2.3 HYBRIDISATION WITH WILD RELATIVES

The possibility that genes from genetically modified crops might spread by hybridisation with wild relatives of the crop has been a central issue in the environmental risk assessment for such crops. The two main concerns are that the gene will alter the biology of the wild species in a way which enables it to become more weedy or invasive (in either the agricultural or wider environment), or that the gene will spread in a way which impacts on biodiversity (either the genetic diversity of the wild species itself or the biodiversity of the community to which it belongs). This latter concern, which at one extreme questions whether transgenes should be transferred to wild relatives under any circumstances (the genetic "pollution" debate), has focused mainly on the potential threat to centres of diversity of crop relatives (e.g. potatoes in South America).

2.3.1 The PGS application

The PGS application assesses the likelihood of movement of transgenes from oilseed rape to wild relatives in Part 2.IV of the 1994 application (especially Sections 4-8). A literature review, experiments done under the BAP and BRIDGE Programmes using genetically modified oilseed rape (see 2.2.1), and an analysis of the physical, ecological and genetic barriers to hybridisation and introgression were detailed in support of the assessment that

"Under natural conditions chances for successful exchange of genetic information are extremely low and practically limited to closely related species (B. campestris (syn. = rapa) and B. juncea."

"The outcrossing frequency of Brassica napus to wild relatives has to be considered extremely low while incrossing from weed species to oilseed rape [seems] somewhat more likely to occur, though [is] still very low."

"Successful introgression of a trait in a wild species is determined to be extremely low."

They also found that genetic modification of oilseed rape makes

" ... no inherent change in the behaviour or capacity to make interspecific crosses."

Such general statements are based on assessing the individual probability of a series of steps in the introgression of transgenes into wild relatives. In Section 2.3.4, recent research findings are compared with information available in 1994/95 under headings equivalent to four such steps:

• Which species of wild relative can hybridise spontaneously with oilseed rape?
• What proportion of hybrid seed is set in such crosses?
• What is known about the fertility and fitness of hybrids?
• Is there evidence that the genetic modification has affected outcrossing rates?

Detailed discussions of the evidence presented in the PGS application in answer to these four questions is deferred until Section 2.3.4.

2.3.2 ACRE's view

The minutes of the ACRE meeting (19 April 1994) record that " ... members discussed ... the transfer of the herbicide tolerant trait to compatible species.". No detail of this discussion is given. However, the view of the Committee was summarised in the 1994/1995 Annual Report as

"The likelihood of any hybrids of herbicide resistant rape and wild relatives surviving and establishing was deemed to be very low, and therefore such hybrids would not pose a risk either to ecosystems or to man's property. This was not a unanimous view. A minority in the Committee felt there were still uncertainties surrounding the extent to which the herbicide genes would spread to wild relatives, and the extent to which this could be regarded as contamination of the gene pool."

The dissenting view, from Ms Julie Hill, is expressed in the minutes as also stemming from uncertainty about the long-term consequences of the genes in wild relatives.

In addition to the evidence provided in the PGS application (itself heavily dependent on a contemporary review by Scheffler & Dale (1994) - see below), DETR and ACRE had access to a 1992 report from the Institute of Terrestrial Ecology published in 1994 as DoE Genetically Modified Organisms Research Report No. 1, Genetically modified crops and hybridisation with wild relatives - a UK perspective, and as a review paper with the same title (Raybould & Gray 1993).

In this review, which pre-dated Sheffler and Dale, oilseed rape was classified as a crop which had an overall "low" probability of gene flow from the crop to a wild species. It was included in a category with crops such as lettuce and barley with wild relatives in the same genus as the crop and with limited sexual compatibility between the crop and wild relative. This was in contrast to two other categories: crops such as sugarbeet, carrot and ryegrass with a "high" probability of gene flow to wild relatives where the wild relatives are effectively the same species as the crop, and crops such as potato, maize and tomato with a "minimal" probability of gene flow to wild relatives because there are no sexually compatible wild relatives in the UK.

The species identified in the DoE Report as the wild relative of oilseed rape with which hybrids were known to occur in the wild in the UK was wild turnip, Brassica rapa. This hybrid was known to occur "sporadically in crops of B. napus" (Stace 1991), and has been described as Brassica x harmsiana.

2.3.3 Other views

The possibility of hybridisation with wild relatives, and the consequent transfer of herbicide-tolerance to weedy species, has formed the basis of several objections to the marketing of the PGS hybrid oilseed rape - both in 1994/95 and subsequently.

In support of its objections in August 1994, the Danish Environmental Protection Agency refers to "recent Danish studies and supplementary literature" which have shown that transfer of genes between oilseed rape and wild turnip "can be substantial" (letter from DEPA to DETR, 10 August 1994). This is a reference to the work of Jørgensen and her colleagues, referred to below. We could not locate the statement that "hybrids have never been found in nature", attributed to the PGS document by DEPA (and suspect it refers to "a viable hybrid between Brassica napus and non-related crop or weed species has never been described or observed" (Part 2.iv.4.1, p.265)). Several references to interspecific hybrids between B. napus and B. rapa in the field (for example, in Western Canada) are reported in the PGS application (which concludes that they have low or zero fertility and do not persist).

Norway also rejected the notification because, among other reasons, of concern about outcrossing into wild relatives.

Greenpeace raised the issue of wild relatives in letters to the Chairman of ACRE (11 April 1994) and the Secretary of State for Environment (4 May 1994). Its main concerns were the potential for rapid evolution of herbicide-tolerant weeds ("resistant super weeds") and the genetic pollution which will arise "when the introduced 'foreign' gene is transferred by cross-breeding to a wild relative". Such transfer is regarded as "genetic contamination of our natural heritage" and would reflect a policy of "industrial genetic pollution". Similar views about the potential consequences of gene transfer to wild and weedy relatives are advanced in a recent and more detailed document by GeneWatch - Genetically engineered oilseed rape : agricultural saviour or new form of pollution? (GeneWatch 1998).