| PAGE
MEASURES AND INDICATORS |
DATA SOURCES AND COMMENTS |
| Extent and Area Intensity of Agriculture | Reinterpretation of Global Land
Cover Characterization Data - 1km resolution satellite data for the
period April 1992 to March 1993 set (GLCCD 1998; USGS EDC 1999a). |
| Agricultural Land Use Balance and Trends | FAO national tabular data for 1965-97 (FAOSTAT 1999). |
| Agroclimatic Factors and Generalized Slope | Global Agroecological Zones
(GAEZ) database at half degree resolution (FAO/IIASA 1999). Includes
climate variables based on 30 years of monthly data (UEA 1998). |
| Percent and Area of
Land Equipped for Irrigation |
University of Kassel global spatial data at half degree resolution (Döll and Siebert 1999). |
| Generalized Agroecosystem Characterization | Combination of PAGE data on extent, agroclimate, slope, and irrigation. |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨ Cropland and
managed pasture detected by satellite interpretation cover some 28
percent of global land surface. Globally, land with greater than 60
percent agriculture occupies 41 percent of the PAGE agricultural
extent, while land with 40–60 percent and 30–40 percent agriculture
occupies 29 and 30 percent of the PAGE agricultural extent,
respectively. Overall, 31 percent of agricultural areas are occupied by crops and the remaining 69 percent are under pasture. Annual cropland is relatively stable at around 1.38 billion hectares, while permanent crops occupy around 131 million hectares (Mha) and show a net growth of almost 2 percent per year. Pasture areas are estimated to be increasing at around 0.3 percent per year. ¨ 91 percent of cropland is under annual crops, such as wheat, while perennial crops, such as citrus and tea, occupy the remainder. Although annual cropland is stable, the harvested area of annual crops is increasing at around 0.3 percent per year. The cropping intensity for annual crops globally currently stands at around 0.8. ¨ Irrigated areas occupy 270 Mha, around 5.4 percent of global agricultural land and 17.5 percent of all cropland. Irrigated area continues to expand, but at a slowing rate, now around 1.6 percent (about 3.3 Mha) per year. This net amount is presumed to allow for irrigated area losses—estimated as up to 1.5 Mha per year from salinization. ¨ 38 percent of the area within the satellite derived global extent of agriculture is found in temperate regions, another 38 percent in tropical regions, and some 23 percent in subtropical regions. |
The resolution and
interpretation of available global satellite data is generally
insufficient to reliably distinguish all types of agricultural land
cover. Extensive pasture, irrigated areas, fallow lands, and farmed
transition areas between cropland and forest are particularly
problematic. Finer resolution satellite data, expected to be available
over the coming years, should improve our ability to reliably detect
major agricultural land cover types. ¨ The utility of remotely sensed data would improve with more frequent interpretations (e.g., every three to five years) which focus on detecting land cover change at a global scale. This would require improved data resolution, more systematic classification processes, and innovative approaches to ground-truthing. ¨ The consistency and reliability of FAO national (tabular) agricultural land use data vary significantly. FAO no longer reports pasture and forest land cover types. Furthermore, national level agricultural land use data provide insufficient spatial disaggregation to help characterize agroecosystems. ¨ Looking solely at year-to-year net changes in land use gives an unrealistically conservative impression of the true dynamics of land use change. Additional information is required on conversions to and from agricultural land as well as between important agricultural land uses. ¨ The University of Kassel has developed the best available digital map of irrigation at a global scale, but the accuracy of the map is variable because of inconsistent scale, age, and reliability of source data. Efforts are continuing to improve this data set (e.g., by linking to published and unpublished elements of FAO’s AQUASTAT databases). ¨ Data on production systems and resource management aspects of land use are extremely scarce at regional and global scales. Proxy measures include the following: statistical data on crop types, areas, and yields; animal populations and products; and the use of labor, irrigation, fertilizers, pesticides, and modern crop varieties. Although such data provide broad notions of land use and management, they are often difficult to obtain at subnational levels, are seldom interrelated, and give no indication of the scale of enterprise, the temporal and spatial arrangements of production systems, conservation practices, and so on. |
