properties after degradation has occurred. Sensitivity and resilience depend on climate and the biophysical structures of the soil, and whether degradation has exceeded a threshold of resilience (such as loss of all organic matter or severe compaction) beyond which recovery is not possible without active intervention (Blaikie and Brookfield, 1987).

     Soil, just like water, is a key resource for agricultural production. Sometimes erroneously subsumed under "land" issues, the availability of soils for growing crops often seem to be taken for granted. Yet in both the developing and the industrialized world, the loss of productive agricultural soils to urban development is enormous. In addition, according to an estimate by the Global Assessment of Human-induced Soil Degradation (GLASOD), degradation had affected 38% of the world's cropland, to some extent as a result of human activity (Oldeman et al., 1991). However, GLASOD did not estimate productivity losses associated with land degradation. In the absence of data on the productivity impacts of land degradation, estimates based on different methods vary widely (Wiebe, 2003).

     The direct influence of agricultural practices cannot be neglected: they account for about a quarter of total soil degradation (GACGC, 1994). AKST is, and always has been, crucial to address these problems both through more classical approaches (e.g., proposing mechanical protection such as bunds and terraces to control surface runoff) and through more comprehensive frameworks aiming at greater integration of water conservation and soil protection and the use of biological methods (Shaxson et al., 1989; Sanders et al., 1999; WOCAT, 2006).

     The impact of nitrates from fertilizers and livestock production on soil and water resources is a related issue. This impact can be described in general terms as the nitrification of the global ecosystem from inorganic fertilizers and alteration of the global nitrogen cycle. Eutrophication as a consequence of nutrient runoff from agriculture poses problems both for human health and the environment. Impacts of eutrophication have been easily discernible in some areas such as the Mediterranean Sea and northwestern Gulf of Mexico (Wood et al., 2000).

     Some agricultural activities have led to a reduction of system productivity. For instance, irrigated agriculture has contributed to water logging and salinization, as well as depletion and chemical contamination of surface and groundwater supplies (Revenga et al., 2000; Wood et al., 2000; CA, 2007). Manure from intensive livestock production has exacerbated the problem of water contamination. Misuse of pesticides has led to contamination of land and water, to negative impacts on non-target species, and to the emergence of pesticide-resistant pests. These problems compound to reduce system productivity (Thrupp, 1998; Conway, 1999). The capacity of coastal and marine ecosystems to produce fish for human harvest is highly degraded by overfishing, destructive trawling techniques, and loss of coastal nursery areas. This is exacerbated by the decline of mangroves, coastal wetlands, and seagrasses with resultant loss of pollutant filtering capacity of coastal habitats.

Biodiversity

Biodiversity underpins agriculture by providing the genetic material for crop and livestock breeding, raw materials for

industry, chemicals for medicine as well as other services that are vital for the success of agriculture, such as pollination. The last century has seen the greatest loss of biodiversity through habitat destruction, for instance through conversion of diverse ecosystems to agriculture. Other factors such as the growing threat from introduction of invasive alien species, fostered by globalization of trade and transport, have further exacerbated the situation. On small islands, introduction of invasive alien species, many through agriculture-related activities, is the main threat to biodiversity. In freshwater systems, an estimated 20% of fish species have become extinct (Wood et al., 2000). Globally, the cost of damage caused by invasive species is estimated to run to hundreds of billions of dollars per year (Pimentel et al., 2001). In developing countries, where agriculture, forestry and fishing account for a high proportion of GDP, the negative impact of invasive species is particularly acute. Globalization and economic development through increasing trade, tourism, travel and transport also increase the numbers of intentionally or accidentally introduced species (McNeely et al., 2001). It is widely predicted that climate change will further increase these threats, favoring species migration and causing ecosystems to become more vulnerable to invasion.

     While agriculture is based on the domestication and use of crop and livestock species, the continuum between (wild) biodiversity and agrobiodiversity has been recognized both in research on plant genetic resources and in conservation efforts for many decades-starting with the hypothesis of "centers of diversity" of crop species proposed by Vavilov in the 1920s. More recently an emphasis on the provisioning services of biodiversity has been added: "Biodiversity, including the number, abundance, and composition of genotypes, populations, species, functional types, communities, and landscape units, strongly influences the provision of ecosystem services and therefore human wellbeing. Processes frequently affected by changes in biodiversity include pollination, seed dispersal, climate regulation, carbon sequestration, agricultural pest and disease control, and human health regulation. Also, by affecting ecosystem processes such as primary production, nutrient and water cycling, and soil formation and retention, biodiversity indirectly supports the production of food, fiber, potable water, shelter, and medicines" (MA, 2005c).

     Agrobiodiversity is the very stuff of food production and an essential resource for plant and animal breeding. Yet it is a resource that is being lost in situ: in farms and agroecosystems (FAO, 1996b; Thrupp, 1998; CBD, 2006). Its conservation is somewhat framed by a paradox: new breeds have boosted agricultural productivity, but simultaneously they displaced traditional cultivars. In response, gene or seed banks have been created to fulfill a double function: to resource plant breeders with the agrobiodiversity needed for further crop development, and to conserve crop diversity that may have disappeared from agricultural systems. Ex situ conservation in seed repositories and gene banks has long been considered to be the central pillar of agrobiodiversity conservation.

     To be effective, agrobiodiversity management needs to operate at several levels: local, national, and international. Against the overall trend of declining diversity in agricultural