![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
ABSTRACT
Indigenous crops are often neglected in development research, largely because they are grown in particular localities and only account for modest shares of agricultural production at a national level. This article aims to rectify this neglect with respect to the Bambara groundnut using a mixed methods study of farmers in Mtwara, Tanzania. The interest is in determining the importance of the crop in local production patterns and livelihoods, as well as potential levers for improved utilisation. Using the Sustainable Rural Livelihoods framework, we show that the crop is popular and recognised for its agronomic and nutritional properties. It is grown as an additional (or marginal) rather than main crop, with most growers reporting meeting consumption and food security needs as their primary motivation. The absence of markets constitutes a strong barrier towards sales of Bambara, and many farmers report being deterred from growing it for that reason.
Acknowledgments
The authors are grateful to the Crops for the Future Research Centre (CFFRC), especially Patrick O’Reilly, Sean Mayes and Festo Massawe, as well as Vincent Leyaro (University of Dar es Salaam) for assistance in facilitating the fieldwork and designing the questionnaires. We are also grateful to the Naliendele Agricultural Research Institute in Mtwara, especially Bakari Kidunda and the team of enumerators, for hosting Basile Boulay and without whose help the survey could not have been implemented. Financial support from the ESRC, and for fieldwork costs from the DTC and the Sustainability Research Priority Area at the University of Nottingham is gratefully acknowledged.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1. These labels refer to groups of crops which are ‘marginal’ at the national or regional level, in aggregate terms, even though they may be important to individual farmers who cultivate them.
2. Bioversity International (Citation2017) notes that only three crops (maize, wheat and rice) account for half of plant-derived calories consumed globally; Van der Merwe et al. (Citation2016) show that under-utilised crops constitute an important part of household diets in South Africa.
3. The fieldwork was facilitated by the Crops for the Future Research Centre (CFFRC, www.cffresearch.org) who had a project with the ARI to develop new varieties of BG with enhanced yield that are easier to cook.
4. Farm size is the total size of land farmed during the 2014/2015 season (it does not include fallow land). Similarly, we only consider plots that were farmed.
5. These account for almost 90% of production in the sample; the rest is made of vegetables (such as tomatoes and okra), sorghum, sesame, and pigeon peas, all of which represent negligible shares individually.
6. In the few instances where a farmer failed to show up to the meeting, the officer and the head of village found a farmer with similar characteristics. In one village we failed to find a replacement farmer and therefore had one less observation. In another three villages an additional farmer was present and could not be turned away.
7. We had a very limited budget and three months to complete the fieldwork. In the absence of a sampling frame we relied on local knowledge. Hence the limited number surveyed in each village and the over-sampling of BG growers.
8. All respondents were either the head or the spouse of the head of household.
9. This could not be fully enforced. In Nanguni, only one woman farmer was present. In Mnazi, the group size could not be kept to six as three additional farmers decided to join and did not leave. In both Mtama and Mwindi the target composition was achieved. Even though focus groups were not sex-segregated, efforts were made at ensuring everyone could speak and express their opinion. Those participants who had not already expressed their opinion were asked to comment on what has previously been said, while those who had contributed repeatedly were politely asked to let others speak. Issues regarding the participation of women were generally not an issue, except in Mtama, where women initially spoke less.
10. Our choice of controls is based on the literature on market participation for smallholders which emphasises the role of household assets, geographic factors, and transaction costs. Asset holdings refers to both productive agricultural assets, which directly increase farm output, and private asset holdings, which can indirectly affect output by easing access to credit for productive investments. The former is accounted for in our regressions by farm size, use of chemical fertiliser, and labour input. For private asset holdings, variables in the vector Assets can be used alongside the dummy for being a cashew grower (as a proxy for access to liquidity as cashew is mainly grown for sale). Owning a mobile phone or a mode of transport can be seen as reducing transaction costs as they facilitate access to information and markets, as suggested in Key et al. (Citation2000). The education of the household head could be used for a similar purpose with search costs being lower for more educated heads (Heltberg & Tarp, Citation2002). Given the study area, geographic characteristics are generally less important than for national-level studies and most variation is accounted for by using ward fixed effects.
11. We also checked if perceptions of BG for women respondents differed based on whether they lived in a man or woman-headed household (See Table A1 in appendix). Perceptions between the two subgroups are very similar to results shown in . The only exceptions are that women respondents from women-headed households are slightly more likely to think it is nutritious and laborious to grow.
12. Housing conditions (types of flooring, walls, roofs, access to water) are similar between growers and non-growers as other indications of wealth. Social capital is more difficult to capture empirically, and we lack data on networks.
13. About 25% of Bambara plots are intercropped, slightly less than for cashews or paddy, compared to 50% or more of maize, cowpeas and cassava.
14. Virtually all unsold produce retained in storage is used for home consumption. Farmers explained in focus groups and informal chats that they kept a few kilograms of seeds in plastic containers to be used for the next planting season, but this represents a negligible quantity relative to harvest.
15. Alternative marketing channels include commercial networks with city traders (through arrangements with farmers to come and pick up their harvest), some commercial deal with exporters, or even a well-functioning inter-village trade operated through intermediary traders, which does not exist for BG.
16. The average farm size in women-headed households in our sample is 2.6 acres, as opposed to 4.6 acres for men-headed ones.
17. Note that awareness of agronomic benefits and awareness of best agronomic practices need not be aligned. Very often farmers were aware of BG’s agronomic benefits (for example, that it requires little fertiliser) but did not know the techniques to be implemented in order to realise these benefits (optimal planting methods).
Additional information
Funding
Notes on contributors
Basile Boulay
Basile Boulay is a researcher in International Development. His research interests include agricultural and development economics, sustainable development and mixed-methods approaches.
Rumman Khan
Rumman Khan is a currently Lecturer in Economics at the School of Economics, University of Bristol. His research interests are mainly based on the microeconometric analysis of household welfare and livelihoods using nationally representative surveys.
Oliver Morrissey
Oliver Morrissey is professor of Development Economics at the School of Economics, University of Nottingham and is director of the Centre for Research in Economic Development and International Trade.