Knowledge for better food systems

Perspectives on agroforestry as a model for sustainable intensification of agriculture

In this in-depth interview, the FCRN’s Marie Persson has invited 3 experts, Roger Leakey, André Gonçalves and Ben Phalan to discuss the role, benefits and challenges with agroforestry. They provide insights and critical recommendations based on experiences analysing and implementing such systems in Brazil, South East Asia and many African countries. Special attention is paid to agroforestry as a possible pathway towards closing the yield gap and creating sustainable intensification.

Roger Leakey, is Professor at the School of Marine and Tropical Biology, James Cook University, Australia, previously Director of Research at the World Agroforestry Centre (ICRAF).

In his recent book “Living with the Trees of Life – Towards the Transformation of Tropical Agriculture” he draws upon his 30 year experience of working in the field of agroforestry and particularly with poor smallholder farmers around, Africa, Asia, Latin America and Oceania.  See also:

André Gonçalves, is Professor of Agroecology at Instituto Federal Catarinense in Brazil and Technical advisor at Centro Ecológico Brazil. A video from the conference “Landscapes in a carbon focused world” where André presented a talk on the topic “Agroforestry and Conservation projects in Brazil: Carbon, Biodiversity, Climate, and People”, is available here.

Ben Phalan, is a postdoctoral fellow at the Conservation Science Group of the University of Cambridge. His current research focuses on the impacts of agriculture on tropical faunas. For more information about his current and previous research, please visit this page.

1. How is agroforestry defined?

Roger Leakey: Agroforestry has been defined as : “a dynamic, ecologically based, natural resources management system that, through the integration of trees in farms and in the landscape, diversifies and sustains production for increased social, economic and environmental benefits for land users at all levels” (Leakey, 1996; adopted by ICRAF, 1997).

Ben Phalan: I use the following description: "Agroforestry is a summary term for practices that involve the integration of trees and other large woody perennials into farming systems through the conservation of existing trees, their active planting and tending, or the tolerance of spontaneous tree regrowth" (Schroth et al. 2004 Agroforestry and biodiversity conservation in tropical landscapes, Island Press, P. 2).

André Gonçalves: An illustration of what these systems can look like might be helpful here. Under the general concept of complex agroforestry systems there are different systems with the basic approach of copying natural patterns (structural and functional). In terms of biodiversity some systems are composed of a few species while others can contain up to 50 species. A simple agroforestry system is a combination of two or three plants, like beans cultivated between the lines of orange trees. An intercropping system where one of the cultures is arboreal perhaps defines a simple agroforestry system better. In a complex agroforestry system we don´t see such “organized” pattern. The varieties of trees are higher and the distribution of the plants on the landscape is more chaotic, following natural patterns. See the image below.

2. The spectrum of perspectives on agroforestry

Ben Phalan: In my view agroforestry should be considered alongside a range of other potential things that can be done to address the needs of people and of other species in particular landscapes. Agroforestry is one way of combining multiple objectives on the same land, but there will also be some things which can’t be combined on the same land. Biodiversity –which is the focus of my research- is a good example: there are many species which cannot persist even in the most benign, complex agroforests. For such species, agroforestry might provide some useful resources, but must at the very least be augmented by adequate protection of natural forests. Agroforestry techniques are best seen as useful tools in the toolbox of those seeking to influence tropical landscapes for the better, but they are not the only useful tools, nor are they always the most appropriate.

Roger Leakey: Agroforestry is a multidisciplinary science which builds on the successes of the Green Revolution, increasing the yields from modern crop varieties and livestock breeds, so closing the yield gap (the difference between the potential yield of a crop variety and the yield actually achieved by farmers) and increasing the returns from the investment in the Green Revolution. The multiplicity of production, environmental, social and economic outcomes from this approach is the expression of sustainable intensification in mainstream agriculture.

André Gonçalves: My view of agroforestry is simpler. Whenever farmers, mainly in tropical areas, add more arboreal endemic plants to their production area, they will be taking steps to promote this land use system. Most of the agricultural models are based on a monocultural pattern, which is completely distant from the patterns of the original ecosystem. Adding trees is a way of trying to copy nature. In my definition agroecology is all about values such as social justice and economic aspects. Otherwise, it would be reduced to the technical dimension.

3. What is the “baseline” – i.e. what was the pre-existing land use prior to the adoption of agroforestry?

André Gonçalves: This depends very much on the region we are talking about. But where we have been working, and what I am referring mostly to, the pre-existing land use was bananas in mono-cropping.

