Meta Evaluation of 6 Hivos Biofuel Projects

Transcription

1 Meta Evaluation of 6 Hivos Biofuel Projects Chapter 3 Jatropha Oil for local development in Mozambique (FACT Foundation) Hivos project number WW205S01 Written by Sona Prakash (Environment & Development Consultancy)

2 TABLE OF CONTENTS I. INTRODUCTION 3 I.1 NATIONAL CONTEXT 3 I.2 LOCAL CONTEXT 3 I.3 PROJECT AND OBJECTIVES 4 I.4 ALLIANCES AND ACCESS TO EXPERTISE 5 I.5 BACKGROUND 6 I.6 EVALUATION METHOD AND DETAILS 7 II. EVALUATION THEMES 7 II.1 CROP PRODUCTION 7 II.1.1 Crop Characteristics and History in the Region 7 II.1.2 Local Conditions and Jatropha Cultivation 8 II.1.3 Planting 8 II.1.4 Maintenance and Inputs 9 II.1.5 Harvest and Yield 11 II.2 LABOUR INPUT 12 II.3 PRODUCTION, INCOME & OTHER BENEFITS 13 II.3.1 Improved Accessions 13 II.3.2 Income, Payment and Profit 14 II.3.3 Flexibility & Diversification 15 II.3.4 Investment & Credit Access 16 II.3.5 Other Benefits for Producers 16 II.4 PROCESSING 17 II.4.1 Oil Extraction 17 II.4.2 Quality of PPO 18 II.4.3 Producing Biodiesel 20 II.4.4 Other Products 20 II.4.5 Investment & Processing Costs; Profit Distribution 20 II.5 MARKETS 22 II.5.1 Protection against Volatile Oil Markets 21 II.5.2 Local Market Development 22 II.5.3 Innovative & Equitable Business Models 23 II.5.4 Flexibility & Diversification 24 II.6 FARMING SYSTEM RELATED ISSUES 24 II.6.1 Risk for Producers 24 II.6.2 Availability & Quality of Land 24 II.6.3 Food Security 25 II.6.4 Traditional Jatropha Cultivation 25 II.6.5 Water Availability 25 II.7 WIDER SOCIAL & ENVIRONMENTAL ISSUES 26 II.7.1 Land Ownership 26 II.7.2 Wasteland 26 II.7.3 Environmental Risks 26 II.8 LOCAL & REGIONAL ECONOMY ISSUES 26 II.8.1 Appropriate Technology 27 1

3 II.8.2 PPO & Local Energy Generation 27 II.8.3 Multifunctional Platforms 28 II.8.4 Other Direct Local Uses of PPO 28 II.8.5 Local Applications of By products 29 II.9 NATIONAL POLICY 30 II.10 OTHER KEY FACTORS FOR GENERAL EVALUATION 30 II.10.1 Diversification & Resilience 31 II.10.2 Gender 31 II.10.3 Sustainability & Future Perspectives 32 III. CONCLUSIONS & RECOMMENDATIONS 32 III.1 ATTAINMENT OF PROJECT OBJECTIVES 32 III.2 GENERAL EVALUATION THEMES & RECOMMENDATIONS 34 III.2.1 Crop Management & Extension Support 35 III.2.2 Labour & Income 36 III.2.3 Processing & Local Use of PPO 37 III.2.4 Marketing 39 III.2.5 Gender 39 OVERALL RECOMMENDATIONS 40 Acknowledgements 40 2