| PAGE MEASURES AND INDICATORS | DATA SOURCES AND COMMENTS |
| Production and
Productivity: Crop and pasture areas (hectares) Yields (metric tons per hectare) |
FAO national tabular data for 1965-97 (FAOSTAT 1999). |
| Intensity of Input
Use: (see Water Services for irrigation indicators) Fertilizer application - Nitrogen (N), phosphorus (P2O5), and potassium (K2O): NPK (kg per hectare) Pesticide application (kilograms per hectare) Labor (agricultural workers per hectare) Tractors (hectares per tractor) |
FAO national tabular data for 1965-97 (FAOSTAT 1999). Yudelman 1998. FAO national tabular data for 1965-97 (FAOSTAT 1999). FAO national tabular data for 1965-97 (FAOSTAT 1999). |
| Value of
Agricultural Production ( VoP): Total VoP ($) VoP per hectare of cropland ($ per hectare) |
FAO national tabular data for 1965-97 (FAOSTAT 1999); Prices: (FAO 1997a). |
| Nutritional Value: Calories, protein, and fat per person |
FAO Food Balance Sheets (FAOSTAT 1999) and FAO World Food Survey (1996). |
| Employment and
Income: Number of agricultural workers Value of production per agricultural worker |
FAO national tabular data for 1965-97 (FAOSTAT 1999). World Development Indicators (World Bank 2000). |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨ Food
production from agroecosystems is valued at around $1.3 trillion per
year (1997) and provides 94 percent of the protein and 99 percent of
the calories consumed by humans. The production process directly
employs some 1.3 billion people. ¨ Food production has more than kept pace with global population growth. On average, food supplies are 24 percent higher per person than in 1961, and real prices are 40 percent lower. Over the same period, the global population has doubled from 3 to 6 billion people. Approximately 790 million people in the developing world are still chronically undernourished, almost two-thirds of whom reside in Asia and the Pacific. ¨ Although the global expansion of agricultural area has been modest in recent decades, intensification has been rapid, as irrigated area increased, fallow time decreased, and the use of purchased inputs and new technologies grew to produce more output per hectare. ¨ The dominant share of the world’s cropland (59 percent) is dedicated to cereal production, but cereal yield growth rates have generally slowed in recent times. ¨ Over the past 30 years, the quantity of livestock products has approximately tripled compared to a doubling of crop outputs. This high rate of growth in livestock demand is expected to continue as, globally, standards of living and average incomes continue to rise. ¨ Increased demand for both crop and livestock products will come predominantly from developing countries, and because of infrastructure, institutional, and other trade and marketing constraints, will often need to be met from improved local agroecosystem capacity. ¨ The generally positive current trends in food production may mask negative trends in the underlying biophysical capacity of agroecosystems, e.g., nutrient mining, soil erosion, and overextraction of groundwater resources. But natural resource data is often too limited in space and time to gauge the full scope of such impacts. ¨ Environmental problems often associated with high-input agroecosystems include salinization of irrigated areas, nutrient and pesticide leaching, and pesticide resistance. Those more associated with low-input and extensive agroecosystems include soil erosion and loss of soil fertility. ¨ The specific mix of inputs and production technology has a direct bearing on the long-term capacity of agroecosystems to provide goods and services. Management practices can change rapidly in response to market signals and new technological opportunities, and can compensate for some aspects of resource degradation. Resource degradation increases reliance on the use of external (often purchased) inputs to maintain production levels. |
¨ The FAO tabular data was
taken from national statistics but consistency and reliability among
countries and years may vary. Additionally, the geopolitical basis of
area and yield data limit their use for agroecosystem assessments that