Ben Phalan: This is a really important question. It's essential to understand how a set of different things has changed in relation to the baseline: productivity, ecosystem services, social outcomes, biodiversity etc.  Also, it's not just a question of what the pre-existing land-use was, but what the other alternatives (“counterfactuals” in the jargon) could be. Some of my research looks at whether, overall, a landscape made up of some natural forest and some high-yielding cropland (such as monoculture plots of oil palm, cassava or maize) might be better for biodiversity than an agroforestry landscape, where both are producing the same amount of food overall. In southwest Ghana, the answer seems to be yes. There are further nuances to that question, and it's certainly not the last word on how that landscape should be managed, but it's the sort of question that very few people are looking at in a systematic and critical way.

Roger Leakey: There are of course a very wide range of land uses across agricultural landscapes, ranging from sustainable fallows on forest margins to totally degraded, treeless and abandoned land with no production potential without total restoration. André discussed the development of complex agroforestry systems at the forest margin. Basically, this is like turning a long natural fallow into an even longer commercial fallow where the “agroforest” mimics a natural forest, but all the trees and much of the other vegetation produces marketable food and non-food products. There is also the restoration of badly degraded land using nitrogen-fixing trees and shrubs and the progressing to the domestication of new tree crops, this is developed further in question 6 below. So, the answer to this question is that agroforestry can be applied on any of these different landuse scenarios anywhere, but it will be done differently depending on the severity of the different biophysical and socio-economic situations.

4. How does agroecology relate to agroforestry?  Does it have to be organic? What are the benefits and disadvantages of agroecology?

Ben Phalan: Agroecology is the study and application of ecological principles in agriculture. It is sometimes interpreted more narrowly, as necessarily implying farming systems with a high diversity of crop species grown together, but one can argue that even simple, species-poor systems such as push-pull maize are agroecological farming systems because of their application of ecological knowledge. Agroecology is not necessarily organic. The ecological principles most relevant to agroforestry systems are the complementary use of resources such as light and nutrients by plant species with different architectures, and – at least in principle – the greater reliability and resilience to pest outbreaks of growing multiple crops together. The benefits of agroecology are clear: farms are ecological systems, so it’s vital for farmers and agronomists to bring ecological knowledge to bear on farm management. The danger is in a narrow interpretation of agroecology that rejects any practices which are “not natural”. How “natural” a farming system looks is not necessarily a good guide to how effective it is at doing the things we want it to do: producing food, supporting people, supporting other species. In my view, we need to get better at assessing and measuring those outcomes, which are the things we’re really interested in, rather than assuming that a farming system is good because it looks “natural”.

André Gonçalves: Agroforestry is by its own nature an agroecological approach. Agroecology, as I understand, is an emphasis on the ecological perspective of agriculture, and ecological in the sense of the natural processes involved in it. As agroforestry tries to get closer to the way nature works, it is intrinsically related with agroecology. It is not necessarily organic, but preferably. I don´t see disadvantages of agroecology, since it is a process of increasing sustainability, food production, environmental services, income for farmers, etc.  In my definition agroecology is all about values such as social justice and economic aspects. Otherwise, it would be reduced to the technical dimension. A couple of years ago we coordinated an extension project and I think it had a very good and representative title: Agroecology: social solutions for environmental problems, and environmental solutions for social problems.

Roger Leakey: Agroforestry is a low input approach to agriculture which is based on applied ecology, indeed it is a way of developing an agroecological succession, akin to the ecological succession that in its mature state is a primary forest. However, agroforestry also recognizes the needs of people for better livelihoods, social justice and the need for economic development. So agroforestry is agroecology and socio-economic development. Many farmers in the developing world are organic by default, because they cannot afford to purchase agricultural inputs like fertilizers and pesticides. Agroforestry provides an almost zero cost approach to restoring soil nitrogen by the use of nitrogen-fixing trees and shrubs. However, soils are typically also deficient in other nutrients and trees cannot provide these (except by moving them from deeper soil horizons to the surface in leaf litter). So to fully close the yield gap farmers’ needs to purchase these inputs, for which of course they need a source of income. Here agroforestry can again help since many of the highly nutritious traditional foods, medicines and other objects of day-to-day use used to come from trees as fruits, nuts, leaves, bark, wood, etc. This traditionally and culturally important resource is rapidly being lost as a result of deforestation. So, agroforesters are working with local communities to rebuild this resource and introduce it into the farming systems. This is being done by “participatory domestication” so that the farmers are the decision makers and the beneficiaries of their innovations. This is the subject of my book “Living with the Trees of Life – Towards the Transformation of Tropical Agriculture”, published by CABI (2012). This participatory domestication programme has been hugely successful and is spreading rapidly through the rural communities of Cameroon – and was awarded an Equator Prize in 2010. So to answer your questions briefly; Agroecology is agroforestry, but agroforestry is more than agroecology. Secondly, agroforestry does not have to be organic, but it is a low input system, which once nutrient deficits are restored should not require fertilizers. Its ecological base means that the need for pesticides should also be minimal as through diversification of the agroecosystem, it promotes agroecological functions (food chains, life cycles, etc) which should keep pests, pathogens and weeds in a near natural balance. And lastly, agroecology is good, but its disadvantage is that it does not provide solutions to all the problems driving land degradation, hunger, malnutrition and poverty.