4 I. INTRODUCTION I.1 NATIONAL CONTEXT Mozambique is one of the poorest and most underdeveloped countries in the world. In terms of human development, it has been placed 172 nd out of 182 countries 1. With rural subsistence farmers constituting 85% of its population 2, the lack of modern energy services in rural areas is perceived as a major impediment to raising agricultural productivity. While 20% of Mozambique s urban population have access to electricity, this is true of only 1 to 2% of the rural population. The country is sparsely populated with large tracts of unutilised land, making grid extension unviable. The prospects for and benefits from investing in small scale biofuel production for local energy generation are therefore considerable. The Mozambican government is actively promoting Jatropha as part of its National Biofuel Policy and Strategy 3 (see section II.9) A National Biofuel Council is being set up to coordinate, supervise and evaluate its implementation. Jatropha cultivation in Mozambique has grown from 7,000 ha in 2008 to 35,000 ha in 2010, and is expected to grow to 170,000 ha in Requests to the Mozambican government to open up land for large scale commercial plantation (mostly for export to the European market to meet the EU s blending target) were followed by concerns over potential pressure on land, water, food production and environmental capital. There was a freeze in land allocation ( ) while agro ecological land zoning was carried out. The first phase of this study reported the availability of 6,966,030 ha (19.4% of total arable land) for large scale agricultural activity 4. Private companies like Sun Biofuel are involved in a big way in biofuel production in Manica province (to provide Jatropha oil to Lufthansa) and other private companies have also set up large scale Jatropha plantations in Inhambane and Sofala. However, this large scale mode of production in itself offers few income generation possibilities for smallholders and little scope for improving local energy access. Apart from the project being evaluated here, the only other small scale Jatropha based activities in Mozambique took place in Manica Province as part of a collaboration between Caritas and ADPP (150 ha Jatropha planted in 2006), but could not sustain momentum because their low input approach did not suit local soil conditions. I.2 LOCAL CONTEXT The project is located in the coastal zone of Cabo Delgado province in Northern Mozambique, an area with very low population density (less than 18 persons/km 2 ). Shifting cultivation dominates in the region. The soils are sandy and no external inputs are used. Draft animals are absent and there is no mechanisation. The major crop and staple food is cassava, followed by maize, rice and millets. Only 22.6% households report producing enough food, and malnutrition is rampant. Cash crop cultivation and paid employment are virtually non existent, though some sesame 1 UNDP Human Development Index Sustainability assessments Jatropha Mozambique, Partners for Innovation (May 2010) 3 Biofuels Policy and Strategy, booklet issued by Republic of Mozambique Ministry of Energy, May Schut at al,

5 seed is now being produced for sale. Liquid fuel consumption in the area is mainly of diesel for mills and kerosene for lamps 5. I.3 PROJECT AND OBJECTIVES (Hivos project number WW205S01) The pilot project Jatropha for local development in Mozambique was run by the FACT Foundation and local counterpart ADPP Mozambique 6. Originally conceived to run from January 2007 to the end of 2009, work on the project started in May 2007 and the project ended in December It aimed to test the feasibility of enhancing rural development by using locally produced Jatropha oil to run local (adapted) diesel engines, as well as for the local production of soap and lamp oil 7. Infrastructure and capacity building (technical and human) were to enable the autonomous upscaling of activities after the termination of the project. Local production of jatropha seed was to be initiated and a local market of end users of oil developed. The creation of capacity among local small farmers and technicians was an important component of the project goals 7. To this end the concept of Farmers Clubs (FCs) was employed whereby a group of 20 to 50 farmers volunteer to work together on a common demonstration plot, subdivided into smaller plots where each farmer can apply new methods learned from project extension workers and ADPP staff. Expressed in different words in the End Report 8 of the project, its objective was to develop a local market for PPO derived from Jatropha seed produced by subsistence farmers in remote rural areas of Mozambique. This broad objective was subdivided into four specific targets: to 500 ha planted with Jatropha. This target of 250 ha was later converted to 250, 000 trees. Because farmers preferred to plant in hedges, a conversion factor of 1000 plants per ha was used to convert the original target (area) into the new target (number of trees). See I.5 for explanation. 2. Obtaining knowledge on: a) The most efficient varieties and their cultivation b) Jatropha as a host for various agricultural pests c) Combating these with natural pesticides 3. Development of a local market for oil, its use in diesel generators for schools and for running corn mills, and as fuel for a few vehicles. More specifically: a) Market created for 25 diesel engines distributed over 25 villages b) Local technicians receive training on converting diesel systems to PPO, with the assistance of experts from abroad. c) Oil pressed at 10 different sites (5 schools and 5 enterprises) to guarantee an initial market for farmers d) Development of other options: use as lamp oil and for soap production 4. Establishment of a training centre on bio fuels in Mozambique: a) Establishing workshop, accommodation buildings and equipment 5 Flemming Nielsen & Jan de Jongh, Jatropha curcas oil production for local development in Mozambique (Banana Hill and Fact Foundation, The Netherlands, 2010). 6 Ajuda de Desenvolvimento de Povo para Povo 7 Project Proposal Summary, May End Report Pilot Project Jatropha Oil for local development in Mozambique, prepared for FACT Foundation by Flemming Nielsen (Banana Hill Consultants) and Jan de Jongh (FACT Arrakis), February

