In summary, depending on the methods used, when investment for production of planting stock is finite, genetic transformation (modification) will probably lead to a reduction of genetic diversity. Compliance costs for the technology must be taken into account when assessing risk and calculating the expected gain.

I. Strategies, costs, and risk:

Vegetatively propagated clones - the "transclone". Each stably transformed cell line using the same plant genotype and the same foreign gene will give rise to a unique transclone. Effective levels of expression of the foreign gene cannot be predicted at present, therefore the performance of transclones must be determined in field trials, often of very long duration. Many transclones, probably 10-20 or more, will need to be field-tested for each plant genotype/foreign gene combination. For a fixed investment, as the number of transclones increases, the number of "host" plant genotypes decreases proportionately, and genetic variability in stands could be substantially reduced.

Transformation techniques currently available for most forest tree species use one of the following 2 options:

1. Proactive transformation - transformation of randomly selected, untested juvenile clones, using some new trait (such as a gene for herbicide tolerance) followed by field testing of all viable transclones that are produced. The value of intrinsic traits of the clones will be approximately the same as the mean for the family, given an adequate sample size.

2. Retroactive transformation - randomly selected, non-transformed juvenile clones are field-tested to determine their growth characteristics, followed by transformation of stored juvenile tissue, using only the best clones. A second round of field testing is necessary to identify the most effective transclones.

These research and development options are only possible with a long-term funding commitment. Calculations have shown that, in order to discover a given number of useful production transclones, the costs for proactive transformation (i.e. option 1) are at least twice that for retroactive transformation (DR Smith, NZ Forest Research Institute, unpublished data). However, practically all current forest tree research follows the proactive option. There is a risk that corporate political expedience could result in a net negative genetic gain because some of the limited number of transgenic clones presently available from research laboratories will perform worse than the family mean for the natural traits.

II. Lower-cost options for deploying transgenics:

Recent developments in rejuvenation of mature pines will significantly improve the economics of transformation of clones that have rare and valuable intrinsic traits. While the technology will not satisfy concerns about reduced genetic variation, I expect this to become a favoured route to tree improvement when the technology is demonstrated with other genera.

Production of transgenic seedlings through transformation of seed orchard clones could also be a viable, relatively low-cost option, but only for a limited range of introduced traits. For instance, herbicide-tolerance genes could be introduced into progeny-tested seed orchard clones, and effective, low-cost screening could be carried out on seedlings in the greenhouse or nursery bed. The surviving seedlings would retain a respectable degree of within-family genetic variation. This option is even more attractive if applied to clones used for control pollination where substantial genetic gain in the progeny has already been demonstrated. Vegetative amplification from transformed seedling stool-beds is a low-cost option that could add extra gain.

The transformed parent approach would not be useful for traits that cannot be selected for in the nursery bed. There is currently no way to test the effectiveness of introduced genes that confer resistance to insects or microbes which become a problem only as trees undergo transition from the juvenile to the mature phase.

An additional layer of costs in the production of transgenic seed may arise from regulations forbidding release of transgenic pollen, meaning that transgenic parent trees must be kept in large GMO houses. As technical problems of producing transgenic trees are overcome, it will be necessary to address the politically-imposed costs. International GMO legislation may well forbid dispersal of transgenic pollen and seed. The use of genes to prevent pollen and seed production in a transgenic tree should be accepted now as a necessary condition of field release of transgenic trees. For vegetatively propagated trees, there are potential gains from increased wood production (estimated at 30% in radiata pine) which makes genetic "immasculation" commercially, as well as proceedurally, attractive.

Dr. Dale Smith
MetaGenetics New Zealand
[email protected]

[To contribute to this conference, send your message to [email protected] For further information on the Electronic Forum on Biotechnology in Food and Agriculture see http://www.fao.org/biotech/forum.asp ]