Local tree knowledge can fast-track agroforestry recommendations for coffee smallholders along a climate gradient in Mount Elgon, Uganda
- First Online:
- 195 Downloads
Arabica coffee (Coffea arabica) is economically important for many smallholder farmers in the Mount Elgon region of East Uganda, but its production is increasingly threatened by climate change. However, ecosystem services (ES) provided by companion trees in coffee agroforestry systems (AFS) can help farmers adapt to climate change. The objectives of this research were to develop agroforestry species recommendations and tailor these to the farmers’ needs and local context, taking into consideration gender. Local knowledge of agroforestry species and ES preferences was collected through farmer interviews and rankings. Using the Bradley-Terry approach, analysis was done along an altitudinal gradient in order to study different climate change scenarios for coffee suitability. Farmers had different needs in terms of ES and tree species at different altitudes, e.g. at low altitude they need a relatively larger set of ES to sustain their coffee production and livelihood. Local knowledge is found to be gender blind as no differences were observed in the rankings of species and ES by men and women. Ranking species by ES and ranking ES by preference is a useful method to help scientists and extension agents to use local knowledge for the development of recommendations on companion trees in AFS for smallholder farmers.
KeywordsClimate change Ecosystem services Smallholder farmer ranking Famer perceptions Local knowledge Shade trees
Coffee is one of the major cash crops for many smallholder farmers in the East African highlands. In Uganda, coffee is the most important export crop with 90% produced by farm households with less than three hectares of land. In fact, Uganda is the second largest coffee producer in Africa, accounting for approximately 2.5% of global coffee production of predominantly Robusta coffee (Coffea canephora) (Chiputwa et al. 2015). Still, Arabica coffee (C. arabica) is of significant economic value for Mount Elgon, Uganda’s larget Arabica growing region in the East (Van Asten et al. 2011).
The performance of Arabica coffee is strongly influenced by climatic variability, and hence particularly sensitive to climate change (Campbell et al. 2014; Craparo et al. 2015; Davis et al. 2012; Vaast et al. 2005). At temperatures higher than the optimum range for Arabica (18–23 °C), coffee growth is reduced while development and ripening of berry pulp are accelerated, often leading to the loss of beverage quality due to an incomplete bean filling (Davis et al. 2012; Vaast et al. 2006). Drought and high temperatures are the major climatic limitations for Arabica coffee production in East Africa. These climatic conditions are expected to be more frequent with climate change, which already impacted Ugandan coffee sector, and is reducing the area suitable for coffee production (Bunn et al. 2015b; Jaramillo et al. 2011). With coffee-growing altitudes ranging from 1000 masl to over 2000 masl, the Mt. Elgon region can be used as a field laboratory to anticipate climate change and evaluate its impact on coffee, as suitable areas will shift upwards with time (Bunn et al. 2015a, b; Läderach et al. 2011).
In Uganda, the vast majority of coffee is grown in agroforestry systems (AFS) with significant presence of trees and bananas (Van Asten et al. 2012). These systems provide a range of ecosystem services (ES) at plot, farm and landscape levels. The ES can be economical (e.g. timber, fuelwood, fruit, and high value niche markets) or environmental (e.g. biodiversity conservation, carbon sequestration and buffering changes in temperature and precipitation), and can enhance the sustainability and resilience of agricultural systems (Cerdán et al. 2012; Nzeyimana et al. 2013; Perfecto et al. 2007; De Souza et al. 2012). However, the extent to which ES are beneficial for coffee productivity depend on the local environment, livelihood strategies of producers, local market conditions (e.g. coffee prices and local wages), and management practices (e.g. use of external inputs) (Van Asten et al. 2010; Cerdán et al. 2012).
