SUMMARY

While there is undoubtedly potential for the application of biotechnology to plantation crops, there are a number of hurdles. First is the time-scale. At the most optimistic, it would take 15 years before any return could be obtained from transformation of oil palm, for example. Private sector plantations are driven by profit, and a project with a pay-back period of fifteen years is very hard to justify. In practice, the pay-back period may be much longer than that, as exemplified by tissue culture propagation of oil palm, which started in the sixties and has yet to make a profit. It is also quite difficult to identify useful transformation targets; herbicide tolerance is not worth pursuing, and pest resistance would probably not last the lifetime of a plantation crop. Improved product quality is probably the most exciting prospect; examples both of the possibilities and of likely problems are given for palm oil.

At present, the application of molecular markers in conventional breeding programmes looks to be the most useful application of biotechnology in plantation crops. There is still much scope for progress by conventional breeding, and marker-assisted selection can help to reduce the time scales.
 
 

INTRODUCTION

I was asked to give a private sector view of the prospects for biotechnology in plantation crops. I will make four points:

0. The time scales for transformation projects are unattractively long.

0. It is difficult to identify useful targets for transformation.

0. There is still much scope for progress through conventional breeding.

0. Molecular markers are one area of biotechnology with immediate application in plantation crops.

TIME SCALES

Any transformation project has to be long term, but the long generation time of some plantation crops is a major additional limiting factor. Figure 1 shows an estimate of the time required to start producing a modified oil from oil palm, after the actual transformation has been achieved. Several man-years of research would be needed for the transformation work, before starting on the 14-year development path.

Figure 1 Minimum time scale for production of genetically modified palm oil

Year 1 Transformation of high yielding clone achieved

Year 3 Field planting of clonal proving trials with transformed lines

Year 6 Fruiting starts - flowering and somaclonal variation evaluated

Year 9 Sufficient yield records compiled to identify best transformed lines

Year 11 Large scale field planting starts

Year 14 First modified oil produced

The private plantation sector is driven by profit; any new investment project has to give an adequate discounted return on capital. It is very difficult to persuade an accountant to back a project with a pay-back period of 15 years, but the time scale in figure 1 is optimistic, and makes no allowance for set-backs and delays. The history of oil palm propagation by tissue culture is instructive in this respect.

Unilever started research on oil palm tissue culture in 1967. By the mid-seventies, plants had been successfully regenerated from culture, and a tissue culture lab was set up in Malaysia in 1976. The first clonal palms were planted out in 1977, and by 1981 many clones were in the field, all flowering and fruiting normally, and exhibiting the expected uniformity. A decision was therefore taken to start large scale commercial production, both in Malaysia and in another lab opened in Britain in that year. In 1985, over 350,000 clonal plants were produced by the lab in Britain, but at the beginning of 1986, severe bunch failure was observed in the first commercial-scale plantings. Subsequent research has shown that this was associated with small changes in the tissue culture protocols, made in the course of scaling up. Meanwhile, though, several hundred hectares of clonal palms had to be replanted. The problems have now been resolved, at least by some laboratories, but the industry’s confidence in the technology has not been restored, and very little clonal planting has yet been done, more than 30 years after the research started.
 
 

TRANSFORMATION TARGETS

Sensible targets for transformation in plantation crops are not easy to find. Herbicide tolerance is not useful in a tree crop after the first year or two. Resistance to insect pests looks attractive at first sight, but it seems likely that pests would develop tolerance to the Bt toxin within the 25-year life of a typical tree crop.

The most promising targets are associated with product quality; relatively simple changes might be needed to produce caffeine-free coffee or tea, for example. Such a product could command sufficient premium in the market to make the investment in transformation worthwhile. The Palm Oil Research Institute of Malaysia (PORIM) have identified high-oleic palm oil as a worthwhile target. Figure 2 shows the pathway of fatty acid synthesis in plants. It appears that reducing the activity of C16-thioesterase, or increasing the activity of KAS II, could increase the level of oleic acid. The significance of this lies in the fact that olive oil, with 70% oleic acid and only 10% palmitic, sells for at least six times the price of palm oil, with 40% oleic and 45% palmitic.

PORIM have successfully transformed oil palms with marker genes (Parveez et al, 1998), but have not yet reported transformation with genes involved in oil synthesis

Figure 2 suggests that reducing C16 thioesterase activity might be all that was necessary to produce high-oleic palm oil, but experience with sunflower mutants shows the sort of complications which may arise. Martinez-Force et al (1999) studied mutant plants (not transgenic)with low KASII and high C16 thioesterase activities. As figure 2 suggests, these mutants have abnormally high palmitic acid levels. However, the seeds also contain 3 unusual fatty acids: palmitoleic (C16:1 D9), palmitolenic (C16:2 D9 D12) and asclepic acid (C18:1 D11; ie. with the double bond in a different place in the chain from oleic acid, which is C18:1 D9).

