not occur in the plant species but are considered desirable by the user of the plant. Breeders have in some cases succeeded in incorporating a single gene from a plant, animal or micro-organism in the plantgenome. Since DNA is a universal code for genetic traits, it is theoretically possible to use the DNA of any living organism to introduce new characteristics in a plant.

With the rise of molecular techniques, plants are increasingly treated as living bioreactors. For example, by making specific changes to the plant genome, the processing industry can obtain plants which produce substances for use by industry.

There are several transformation methods (the transfer of the desired gene to the plant). Basically, to make a transgenic plant, a gene expressing the desired trait is isolated at the DNA level (one step further than the level of the cell which - as stated earlier - is the most elementary unit of life). Cells or tissues of the recipient plant are brought into contact with the desired DNA fragments. The gene construct is introduced into the host cells by way of a vector (a plasmid). Growth hormones are then used to replicate the altered cells. These eventually grow out to miniature plants in test-tubes which are transferred to soil when they are large enough.

Contrary to the situation with crossing and some tissue culture methods, in which existing genetic material is recombined naturally, DNA fragments are inserted at random. The incorporation of new DNA always alters the structure of the host genome and may lead to (in)visible mutations regarding other characteristics. The latter mutations only become apparent in progeny, after selfing. These mutations typically occur in 5 to 50% of transgenes, depending on plant species, transformation method, tissue culture method and the type of DNA fragment. Other mutations also occur, but these do not noticeably affect the plant. It is assumed that subsequent selections of transgenic plants in a crossing programme will eliminate unwanted mutants.

             Ecological objections
Organic agriculture's ecological objections to gene technology pertain not only to ecological risks but also to the concept of a plant, not merely as a sum of isolated characteristics but as an interconnected whole (Kunz, 1998). From this point of view, the idea that genetic engineering could change one specific characteristic without changing any other aspect of the plant is unecological, reductionist thinking.

Living nature is a complex of interactions in which each change in an organism affects other levels within and outside that organism (Ho, 1997; Wirz, 1997). The mutations which arise during the tissue culture step, but also the unexpected side effects which first occur on the trial plot (for example, the gene might be present but is not expressed), only go to prove this point further (Parr, 1997). These might be dismissed as mere technical flaws by those working on developing these techniques, but from the viewpoint of organic agriculture they are the inevitable result of underestimating the complexity of living nature. So much of ecology (genotype-environment interactions) is still a black box, that those in organic agriculture are convinced that genetic modification acts at too low a level of incorporation in the breeding process and that it is not (yet) ecologically wise to apply such techniques in organic breeding. There are huge gaps in our knowledge of what happens when foreign genes are transferred to a host and what consequences this has for the host genome, the environment and consumers (Schiffelers, 1998; Stotzky, 1997), and what may be worse we are completely ignorant about the long-term effects of such techniques (Eckelkamp et al., 1997). In these circumstances, the application of genetic modification techniques in practice is scientifically unsound.

Organic farming organisations are convinced that genetic modification is not vital to the development of organic farming. Currently, the biotechnology industry is focused mainly ondeveloping transgenic herbicide-resistant varieties. As herbicides are not used in organic farming, the worthlessness of such varieties for organic farming should be evident.

We also reject the proposition that biotechnology is crucial if organic varieties with disease resistance are to be developed. Many diseases can be prevented by extended crop rotation and the use of organic fertiliser, as the organic control of beet eelworm has shown. In other words, organic agriculture is not interested in varieties that are resistant to pests such as beet eelworm. Of course organic farming systems cannot control all diseases and pests, even though there is less pathogenic stress than in conventional farming. Asexually propagated varieties, such as potatoes affected by blight and apple affected by scab, remain vulnerable. However, breeding for resistance at low levels of plant organisation is a short-term solution along the same lines as the control of symptoms. This is not the way organic agriculture should go. A more fundamental approach is called for, but this is yet to be developed through research at higher incorporation levels. Only then will the current lack of knowledge and restrictions of one-dimensional, overbred varieties be overcome and will we gain a true understanding of the complexity of these diseases and develop structural solutions.

Another ecological argument against the use of transgenic plants is the economic necessity of incorporating a gene construct in large numbers of crops and varieties, as with the Bacillus thuringiensis gene, for it to have any effect. As a result, however, the tendency towards monocultures at the crop level and erosion at the variety level thus also reaches the DNA level. The effects of such strategies are generally shortlived. Because the Bt gene is embedded structurally in plants, insects are bound to adapt to the gene sooner, thus rendering its value as an insecticide worthless (Holmes, 1993). Such a solution conflicts with organic farming's aims of biodiversity and self-regulation. Organic farmers are content to resort to an organic spray containing B. thuringiensis when absolutely necessary. Moreover, the use of sprays allows more flexibility in using diverse strains and combinations of B. thuringiensis, thus delaying selection of insect pests that are resistant to this biological control agent. Indeed, if monocultural use of plants with B. thuringiensis incorporated in the genome leads to rapid loss of effectiveness of the agent for all growers, then we have an ethical issue. That is, if it seems likely that incorporation of a useful resistance character into a crop genome will lead to enhanced selection for a pathogen or pest that is not affected by the resistance, and that this therefore jeopardises wider use of the resistance character, then use of the gmo may need to be discouraged or banned.