As trees in coffee AFS provide a range of ES that are not directly related to coffee production, they are usually referred to as companion trees and increasingly promoted as a “climate-smart” practice, improving the resilience of coffee production by creating favourable microclimate conditions, as well as improving soil health (Van Asten et al. 2010; Beer et al. 1998; Campbell et al. 2014; Nzeyimana et al. 2013; Vaast et al. 2006). Trees in AFS increase the biodiversity that enhances biological control of P&D (Staver et al. 2001; Bos et al. 2007; Perfecto et al. 2007; Tscharntke et al. 2011). The role of companion trees in climate change adaptation and mitigation needs to be better documented and compared to mono-cropping systems (Cerdán et al. 2012; Harvey et al. 2014; Rahn et al. 2013; Vaast et al. 2005, 2006). Research on coffee companion trees has been mainly concentrated in Latin America and India, and mostly on their effects on coffee. More research is needed to provide site-specific insights in the benefits and constraints for East Africa as well as to better address the livelihood needs of coffee communities. In this regard, the use of local knowledge on agroforestry can be helpful whilst avoiding expensive and time-consuming trials.
Local tree knowledge plays an important role in traditional agroforestry design, because farmers customise their AFS according to their knowledge and preference of trees and ES that these trees are providing (Cerdán et al. 2012; Valencia et al. 2015). Preferences and management decisions regarding companion trees can be gender-specific, especially with coffee, as it is traditionally a men’s cash crop (Kasente et al. 2002; Kelemen et al. 2015; Kiptot et al. 2014; Kiptot 2015). Yet, according to Lecoutere and Jassogne (2016) and Villamor et al. (2014), literature on gender in coffee systems is scarce, and hence there is a need to fill this gap, particularly on local tree knowledge. Local knowledge is, as opposed to indigenous knowledge, not geographically or ethnically specific and can be similar at different locations if within similar agro-ecological context (Sinclair and Joshi 2000). While Albertin and Nair (2004) and Soto-Pinto et al. (2007) have reported on local knowledge on tree diversity in coffee AFS, it is only in recent years that research has documented on both local knowledge and ES provision in coffee AFS and landscapes (Cerdán et al. 2012; Lamond et al. 2016; Smith-Dumont et al. in press).
The objectives of this research were (1) to develop agroforestry species recommendations along an altitude and climate gradient, using local tree knowledge on ES provision, and (2) to align these recommendations to farmers’ needs by ranking a group of key ES according to farmers’ preferences and livelihood strategies. Potential differences between men and women were taken into consideration for both objectives.
Materials and methods
The mean annual rainfall in the study area ranges from 1200 to 1400 and 1800 mm and the mean annual temperatures are 23, 21 and 18 °C, at respectively low, mid and high altitudes (Hijmans et al. 2005). Although there is a significant local climate variability, Mbogga (2012) suggested that the Mt. Elgon region had experienced an increase in temperatures between 0.4 and 1.2 °C when comparing the 2001–2011 period to the 1961–1990 period. Furthermore, dry seasons are getting longer, causing crop water stress, particularly at lower altitudes (Bunn et al. 2015b; Mbogga 2012).
Due to the relatively high population density of approximately 250 and 300 inhabitants per km2 for Bulambuli and Kapchorwa, respectively (Uganda Bureau of Statistics 2016), the landscape is mainly made of smallholder farms (<2 ha) with generally intensive and mixed coffee (C. arabica) x banana (Musa spp. AAA-EA) based agricultural systems. Coffee is grown under varying levels of shade provided by various companion tree species and banana. Traditional East African Arabica coffee varieties are predominantly grown, including SL 14, SL 28 and Nyasaland (locally known as Bugisu Local). Coffee productivity has been shown to be substantially lower than its potential due to low soil fertility, high P&D pressure and poor land and coffee tree management practices (Wairegi and Van Asten 2010; Wang et al. 2015).
Surveys and data collection
In total, 301 farmers were selected from 196 farms, including 183 men (M) and 118 women (F). Most of the women (105) were from the same households as the men that were interviewed. For the low, mid and high altitude zones there were 103 farmers (63 M, 40 F), 97 farmers (61 M, 36 F) and 101 farmers (59 M, 42 F), respectively. Questionnaires were digitised and used together with plasticised ES pictograms and tree species fact/technical sheets with clear recognizable pictures of the tree shape, mean features of crown, bark, fruits, and leaves.