6 b) Capacity building of TC head, TC teachers, and future experts teachers c) Development of Training Centres for Rural Technologies A lot more information on the project can be found on the website I.4 ALLIANCES AND ACCESS TO EXPERTISE The agricultural research component was designed into the project from the beginning, developed and managed by Flemming Nielsen (of Banana Hill Consultants in the Netherlands). He was technical adviser to the government of Mozambique at the outset of the project, had worked within the national agricultural research system (at IIAM 9 and earlier at ICRAF Mozambique 10 ) and was supported by Danida (which allowed him to spend part of his working time on this project). Jatropha variety trials were originally begun in the Mandonge research forest outside Sussundenga, implemented by IIAM as part of the research component of the project. These were not followed up however, and many plants were lost due to lack of maintenance and weeding. FACT then discontinued support and the trials were abandoned by IIAM. After this some variety trials were set up in Cabo Delgado by FACT/Banana Hill. A survey of insects infesting Jatropha in five provinces (including Cabo Delgado) and pest control trials were undertaken by students from Universidade Eduardo Mondlane (UEM) as part of their thesis work. These were carried out at the project s experimental plots located at EPF Bilibiza 11, from where the project was also administered. Soil nutrient and plant development analysis was done by a student from the University of Wageningen. Finally, the Institute of Geography of the University of Copenhagen took the lead in research on energy and carbon balance of the Jatropha cultivation system. Details on this research and the rest carried out by FACT/Banana Hill can be found at the project website 12. Apart from the institutions mentioned, the provincial and district agriculture departments were involved at the monitoring level. The processing facility and workshop (referred to as the Bilibiza Biofuel Centre, or BBC) were also located on the campus of the EPF Bilibiza. Research on the processing component was designed by Jan de Jongh (of Arrakis in the Netherlands) who was project coordinator for FACT; he was assisted by Christian Fenger (of Gaia Movement, Switzerland). It was carried out in collaboration with Dajolka (Denmark), PPO Groeneveld (Netherlands) and Evretz (Mozambique). Arrakis, in the person of Jan de Jongh, was closely involved in all aspects of the processing stages, including selection of technology and quality control of oil. Niels Anso of Dajolka was responsible for modifying the diesel engines currently in use at the BBC (plus one or two car engines) and training other personnel on this. PPO Groeneveld (Ger Groeneveld) and Evretz were involved in developing earlier engine modification kits. Krishna Raghavan (Arrakis associate) completed the processing 9 Instituto de Investigação Agrária de Moçambique 10 ICRAF had also agreed to be a research partner in the project, but closed its operations in Mozambique due to financial constraints before the research component of the project was started. 11 Escola de Profesores de Futuro, one of several teacher training colleges in Mozambique, established and run by ADPP Mozambique and the Ministry of Educaton of the Government of Mozambique program 5