5. How far is the definition of agroforestry a production-mode one and how far does it encompass particular social or economic approaches to production?

André Gonçalves: I understand agroforestry as a production system combining multiple crops (annual and perennials) mediated by cultural, economic, and technical circumstances. One modest example is that farmers with indigenous roots in Brazil find it easier to cultivate in an agroforestry model than ones growing up cultivating monocrops such as sugarcane or soybean. In this case the cultural aspect therefore influences the way the farmer will manage the system.  Another example is the home gardens, which are generally productive areas managed by women. Most of these systems follow an agroforestry pattern.

Roger Leakey: Agroforestry is a land use system that can achieve social, economic and environmental sustainability – by combining “food production, income generation, and promotion of environmental services”. An agroforest is a man-made replica of a natural forest in structure and function, so an agroforest is an agroecosystem which, through the inclusion of trees, goes through an ecological succession akin to that in natural ecosystems – becoming more complex and biodiverse as it matures. Agroforests have been found to contain about 75% of the species found in natural forests (cf. a monocultural plantation has less than 5%). Thus they are good for wildlife conservation and for the maintenance of healthy agroecological functions. Because the farming systems contain a high proportion of mature trees they also sequester carbon in the biomass in the same way as natural forests. Agroforestry encompasses ideas about social and economic approaches to production, including social justice (farmers’ rights, intellectual property rights of local communities, gender equity, etc).  Within this, it also promotes business and employment opportunities (including access to micro-finance) in rural communities in processing, value-adding, packaging and marketing of agroforestry tree products (AFTPs). Some of the poverty issues in developing countries are the result of having 80% of the rural population in not very productive and very small farms. Getting some of farmers out of farming and into rural cottage industries is a good way of giving them a pathway out of poverty and into the cash economy.

Ben Phalan: Agroforestry is the integration of trees into farming systems. This can take many different forms. There are “rustic” agroforestry systems, which involve clearing the forest understorey but keeping many of the larger original forest trees in place – crops such as cocoa and coffee can be grown under shade in this way. There are “technified” agroforestry systems, which involve planting a canopy of one or a few species of specially selected trees amongst the crops. The trees might be fast-growing, nitrogen-fixing, or have other desirable properties (they may also sometimes be non-native species). There are forest gardens, which involve growing a diverse mixture of trees which provide fruit, nuts and other forest products. There are silvopastoral systems, with livestock – such as the dehesas and montados of southern Europe. And so on – agroforestry is not just one thing. I think it would be more useful if agroforestry and agroecology were defined purely in agronomic terms, because it would make communication clearer. Agroforestry would then simply refer to the integration of trees into farming systems and agroecology to the study of ecological interactions in agriculture, regardless of who is doing the farming. However, the term 'agroecology' in particular is often loaded with connotations about social justice and political empowerment. Such issues are extremely important in their own right, but they are quite distinct from questions of agronomy and agricultural science. I don't find it useful to conflate such issues, because doing so can limit and confuse people's thinking.

6. How does agroforestry try to ease the tension between increased food production and environmental problems? What are the results?