For each altitude zone, the 20 most abundant species were selected for the ranking exercises, covering more than 98% of the recorded trees. At low and mid altitude, this yielded 23 and 22 species respectively, as the 20th place was shared by several species. At high altitude, only 19 species were recorded and hence all were selected. Since coffee is predominantly intercropped with banana, also providing shade, it was included in the ranking exercise along with tree species. Of those 20 species, the farmers were asked to identify the 10 species they know best for the ranking exercise. The authors considered based on expert knowledge that people can rank to a maximum of 10 items, without losing ranking quality.
Twelve ecosystem services (ES) attributed to companion trees by farmers are grouped into two MEA categories and six subcategories: microclimate (buffering temperature extremes and conserving soil moisture), soil fertility (producing mulch and controlling erosion), pests and diseases control (decreasing incidence of White Coffee Stem Borer (WCSB) and Coffee Leaf Rust (CLR)), weed control (weed), coffee production (yield improvement/stabilisation and coffee tree life expectancy increase) and tree products (food, timber, fuelwood)
Pests and diseases
The questionnaires consisted of three questions: (1) identify 10 tree species that you know best out of the top 20 most abundant species at your specific altitude, (2) rank these species according to each of the 12 ES (tree species with the same rank were allowed, i.e. ties), and (3) rank the 12 ES according to importance for your livelihood.
Ranking tree species and ecosystem services with the Bradley Terry model
A ranking is a relationship between a set of items such that, for any two items, the first is either ‘ranked higher than’, ‘ranked lower than’ or ‘ranked equal to’ the second. Ranking should not be confused with rating, where the items are scored with absolute values on a scale for example from 1 (bad) to 5 (good). In ranking however, items are scored relative to the other items, hence their scores have no meaning out of the context. Ranking is therefore a subjective evaluation that is useful in pair-wise comparisons.
There are some well-known non-parametric statistical methods based on ranks, i.e. the Kruskall–Wallace test and Friedman test. However, they are not suitable in the context of this study as they do not account for ties in partial rankings (items have the same ranking score). For this reason, the Bradley and Terry (1952) approach was used to analyse the ranking data of tree species and ES. Hence, the data was analysed in R (R Core Team 2015) using the BradleyTerry2 package (Turner and Firth 2012) and three other R functions as explained Van der Wolf et al. (2016).
The Bradley Terry analysis yielded ranking estimates from each individual farmer and used those to create one combined ranking for all the farmers. The analysis also yielded p-values from Wald comparison tests to indicate how significantly different these ranking estimates were from each other. As argued by Van der Wolf et al. (2016), the confidence intervals could vary quite significantly between tree species so they were plotted not solely based on the order of decreasing ranking estimates, but on a combination that includes both the estimate and the size of the confidence interval.
Results and discussion
Tree species ranking by ecosystem service and altitude
Cordia africana was exceptionally well ranked for all altitudes and ES (except food) and was followed by the four Ficus species and A. coriaria. Literature confirms that some of these species are well known for providing several of the presently targeted ES (Meunier et al. 2010; Ndenecho and Lambi 2010; Orwa et al. 2009). Musa spp. were ranked from medium to low for all ES other than food, which illustrates that farmers deliberately intercrop coffee and banana very commonly in the very densely populated area of Southern Uganda for food security reasons (Van Asten et al. 2011; Ellis and Bahiigwa 2003). Farmers need to secure food for their households irrespective of any other potential ecosystem service or disservice of banana compared to companion trees.
Overall, different categories of tree species can be identified based on the ranking results. Species such as C. africana and F. mucuso were ranked high, irrespective of altitude and ES. Likewise, Spathodea campanulata and Psidium guajava were predominantly ranked medium and low, respectively. Then there were species whose ranking performance is dependent on altitude or ES. For instance, Ficus natalensis, Ficus ovata and Terminalia ivorensis were generally ranked higher at high altitudes and lower at low altitudes, whereas it was the opposite for Aleurites molucana and Faidherbia albida. Species that were ranked depending on the ES, were highly ranked only for specific ES. However, their rankings were also dependent on altitude to some extent. For example, Musa spp. were consistently ranked high for food, while they were ranked low for the other ES at low altitude. Similarly, Milicia excelsa was only ranked high for fuelwood and timber at mid and high altitudes, but ranked low for fuelwood and all other ES at low altitude. Finally, Eucalyptus grandis was consistently ranked low for all ES except for fuelwood and timber, but also ranked low for these at low altitude.
