which annual average stocking rates are less than ten temperate livestock units per hectare of agricultural land.
• Rainfed mixed farming systems, in which more than 90% of the value of non-livestock farm production comes from rainfed land use, including the following classes.
• Irrigated mixed farming systems, in which more than 10% of the value of non-livestock farm production comes from irrigated land use.
The grassland-based and mixed systems are further categorized on the basis of climate: arid-semiarid (with a length of growing period < 180 days), humid-subhumid (Length of Growing Period or LGP > 180 days), and tropical highlands/temperate regions. This gives 11 categories in all. This system has been mapped using the methods of Kruska et al. (2003), and is now regularly updated with new datasets (Kruska, 2006). For land-use/cover, we use version 3 of the Global Land Cover (GLC) 2000 data layer (Joint Research Laboratory, 2005). For Africa, this included irrigated areas, so this is used instead of the irrigated areas database of Döll and Siebert (2000), which is used for Asia and Latin America. For human population, we use new 1-km data (GRUMP, 2005). For length of growing period, we use a layer developed from the WorldCLIM 1-km data for 2000 (Hijmans et al., 2004), together with a new "highlands" layer for the same year based on the same dataset (Jones and Thornton, 2005). Cropland and range-land are now defined from GLC 2000, and rock and sand areas are now included as part of rangelands.
The original LGP breakdown into arid-semiarid, humid-subhumid and highland-temperate areas has now been expanded to include hyper-arid regions, defined by FAO as areas with zero growing days. This was done because livestock are often found in some of these regions in wetter years when the LGP is greater than zero. Areas in GLC 2000 defined as rangeland but having a human population density greater than or equal to 20 persons per km2 as well as an LGP greater than 60 (which can allow cropping) are now included in the mixed system categories.
The landless systems still present a problem, and are not included in version 3 of the classification. Urban areas have been left as defined by GLC 2000. To look at possible changes in the future, we use the GRUMP population data and project human population out to 2030 and 2050 by prorata allocation of appropriate population figures (e.g., the UN medium-variant population data for each year by country, or the Millennium Ecosystem Assessment country-level population projections). LGP changes to 2030 and 2050 are projected using downscaled outputs of coarse-gridded GCM outputs, using methods outlined in Jones and Thornton (2003).
A.5.7.3 Application
The mapped Seré and Steinfeld (1996) classification was originally developed for a global livestock and poverty mapping study designed to assist in targeting research and development activities concerning livestock (Thornton et al., 2002; 2003). Estimates of the numbers of poor livestock keepers by production system and region were derived and mapped. This information was used in the study of Perry et al. (2002), which was carried out to identify priority re- |
|
search opportunities that can improve the livelihoods of the poor through better control of animal diseases in Africa and Asia. Possible changes in livestock systems and their implications have been assessed for West Africa (Kristjanson et al., 2004). The methods have recently been used in work to assess the spatial distribution of methane emissions from African domestic ruminants to 2030 (Herrero et al., 2008), and in a study to map climate vulnerability and poverty in sub-Saharan Africa in relation to projected climate change (Thornton et al., 2006).
A.5.7.4 Uncertainty
Uncertainties in the scheme are outlined in Table A.5.12, together with levels of confidence for scenario calculations in Table A.5.13.
A.5.8 Global Methodology for Mapping Human Impacts on the Biosphere (GLOBIO 3)
A.5.8.1Introduction
Biodiversity as defined by the Convention on Biological Diversity (CBD) encompasses the diversity of genes, species, and ecosystems. The 2010 target agreed on by the CBD Conference of the Parties (COP) in 2002 specifies a significant reduction in the rate of loss of biodiversity.
Biodiversity loss is defined as the long-term or permanent qualitative or quantitative reduction in components of biodiversity and their potential to provide goods and services, to be measured at global, regional and national levels. A number of provisional indicators of biodiversity loss have been listed for use at a global scale and suggested for use at a regional or national scale as appropriate (UNEP, 2006). These indicators include trends in the extent of biomes/ecosystems/ habitats, trends in the abundance or range of selected species, coverage of protected areas, threats to biodiversity and trends in fragmentation or connectivity of habitats.
The GLOBIO3 model produces a response indicator on an aggregated level, called the Mean Species Abundance (MSA) relative to the original abundance of species in each natural biome. The model incorporates this indicator in scenario projections, being uniquely able to project trends in the abundance of species (SCBD/MNP, 2007). A large number of species-climate or species-habitat response models exist, which examine the response of individual species to change. GLOBIO3 differs from these models as it measures habitat integrity through the lens of remaining species-level diversity, rather than individual species abundance.
A.5.8.2 Model structure and data
The GLOBIO 3 model framework describes biodiversity by means of estimating remaining mean species abundance of original species, relative to their abundance in primary vegetation. This measure of MSA is similar to the Biodiversity Integrity Index (Majer and Beeston, 1996) and the Biodiversity Intactness Index (Scholes and Biggs, 2005) and can be considered as a proxy for CBD indicators (UNEP, 2004).
The core of GLOBIO 3 is a set of regression equations describing the impact on biodiversity of the degree of pressure using dose-response relationships. These dose-response relationships are derived from a database of observations of species response to change. The database includes separate |