André Gonçalves: Many agroforestry systems that I have visited in these years are much more productive than conventional production systems. The conventional production systems I am referring to are those managed with many external inputs such as chemical fertilizers and pesticides, which are in general considered more productive. The classical examples are soybean, sugarcane, poultry, swine, etc. If we consider total food production agroforestry systems are much more efficient. I think there is a general ignorance about the capacity of these systems in combining food production and environmental services.  In agroforestry systems, the liquid production (outputs minus inputs) are in general higher than conventional production systems. In addition, agroforests promote a number of environmental services such as carbon sequestration and biodiversity protection that are usually not taken into account. I don´t see a more efficient way of combining food production and ecosystem services.

Roger Leakey: Agroforestry eases the tension between increased food production and environmental problems (soil and land degradation, erosion, water runoff and flooding, loss of biodiversity, breakdown of agroecological function, increased GHG emissions, etc) by harnessing biological nitrogen fixation and agroecosystem diversification. The introduction of trees reduces yield of those crops with which they are in direct competition for light, water and nutrients (ie. those very close to the tree). However, the benefits to soil fertility, soil structure, water infiltration, and the agroecosystem outweigh these losses across the field system. Perhaps even more importantly, the tree products (fodder, wood, fuel, fruits, nuts, resins, gums, extractives, medicines, etc) have a much greater market value than the lost yield, so overall the system is economically more beneficial. In some farming systems (e.g. cocoa, coffee, etc) the shade of the trees is also beneficial to the yield and sustainability of the crop. Adding productive trees as shade, instead of unproductive ones, adds additional sources of income. There is also the benefit of risk alleviation from all this diversification. Farmers trying to support their families on 1-2 ha of land in the absence of a social services support system, can’t have all their eggs in one basket. Modern intensive farming systems (close to monocultures) are very risky if the farmer does not have access to the full Green Revolution package of agrichemicals and does not have alternative sources of income.

If we want to improve agricultural productivity while raising incomes there are some specific recommendations to suggest. I believe that the first step towards improved farm productivity is to rehabilitate degraded land and to improve soil fertility by the use of nitrogen fixing trees and shrubs. This can improve cereal yields two- to three-fold and so initiates the closure of the yield gap. It also starts the diversification of the agroecosystem and improves some agroecological functions. The second step is to further diversify the farming system by introducing useful trees into the farmland that also produce marketable products, including nutritious traditional foods and traditional medicines, so starting the creation of agroforests. One example here is the formation of cocoa agroforests in which the shade canopy above the cocoa is made up of species producing marketable food and non-food products, including timber.  Within this second step much work has been done over the last 20 years to domesticate these traditionally-important and useful trees, using simple horticultural techniques for the development of clonal cultivars. This is implemented in participatory mode in partnership with NGOs and local communities so that the farmers are the beneficiaries of their own innovations. These partnerships provide training in tree nursery management, agroforestry, product marketing, use of microfinance, and community development. One of the impacts of the participatory domestication programme is that young people are starting to say that they can now see a career within their village community and so they do not need to to migrate to local towns and cities in search of work. If scaled up, this could perhaps help to reduce crime, drug abuse, even terrorism. The cultivation of trees producing marketable products initiates the third step to more sustainable and productive agriculture (further closing the yield gap) which is to generate income and create opportunities in the communities for product processing and value-adding; so opening up business, employment and trade opportunities in the rural and urban communities. Experience indicates the income generated in this way is typically used by the participating communities to develop local infrastructure (wells, piped water, store houses, roads and bridges, etc.); purchase agricultural inputs and livestock; send children to school, and pay for medical treatment. Together, I believe, these three steps to closing the yield gap – agroforestry practices for land rehabilitation, tree domestication and product commercialization - represent a highly adaptable generic model for the rehabilitation and improvement of smallholder farming practices in the tropics. Evidence from on-going projects, especially in Africa, show that within 5-10 years the livelihoods of the participating communities are being greatly improved year on year. Among other options, the income from the sale of nursery plants and tree products can be used to purchase fertilizers and pesticides to further close the yield gap.

Ben Phalan: I agree with Roger that using agroforestry to rehabilitate degraded land offers a very promising way forward in some parts of the world (although we do need to be careful about definitions: logged forest, for example, is sometimes considered to be “degraded land” but often retains much of its original biodiversity, and could be restored rather than converted to agroforestry). Regarding the second step that Roger mentions and effects of trees on yields and biodiversity, I think the picture is more complex than a straightforward win-win.  N-fixing trees can add nitrogen and organic matter to the soil, so seem likely to enhance yields. However, I don't know of much evidence that they provide any real biodiversity benefit. There is some evidence for crops such as coffee that some degree of shade can enhance yields, at least when farmers do not have access to fertiliser. But when high-yielding varieties and fertilisers are used appropriately, the highest yields appear to be in unshaded or minimally shaded systems. There is a need for more detailed, systematically collected data to understand this properly. It's important to distinguish between data that show highER yields compared to current practices (which are often very low-yielding in many parts of the world) and yields which are high when assessed in relation to attainable yields under best management practices.