There are biophysical factors (e.g. climate, soil characteristics) and socio-economic factors (e.g. distance to markets, population pressure) that influence the farmer’s appreciation and ranking of a companion tree species in different contexts (Lamond et al. 2016). These factors have not been identified and studied as such, but the authors recommend taking them into account in further ranking research. Under the widely accepted assumption that climate is one of the determining factors, it is expected that the aforementioned tree species and their respective ES provision will play different but complementary roles in climate change adaptation. It is suggested that species that were ranked high for all altitudes, are climate change resilient. Species whose ranking performance is altitude dependent and that were currently ranked high at low altitude, can be useful in the future for adaptation at mid to high altitudes. On the other hand, those species that were ranked low at low altitudes but high at higher altitude, will probably become decreasingly suitable in a changing climate.
Although biodiversity conservation was not included in the list of ES selected by farmers in the present study, it is still important to take into account when recommending (companion) tree species in general. In the present case, Prunus africana and M. excelsa were rated by the IUCN as ‘vulnerable’ and ‘near threatened’ respectively (World Conservation Monitoring Centre 1998a, b). As such, tree species recommendations could become part of a conservation strategy for Mt. Elgon region.
Ecosystem service ranking by altitude
Figure 4 shows that the ES rankings at higher altitude are more differentiated as opposed to those at lower altitude with seven, five, and three distinct groups, respectively. In other words, farmers at high altitude were more in agreement to distinguish the ES than the farmers at low altitude.
The provisional services (weed, food, timber, and fuel) were ranked second from last at all altitudes. Figure 4 shows that within these provisional services rankings, food was preferred over weed at high altitude, and both were preferred over fuelwood and timber for high and mid altitude. For low altitude, these four ES were ranked equally important and relatively better than at high and mid altitudes. At high altitude and far away from markets, households rely much more on their own food production for auto-sufficiency and have less access to markets for timber, fuelwood and fruits compared to low altitude farmers closer to an urban demand. Therefore, it makes sense that food and weed were relatively more important at high altitudes, and that timber and fuelwood increased in relative importance at low altitude.
The consistent and significant low ranking of P&D is attributed to the lack of knowledge, as previously discussed. Still, farmers were able to rank CLR as significantly more important than WCSB at high altitude, probably because incidences of CLR have been found to be higher in unshaded systems which predominate at high altitude (Soto-Pinto et al. 2002).
Lastly, it has to be noted that mulch was the most preferred ES over all altitudes and in particular at high altitude. This preference could be explained by its multi-functionality. Mulch can provide a series of ES, such as soil moisture conservation, erosion control, yield improvement (through organic matter addition and nutrient cycling) and weed control. At high altitude, compared to low altitude, the need for on-farm trees is lower because people harvest tree products from the nearby forest reserve (Chhetri et al. 2003). Hence, there are fewer trees providing canopy cover and mulching material, which could explain the relative importance of mulch as an ES at high altitude.
Tree species and ecosystem service rankings by gender
The absence of gender differences in ES preferences can be explained by the focus of this study on coffee plots and not on the whole farm that generally comprises plots with food crops for which women have more power on management decisions. The absence of gender differences in the tree species rankings, however, is an indication that local knowledge is not only ethnically and geographically blind (Sinclair and Joshi 2000), but also gender blind and therefore adding value to the use of local knowledge in agroforestry research and development.