require such data be reaggregated into agroecosystem units. ¨ Enormous regional disparities exist among the indicators of yield, nutrition, value, and income. More local data is essential to better appreciate the nature and source of these differences and to target appropriate food and environment related policy and technology interventions. ¨ Soil nutrient balances complement yield data to provide a better understanding of agroecosystem condition. Soil nutrient balance trends, characterized by production systems (not simply by commodity) and by nutrient source (both inorganic and organic), would greatly enrich debates on the design and targeting of appropriate, integrated nutrient strategies. ¨ Relative to the other goods and services considered in this study, data on food, feed, and fiber is the most complete. ¨ |
| PAGE MEASURES AND INDICATORS | DATA SOURCES AND COMMENTS |
| Inherent Soil
Constraints: Dominant soil fertility constraints Area free of soil constraints |
Fertility Capability Classification (FCC) (Sanchez et al. 1982; Smith 1989; Smith et al. 1997). FCC modifiers assessed for each soil mapping unit (SMU) of FAO’s Digital Soil Map of the World at a resolution of 5x5 km (FAO 1995). SMUs comprise multiple soil types whose area shares but not location are known. This limits the spatial interpretation of such data. |
| Status and Change in
Soil Quality: Severity (extent and degree) of soil degradation |
Global and regional (South and Southeast Asia) assessments of human-induced soil degradation (GLASOD: Oldeman et al. 1991a; and ASSOD: Van Lynden and Oldeman 1997) at scales of 1:10m and 1:5m respectively. Data based mainly on expert opinion for interpretation at regional and global scales (GLASOD) and more national and quantitative information (ASSOD). |
| Soil Organic Matter
(SOM): Organic carbon content in the upper 100 centimeters of soil (metric tons per hectare) |
WISE global data set of soil profiles interpreted for carbon content and soil type (Batjes 1996; Batjes 2000), then applied to FAO’s Soil Map of the World mapping units (FAO 1995). |
| Soil Nutrient
Balance: Net nutrient flux (kg NPK per ha) |
IFDC country- and
crop-specific estimates of nutrient balances for Latin America and the
Caribbean: mid-1980’s and mid-1990’s (Henao 1999). |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨ Combining the
GLASOD soil degradation map with the PAGE agricultural extent suggests
that human-induced degradation since the mid 1900s is more severe
within agricultural lands. Over 40 percent of the PAGE agricultural
extent coincides with GLASOD mapping units that contain moderately
degraded areas, and 9 percent of the extent coincides with mapping
units that contain strongly or extremely degraded areas. Strong or
extreme degradation implies that soils are very costly or infeasible to
rehabilitate to their original (mid-1900s) state. This degradation is
estimated to have reduced crop productivity by around 13 percent. No
global estimates of improving soil quality are known to exist. ¨ Over three quarters of the PAGE agricultural extent contains soils that are predominantly constrained (>70 percent of area has some soil fertility constraints). Just over half the agricultural extent is in flatter lands (up to 8 percent slope). Only 6 percent of land within the PAGE agricultural extent is both flat and relatively free of soil constraints. Most of this land lies in temperate regions. ¨ Depletion of SOM is widespread, reducing fertility, moisture retention, and soil workability, and increasing CO2 emissions. Good land use practices can rebuild SOM levels. ¨ Salinization (an accumulation in the soil of dissolved salts) includes both natural and human-induced (secondary) salinization and occurs on agricultural and nonagricultural, and irrigated and nonirrigated land. Although salinization data is poor, rough estimates indicate about 20 percent of irrigated land suffers from salinization. Around 1.5 million hectares of irrigated land per year are lost to salinization and about $11 billion per year in reduced productivity, or just under 1 percent of both the global irrigated area and annual value of production. Salinization also affects water quality. ¨ Regional analysis of soil nutrient dynamics in Latin America and the Caribbean suggests that for most crops and cropping systems the nutrient balance is significantly negative, although depletion rates appear to be declining (because of substantial growth in fertilizer application in some countries). Previous analyses have demonstrated similar negative balances in Africa. |
Relative to its economic and
environmental importance, the current lack of comprehensive, reliable,
and up-todate global soil quality data is woeful. Developing reliable,
cost-effective methods for monitoring soil degradation is imperative to
help to mitigate further losses of productive capacity as well as to
