Is cultivable land a scarce resource?
Many publications warn about the scarcity of cultivable land and how it is being degraded and diminished due to various development types (including urbanization), in the context of an increasing human population with expanding needs (Millennium Ecosystem Assessment, 2005; Oldeman et al., 1991). Research aimed at assessing global and regional areas of cultivable land, whether in use or not, is even scarcer still (Fischer et al., 2002; Ramankutty et al., 2005).
There are four reference databases available on this subject: the FAO's FAOSTAT, which provides information on the area under cultivation; the fieldwork element of the Global Agro-Ecological Zones (GAEZ) study by the International Institute for Applied Systems Analysis (IIASA) and the FAO, which provides information on cultivable land; the database of the Center for Sustainability And the Global Environment (SAGE) of the University of Wisconsin and also that of the University McGill which contains data on cultivable and cultivated areas.
This article focuses on a comparison of the GAEZ and FAOSTAT databases.
Cultivable and cultivated land: definitions and uncertainties
The GAEZ study evaluates global land suitability for the cultivation of 154 plant varieties and the potential yields in function of three main crop management methods - "advanced", "improved" and "traditional" - and depending on whether crop production is rain-fed or irrigated. The study classifies land into five categories according to its capacity for cultivation, by comparing an area's yield potential with the maximum yield attained in the same climatic zone. On this basis, land is classified as: very suitable if the yield potential exceeds 80% of the maximum yield attained, suitable if this ratio is between 60% and 80%, moderately suitable between 40% and 60%, marginally suitable between 20% and 40%, and not suitable below 20% (Fischer et al., 2002). The GAEZ database shows corresponding areas of land, as well as those under forest cover, and indicates the infrastructure range.
The FAOSTAT database provides estimates of areas of "arable land
" and land under "permanent crops
", the sum of which we call "cultivated land".
There are of course many uncertainties in the FAOSTAT and GAEZ databases. A. Young, soil scientist, claims that cultivable land might be overestimated by 10% to 15% and that cultivated land might be underestimated by 10% to 20% (Young, 1999).
From these two databases, land currently under cultivation represents nearly 40% of useable land in rain-fed agriculture (irrigation not necessary) worldwide, i.e. about 1,560 million hectares out of 4,150 million. By taking into account the maximum error suggested by A. Young, this proportion would increase to 53%, or 1,874 million hectares out of 3,529 million.
Three hypotheses for the expansion of cultivated land, based on IIASA and FAO data
With knowledge of the various land category areas that can be used in rain-fed agriculture (GAEZ study), and the areas of cultivated land in 2005 (FAOSTAT), one can calculate the theoretical possibilities for cultivated land expansion. With this intention, we considered three hypotheses: 1) a very restrictive scenario, which assumes that land classified as very suitable, suitable and moderately suitable could be cultivated, except for forested land and land required for urban and other infrastructure
; 2) a less restrictive scenario, which includes marginally suitable land as well as the three categories of the first hypothesis, except for forested land; and 3) a least restrictive hypothesis which also integrates all arable land under forest, i.e. one third of the world's forests.
Results and interpretation
Some aspects of this type of analysis tend to lead to an overestimation of the expansion possibilities for cultivated areas. This would occur mainly in areas currently classified as "grassland", "shrubland" or "permanent meadows and pastures", some of which are possibly located in protected areas, although it is not possible to distinguish this from the data. In addition, to be classified as suitable, an area of land only has to be suitable for growing one of the 154 species considered. However, other aspects of the analysis tend to lead to an underestimation of the expansion possibilities: the GAEZ study assumes low-yielding land to be unsuitable for cultivation and does not give due consideration for possible improvement by human intervention that could render some land cultivable.
Thus, the expansion of cultivated area on a global scale was calculated according to the three hypotheses; the results were, respectively: 1) approximately 1000 million hectares, which leads to 1.7 times more than the present area; 2) 1450 million hectares, 2 times more, and 3) 2350 million hectares, 2.5 times more.
The potential for expansion, however, varies considerably from one region to another (see figure), for example in hypothesis 1 South America and sub-Saharan Africa have a particularly high capacity for expansion, while in Asia there is almost none due to the fact that the currently cultivated area already exceeds that which could be cultivated according the three hypotheses analyzed. This means that for Asia at least, the IIASA methods clearly tend to underestimate the potential of land for cultivation, and by implication the ingenuity of people to develop land to make it cultivable.
The GAEZ study also examines the consequences of seven global warming scenarios on the areas cultivable with wheat, maize or rice: all of which predict a limited extension (1% to 6%) of cultivable land globally for these crops, though a decrease in developing countries (1.3% to 11%) and an increase in developed countries (11% to 25%).
Agricultural development and public policy
Thus, according to IIASA and FAO data, land that can be used in rain-fed agriculture and is presently uncultivated is not, and will not be soon, a globally scarce resource. This is true even if all forests and currently protected areas remain uncultivated and when the possible effects of global warming are taken into account. However, uncultivated cultivable land is indeed a limited resource, or even almost non-existent in the Middle East and Asia. In addition, only local ecological, technical, economic and social studies can provide a full understanding of whether uncultivated cultivable lands are actually available for cultivation, whether they are accessible and by whom they could be cultivated.
The fact remains that those responsible for agricultural public policy, whether national or stemming from international cooperation, have flexibility in the type of international agricultural development they will prioritize. A first option, to which most institutions currently adhere, is to pursue policies and practices that are conducive to a competitive mode of agricultural development that is uneven and indeed contradictory. Such an approach has entailed a sharp increase in labour productivity and crop yields for some of the world's production units, while hundreds of millions of other farmers have fallen into poverty, malnutrition and possibly exodus and emigration. In some areas, added to these serious social issues are environmental problems, such as salinity, pollution, biodiversity loss and more (IAASTD, 2008; Mazoyer, Roudart, 2006).
Given that in many parts of the world it is possible to cultivate larger areas than at present, an alternative solution would be to promote diversified agriculture, with relatively low yields, low inputs, that use little or no fossil fuels, have few negative effects on the environment or perhaps even positive effects, and that could ensure a decent livelihood for nearly three billion people - the world's farming population - in the true spirit of the "Doubly Green Revolution".
For this approach to succeed, the right problem must be addressed: the main challenge is political, economic and social, it concerns the development of new types of agricultural policies; it is not about the scarcity of cultivable land.
Can cultivated land be extended?