For every tree a reason — research “in” rather than “for” agroforestry development


Farmer Athanus Makokha (centre) explains uses of various trees on her farm to ICRAF’s John Nyaga (right) and field assistant Robert Situma (left) in Trans Nzoia County, Kenya. The indigenous tree Croton macrostachyus grows on the crop field to the left, while other C. macrostachyus trees as well as Grevillea robusta, Cupressus lusitanica and Eucalyptus spp grow around the homestead area in the background. Credit: J. Nyaga

In their ground-breaking article about the importance of inserting research into agroforestry development, Ric Coe, Fergus Sinclair and Edmund Barrios of the World Agroforestry Centre (ICRAF) emphasize how important this small but extremely significant change in wording — R in D rather than Research for Development — will be in efforts to expand agroforestry in landscapes.

“In addressing current calls for rapid scaling-up of agroforestry,” they write, agroforestry researchers “need to avoid the pitfalls of what might be called bright-side science, where positive evidence is given more weight than negative and data that could contradict prevailing enthusiasms are given limited attention or not collected at all”.

Evoking the lessons learned from the overselling of alley cropping technology in the 1990s, they note that R in D is the best approach of assessing the most appropriate, workable and acceptable agroforestry systems, or “options”, for specific “contexts”.

And what research in development shows is that agroforestry, the adoption and management of trees in agricultural fields, takes many forms to suit different people and places. Take, for example, the findings of a recent study in Kenya’s Rift Valley of the factors that shape farmers’ decisions about which and how many trees to plant, at what densities, where, and in what locations and arrangements on their land.

Farmers make trade-offs in their decisions about trees

The study, undertaken by researchers from ICRAF, Jomo Kenyatta University of Agriculture and Technology, and the Swedish University of Agricultural Sciences, zeroes in on smallholder farms in five settlements in the Trans Nzoia County in western Kenya. The settlements were formerly large-scale maize farms under a single owner before the land was subdivided after Kenyan independence in 1964.

The researchers analyze how three factors — household resource endowment, land tenure security, and length of time the farm had been under the current management — influence the adoption of different agroforestry options.

The farms are small, with a mean size of less than one hectare, but with substantial tree cover — an average of more than 80 trees of eight different species per farm. All the households cultivate maize and 80% of them also keep cows, on average 2.4 per household. The farms include a wide range of tree species, 44 in total, more than half of which are indigenous, used in a wide range of agroforestry practices from boundary planting to dispersed trees in fields, hedgerows and woodlots.

Farmer Mr. David Kisaka (right) details the contributions of different tree species on his farm to agroforestry researchers in Trans Nzoia County, Kenya. Mangifera indica and Persea americana (fruit) trees are dispersed around the homestead, which also acts as a livestock holding area. Eucalyptus spp are planted on the farm boundary. Credit: J. Nyaga

Households with fewer resources plant more fruit trees to supplement family nutrition and income and also have higher tree diversity than do wealthier families, even though their farms are smaller. Fruit trees tend to be cultivated around homesteads, rather than in or around fields where many other species are planted to enhance soil fertility, provide fodder, as boundary markers and in woodlots to produce timber and firewood.

Because of the high demand for fuelwood and the high value of timber in supplementing farm income, all households regardless of their level of resource endowment, planted trees to provide these important products.

Land tenure also influences the adoption of different agroforestry options; farmers with secure land tenure look to longer-term farm investment with increased tree cover and timber species. Those with less secure tenure choose more fast-growing multipurpose and nitrogen-fixing tree species, opting for more diversity of species, a kind of insurance against risk.

The researchers found that the three factors — household wealth, land tenure and time under current owners — determined “the establishment of a diverse tree cover” in an area that 23 years earlier had been largely bereft of trees . And this finding, they note, is important because it “contradicts the discourse that poorer farmers are always the main reason behind deforestation”.

‘This implies that in the short term we need to develop and promote different agroforestry options that suit farmers,” says lead author of the study, ICRAF’s John Nyaga. “Farmers have different resource profiles and security of tenure, so there can be no one size to fit all. In the longer term, we need to expand our notion of what the options are and include changes to tenure as we try out various options with farmers.”

On the farm of Convic Mbali in Trans Nzoia County, Markhamia lutea and Grevillea robusta trees have been planted in cropping fields, while Eucalyptus spp are incorporated near the homestead area. Credit: J. Nyaga

On the farm of Convic Mbali in Trans Nzoia County, Markhamia lutea and Grevillea robusta trees have been planted in cropping fields, while Eucalyptus spp are incorporated near the homestead area. Credit: J. Nyaga

No silver bullets please, we’re working with smallholders

Studies like this, which put the research firmly into development, highlight the importance of understanding the structure, densities and multiple uses of tree populations in agricultural landscapes. Such an understanding counters any temptation to search for narrow, one-size-fits-all technological fixes, the proverbial “silver bullets” that have no place when working with smallholder farmers, whose land-use systems are complex, variable, integrated and multi-disciplinary systems.

Not surprisingly, R in D was front and centre in discussions at a recent gathering of ICRAF scientists in Bogor, Indonesia. In a session devoted to R in D, they profiled the way they’re integrating research and development in projects to identify appropriate options for scaling up agroforestry on three continents. These included: Agroforestry and Forestry (AgFor) in Sulawesi, Indonesia; sustainable cocoa in Peru; Agroforestry for Livelihoods of Smallholder Farmers in Vietnam; Vision 4 Change cocoa farming in Côte d’Ivoire, and; agroforestry solutions to ease pressure on Virunga National Park, Democratic Republic of Congo.

A key message that emerged from an ICRAF-led programme on Drylands Development in five African countries was that there is “no single solution”. ICRAF’s Karl Hughes, who profiled the programme, offered other rules of thumb for successful R in D, namely “averages can mislead, there is a need to customize options and offer a range of options to farmers, and to avoid overly simplistic messages”. This reinforces the message from Coe, Sinclair and Barrios, who say that scaling up of agroforestry needs more than the “wide scale promotion of a few iconic agroforestry practices” to embrace the fine scale variation in context that they are finding everywhere that they look.

Fergus Sinclair, who led the session on R in D, expressed excitement about the progress ICRAF is making in putting research squarely into development to develop a strong evidence base on what options suit different contexts, while getting on with imperative of improving farmers’ access to agroforestry technologies.

“By working with development partners in both the government and NGO sectors” he says, “we’re combining the large resources needed to learn what agroforestry options suit different farmers, with research methods that make this learning efficient. This is generating robust evidence that can inform investment decisions.”

For more information, please contact Dr. Fergus L. Sinclair at:

Further reading:

Coe R, Sinclair FL and Barrios E. 2014. Scaling up agroforestry requires research ‘in’ rather than ‘for’ development. Current Opinion in Environmental Sustainability 6:73–77. doi:10.1016/j.cosust.2013.10.013

Mamo A. 2013. One small change of words – a giant leap in effectiveness. ICRAF Agoforestry World Blog.

Nyaga J, Barrios E, Muthuri CW, Öborn I, Matiru V and Sinclair FL. 2015. Evaluating factors influencing heterogeneity in agroforestry adoption and practices within smallholder farms in Rift Valley, Kenya. Agriculture, Ecosystems and Environment 212: 106–118. Available at:

Joan Baxter

Joan Baxter is a communications specialist, journalist and award-winning author, with many years of experience in development and environmental research and writing in Africa and elsewhere.

You may also like...