                        Proposition:    Organic agriculture does not allow gmos in its production systems since these conflict with the principles of organic farming. From the organic point of view, genetic modification affects the plant at too low a level of organisation, carries too many unidentified ecological and (public) health risks, encourages large-scale monoculture and is the result of unsustainable, short-term thinking.

             Socio-economic objections
We have already shown that those developing gene technology use patents to guarantee returns on their investments. Modern laboratories with a high labour input and expensive equipment are requisites for gene research and the use of gene technology in practice. In addition, a licence is often required to use these techniques commercially, as the techniques are often patented too, as well as genes themselves. In short, genetic engineering techniques are highlycapital intensive.

Organic agriculture strongly objects against the practice of patenting genes, but it is also concerned about the consequences of such patents. Varieties containing patented genes may be freely used by other breeders as cross-parents, but if it appears that the patented gene has been incroporated into the new variety, the patent-holder has the right to demand payment. A firm's position is determined by the patents it holds (IKC, 1997).

The enormous financial investment required to develop gene technologies is partly responsible for the increasing number of mergers between chemical firms (agro-chemicals) and breeding companies. It is not unthinkable that these new biotechnology firms will get increasingly more influence on the course of future breeding activities. For example, breeding activities for primary agricultural crops (cereals and maize) in our countyr (the Netherlands) are already controlled by large foreign companies (IKC, 1997).

Another concern is the high priority given to the development of herbicide-resistant varieties, so that when a farmer buys new seed he has little other choice than to buy the company's herbicide at the same time. The marketing of these varieties is particularly profitable to multi-nationals. The organic sector feels that new breeds should be developed on the basis of farmers' needs and not multi-nationals' desire for economic gain. If breeding were left to multi-nationals, regional differentiation and unicity would be bound to suffer. There would be fewer breeders for a given crop, resulting in genetic erosion of the cultivar, a development which the organic sector wishes to avoid at all costs.

                        Proposition:    Organic agriculture is against the use of gmo's in its production systems because they are not appropriate to the socio-economic objectives of organic farming. Genetic manipulation is too closely associated with factory farming and the practice of patenting (parts of) organisms and genes, which may result in the genetic erosion of our cultivated crops.

         4.2.5.2    Diagnostic techniques

Genetic engineering also allows for selection at the DNA level without altering DNA as in genetic modification.

Biochemical and molecular methods are often used for indirect selection. Isozymes, which have a slightly different composition from common enzymes, are made visible by gel electrophoresis. A certain band expressing the desired trait must be found on the gel. Isozyme analyses in seedlings may replace extensive tests for diseases and baking quality. In effect, a preliminary selection can be made using indirect selection techniques.

Molecular markers are a recent development in DNA technology. Variation in DNA sequences around the gene or genes that express the desired trait can be shown by isolating DNA and colouring the relevant fragments of the DNA base sequences. DNA is fragmented using restriction enzymes and placed on a gel. The characteristic pattern of the DNA fragments can be used to find patterns related to the desired trait (DNA fingerprinting).

In the process of identifying such molecular markers for indirect selection, it has become clear that one trait typically links up with several patterns. In other words, multiple genes which may code for multiple enzymes in biosynthesis may be involved in the expression of a single trait. These results further breeders' understanding of the desired characteristic and lead togreater selection efficiency, especially when different genes have different responses to, for example, environmental conditions. In some cases, this technique may supplement other selection techniques, but it should always be placed in its proper (limited) perspective. In an organic plant breeding system, which stresses the plant-environment interaction, there is no alternative to selection in the field.

                        Proposition:    In an organic breeding system, DNA diagnostic methods could supplement existing selection methods.

         4.3    Summary and conclusions

Conventional, commercial breeding has always concentrated on developing highly productive and uniform varieties which perform well in standardised growing conditions. As a consequence of the ever-increasing specialisation in agriculture, a cleft has split farming and breeding activities. Breeders' economic interests have had a major effect on the direction of breeding developments. In agricultural research, breeding scientists are burrowing ever deeper, concentrating on lower levels of plant organisation and depending increasingly on molecular, biochemical and physiological knowledge. The breeding process has shifted from farmers' fields to trial plots to greenhouses and to laboratories. The (regionally determined) plant - environment interaction which plays such a crucial role in organic farming has been more or less relegated to the backseat.

Organic agriculture needs a breeding system which, as a primary objective, takes into account the complexity and, biodiversity of agro-ecological systems and which works at high levels of plant organisation. In some cases, techniques will have to be (further) developed to meet these criteria.

The organic sector wants breeders to respect the ecological and socio-economic principles of organic agriculture. To this end, breeders and farmers should maintain close ties and work together towards the development of regional, participatory plant breeding systems.

The following conclusions can be drawn from the preconditions:

-    crossing and selection techniques at the plant level and the crop level are appropriate for organic plant breeding;

-    different crossing techniques are appropriate in an organic plant breeding system provided that pollination and seed-formation occur on the plant;

-    hybrids may have a role in organic farming, provided that the F1 is fertile and that the parent lines may be maintained in natural circumstances (cms is not applied without a restorer line);

-    techniques at the cell level are not really appropriate for organic breeding, although their use is common. Alternative techniques must be developed;

-    genetic modification conflicts with the ecological and socio-economic principles of organic farming;

-    DNA diagnostic methods might supplement other selection methods used in organic agriculture.

Table 5 lists the various breeding techniques and indicates which are (in)appropriate in an organic plant breeding system.

Table 5. Suitability of breeding techniques at different levels of plant organisation.