target rehabilitation efforts. ¨ The long-term monitoring of SOM and soil biota are increasingly viewed as a strategic means of measuring progress toward achieving sustainable agriculture and of keeping abreast of degradation trends. ¨ Efforts are underway to apply remote sensing methods to soil organic matter monitoring but an adequate ground-truthing capacity still needs to be developed. ¨ National soil research and survey agencies need to strike a more appropriate balance between traditional soil survey activities and fostering scientific and farmer capacity to monitor soil condition on an ongoing basis. ¨ Much more work is needed to develop indicators (and underlying scientific evidence) that relate land use and management, soil quality indicators, production, and economic and environmental outcomes. This is particularly so for better articulating the role of soil quality in providing environmental goods and services from agroecosystems. ¨ Soil nutrient balance data is available at a national level for Latin America and the Caribbean and Africa (Henao 1999; Henao and Baanante 1999) and subnationally for Sub-Saharan Africa (Smaling et al. 1997). Further efforts to make consistent nutrient balance assessments would provide greater insights into the spatial and temporal patterns of agroecosystem productivity. |
| PAGE MEASURES AND INDICATORS | DATA SOURCES AND COMMENTS |
| ¨Water Supply
for Rainfed Agriculture: Rainfall (mm per year) Length of growing period (LGP in days) LGP variability |
Global estimates of rainfall based on spatial extrapolation of monthly data from rainfall and climate stations over 30 years (UEA 1998). Global length of growing period (LGP) estimates based on University of East Anglia (UEA) data using a water balance model (FAO/IIASA 1999). |
| Water Use for
Irrigation: Area equipped for irrigation (percent) |
University of Kassel 1999 global
map at half degree resolution (Döll and Siebert 1999);Global Land Cover Characterization Data set (GLCCD 1998; USGS EDC 1999a). |
| Irrigation depth (mm
per year) |
Estimates of national water
extraction for agriculture (WRI 1998) and area irrigated (FAOSTAT 1999;
Seckler et al. 1998). |
| Efficiency (ratio of crop water use to amount extracted) | Country and crop-specific factors (Seckler et al. 1998) applied to the WRI and FAO data, used to determine irrigation depth. |
| Effects of
Agriculture on Water Supply: Proportion of major watersheds occupied by agriculture |
Global digital map of watersheds (UNH – GRDC 1999). |
| Potential Water
Quality Effects: Soil salinization |
Global salinization estimates (Postel 1999; Ghassemi et al. 1995). |
| Fertilizer application rates | National fertilizer consumption (FAOSTAT 1999b). |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨¨ A fifth
of the agricultural extent is in arid or dry semiarid, a quarter in
moist semiarid, over a third in subhumid, and a fifth in humid regions.
Much of the agricultural extent suffers from high rainfall variability.
Global climate change is likely to affect rainfall distribution
significantly. ¨ Irrigation accounts for 70 percent of the water withdrawn from freshwater systems for human use. Of that only 30-60 percent is returned for downstream use, making irrigation the largest net user of freshwater globally. ¨ The 17 percent of global cropland that is irrigated produces an estimated 30-40 percent of the world’s crops. The share of cropland that is irrigated has grown quickly, increasing 72 percent from 1966–1996. ¨ Competition with other water uses, especially drinking water and industrial use will be most intense in developing countries, where populations and industries are growing fastest. Agriculture may increasingly depend upon water recycled from domestic and industrial uses. ¨ There is an urgent need to increase irrigation water use efficiency. If efficiency can be improved, less water would need to be extracted from rivers and aquifers per ton of food, feed, or fiber produced. If the excess water is not used to expand food production, improvements in efficiency could help mitigate negative environmental effects of water extraction. ¨ Over 50 percent of total river basin area is under agricultural cover in the major watersheds of Europe and South Asia; over 30 percent of total basin area is under agricultural cover in large parts of the United States, South America, North Africa, Southeast Asia, and Australia. ¨ If not carefully managed, agricultural intensification in high external input agroecosystems can result in leaching of mineral fertilizers (especially nitrogen), pesticides, and animal-manure residues into water courses. ¨ Inadequately managed intensification particularly on more sloping lands with lower quality soils tends to increase soil erosion as well as the effects of sediment on aquatic systems, hydraulic structures, and water usage. |
