for food and agriculture, (4) assessing the impact of new technologies, such as modern biotechnology in general and Genetic Use of Restriction Technologies (GURTs) in particular. The work program also has cross-cutting initiatives for conservation and sustainable use of pollinators and soil biodiversity, studies the impacts of trade liberalization on agricultural biodiversity, identifies policy to promote mainstreaming and integration of biodiversity into sectoral and cross-sectoral plans and programs. But the CBD is a framework, or umbrella agreement that requires its constituent Parties to adopt policies and enact legislation for effective implementation of its Decisions.

     Even if its Decisions are adopted and implemented fully at the national level, there is a danger that the CBD, like many other policy instruments, will be continually "running behind the future," (e.g., the CBD 2010 Target) to significantly reduce the rate of biodiversity loss. Historically, the principal policy instrument has been the establishment of protected areas, although this has been ineffective where prime agricultural land and high biodiversity compete, as can be seen by the underrepresentation of lowland, fertile land in the majority of current national protected area systems (WCMC, 2006).

     Broadly, natural habitats around the world can be divided into three categories, each requiring different, but overlapping or integrated sets of policies to ensure their survival in the long-term (Chomitz, 2007).The first category can be defined as wilderness: the majority of the land (or aquatic) area is natural, and anthropogenic land use has had a minor impact. With the exception of the major tropical rainforest regions of Amazonia, the Congo, Indonesia and Papua New Guinea, the majority of these areas are in temperate regions and do not harbor high levels of biodiversity, although they may provide valuable ecosystem services, especially in terms of water supply and carbon sequestration. Policies that promote establishment of protected areas in these regions are still feasible due to lack of pressure from alternative land use, but even in these areas, protected area design must consider the external threats arising from climate change (e.g., increased wildfires, and global transport of pollutants).

     The second major category of land could be termed frontier: land potentially suitable for agriculture that is close to an expanding agricultural system. Effective policies for the sustainable management of natural resources in these areas are difficult to design and implement. In most countries, traditional concepts of agriculture are used to develop protection policies based on the ecosystem representation and species richness as sole criteria. However, (sensu Peres and Terborgh, 1995) the development of sustainable natural resource management policies in terms of local community support and resilience in the face of climate change will be critical in coming years. Also critical will be the acknowledgement that appropriate policies and institutional arrangements (e.g., providing positive incentives to farmers to adopt sustainable soil management practices in areas where soils are depleted) can ultimately result in improved natural resources quality through agricultural use (Izac, 1997).

      Increasingly, improved methods of measuring and mapping total ecosystem value of natural land are allowing landchapter

use planners and landholders to make informed economic decisions based on a broader range of criteria than agricultural production alone (Troy and Wilson, 2006). This is allowing policy makers to introduce land-use planning "rules" (zoning) and economic incentives to better conserve natural environments in complex agricultural land-use mosaics.

     At a relatively large scale, this kind of planning is increasingly emerging in the Brazilian Amazon and Atlantic rainforests, (Campos and Nepstad, 2006; Wuethrich, 2007, respectively), where government and landholders are slowly forging agreements on establishment of a complex mosaic of protected areas, sustainable use forests and agricultural land. This represents a shift in policy away from prescriptive land use decisions made by the imposition of protected area on unwilling land-users towards the use of incentives, including payments for conservation. Auction bids for direct payments for conservation services such as native forest protection, reforestation, and restoration of riparian vegetation can further improve efficiency (Chomitz et al., 2006). Under this type of policy, eligible landowners voluntarily decide whether to apply for participation, and the resultant conservation network emerges as a consequence of many independent choices about participation. Similar incentivebased schemes may be found in the US Conservation Reserve Program (CRP), the Bush Tender program in Australia and the Costa Rica Environmental Services Payment program (see references in Chomitz et al., 2006).

     In the more "crowded" landscapes of Europe and the west coast of the USA, where remaining natural land exists in an agricultural and urban matrix rather than the converse, similar trends towards land use planning based on ecosystem service valuation and "multifunctionality" are being explored (Zander et al., 2007). In California, a spatially explicit conservation planning framework to explore trade-offs and opportunities for aligning conservation goals for biodiversity with six ecosystem services (carbon storage, flood control, forage production, outdoor recreation, crop pollination, and water provision) has been used. Although there are important potential trade-offs between conservation for biodiversity and for ecosystem services, a systematic planning framework offers scope for identifying valuable synergies (Chan et al., 2006).

      In Europe, agroenvironmental subsidies have been used as incentives to maintain and promote biodiversity-friendly land use on agricultural land. There has been some criticism that the schemes do not deliver all of the environmental and biodiversity benefits for which they were designed, especially as the scale of implementation becomes too small and fragmented (Whittingham, 2007). One option that avoids this situation is the adoption of regional planning approaches (e.g., the OECD environmental farm plan programs) to generate more coordinated land use patterns across larger landscapes (Manderson et al., 2007).

      A recent summary (Chan et al., 2006) of the policies for sustainable development at the interface between tropical forest and agriculture shows how these can be used to promote the trends described above:

     At the international level:

• Mobilize carbon finance to reduce deforestation and promote sustainable agriculture.