7 workshop and started the operation, including tests, during 2 months in Bilibiza at the end of I.5 BACKGROUND As a cultivated crop, Jatropha was new to the area and there was little and often conflicting information on best agronomic practices. Formal research was part of the project from the beginning, but planting could not wait for its results. An action research approach was used, with research and implementation carried out in parallel. Observation plots were set up at different locations and monitored regularly. In the end the Jatropha was planted as hedges, with a spacing of around 1m between plants. In order to enable comparison with the original project goals expressed as area (250 to 500 ha), a 3X3 configuration (1000 plants per ha, with 3m between plants) was taken as the equivalent in a field. This was based on the fact that the level of competition for nutrients and the yields possible per tree were roughly the same for the two cases (1m spacing in hedges and 3X3 in fields). For consistency with the actual planting scheme, we will use only the number of trees as a measure of amount of the Jatropha planted, with a target of between 250,000 and 500,000 trees. The number of seedlings distributed by the project was: (in 2007), (in 2008) and (in 2009), a total of 250,000. In addition, farmers also planted some themselves from cuttings (in the nurseries of their respective FCs) as well as seed and seedlings. The live plant counts compiled in 2008 and 2010 in the 5 various districts covered by the project are indicated in Table 1. Table 1: Jatropha planted and farmers mobilised under project District Farmers Live Plant Count Live Plant Count (May 2010) mobilised (November 2008) 13 Farmers fields Demo fields Quissanga ,434 4,845 1,756 Pemba Metuge ,128 15,061 5,537 Meluco , ,903 5,639 Ancuabe 400 4,376 26,279 2,792 Macomia , ,065 20,071 TOTAL , ,353 35,795 TOTAL STANDING TREES (in 2010) 608, 148 A total of around 1850 farmers were mobilised for the project, as shown in Table 1. These 1850 belonged to 36FCs, all based inside Quirimbas National Park. A total of 750 of these farmers were women. Only one person was counted per family as an official participant in the project. Between the termination of the project in 2010 and the evaluation visit in April 2011, another 1000 farmers were mobilised into the program, involving another 20 FCs in Chiure, outside Quirimbas. 13 As reported in the ADPP Fact progress report of November

8 Farmers interviewed in Primer de Maio (Meluco District) and Koko (Macomia District) villages had planted in 2007, and therefore had a harvest. The farmers of Bilibiza FC had planted only in 2010 and therefore had no harvest. I.6 EVALUATION METHOD AND DETAILS The evaluation started in Bilibiza, Cabo Delgado, on April 15. In terms of project staff, only the (agricultural) technical project leader, Bachir Afonso was available for discussion 14. We also inspected the nurseries and experimentation plots of the project at Bilibiza. In addition, the following three groups of producers were interviewed: (1) Eleven (all men) in the village of Primeiro de Maio, Meluco District. (2) Six (4 men, including president of FC, and 2 women who were wives of recruited producers, not members themselves) in Koko village, Macomia District (an FC that consists in total of 16 women and 34 men). (3) Twenty two (21 women and 1 man who was the president) from Bilibiza FC. Note that this FC was recruited later, and is technically not part of the project being evaluated. The size of the total amount of land (for all crops) owned by the farmers interviewed varied between 0.5 ha and 8 ha. The number of Jatropha trees they owned varied between 1000 and 4000 (estimate from farmers and from surveying some of the fields), but it was hard for most farmers to give a number. Local staff had not done any planning for the evaluation visit, so it was difficult to get producers together for interviewing; many were working in the fields and were hard to reach. On the processing side, apart from an inspection of the BBC processing facility, a brief discussion could be held with Itai Chikanya (the last technician trained in processing who left Bilibiza thereafter) at the conclusion of the visit (on April 18). On April 19, a short meeting was held in Maputo at ADPP headquarters with Erik Schurmann. A lot of the information needed for this report was obtained later (after the evaluation visit) by interviewing (and via document exchange with) Jan de Jongh (Arrakis) and Flemming Nielsen (Banana Hill Consultants) in the Netherlands. The questions and discussions with all stakeholders were of a semi structured nature, and derived from the report prepared earlier during the planning phase, along with additional themes mentioned in the ToR and the proposal presented by the lead evaluator. Many questions were modified and others added during the course of the evaluation. II. EVALUATION THEMES II.1 CROP PRODUCTION Main Question: Which factors influence a good production of seed and a good oil content? II.1.1 Crop characteristics and history in the region Though most of the farmers interviewed in Meluco and Macomia districts said they had never cultivated Jatropha before the project or indeed even seen it Jatropha 14 The main project leader for ADPP, Henderson Maposa from Malawi, had left after the project ended 7