¨ Hydrological monitoring,
especially for groundwater levels, river flow, and water quality, is
inadequate and, in many cases, has declined in recent years. More
cost-effective methods of water resource assessment and hydrological
monitoring are needed. ¨ Even with reliable water quality data it is often difficult to relate water quality changes directly to agricultural activities (e.g., the effects of pesticides from private and public gardens, and the contribution of nutrients from sewage and industrial processing). ¨ The University of Kassel irrigation map is the best currently available but accuracy is variable because of inconsistent source data (Döll and Siebert 1999). The satellite interpreted Land Cover Characterization Data set (GLCCD 1998; USGS EDC 1999a) is unreliable in detecting irrigated areas, particularly in South America, Africa, and Oceania. Better data is needed on irrigation location, type, and water use (e.g., FAO’s AQUASTAT (2000a)). ¨ Satellite-derived data on rain-use efficiency is now available (University of Maryland 1999) and its suitability as a regional scale indicator of water supply for rainfed agroecosystems merits further investigation. Related indicators of farmer use and management of rainwater need to be developed and tested. |
| PAGE MEASURES AND INDICATORS | DATA SOURCES AND COMMENTS |
| C¨o nservation
of Natural Habitat: Proportion of habitat area occupied by agriculture |
WWF-US Global 200 Ecoregions
(Olson et al. 1999). Major habitats based on broad environmental characteristics and expert opinion. |
| Pressure on
Protected Areas: Proportion of protected areas occupied by agriculture |
Global protected areas database (WCMC 1999). |
| Habitat Quality in
Agricultural Areas: Proportion of tree cover within the PAGE agricultural extent |
Global percentage tree cover
data set based on 1 km resolution AVHRR satellite data (Defries et al. 2000). |
| Agrobiodiversity: Species Diversity: Crop, livestock, tree crop species |
Crop and livestock species (FAO); tree crop species (Kindt et al. 1997). |
| Share of Modern Varieties | Global estimates of the adoption of modern crop varieties (Byerlee 1996; Smale 2000). |
| Germplasm
Conservation: Land races and wild species in ex-situ and in-situ collections. |
Data from major germplasm
holding institutions: CGIAR China, the United States, and Russia and global estimates of ex-situ holdings (Evenson et al. 1998). |
| Area Planted to Transgenic Crops | Percentage estimates for select
countries compiled from agricultural survey and census (James 1999). |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨¨
Agricultural land, which supports far less biodiversity than natural
forest, has expanded primarily at the expense of forests. ¨ About 30 percent of the potential area of temperate, subtropical, and tropical forests has been converted to agriculture. ¨ Many of the areas established to protect biodiversity fall in or around agricultural lands, increasing the difficulties of effective protection. ¨ Biodiversity loss is often considerable within high-input agroecosystems, but low-input and extensive systems can also bring about significant biodiversity loss through increased conversion of natural habitats. ¨ Although tree cover is fairly low in agricultural lands of many parts of the world, a majority of rainfed agricultural land in Latin America, Sub-Saharan Africa, and South and Southeast Asia has significant and increasing tree cover, which enhances habitat for wild biodiversity. ¨ A number of agricultural systems and management strategies, such as fallowing, agroforestry, shaded coffee, and integrated pest management, can encourage diversity as well as productivity. ¨ Of the 7,000 crop species used in agriculture, only 120 are important at a national level. An estimated 90 percent of the world’s calorie intake comes from just 30 crops. The number of livestock breeds has declined greatly in the past 100 years. The number of domesticated tree crops has increased. ¨ In the early 1990s, crop area sown to modern varieties of rice and wheat in developing countries had reached around 75 percent, and for maize 60 percent. ¨ The global area planted with transgenic crops, some 82 percent of which was in OECD countries, increased from only 1.7 million hectares in 1996 to 39.9 million in 1999. The seven principal transgenic crops grown in 1998 were soybean, maize, cotton, canola (rapeseed), potato, squash, and papaya. |
¨ The currently published
national data on land use show net land use changes and, therefore,
understate the true scale of agricultural conversion impacts on
