World Agroforestry Centre - Transforming Lives and Landscapes

A Major land use - The Proof

A World Agroforestry Centre study used remote sensing data to analyse the extent of tree cover on agricultural land, and its relationship with population density and climate. Over 1 billion hectares of agricultural land – or 43% – have more than 10% tree cover, and these areas are home to almost a third of the 1.8 billion people who live on agricultural land. Some 0.6 billion hectares of agricultural land have more than 20% tree cover, and 160 million hectares more than 50%. "Before we conducted the study, the only figures available were guesstimates,” explains Richard Coe, co-author of Trees on Farm: Analysis of Global Extent and Geographical Patterns of Agroforestry. These varied wildly, with one as low as 50,000 hectares and another of over 307 million hectares, the latter figure being based on the assumption that 20% of agricultural land is covered with trees. "There are limitations to our study,” continues Coe, "but it is a significant step in the right direction.”

Agroforestry is a feature of agriculture landscapes throughout the world, but the extent to which it is practised varies from region to region. It is particularly significant in Central America; less so in East Asia. There is a strong positive correlation between tree cover and humidity, but the relationship between tree cover and population density is less clear. This is presumably because other factors, such as markets, government policies, development programmes and local history, also influence the level of tree cover on farmland. The study has several limitations. For example, tree cover estimates are based on computer analysis of remote sensing of one kilometre square pixels. Fifty per cent tree cover in a square kilometer could mean one large block of trees – in other words, a small forest – or an even scattering across farmland. And the analysis provides no information about the nature and use of trees on farmland.

The global figures for tree cover are almost certainly conservative. There are large areas of agroforestry that are excluded from agricultural land, such as the jungle rubber systems in Indonesia and cocoa agroforestry in West Africa. In global land cover databases these areas are usually classified as forest, not as agricultural land. Trees on Farm contains some important messages for politicians, climate-change negotiators, development specialists and others in a position to influence policy. It provides firm evidence that large areas of agricultural land contain significant tree cover; it also suggests that certain areas – for example, along the fringes of the Sahara desert – could support many more trees on farms than they currently do. "What is needed now is a series of much more detailed analyses that provide a better understanding of where people plant trees, why they keep them and how they use them,” says Coe. Recent research conducted by the World Agroforestry Centre in India (see box) and Indonesia is beginning to do precisely that.

“Before we conducted the study, the only figures available were guesstimates.” Richard Coe

Over 1 billion
hectares of
agricultural land –
almost half of the
world’s farmland
– have more than
10% tree cover;
160 million
hectares have more
than 50% tree
cover.

Focus on India

"If you know how many trees there are on agricultural land, that’s useful,” explains Pal Singh, the World Agroforestry Centre’s Regional Coordinator for South Asia. "But it’s much more useful if you know which species they are, and what they provide to farmers.”

A recent study conducted by Pal Singh and AN Singh provides the most thorough analysis to date of the extent of agroforestry in India. The scientists looked at satellite imagery analysis carried out by the Forest Survey of India for 120 selected districts and the Punjab state. Detailed analysis was conducted for Yamuna Nagar district in Haryana, and a number of villages in Lucknow district of Uttar Pradesh. The scientists used different methods of sampling on remotely sensed data to analyse the nature and extent of linear plantations, such as avenues along canals and roads, block plantations and scattered trees, at different levels.

Countrywide, the most important agroforestry tree was mango, followed by neem and coconut. Not surprisingly, there was considerable variation between states, with just 0.3% tree cover on farmland in Sikkim to 13% in the Lakshadweep. In Punjab, almost half the trees on farms are eucalypts and poplars. In Kerala, mango, coconut and other fruit trees predominate.

But does this have any implications for policy makers? "Studies like this will provide important information to central government and the states,” says Pal Singh, "and they will certainly be useful to the Greening India Programme.” Under this programme, central government has stipulated that all states must have 33% tree cover by the year 2020. This, it is hoped, will encourage carbon sequestration and restore degraded lands.

Some states will be able to achieve their targets by planting more trees on state-owned forest land, but for those lacking forest land, the increase will have to come from planting trees on agricultural land – in other words, through agroforestry.