7. Where is it common and where is it successful - and under what circumstances?

Ben Phalan: My impression is that it is commonest in tropical smallholder landscapes, often those in more remote or forest frontier areas rather than in well-connected and long-established croplands. But there are exceptions. Homegardens are maintained in densely populated village landscapes, and dehesas are found well outside the tropics, in Spain and Portugal.

Agroforestry is widely promoted, especially in tropical countries, as a way of reconciling multiple objectives within a single farming system, but it is important to recognize that just because agroforestry can work well in some contexts does not mean that it is the best landscape option everywhere for everything. Agroforestry can also mean very different things to different people.  Talk to different people in a tropical landscape, and you’ll find some very different perceptions of what agroforestry is. Some of these perceptions are a long way from textbook definitions, yet these mutated meanings can have a great deal of influence as to what forms of agroforestry are taken up locally and what its actual impacts are on the ground. To give some examples, I have met Ghanaian farmers who insisted that their monoculture plots of teak trees were examples of agroforestry; self-styled agroforestry projects promoting neem (Azadirachta indica) trees in West Africa (neem might be a “wonder tree” in its native India, but it can be a seriously invasive species elsewhere); and those describing  oil palm plantations as agroforestry because the farmers integrate other crops such as cassava under the oil palm trees during the first few years of plantation establishment. The consequences, good and bad, of planting neem trees outside their native continent are very different to the consequences of maintaining a diverse mix of native tree species, yet both are easily filed under the same label, “agroforestry”.

Although agroforests are not a substitute for natural forests, it might seem obvious that a diverse mix of native tree species is better for biodiversity than a monoculture. This is not necessarily the case. Where the yields of agroforestry systems are lower than those of alternative forms of farming, more land will be required to produce a given amount of food, leaving less potential to conserve natural forests. In such a scenario, a landscape of diverse agroforestry systems might (or might not) be a poor compromise overall, compared to a landscape where the same amount of food is produced from less land, and a greater area of natural forests are protected. To understand if that is the case, detailed data on both yields and biodiversity need to be collected. However, most of the analyses that have been done to date use simplistic metrics such as species richness (for example, the figure of 75% of species mentioned earlier in the interview) which tell us very little about conservation value. Metrics such as species richness fail to detect, for example, a transition from a forest with healthy populations of endemic species (a species that is not found nowhere else) to an agroforest dominated by widespread generalist species, which can adapt to a range of different conditions, and therefore can miss large changes in conservation value. Furthermore, examples of rare or threatened species being found in complex agroforests are not evidence that the agroforests are able to support those species in the long term.

Where agroforestry is as (or more) high-yielding as alternatives, it could be used as part of a strategy to ensure that the footprint of farming is minimised, and that natural forests are “spared” from cultivation. In my view, this is the area where agroforestry has most potential to deliver good outcomes for both food security and biodiversity conservation. But it will not be enough to develop and promote the agroforestry part of the equation. Sparing land for nature will require not only that high-yielding agroforestry methods are promoted, but also that specific measures are put in place to ensure that natural forests are protected. (NB: Question 8 below goes into more detail on the governance challenges relating to deforestation).

There are probably many places where agroforestry techniques would be an improvement on current practices. Certainly perennial crops – whether grown in agroforests or not – often offer advantages over annual crops in tropical climates, such as reduced soil erosion. There are some contexts where high-yielding agroforestry may be particularly attractive to farmers: for example in parts of the world where people rely mainly on food that comes from their own land, and where they don’t have ready access to capital and markets. However, it would be a mistake to think that what seems best for farmers is necessarily the most sustainable. There are often trade-offs that need to be considered. An example of such a trade-off would be where the tree species most favoured by farmers make up a small fraction of the diversity found in a natural forest. Over time, farmers will tend to thin out the less "useful" trees and promote the "useful" species (which often include non-native species that are typically of lower value to native insects and other biodiversity). The system will shift towards a mix of tree species that provides maximum benefits for the farmer, but will be of ever-diminishing value to other species. An example of this is in the Western Ghats of India, where one of the trees most popular with farmers - the silver oak (Grevillea robusta)- is coming to replace many of the native species in agroforestry systems.