Limitations of the study approach
The approach used in this study has the advantage that data collection is relatively quick and easy, the farmer’s options by context are considered and recommendations can be readily made with the ranking results. However, some limitations have to be considered. First of all, the list of 12 ES is not exhaustive. It was decided to focus on those ES that were relevant to coffee farmers and as such taking into account climate change, food security and livelihood. However, an important ES not presently assessed, is shade quality in terms of light interception over the production cycle particularly. It is recommended to complement the current data base on a wider range of ES in further documentation of farmers’ local knowledge. Secondly, it has to be emphasised that the ranking results are relative and as such the ranking of individual species should always be considered in relation to the other species concurrently used in the ranking exercise. For instance, a low ranked species is not necessarily a bad species and could be higher ranked when considered in another context with a different set of species and ES. Thirdly, although local knowledge is site specific and takes into account livelihood for instance, it should be validated and used complementary with expert knowledge of trees and their ES (Lamond et al. 2016; Smith-Dumont et al. in press; Van der Wolf et al. 2016). Furthermore, from a more holistic and realistic farmer’s perspective, final recommendations by agricultural services have to take into account the biophysical and socio-economic contexts, both at farm and landscape level. As such research on local knowledge should also take into account not only the coffee plots, but the whole farm. Lastly, it is possible that the abundance of tree species influences the selection of trees and maybe to some extent their rankings. Locally abundant species are better known than rare ones and might therefore be selected and ranked by more farmers. Consequently, it is possible that unknown and non-abundant but potentially interesting species are ranked with more uncertainty or might even be left out of the analysis.
This paper describes how local knowledge can be used to rank tree species for a range of ES and rank those ES by the farmer’s preference. These rankings can then be used in order to make recommendations for companion trees in coffee agroforestry systems along an altitudinal gradient that are tailored to the farmer’s needs. Species can be categorised according to their capabilities for climate change adaptation, in terms of their capacity of providing ES under different climates. As farmers are increasingly experiencing difficulties growing Arabica coffee at low altitudes, they need a large set of ES that support their coffee production’s resilience by e.g. contributing to soil fertility, microclimate, and coffee production. Despite the known impact of P&D on coffee productivity and the potential P&D controlling services of companion trees, farmers lacked knowledge on how species affect P&D dynamics. Moreover, it was expected that men and women would rank tree species and ES differently. However, the analysis showed that gender does not influence preferences and management decisions on companion species and the ES they provide, so it can be concluded that local knowledge is gender blind, adding value to its use in agroforestry research and development. A ranking approach has shown to be an appropriate and quick method for scientists to use local knowledge for the development of recommendation tools for coffee farmers and their associations. The ES preference rankings at each altitude are useful to understand the farmer’s needs and their options by context since they inherently include site-specific and livelihood needs. Some limitations to the approach include the non-exhaustive list of ES and the relativity of the ranking results. The plot-level approach is especially useful if complemented by expert knowledge at farm and landscape level.
This research was conducted under the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS) and under the Program Forestry, Trees and Agroforestry (FTA). The study was supported by the Federal Ministry for Economic Cooperation and Development of Germany (BMZ). The authors of this paper wish to show their appreciation and gratitude for Franco Manget and Wilberforce Wodada for their valuable assistance in the field, Theresa Liebig for her help with the baseline data collection and advice on pests and diseases, Dr. Richard Coe (ICRAF) for his advices on Bradley Terry ranking analysis in R, Allan Heinze for the ranking analysis functions in R, Jenny Ordonez (ICRAF) for her contribution to the methodology, Metajua for digitising the surveys, Ewaut Kissel for his significant help in R programming, and Mandy Malan for her daily support and endless reviews.
