ciated with productivity enhancing technologies concluded that empirical evidence for these associations only exists for three scenarios-salinity, lower soil fertility, and pesticides and health (Maredia and Pingali, 2001). Furthermore, many of the best documented environmental costs from agriculture are related to the misapplication of technologies or over-use of resources rather than to the direct impacts of technology per se. Examples of this include the subsidydriven exploitation of groundwater for irrigation (Pimentel et al., 1997) and a lack of a complementary investment in drainage to reduce salinity problems in irrigated areas with poorly-drained soils (NAS, 1989). Some authors highlight the need for a counterfactual argument, i.e., what would have happened in the absence of yield enhancing technologies (e.g., Maredia and Pingali, 2001). For example, how much extra land would be required if yield levels had not been enhanced? Estimates suggest that at 1961 yield levels, an extra 1.4 billion ha of cultivated land would be required to match current levels of food production (MEA, 2005).

Resource-conserving technologies may reduce or eliminate some of the environmental costs associated with agricultural production with mixed results in terms of yield and overall water use.

Resource-conserving technologies (RCT) such as reduced tillage and conservation agriculture systems have been widely adopted by farmers in the last 25 years. For example, no-till systems now occupy about 95 million ha, mostly in North and South America (Derpsch, 2005), with current expansion in the Ingo-Gangetic Plain of South Asia (Hobbs et al., 2006; Ahmad et al., 2007). In general, no-till systems are associated with greatly reduced rates of soil erosion from wind and water (Schuller et al., 2007), higher rates of water infiltration (Wuest et al., 2006), groundwater recharge, and enhanced conservation of soil organic matter (West and Post, 2002). Yields can be increased with these practices, but while the physical structure of the surface soil regenerates, there can be significant interactions with crop type (Halvorson and Reule, 2006), disease interactions (Schroeder and Paulitz, 2006), surface residue retention rates (Govaerts et al., 2005), and time since conversion from conventional tillage. Other resource conserving technologies such as contour farming and ridging are also useful for increasing water infiltration, and reducing surface runoff and erosion (Reij et al., 1988; Habitu and Mahoo 1999; Cassman et al., 2005). Evidence from Pakistan (Ahmad et al., 2007) suggests that while RCT results in reduced water applications at the field scale, this does not necessarily translate into reduced overall water use as RCT serves to recharge the groundwater and then be reused by farmers through pumping. The increased profitability of RCTs also results in the expansion of the area cropped.

Modern agriculture has had negative impacts on biodiversity.

The promotion and widespread adoption of modern agricultural technologies, such as modern crop and livestock varieties and management practices, has led to a reduction in biodiversity, though this is contested for some crops (Maredia and Pingali, 2001; Smale et al., 2002; Dreisigacker et al., 2003). Although biodiversity may have been temporally reduced, genetic diversity is now increasing in major cereal crops. The CGIAR and other research centers hold in trust large numbers of crop plant accessions representing diversity.

Land degradation is a threat to food security and rural livelihoods through its effects on agricultural production and the environment.

Land degradation typically refers to a decline in land function due to anthropogenic factors such as overgrazing, deforestation, and poor agricultural management (FAO/ UNEP, 1996; www.unep.org/GEO/geo3). Degradation affects 1.9 billion ha and 2.6 billion people and with varying degrees of severity, and potential for recovery, encompasses a third of all arable land with adverse effects on agricultural productivity and environmental quality (Eswaran, 1993; UNEP, 1999; Esawaran et al., 2001, 2006). Inadequate replenishment of soil nutrients, erosion, and salinization are among the most common causes of degradation (Guerny, 1995; Nair et al., 1999). The GEO Report foresees that by 2030 developing countries will need 120 million additional hectares for agriculture and that this will need to be met by commercial intensification and extensification, using lands under tropical forest and with high biodiversity value (Ash et al., 2007). The restoration of degraded agricultural land is a much more acceptable option. Restoration techniques are available, but their use is inadequately supported by policy. The recovery potential of degraded land is a function of the severity, and form of degradation, resource availability and economic factors. Soil nutrient depletion can be remedied by moderate application of inorganic fertilizer or organic soil amendments, which can dramatically improve grain yields in the near-term, although responses are sensitive to factors such as soil characteristics (Zingore et al., 2007). Low-input farming systems, which are characterized by diversification at the plot and landscape scale can reverse many of the processes of land degradation, especially nutrient depletion (Cooper et al., 1996; Sanchez and Leakey, 1997; Leakey et al., 2005a).

Global livestock production is associated with a range of environmental problems and also some environmental benefits.

The environmental problems associated with livestock production include direct contributions to greenhouse gas emissions from ruminants and indirect contributions to environmental degradation due to deforestation for pastures, land degradation due to overstocking, and loss of wildlife