André Gonçalves: We have examples of agroforestry systems in many parts of my country. I believe that in all main Brazilian Biomes, i.e., Amazonia, Atlantic Forest, Caatinga (semi-arid), and Cerrado (savanna-like vegetation) we have occurrence of agroforestry systems. The main cause of success is the farmer. When the producer is convinced about the possibility of producing differently than the majority of farmers, she/he develops beautiful systems. Obviously other aspects count, such as market for the products, access to credit lines, extension service sensitive to this way of producing, organizational structures, etc. One successful example from Centro Ecológico might serve to highlight the possibilities of these systems. It is a project of producing açaí berries, which is a local species, typical from the Atlantic Forest. We plant it mixed with bananas and endemic trees. Several farmers have been implementing and managing these complex agroforestry systems, showing that it is possible to combine food production, income generation, and promotion of environmental services.  

In Brazil however, we have a number of challenges such as technical capacity of the extension services, adequate credit lines, market barriers, and even our legislation, in many ways prevent the adoption of complex agroforestry systems. For farmers it is very difficult to be different from their neighbors. Farmers doing agroforestry are still considered a bit weird in their communities. I would not say that is unsuccessful. I would say that we don´t have proper conditions to expand and scale up agroforestry systems. The challenges that we have prevent the expansion of AFS.

Roger Leakey: With regards to what André just mentioned on finding commercially viable products and marketing strategies to create a successful and sustainable agroforestry system: the complex agroforests that André describes are approaching maturity and this is what confers their ecological/environmental sustainability as production systems. As he says, to improve the economic benefits from these systems they need to be enriched with marketable species. Thus, for the smallholder farmer trying to support a household on a small area of land, the trick is to fill as many as possible of the ecological niches in these agroforests with useful and commercially marketable species producing timber and non-timber tree products. In South East Asia these agroforests (3.5 million ha in Indonesia) are typically rich with commercially important species, producing useful and marketable products for 80+ years, and sometimes indefinitely. Many of these products have traditional significance and so are important culturally, while others may be exotic species (eg. rubber) grown purely for commercial reasons. The continuous stream of food and non-food products from these smallholder systems makes the economic returns much better than juxtaposed conventional modern farming systems. As André mentions Latin American agroforests have many of these same qualities.

I believe that the greatest benefits from agroforestry are in the tropics and sub-tropics where the farmers, for financial reasons, do not have access to the full Green Revolution package (including capital intensive mechanization, etc), and where farm  size is small. Having said that, there are also good examples from Europe and USA. The multiplicity of production, environmental, social and economic outcomes from this approach is the expression of sustainable intensification in mainstream agriculture. This is especially important in Africa where modern agriculture has been least successful. I believe it is also crucially important for the mitigation of climate change. There are said to be about 900 million ha worldwide which could be raised from <5 tonnes to between 90-150 tonnes carbon per hectare by conversion from degraded land to an agroforest.

There are also examples of commercial monocultures being switched to agroforestry. Probably the most common is the conversion of rubber plantations to “jungle rubber” in Indonesia. Around 25% of world rubber now comes from agroforests. Likewise much cocoa and coffee production is now from smallholder production systems that are agroforests.

8. One important issue as mentioned above to reduce deforestation is governance. Is there any evidence that agroforestry reduced pressures on forest?


Ben Phalan: In many parts of the world, one of the key priorities for biodiversity conservation is to prevent the conversion of forests and other natural landcovers to farmland (whether agroforestry or other types of farming). There are various ways in which this can be done, but it mostly comes down to land-use zoning. If done by governments this can include protected areas or forest reserves; if done by local communities it can involve community forests; if done by companies it can involve High Conservation Value areas within a commodity crop certification framework; if based on market mechanisms (Payments for Ecosystem Services) it might include individual agreements with landowners. In general terms, it is fairly clear that relying exclusively on changes within the agricultural sector to reduce deforestation is unlikely to be very effective. If farmers adopt high-yielding agroforestry without there being any specific protection in place for forests, then it is quite probable that deforestation will continue or even accelerate. So specific policies and incentives are needed to ensure that forests are protected. A common misperception of land sparing is that it relies on the Borlaug effect, i.e. that if yields are increased, demand for agricultural land will fall. Land sparing is not synonymous with the Borlaug effect. The assumption that I make is not that demand for agricultural land will fall, but that with high yields the need for agricultural land will fall, and therefore high yields open up the possibility of sparing land for nature. The question that I think is most interesting therefore is not the descriptive one, as to whether agroforestry has reduced pressures on forest, but a more proactive one: how can high-yielding agroforestry be actively incorporated into strategies to reduce pressures on forest?

