catch database is used for each year from 1950 to present to fit the catches. Once the model has been tuned and is deemed to perform satisfactorily, a series of future-oriented evaluations is performed. The EcoOcean model is not linked to any of the other models and is only used to add insights about the future quality of the marine systems.

     IMPACT provides broad insights into socioeconomic consequences. The regional GEN-CGE and CAPSiM models provide added insights into distributional consequences and gender for China and India, respectively. The WATERSIM model is used to provide more insights into the water bal­ance. Water demand for irrigation, domestic purposes, in­dustrial sectors, livestock and the environment is estimated on a basin scale. Water supply for each basin is expressed as a function of climate, hydrology and infrastructure. The model iterates between basin, region and globe until the conditions of economic equilibrium and hydrologic water balance are met.

5.2.2.2 Policy experiments IMPACT is the core modeling tool used to assess agriculture and AKST for the IAASTD. It is therefore represented in most of the policy experiments. Investments requirements in AKST and changes in diets (both organic and nonmeat) are simulated by IMPACT. Climate mitigation and bioen-ergy policies are implemented first in IMAGE. Trade liber­alization is performed by GTEM, simulating changes in the economic structure. IMPACT then picks up changes in the economic structure and simulates the consequences for food supply and demand. The regional consequences are exam­ined through the regional GEN-CGE and CAPSiM models for India and China. Finally, changes in water productivity are assessed by WATERSIM.

5.3 Description of Reference World, Including Quantification

5.3.1 Rationale of reference world

The reference case imagines a world developing over the next decades as it does today, without anticipating delib­erate interventions requiring new or intensified policies in response to the projected developments.  Current policy pathways are expected to continue out to 2050. This con­tinuation of the "real world" is plausible. In subchapter 5.4 some of the major questions affecting the future of agricul­ture and AKST are simulated and results are compared to the reference world.

5.3.2 Inputs into the reference world

5.3.2.1 Population

In the reference case the global population increases from slightly more than 6.1 billion in 2000 to over 8.2 billion in 2050. Most of the growth is concentrated in middle-income and low-income countries like Brazil, India, China and Russia and the rest of the world (Table 5-2). Population growth con­tinues to slow in high-income countries. Population growth drives changes in food demand and is an indirect driver for AKST. The data for population changes are taken from the medium variant projections of the UN (UN, 2005), based on an assessment of previous studies (see also Chapter 4).

5.3.2.2 Overall economic growth

Economic growth assumptions are loosely based on the Tech-noGarden (TG) scenario of the Millennium Ecosystem As­sessment (MA, 2005). Incomes are expressed as MER-based values. The economic growth assumptions of the TechnoGar-den scenario are near the mid-range growth scenarios in the literature for the world as a whole and most regions. In some regions the scenario is a relatively optimistic scenario (e.g., sub-Saharan Africa). A comparison of economic growth pro­jections in other scenarios is made in Chapter 4. Information at the regional level is provided in Table 5-3.

5.3.2.3 Agricultural productivity

Agricultural productivity values are based on the Millennium Ecosystem Assessment (MA) TechnoGarden (TG) scenario and the recent FAO interim report projections to 2030/2050 (FAO, 2006a). MA assumptions have been adjusted from the TG assumptions to allow for conformity to FAO pro­jections of total production and per-capita consumption in meats and cereals, and to our own expert assessment. The main recent developments regarding technological change with continued slowing of growth overall have been taken into account. Growth in numbers and slaughtered carcass weight of livestock has been adjusted in a similar fashion.

5.3.2.4 Nonagricultural productivity

Growth in nonagricultural sectors is projected to be low­er than in agriculture in the reference case. The nonagri­cultural GDP growth rates are likewise based on the MA TechnoGarden scenario but with adjustments to align with World Bank medium-term projections. While the relatively higher productivity in agriculture reflects largely the de­clining trends in the agricultural terms of trade, this is not translated into higher output growth in agricultural sectors relative to nonagricultural sectors.

     Disparities in growth rates among countries in the devel­oping world are projected to continue to remain high while more developed regions will see more stable growth. Devel­oped regions will see relatively low and stable to declining growth rates between 1 and 4% per year out to 2050. Most of NAE falls into this category while several countries in ESAP (East and South Asia and the Pacific) (South Korea, Japan, New Zealand, Australia) and South Africa are quite similar in growth patterns. The LAC region will also see stable growth rates through the projection period, though slightly higher than for developed regions between 3.5 and 4.5% per year out to 2050.

     East and Southeast Asia will also see stable to declin­ing GDP growth rates through the projection period, but the rates will remain relatively high between 4 and 7% per year. In particular, China's economy will be slowing from the 10% growth in recent years to a more stable rate of 5.6% per year on average out to 2050. On the other hand, growth in South Asia will follow the strong reforms and initiatives in India focusing on macroeconomic stabilization and market reforms and should lead to projected improved income growth in that subregion of 6.5% per year out to 2050. CWANA will also see an increase in GDP growth rates through the projection period though the rates are a bit more modest and will lead to an average 4% per year out to 2050 for the region.