The main points of the Act are the following: a) the end of the monopolies for performance testing and insemination. This activity was completely liberalised but a universal service for land use and conservation of genetic diversity was established; b) a "Ruminant" interprofession, France Génétique Elevage, was recognized. This interprofession brought together the most representative organisations with the intent to replace the existing co-management between the state and the agricultural profession and to manage the organisation of the French scheme, in particular the management of the information systems; Inra and Idele are associate members of this interprofession; c) a new breed organisation was defined i.e. the Selection Organisation (OS); d) the National Commission on Genetic Improvement (CNAG) was renovated to play a role in the certification of the OS and call offering processes.

INRA's role as manager of the national information systems and the genetic evaluation of breeding animals was maintained for the following reasons:

• The selection market, whether national or international, is a competitive market. It is therefore important that products that are placed on the market be certified and objectively compared. The effort to standardise the procedures and define genetic values for breeding animals should be done by an independent arbiter. The genetic evaluations are unique, objective, independent and of high quality for specific species and sectors. In addition, the competing breeders use the same information collected from populations and gathered in the national database. The perennity of this nationally important database and its access need to be guaranteed. This is a long-term stake, guaranteeing a loyal competition between breeders.

• The livestock animal populations are part of society's heritage. For this reason, one of the government's stakes is to conserve these resources and to guide them towards the best satisfaction of society's needs. Different species should be treated differently. However there is a preexisting foundation from which the government cannot disengage itself.

Maintaining biodiversity is one of society's objectives: the government has an obligation to preserve the future by avoiding an excessive use of the available resources. However, today's economic short-term logic does not favour the preservation of genetic diversity. Even though farm species are not endangered, it is important to maintain breed diversity (preventing breeds with small populations and that are poorly profitable from disappearing in favour of "more important" breeds) and within-breed diversity (preventing intense selection to lead to the disappearance of alleles that are not important today but can be in the future). The government should participate in maintaining biodiversity by establishing rules and recommendations and specific support. It is also the government's responsability to lead debates on the use of rapidly emerging biotechnologies which raise many questions, possibly through reasoned use and rules. Finally, the status of genetic resources should be defined.

In species with short life cycles like poultry, the biological properties (reproduction, generation intervals) are favourable thus the producer's, processor's and consumer's constraints can be rapidly transfered to the breeder in a purely economic context. The specific role of the state for guaranteeing the populations is limited to biodiversity. Public research, which has a limited role in selection management, is not much involved and without any specific institutional partner.

Ruminants have much less favourable biological traits and this has major consequences on selection: in these species, selection is slow - it takes from 10 to 20 years for the results of selection to reach producers and consumers. This situation is responsible for the opposition between short-term (ie profitability of selection) and long-term objectives (ie sustainability of selection). If the state completely withdraws and there is a more competitition, short-term objectives could completely replace long-term objectives.

> What do long-term objectives imply?

• In the definition of the objectives of selection: the state must guarantee that all society's interests are considered, even if there is no immediate market value. These interests include economic profitability of producers and sectors but also nutritional and health quality of food for consumers, sustainability of farms, overall functionality (longevity, disease resistance, . . .), lower environmental impact, and welfare of animals in their environment. In many cases, these traits are complex, difficult to measure and to select for. The state should play an active role including making proposals for working solutions.

• Diversity often causes specific problems in these species. Indeed, when it takes much time to set up measures, we have a tendancy to underestimate the rapid erosion of genetic variability that this causes. Often, selection becomes greater in species with long life cycles than in species with short life cycles. The government should therefore contribute to maintain this breed and within-breed diversity by making recommendations for the selection of large species and by giving specific support to breeds that are endangered.

Ruminants play a major role in land management. In grasslands and in harsh or mountainous areas, they maintain land that is also the state's heritage. It is therefore important for the government to guarantee that this activity be maintained. To do this, the state should guarantee that selection objectives include constraints for the adaptation of populations to their environment and that genetic schemes are well integrated into the policy of farm and land development. Moreover, the liberalisation of selection should not lead to the disappearance of programs in some unfavoured zones. Indeed, if this happens, the concerned populations would probably decrease in number or completely disappear since selection would not be possible elsewhere due to important genotype x environment interactions.

In conclusion, an important stake of the Framework Act is to transfer the French genetic program to its main actors, the breeders, while guaranteeing society's interests on the long-term. By maintaining INRA's missions and the corresponding financial means, the state has created healthy competition between breeders and is guiding populations towards desired objectives for a more sustainable production, more harmonious land use and an improved quality of food products.