For anyone interested in getting into these issues in more depth, I highly recommend two books – Agricultural Technologies and Tropical Deforestation edited by Angelsen & Kaimowitz (2001), and Agroforestry and Biodiversity Conservation in Tropical Landscapes edited by Schroth et al. (2004).

Roger Leakey: In both SE Asia and Latin America agroforests are generally found on the margins of areas of natural forest - a resource that is in decline worldwide, especially in Africa. Consequently multistrata agroforests are much less common in Africa and the challenge for agroforesters is to find ways of recreating an agroecological succession on degraded land that, negating the need to clear forest, meets the urgent needs for food and nutritional security, and then reduces poverty and delivers social benefits for farmers. In other words the need is to sustainably intensify the badly degraded farmland and deliver public goods and services like the mitigation of climate change and the protection of biodiversity. Great progress has been made in that direction, led by the World Agroforestry Centre, one of the CGIAR research centres. I cannot do justice to that work in a short comment like this. Suffice it to say that there are many ways in which trees can be integrated into farms and landscapes to provide benefits to farmers, local communities and to the local and global environment, including land rehabilitation, whether in drylands, savannah, montane woodlands or tropical rainforest.

9. If agroforestry has multiple benefits, according to its advocates, why isn’t it being more widely adopted?  What are the practical and ideological constraints?

André Gonçalves:  There are currently two different segments in Brazil illustrating a great complexity in agricultural strategy. First there is the structural intensification and large export of biofuels from Brazil to Europe and on the one hand the agoforestry systems. Under these two extremes – agribusiness and family farmers – we have an immense variety of production systems. But in general terms there is a clash of interests between these two segments. Even though in the last few years we have seen many public policies to support the small holder sector, most of the investments are oriented to the agribusiness sector. Therefore, it is difficult to harmonize different interests.  

Ben Phalan: Agroforestry has actually been quite widely adopted in tropical countries. It is clearly an attractive option for smallholder farmers in some circumstances. One reason it is not more widely used in the West may be that growing multiple crops together makes mechanisation more or less impossible. Everything has to be done by hand. As long as fossil fuels are cheap and labour is expensive, that will tend to make agroforestry less competitive.

Roger Leakey: Regarding the clash that André mentions between agribusiness and sustainable agriculture, business and agroforestry initiatives are in my opinion mutually dependent on each other (as I mention in step 3 in my answer to Question 5 above). In fact there are currently some interesting Public Private Partnerships in progress between local communities engaged in agroforestry and multinational companies. Two examples are the development of a new edible oil crop (Allanblackia species) for Africa by Unilever in partnership with local communities and other stakeholders in Ghana, Nigeria, and Tanzania and the cultivation of raw materials for car manufacture (Daimler AG) using new alternatives to fiberglass made from oils, resins, fibres etc. grown in local community agroforestry systems in Brazil. These are used in the factory in exchange for the company’s involvement in providing health and education. Hopefully agribusiness will see that the closure of the yield gap brings hundreds of millions, or even billions, of farmers into the cash economy where they will be able to afford agricultural inputs, so becoming new customers. The closure of the yield gap would also remove the socio-economic constraints to the purchase of the products of biotechnology. I would like to see the agribusiness companies putting their weight behind more sustainable farming systems based on agroforestry and what I have described as Multifunctional Agriculture. This might then create billions of new customers for fertilizers, pesticides and even some wisely applied products of biotechnology.

Perhaps even more importantly, closing the yield gap also greatly enhance the importance of genetic advances addressing the adaptation of crops to climate change – advances which currently are of doubtful relevance given that farmers can’t even afford to buy fertilizers. I suggest that this mutual dependency with agribusiness means that agroforestry can help to mend some of the broken bridges in the agricultural industry and also address some of the disconnects between agroforestry and policy makers. André’s interesting and important work with the NGO, Centro Ecológico, in Brazil along the Rio Grande do Sul on how to manage and enrich agroforests in Latin America is crucial to developing a better understanding of how to adapt and enhance new approaches to sustainable intensification that will be applicable and locally relevant across the world.