biodiversity. Higher resolution data on land conversion (both spatially
and temporally) is needed. ¨ The size of many protected areas is known but often their precise geographic boundaries are unknown. Increased precision in the delineation of protected area boundaries (as well as in the spatial delineation of agricultural extent) are needed to enhance the quality of the protected area pressure indicator. ¨ Improved road network and land use/cover data would help determine the level of habitat fragmentation in agricultural landscapes. ¨ Improved data on production systems diversity could be used as a proxy for the potential area and quality of wildlife habitat within agricultural areas. ¨ Data on the rate and extent of adoption of modern varieties is relatively sparse, regionally inconsistent, and limited to the major food crops. ¨ International initiatives are well in hand to georeference germplasm accessions (e.g., WIEWS and ICIS databases). This should considerably enhance agrobiodiversity assays. ¨ There is insufficient data on the abundance of wild flora and fauna in and around agricultural production areas, and on the impacts of specific crop combinations and management changes on wildlife populations. |
| PAGE MEASURES AND INDICATORS | DATA SOURCES AND COMMENTS |
| V¨ egetation
Carbon in the PAGE Agricultural Extent: Above- and below-ground live vegetation carbon (metric tons per hectare) |
Estimates of carbon density in
above- and below-ground live vegetation by ecosystem type (Olson et al. 1983; USGS EDC 1999b) applied to GLCCD (1998) land cover vegetation class extents. |
| Soil Carbon in the
PAGE Agricultural Extent: Soil carbon in the first 100cm of the soil profile, excluding litter (metric tons per hectare) |
WISE global data set of soil
profiles interpreted for carbon content and soil type (Batjes 1996; Batjes 2000), then applied to FAO’s Soil Map of the World mapping units (FAO 1995); Latin America and the Caribbean (LAC) example uses the Latin America and the Caribbean Soil and Terrain (SOTER) database (ISRIC 1999; FAO 1998). |
| Carbon-based
Greenhouse Gas (GHG) Emissions from Agriculture: Carbon dioxide (CO2) emissions Methane (CH4) emissions |
CO2 land use change emissions
(Houghton and Hackler 1995); CH4 emissions (Stern and Kaufman 1998); Land conversion, livestock, and paddy rice trends (FAOSTAT 1999). |
| CONDITIONS AND TRENDS | INFORMATION STATUS AND NEEDS |
| ¨
Agroecosystems’ share of carbon storage is estimated to be 18-24
percent of global total. ¨ In agricultural areas, the carbon stored in soils is generally more than double that stored in the vegetation that these soils support (102 metric tons/hectare versus 17-47 metric tons/hectare). ¨ Regional studies for Latin America and the Caribbean show that almost half of the soil carbon is stored in the top 30 cm of the soil profile¾the depth most accessible to the rooting systems of annual crops and pasture. ¨ The primary sources of agriculture-based carbon emissions are biomass burning and methane emissions from livestock and paddy rice production. ¨ CO2 emissions related to land use change have increased dramatically since the mid 1800s. Most land use change CO2 emissions are now taking place in developing countries. ¨ Livestock is the largest agriculture-related source of GHG emissions; the growth in livestock populations is also taking place primarily in developing countries. ¨ Cropland and pasture management strategies that result in improved soil organic matter content also increase carbon sequestration capacity and, thus, help reduce agriculture-induced GHG emissions. |
¨ Data on carbon storage
could be improved by better characterization of agricultural land cover
and vegetation carbon content, and by a denser network of soil profile
data from agricultural soils. ¨ Efforts are underway to apply remote sensing methods to soil organic matter monitoring but adequate groundtruthing capacity would still need to be developed. ¨ It is possible to use satellites to monitor the incidence and severity of fires and hence estimate CO2 and CH4 emissions from biomass burning. However, attributing specific fire incidents to agricultural practices is difficult. ¨ Monitoring trends in paddy rice areas, ruminant livestock numbers and feed quality, and total livestock provides reasonable proxies for trends in methane emissions from agricultural sources. ¨ FCCC related research activities into the potential for emission reduction credits from land use change is already generating, and will likely continue to generate, much relevant data and a greater understanding of soil carbon dynamics in agroecosystems. |