Researchers at Scotts Miracle-Gro have a vision of a greener future. The lawn-care company, based in Marysville, Ohio, wants to develop a dwarf grass that needs less frequent maintenance than standard Kentucky bluegrass. But there is a catch: such grass is unlikely to stand up to weeds. No problem, the company reasons, it will make a dwarf grass that is resistant to herbicide to help homeowners to nip those weeds in the bud.

Development of this genetically modified (GM) Kentucky bluegrass made headlines this month when the US Department of Agriculture (USDA) told Scotts that it did not have the authority to regulate it (see page 274). As a result, Scotts is free to start selling its new crop without oversight.

The reason for this is historical. US regulation of GM crops relies on its authority to control plant pests, and so the USDA has regulated crops on the basis of the way plant-pest-based tools are used to make them. It is a bizarre approach, given the low pest risk from the tools. But it had some merit when it was first developed because foreign genes were often inserted into the plant genome by a bacterium that can be lethal to some plants. Once in place, the expression of the foreign gene was guided by a series of genetic elements pulled from plant viruses.

In the United States, genetic-modification regulation rests not on the final product but on the methods used.

To get around this, researchers at Scotts made GM grass without using plant pests. It took more work, but the company reasoned that the streamlined regulation — as well as possible greater consumer acceptance and relief from the patent stranglehold on more traditional genetic-engineering methods — would make it worthwhile. So they mined the wealth of plant genomic data now available, snipped a herbicide-resistance gene from the model plant Arabidopsis thaliana, sewed it to genetic elements pulled from maize (corn) and rice to drive the gene's expression, and used a gene gun to blast it into the Kentucky bluegrass genome.

This technique is not the only GM method likely to fall outside USDA regulations. Plant biologists have made tremendous strides since the current rules were cobbled together in 1986, advancing both our fundamental understanding of plant genetics and the technical know-how in manipulating gene expression. Genetic changes can now be made at specific sites in the genome, and foreign genes can even be expressed in plant cells without integrating them into the genome at all. And gene expression can be regulated using RNA molecules — including, in some cases, ones made by the plant in response to attack by a pathogen.

Many of these advances are still years from commercialization. But regulators must prepare the ground. Monsanto GM soya beans, which use RNA interference to modulate the expression of endogenous genes, are already awaiting a decision from the USDA.

The USDA and others need to reconsider how they define and control GM species. If a crop developer uses genetic engineering to delete a discrete segment of a plant genome, how much regulation does that require? Would those same guidelines be appropriate for a crop that expresses half-a-dozen foreign herbicide- and insect-resistance genes, engineered without the use of plant pests? Such questions are particularly important where — as in the United States — GM regulation rests not on the final product of genetic engineering, but on the methods used in the process.

The European Commission is tackling the issue, and has commissioned a study into how new plant techniques fall under the rubric of the European Union definition of GM crops. Similarly, the USDA's Advisory Committee on Biotechnology and 21st Century Agriculture has raised the problem as a point of concern. But the USDA's proposed changes to its GM regulatory powers, released in draft form in 2008, failed to address challenges posed by new technologies.

The USDA's Kentucky bluegrass ruling comes at a crucial time for agricultural biotechnology. Some estimate that the world must increase the rate of growth in agricultural productivity by 25% per year to meet growing worldwide demand for food and biofuels. Many argue that advances in agricultural biotechnology, some of which may come from GM crops, will be needed to meet this demand. Industry, particularly smaller companies, needs to know how these crops will be regulated before they will invest to develop new techniques.

The new breed of GM crops could help gain wider acceptance for the technology, by settling long-standing unease about the use of foreign genes and the inability to target such genes to a specific location in the genome. But it is doubtful that dubious consumers are ready for GM crops to escape regulation altogether.