In the Background Document prepared for this conference, on the appropriateness of currently available biotechnologies for the fishery sector in developing countries, a brief coverage of some main biotechnologies was provided. These included the use of protein or DNA markers, triploidisation, sex-reversal, hybridisation, selective breeding, freezing of male gametes, genetic modification of fish and, finally, DNA-based technologies to diagnose and characterise fish pathogens and to develop vaccines. They were discussed in the context of 3 main areas: fishery management, aquaculture and conservation.
However, participants in the conference focused to a large degree on a single biotechnology, the use of genetic modification, in a single main area, aquaculture. Of the 26 messages posted during the conference, 19 dealt only with this theme. Apart from genetic modification, the technology of triploidisation was also much discussed, but only in the context of its application to genetically modified (GM) fish.
A range of factors (such as the impact on human health, the status with respect to intellectual property rights, the costs or capacity-building required) that might influence the appropriateness of the different biotechnologies were also mentioned in the Background Document. But again, one factor dominated the discussions: the potential ecological risk or environmental impact of GM fish.
Section 1 of this document attempts to summarise the main elements of the discussions. Specific references to messages posted, giving the participant's surname and the date posted (day/month), are included. Section 2 provides some information about participation in the conference and Section 3 gives the name and country of the people that sent referenced messages.
1. Topics discussed in the conference
1.1 The nature of GM fish
There was some basic disagreement about how different GM fish were from non-GM fish. Muir (1/9) maintained that GM fish were very different as they could retain all the benefits of the wild species, while the transferred gene (the transgene) could potentially confer major advantages on the individual fish, such as being able to spawn at different times or invade new habitats. Conversely, the transgene could also make individuals less fit than wild types by affecting traits such as juvenile survival (Muir, 30/8). Moav (4/9 and 28/9) maintained that GM fish lines were similar to the domesticated parental lines which created them and that their genetic superiority for traits such as growth rate or disease resistance would be similar to that achievable through many years of conventional selective breeding.
1.2 Production of GM fish in developing countries
Currently, there is no commercial growth of GM fish, either in the developed or developing world. Norris (23/8), however, predicted that within the next five years or so, production of GM fish for human consumption would be a reality. She argued that there were two reasons why it might happen in a developing country such as Chile. The first is that Chile is an important producer of farmed fish, thus representing a major potential market for the technology. Secondly, consumer opposition to GM organisms in general is far tamer than in developed countries, a point also made by Mair (15/9). Halos (12/9) emphasised that in densely populated developing countries with rising population numbers the priority is providing people with food as "poor people do not care to save for tomorrow since they fear tomorrow may not come for them, anyway". Mair (15/9) concluded that concerns about human health and environmental aspects of GM fish would inevitably be weighted lower when food security was a major issue, which could result in GM fish in aquaculture being adopted first in developing rather than developed countries.
1.3 Potential environmental impact of GM fish
As mentioned earlier, this was the major topic taken up by the participants during the conference. Discussions touched on four main areas
a) the potential impact when GM fish are introduced into ecosystems where the wild species already exists
b) the potential impact when GM fish are introduced into ecosystems where the wild species does not exist
c) whether triploidisation (and thus sterilisation) of the GM fish would reduce the ecological risk
d) biosafety in developing countries
a) Growing GM fish where wild relatives exist
Muir (30/8) pointed out that, unlike the situation of domesticated animals, domesticated GM fish might escape into an ecosystem where the wild non-GM members of the same species are found (e.g. a hypothetical case might be production of transgenic Atlantic salmon in the Atlantic Ocean). He argued that this was a major concern because a) the wild relatives are likely to be an integral part of the ecosystem and disruption of the species could affect the entire ecosystem and b) the escaped individuals can establish themselves by interbreeding with the wild relatives.
For this situation, Muir (30/8 and 1/9) summarised results from a paper he co-authored in 1999 which, using a theoretical model, considered the potential consequences of a small number of GM fish escaping and mating with their wild relatives. His results showed that if the transgene increased mating success but reduced the viability of transgenic offspring, then the local fish population could be driven to extinction. Halos (31/8) pointed out that the introduction of a new fish strain or of a superior conventionally bred strain might have the same consequences on a wild fish population and that this phenomenon was thus not unique to GM fish. She reported that this had already happened with the native catfish strain in the Philippines.
Regarding fish escapes, Halos (12/9) and Mair (15/9) both described the problems, especially due to large environmental extremes, of enforcing risk management strategies in developing countries. Mair (15/9) concluded, based on his practical experiences, "I would never like to guarantee that any domesticated fish cannot escape from an aquaculture facility".
