Science of desirable weeds

It is not clear at what readership this volume is aimed. It is mainly concerned with weeds and virus diseases. It may be too verbose and repetitive, especially in attempting to explain weeds and weediness, to be offered to decision makers, and not sufficiently far-reaching in discussing the nature of "viruses" and virus diseases to offer to the lay reader a usefully provocative assessment of the threats of technologically scattered fragments of DNA or RNA. Reverse transcription, unmentioned, is not only fundamental to genetic engineering but is also the means by which retroviruses such as HIV probably evolve.

The distinction between traditional plant breeding and biotechnological breeding is not very fairly made. "Biotechnology cannot create new genes; it can only find and move them." Is that true? Genetic engineering can also certainly recombine genes from diverse organisms and thereby create organisms with properties quite abnormal for the apparent beast or plant in question. It is true that conventional plant breeding allows far more restricted recombination within related families and usually within genera and mostly within species. Therein perhaps lies a real and important distinction. Many primitive relatives of crop plants contain toxins and even more present "weedy" attributes mostly through more efficient natural dispersion and bioprotection mechanisms. The use of wild relatives to restore some desirable traits such as disease and pest resistance is commonplace in conventional plant breeding but can sometimes be risky too.

Such arguments are relatively trivial in relation to the overall concerns which many feel towards genetic engineering. Do we actually need the "hundreds of kinds of transgenic vegetables, grains, fruits, trees, fibre crops and ornamentals" that could be produced by the turn of the century and "will this facilitate or retard a global transition to a sustainable agriculture"? Sadly, such basic questions are deemed beyond the scope of this book.

One might have expected to find in a book trying to explain the processes of crop improvement to lay people, an account of the gene introgression linked with disruptive selection. Position effects also receive no mention. The architecture associated with chromosome behaviour will remain the structural genetic basis of stable organisms. The insertion of new or alien DNA by "bioballistics" or other means into engineered organisms, in ways that have not been screened through the long processes of evolution, selection and domestication and may not fit into well-stabilised and "balanced" heritable chromosomes, can be viewed as hazardous in a broad ecological perspective.

The authors really labour to define plants out of desired place or time as weeds. They might have created food for thought by describing nucleic acids out of place as potential genetic weeds! If we had then understood the nature of DNA, RNA and of transcription and translation, would the term virus ever have been coined?

Its centrepiece is a supposedly new and highly structured approach to environmental risk assessment, due to Robin Manasse and already published elsewhere as cited in the bibliography. The Union of Concerned Scientists would wish there to be set up a new federal regulatory programme administered by a new agency with an "Environmental Mission". In their view the programme should be based on this new approach to risk assessment. New policy should be based on clear structures so "an educated public can evaluate the risks and weigh them against the benefits". How politically correct can you get?

A new regulatory infrastructure would certainly be job-creating and would hinder some possible excesses of the corporations of the free-market economy. but on the showing of this book I doubt if it would substantially save us from ourselves or help even a well-educated public to do so.

Colin Leakey is an applied biologist working mainly on agricultural, food and industrial feedstock aspects of pulses and leguminous oilseeds.