Subzone data for farm area and output for the larger countries are relatively accessible for Australia, Brazil, China, and the United States but not for Canada, India, and especially Russia.
For Australia, the Australian Bureau of Statistics (ABS 2005) provides data on farm area and value of agricultural products sold by province, and for all the provinces there is an obvious mapping to the specific geographical subzone (although for two provinces, the amounts are divided evenly between the zones they straddle).
For Brazil, data on farm area and output are available by major geographical region from the agricultural census of 1995-96 (IBGE 2006). With minor reallocations, these regions correspond to the subzones used here (see appendix D).
For China, a special compilation by the US Department of Agriculture (ERS 2006a) provides data on production (in 2003) and cultivated area (in 1996) by province, and for almost all provinces there is an obvious map ping to the specific geographical subzone (although for one province the amounts are divided evenly between the two zones it straddles).
For the United States, the agricultural census of 2002 provides statelevel data on farm area and value of agricultural products sold (USDA 2004). The states map relatively clearly into the subzones used here.2
In all four countries, the proportional distributions of farm area or output data from the source in question across the subzones is applied to the country total farm area and output estimate from this study to obtain the subzone farm area and output estimates.
For Canada, Statistics Canada (2007) reports farm area by province from the 2001 agricultural census. The provinces broadly conform to subzones used here (but Alberta is divided evenly between Canada Southeast and Canada Pacific Coast). Because the census reports zero farms or farm area above latitude 60°N, farm area and output for this study's subzones Canada Arctic and Canada Northwest Territories are set to zero. In the absence of data on output in each province, it is assumed that the percentage distribution across provinces is identical for output and farm area.
For India, the US Department of Agriculture (ERS 2006b) indicates that almost the entire land area of the country is in cropland. Accordingly, the direct estimate of the land area in each subzone from the mapping of grid cells in this study provides the basis for the proportional distribution of farm land, and the corresponding hectare amounts are obtained by applying these proportions to the census estimate of India's total farm area (FAO 2006a). Land area is once again used as a proxy for output, which, as for Canada, results in a uniform estimate of output per hectare for each subzone.
Subnational data are least available for Russia. In this case, a rough approximation is made by applying broad ranges for the density of permanent crops plus arable land indicated in a country mapping summary by the FAO (2006b). These ranges are wide, and the point estimate chosen for each region is set within the respective range such that the aggregate farm land area for Russia as a whole equals a target estimate.3 This estimate in turn is set at twice the FAO (2005) estimate for arable land plus permanent crops. Arable land and permanent crops are placed at only 7 percent of total land area, so even the doubling to include pasture (approximately the relationship in the United States) leaves a relatively low estimate for the amount of farm land in Russia.
2. Note, however, that Colorado, Kansas, and Utah are each divided evenly between subzones US Rockies and Plains and US Southwest and Plains (see table 4.2); Missouri between Lakes and Northeast and US Southeast; and Oklahoma between US Southeast and US Southwest and Plains.
3. The resulting fractions of total land area in farms for the subzones used in this study are as follows: Caspian Black Sea, 55 percent; Far Eastern, 5 percent; North European, 20 percent; North Urals Siberia, 5 percent; Northeast Siberia, 5 percent; South Urals Siberia, 25 percent; and Southeast Siberia, 15 percent.
Was this article helpful?