Halos (31/8) argued that if GM fish mated with wild relatives, this might increase genetic diversity in the wild population. Muir (1/9), however, refuted this, concluding that GM fish (or exotic non-GM fish species) might in the short term add diversity but, in the long term, they decrease it because they eliminate competitors.
b) Growing GM fish where wild relatives do not exist
Moav (4/9) pointed out that in his country, Israel, carp had been imported from Europe and that the transgenic carp (with increased growth rates) that they had developed in Israel would not present such potential problems as there was no native carp population. Muir (5/9) suggested that the issue of the production of GM fish in regions where the wild species does not exist was of great importance and that there was a range of other potential examples, such as the production of transgenic tilapia in Cuba or transgenic Atlantic salmon in the Pacific. He had two major concerns about such potential initiatives:
1) He argued that the introduction of exotic non-native species usually resulted in ecosystem disruption and so should be treated with extreme caution - an argument that would be valid for both GM and non-GM introduced species. A much-quoted example of the potential hazards was the introduction of the (non-GM) grass carp, Ctenopharyngodon idella, species from Asia to the United States to control aquatic weeds (Kapuscinski, 22/9). The species caused much ecological damage and its escape into new ecosystems in the country has exacerbated the problem (Muir, 7/9). In this context, Ashton (25/9) also noted that the movement of African fish species between regions had caused serious environmental disruption to aquatic systems in Africa. Halos (12/9) underlined that the increasing need for food in developing countries has been a driving force behind the practice of introducing exotic fast-growing aquatic species. She argued that if GM fish gave higher yields per unit area and at a lower cost then they might be a better alternative.
2) Muir's (5/9) second concern was that such initiatives might give a false sense of security, as there was nevertheless a danger of the GM fish being later transported and introduced to other regions of the world where wild relatives are found - something which might be driven by economical imperatives. Mair (15/9) supported this point saying that "you can be fairly sure that if a fish is considered superior for aquaculture, its movement will be impossible to control completely". Muir (5/9) wondered who would be liable for any environmental damage in these situations.
c) Triploidisation
With such concerns expressed about the potential ecological risk of GM fish mating with non-GM wild relatives, the potential application of triploidisation to GM fish to ensure their sterility was raised (Ibarra, 6/9). Benfey (6/9) pointed out that reliable technologies for making GM fish triploid exist for salmonids and that this was a simple way to ensure they would not breed if they escaped into the wild. He also suggested that companies producing transgenic fish might want to only sell sterile fish, in order to protect their investment.
In theory, each individual GM fish could be tested to ensure it was triploid before being released, a procedure already established in some situations with the grass carp in the United States (Benfey, 7/9; Kapuscinski, 22/9). Chevassus (11/9), however, pointed out that it is possible to test for triploidy but not for sterility and that in some species, although not in salmonids, a few or large number of triploid individuals may in fact be fertile. Muir (6/9) also argued that it is actually hard to quantify how successful a sterilisation technique such as triploidisation may be if the true probability of failure is very low (e.g. 1 in a million), because, to reliably quantify it, an extremely large number of fish, more than normally tested, may be required.
Muir (11/9) also pointed out that even though triploid males might be sterile they may still mate with fertile females of the wild species, thus interfering with reproduction and breeding of the wild population. To avoid this potential problem, he thus proposed that GM fish, in addition to being made triploid, should also be sex-reversed so that only females be grown. Mork (11/9) reported that a Working Group on the Application of Genetics in Fisheries and Mariculture, belonging to the Mariculture Committee of the International Council for the Exploration of the Sea, had considered the issue of triploidisation at various times throughout the 1990's. An impetus for this work was the finding that some previously triploid Pacific oysters (Crassostrea gigas) introduced to the east coast of the United States reverted back to the diploid state. Their conclusion in a 1995 report was that "no current mass triploidisation/sterilisation technique is guaranteed 100% effective".