10. What are the motivations for farmers to practice agroforestry –how much are they “traditional” practices and how much are they the result of interventions by, for example, ICRAF and other promoters of agroforestry?

Ben Phalan: I would tend to assume that where the benefits of certain agroforestry systems to farmers themselves are high, they are likely to have already adopted them. But some of the benefits of agroforestry – such as enhanced carbon storage and biodiversity relative to other farming systems – accrue to society more widely rather than to individual farmers, so there may be instances where it would be desirable to promote it even when farmers don't see a particular advantage. Of course, that makes it important to understand whether particular agroforestry systems are really as wonderful as they're made out to be.

André Gonçalves: In general, the motivations come from incentives of organizations (mainly nongovernmental) working with issues of rural development. More recently, the government has been giving some support such as policies to market the products, credit lines, and some extension service. Some farmers are motivated when they see concrete results. We do have some “traditional” systems, such as the cabruca (cocoa mixed with tropical trees), shaded coffee, rubber trees, brazilian nuts, etc., but I would say that most of the examples that we have currently, come as a result of interventions.

Roger Leakey: In many parts of the developing world farmers have been told to practice modern intensive agriculture, which is based on a system that works in temperate climates where people are wealthy and most (over 97%) are not engaged in farming. In a few places, notably parts of SE Asia and Latin America, local people have developed their own alternative to shifting cultivation when faced with growing population pressure and inadequate land resources for everyone to practice the traditional swidden agriculture. These are the areas with well-developed complex agroforests. In many other places, where people have had to become sedentary farmers with only a few hectares on which to support their families, they have tried to adopt intensive agriculture and have found themselves trapped in poverty, hunger and malnutrition. These people now don’t have much trust in “experts” from outside. A good example of this maybe the early efforts of agroforesters to push alley cropping. It was poorly adopted because it was too labour intensive and so did not meet people’s needs. In contrast, when we visited poor farmers and asked them what they would like to be able to do and to grow they, pretty much unanimously, said they would like to grow their traditional food species, many of which are fruits, nuts and leaves from trees. That was the start of what has now become the participatory tree domestication programme. It is being incredibly well adopted, I believe, because it is what these people said they wanted, it builds on tradition and culture, and it builds on existing local and regional markets. Why had these people not done this for themselves? Well, they lacked the basic horticultural techniques of rooting cuttings, grafting and marcotting. This has been overcome using low-tech, inexpensive (almost no cost) propagators that do not require running water or electricity. So, to answer the question: yes, agroforestry can be built on tradition. This is motivating. However, much of the success has come from not deliberately promoting agroforestry (or any other form of agriculture), but from helping people to do what they would like to do by providing training and skills. This too is motivating.

11. What are the priority research questions with regards to agroforestry?

Roger Leakey: I would identify two major needs. First of all we need much better understanding of how diversification (involving both the planned and unplanned biodiversity) promotes agroecological function. Secondly we need to discover how best to scale-up the agroforestry successes within widespread development programmes. To promote this up-scaling we need much better adoption and impact studies to persuade policy makers and donors that agroforestry deserves their support.

André Gonçalves: For the food climate research community at large and as far as agriculture is concerned, the main challenge is to combine food production for a growing population and at the same time promote environmental enhancement. Involving and engaging millions of poor farmers in this endeavor is also a big challenge.  At the organizational level we have been trying to demonstrate, at least at the local level, that it is possible to achieve environmental protection and food production at the same time. I believe that what we have done is to show, at a modest scale that we do have solutions for the crisis food-environmental challenges we face. Implementation of public policies, designed to scale up these experiences is necessary to overcome the impasse that we are living, were we need to have a more sustainable food system but the actions taken to achieve it are few.  

Ben Phalan: In my view, there are three important research themes which need to be looked at in a thorough and systematic way. First, how acceptable are the compromises reached between multiple objectives in agroforestry (food, income, biodiversity, carbon, etc), compared with alternative uses of the same land? Second, how does the adoption of agroforestry affect land-use change dynamics, especially rates of deforestation? And third, what practical incentives or policies would be effective in shifting the landscape towards a more desirable state (and conversely, what barriers are preventing that from happening)? The answers to these questions will be context-specific.


For more information on the issue of sustainable intensification, please see the joint publication of FCRN and Oxford Martin Programme on the Future of Food “Sustainable intensification in agriculture. Navigating a course through competing food system priorities” on our website where it is also available for direct download.