- Bradley RA, Terry ME (1952) Rank analysis of incomplete block designs: the method of paired comparisons. Biometrika 39(3–4):324–345Google Scholar
- Chhetri P, Mugisha A, White S (2003) Community resource use in Kibale and Mt Elgon National Parks, Uganda. Parks 13(1):28–38Google Scholar
- De Souza HN, de Goede RGM, Brussaard L, Cardoso IM, Duarte EMG, Fernandes RBA, Pulleman MM et al (2012) Protective shade, tree diversity and soil properties in coffee agroforestry systems in the Atlantic rainforest biome. Agr Ecosyst Environ 146(1):179–196. doi:10.1016/j.agee.2011.11.007 CrossRefGoogle Scholar
- Jaramillo J, Muchugu E, Vega FE, Davis A, Borgemeister C, Chabi-Olaye A (2011) Some like it hot: the influence and implications of climate change on coffee berry borer (Hypothenemus hampei) and coffee production in East Africa. PLoS ONE. doi:10.1371/journal.pone.0024528 PubMedPubMedCentralGoogle Scholar
- Kasente D, Lockwood M, Vivian J, Whitehead A (2002) Gender and the expansion of non-traditional agricultural exports in Uganda. In: Razavi S (ed) shifting burdens: gender and agrarian change under neoliberalism. Kumarian Press, Bloomfield, pp 35–65Google Scholar
- Kelemen E, Potschin M, Martín-López B, Pataki G (2015) Ecosystem services: a gender perspective. OpenNESS Ecosystem Service Reference BookGoogle Scholar
- Läderach P, Haggar J, Lau C, Eitzinger A, Ovalle O, Baca M, Lundy M et al (2013) Mesoamerican Coffee: building a climate change adaptation strategy. Int Center Trop Agric 2:1–4Google Scholar
- Lecoutere E, Jassogne L (2016) “We’re in this together”: changing intra-household decision making for more cooperative smallholder farming. Universiteit Antwerpen, Institute of Development Policy and Management (IOB), AntwerpGoogle Scholar
- Leemans R, de Groot RS (2003) Millennium Ecosystem Assessment: Ecosystems and human well-being: a framework for assessment. Island Press. http://library.wur.nl/WebQuery/wurpubs/wever/326575
- Liebig T, Jassogne L, Rahn E, Läderach P, Poehling H-M, Kucel P et al (2016) Towards a collaborative research: a case study on linking science to farmers’ perceptions and knowledge on arabica coffee pests and diseases and its management. PLoS ONE 11(8):e0159392. doi:10.1371/journal.pone.0159392 CrossRefPubMedPubMedCentralGoogle Scholar
- Mbogga MS (2012) Climate profiles and climate change vulnerability assessment for the Mbale region of UgandaGoogle Scholar
- Meunier Q, Lemmens R, Morin A (2010) Alternatives to exotic species in Uganda: growth and cultivation of 85 indigenous trees. GraphiConsult (U) Ltd., KampalaGoogle Scholar
- Ndenecho E, Lambi C (2010) Cameroon arid lands in transition: a case study of a fragile environment at the ante-room of desertification. J Appl Soc Sci 8(1&2):141–154Google Scholar
- Orwa C, Mutua A, Kindt R, Jamnadass R, Anthony S (2009) Agroforestree database: a tree reference and selection guide version 4.0Google Scholar
- Perfecto I, Armbrecht I, Philpott SM, Soto-Pinto L, Dietsch TV (2007) Shaded coffee and the stability of rainforest margins in northern Latin America. In: Tscharntke T, Leuschner C, Zeller M, Guhardja E, Bidin A (eds) Stability of tropical rainforest margins: linking ecological, economic and social constraints of land use and conservation. Springer, Berlin, pp 227–263Google Scholar
- R Core Team (2015) A language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
- Sinclair FL, Joshi L (2000) Taking local knowledge about trees seriously. In: Lawrence A (ed) Forestry, forest users and research: new ways of learning. European Tropical Forest Research Network (ETFRN), Wageningen, pp 45–61Google Scholar
- Smith-Dumont E, Lamond G, Nansamba R, Gassner A, Sinclair FL (in press) The utility of farmer ranking of tree attributes for selecting companion trees in coffee production systems. Agric SystGoogle Scholar
- Uganda Bureau of Statistics (2016) The national population and housing census 2014: main report. Uganda Bureau of Statistics, KampalaGoogle Scholar
- Vaast P, Kanten R, Siles P, Dzib B (2005) Shade: a key factor for coffee sustainability and quality. Conf Coffee 4:887–896Google Scholar
- Van Asten PJA, Wanyama I, Mukasa D, Nansamba R, Kisaakye J, Sserubiri I, Jassogne L et al (2012) Mapping and evaluating improved intercrop and soil management options for Ugandan coffee farmers. Technical reportGoogle Scholar
- World Conservation Monitoring Centre (1998a) Milicia excelsa. The IUCN Red List of Threatened Species 1998: e.T33903A9817388Google Scholar
- World Conservation Monitoring Centre (1998b) Prunus africana. The IUCN Red List of Threatened Species 1998: e.T33631A9799059Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.