Mair (15/9) pointed out that there was an additional reservation about the application of triploidisation in aquaculture of GM fish in developing countries, i.e. that "the application of triploidy in commercial stocks (mainly salmonids and grass carp) has been limited to species that are habitually bred using artificial fertilization and incubation. For most of the important species in developing country aquaculture (namely tilapias and carp) artificial fertilization is rarely used and therefore application of triploidy on a commercial scale would be very unlikely to be viable".
d) Biosafety
Such discussions on the potential impact of GM fish escaping into the wild and the use of technologies such as triploidisation to minimise potential risks, brought up the main issue of biosafety which, broadly defined in relation to GM organisms, involves assessing and monitoring the effects of possible gene flow, competitiveness and the effects on other organisms, as well as possible deleterious effects of the products on health of animals and humans. Ibarra (6/9) noted that in developing countries there was a substantial lack of human resources in the fishery sector trained in genetics and that this could lead to the situation where "potentially high-risk biotechnologies will become implemented without a careful evaluation". Norris (23/8) also expressed the fear that GM fish might be introduced in developing countries "without even considering risk assessment for such introductions". Ashton (25/9) insisted that, prior to release of any fish, GM or not, in developing countries, there was a need for adequate biosafety protocols, legal instruments, liability procedures and a clear thread of responsibility for any damage that might be caused to the countries by their release. Del Valle Pignataro (27/9) lamented the fact that, in relation to introducing non-native species (GM or not) to developing countries, it was not possible in most cases to establish strict regulatory/monitoring systems, due to factors such as low economic priority or the lack of qualified human resources.
Gjoen (5/9) argued that it was difficult to foresee all the risks involved with GM fish and that the precautionary principle should be given priority, a view that was shared by Ashton (25/9). The need for carrying out risk assessment in a scientifically sound manner was emphasised in a few messages (Moav, 4/9; Muir, 5/9; Gjoen, 5/9; Moav, 28/9).
1.4 Use of genetic modification versus other alternatives
Genetic modification dominated discussions in the conference. Nevertheless, some participants did consider other biotechnologies and other aspects of aquaculture in developing countries. Doering's (25/9) perspective was that, with few exceptions, the fish species cultured today are wild and that enormous gains for traits such as productivity, growth rate or survival can be achieved by selective breeding, assisted by molecular methods. He argued that, apart from concerns about the potential environmental impact, "transgenic aquatic animals are not sensible or cost-effective in the genetic background of a wild animal and the enormous productivity gains to be made by intensive selective breeding". Norris (23/8) also emphasised that many developing countries were "in need of practical help and advice in developing good aquaculture breeding and husbandry practices which would benefit their programs greatly".
Ibarra (6/9) suggested that most of the currently available genetic biotechnologies are very appropriate for developing countries and that the main reason for their under-use was the "lack of human resources within the fishery and aquaculture sector trained in the adequate use of those genetic biotechnologies".
Doering (25/9) emphasised that many of the current problems in aquaculture in developing countries have low-technology solutions and that "the species appropriate for culture in developing countries generally do not have the production economics to justify many high cost inputs such as vaccines and artificial larval feeds". He argued that policy makers and scientists can become over-enthusiastic for molecular techniques, ignoring the large capacity-building needs that these technologies require. His conclusion was that "investments in developing countries on farmer education, reducing culture stress and improving water quality as well as domestication will yield higher returns than investments in high technologies".
Ashton (25/9) argued for the prioritisation of local solutions in developing countries, and that management systems which secure the protection, husbandry and sustainability of native species should first be put in place before any fish, GM or not, are introduced. Del Valle Pignataro (27/9) supported this viewpoint. She suggested that prioritisation should be given to domestication, culture and (eventual) genetic improvement of native fish species that are already exploited and that have good consumer acceptance in developing countries. She gave a summary of their ongoing marine fish efforts in this direction in Mexico, which will eventually involve the use of selective breeding with medium-level biotechnologies
2. Participation in the conference
The conference ran for just over 2 months, from 1 August to 8 October 2000. There were 149 participants and they submitted a total of 26 messages. These numbers were the lowest among the first five conferences. Messages came from 16 individuals (11 % of all registered) living in 12 different countries. Six of the individuals worked in universities, four in research institutes, two each were from NGOs and development agencies while there was one each from the private industry and a government ministry. The countries contributing most messages were the United States (9) and Norway (3). Participants in North America, Europe and Asia accounted for roughly 40, 30 and 20% respectively of all messages posted. Forty-two percent of messages were posted by participants living in developing countries.
3. Name and country of participants with referenced messages
Ashton, Glenn. South Africa
Benfey, Tillmann. Canada
Chevassus, Bernard. France
Del Valle Pignataro, Gabriela. Mexico
Doering, Don. United States
Gjoen, Hans Magnus. Norway
Halos, Saturnina. The Philippines
Ibarra, Ana. Mexico
Kapuscinski, Anne. United States
Mair, Graham. Thailand
Moav, Boaz. Israel
Mork, Jarle. Norway
Muir, Bill. United States
Norris, Ashie. Ireland