The additional reactions which the authors hypothesise to occur are shown in figure 3.
 
 

It appears that C16-ACP accumulates to high levels, despite the higher C16 thioesterase activity, and the enzyme (D9 desaturase) which normally desaturates C18-ACP to C18:1, also acts on C16-ACP, to give C16:1. Although KASII activity is reduced, this enzyme acts on C16:1-ACP, adding 2 carbons to extend the chain to C18:1, but because the chain elongation has occurred after the desaturation instead of before, the double bond appears in the wrong place (D11 instead of D9). The third unusual fatty acid, C16:2, is formed when oleate desaturase, which normally desaturates C18:1 to C18:2, acts also on C16:1.

The lesson is that a single, apparently simple, change in a synthetic pathway may have unexpected side effects. Successful transformation programmes have been those where the transformation work is integrated into a conventional breeding programme, allowing individuals with the required phenotype to be developed from a range of transformants.
 
 

CONVENTIONAL BREEDING

In many plantation crops, conventional breeding still has much to offer, without the need to bring in new forms of variation by transformation. Breeding progress has been particularly slow in some vegetatively propagated crops. Identifying and cloning good individuals from existing commercial seedling fields offers the illusory prospect of much faster progress than a long-term breeding programme. The sterility of this approach is illustrated by experience with tea, though: the best clones in most tea-growing countries were released in the fifties and sixties, and have not been improved on since.

Table 1 shows results from a tea breeding programme in East Africa, illustrating the sort of progress which can be made by breeding, in contrast to conventional field selection. The data in this table were from the first year of production only, but experience shows that yields of tea clones in the first year are highly correlated with yield over longer periods.

Table 1 Tea clone development by different methods

Source of clones Clones Percent outyielding

tested standard clone

Traditional selection in commercial seedling fields 270 4

Selection from crosses between standard clones 140 46

Selection from crosses between best available clones 48 94
 
 

MOLECULAR MARKERS

The one area of biotechnology with clear applications in plantation crops is that of molecular markers. I have already noted the scope still offered by conventional breeding, but generation times in tree crop breeding programmes are very long (about 8 years in cocoa and tea, 8-10 years in oil palm, up to 15 years in rubber and coconut). Marker-assisted selection offers prospects of accelerating the process.

DNA markers fall into two distinct classes: random markers, and markers linked to useful characters. Random markers are relatively easy to develop; they can be used for ‘fingerprinting’, to determine the identity of clones, or the legitimacy of progenies (Mayes et al, 1996). They can also be used in studies of genetic diversity and genetic distance (Shah et al, 1994; Mayes et al, in press), for planning crossing programmes, and in conservation and prospection projects. These uses all contribute to increasing the accuracy and efficiency of a breeding programme, but have little effect on time scales. To accelerate the programme, one needs linked markers.

An important first step towards developing linked markers is the construction of a linkage map. Maps have been published for oil palm (Mayes et al, 1997), cocoa (Lanaud et al, 1995), and coffee (Paillard et al, 1996), and some work has been done on rubber (Seguin et al, 1996) and tea (Unilever, unpublished). A map allows the selection of markers which are evenly distributed over the genome, thus enhancing the probability of finding markers linked to quantitative traits. In cocoa, markers associated with growth and flowering characteristics have been identified (Crouzillat et al, 1996), and some preliminary results from oil palm have been published (Jack et al, 1998). Mayes et al (1997) found a marker linked to the shell thickness gene in oil palm; this will allow selection of specific fruit types in the nursery, before fruiting starts.

In general, linked markers will allow selection for characters which are not expressed, such as disease resistance in an environment where the disease is not present, and selection for mature characters in immature plants. The latter possibility, particularly, could significantly accelerate breeding progress.
 
 

CONCLUSION

To summarise, I think biotechnology will make significant contributions to the development of plantation crops within the next few years, but only if it is integrated with conventional breeding programmes. The most significant area will be that of molecular markers. Transformation to improve product quality is a possibility, but the time scale is such that it will be many years before any products are available. For the immediate future, ‘genetic modification’ of plantation crops will continue to be done by traditional plant breeders, not by genetic engineers.
 
 

REFERENCES