9 was known in the region as a medicinal plant and was already being used for fencing homesteads and fields. While it can be effective as a live fence in preventing some wild animals from entering the fields and destroying crops, it cannot keep away elephants, wild pigs and baboons, which are the main threat in the area. As a cultivated crop, however, Jatropha was new to the area. The project involved cultivation of locally available accessions of Jatropha Curcas. II.1.2 Local conditions and Jatropha cultivation Cabo Delgado province gets around 800 mm rainfall annually. The bedrock is close to the surface, making water logging common in low lying areas during the rainy season, despite the fact that the soils are mostly sandy. However, Cabo Delgado was more suitable for Jatropha production than Manica Province, that was originally included in the project, and where Jatropha plantation trials were not successful due to pest (flea beetle) infestation and water logging, exacerbated by heavier soil and annual rainfall levels of above 1200 mm. In terms of pest infestation, according to a survey done by Flemming Nielsen and UEM 15, Cabo Delgado was the most suitable area for growing Jatropha in Mozambique. Soil research done by a student from the University of Wageningen 16 focused on whether the differences in plant development could be explained by variations in soil nutrient levels. The findings indicate that nitrogen and potassium levels did not influence plant development but that higher phosphorus levels led to more branching and higher plants. However, there is little scope for manipulating phosphorus levels within the current farming systems in Cabo Delgado, somewhat unfortunate for this project. II.1.3 Planting Times of planting: Seedlings were prepared towards the end of the dry season (June/July), and were ready for transplanting at the beginning of the rains. Configuration and Spacing: The project initially focused on plots of Jatropha in 2X2 or 3X3 configurations. However, it was eventually decided to follow the farmers preference for growing it as hedges around fields, to delineate their territory since it is easily distinguishable from other plants. Besides, national policy also discourages Jatropha plots (including intercropping) within Quirimbas National Park (for reasons of food security and nature conservation). Of the farmers interviewed during the evaluation, all had planted Jatropha as hedges. Only one farmer had tried intercropping (with maize, cassava, groundnuts, sesame, beans) and found that plants growing close to Jatropha did not do as well. Farmers seem to have settled for a planting distance of around 1.0m in hedges, not dense enough for the hedges to function as live fences. Though the intention had originally been for the Jatropha hedges to double as live fences for protecting fields and homesteads from animals, they were not effective in keeping away baboons, wild 15 Implemented by Pomme Christiane Gagnaux ( ), Superviser Prof. Dr. Luisa Santos, UEM. 16 MSc. thesis Josema Albeniz Larrauri, WUR; see 8

10 pigs and elephants, that are the main threat to crops in the project area. The farmers then decided to concentrate instead on getting better Jatropha yields. Trials with different spacings were established in demonstration plots, including plants from cuttings, direct sowing and seedlings, so farmers could compare the results of different techniques. Data was collected from the plots by researchers but is as yet insufficient to draw conclusions. Direct Sowing vs. Cuttings and Seedlings: Though cuttings were tried successfully, direct sowing was chosen originally since plants from cuttings don t develop taproots, an especially valuable aspect for this area with its poor soils and water scarcity. Informal testing of direct sowing has given good results in most localities. The farmers interviewed said they originally used seedlings from their nurseries, planted from seed and using plastic bags provided by the project. Each FC has its own nursery and many are equipped with a well and hand rope pump facilitated by the project, though water availability remains a problem. But now many farmers are using cuttings because they grow faster and are easy to plant and many farmers are familiar with propagation methods. They also replant seedlings that germinate around the original plants. II.1.4 Maintenance and Inputs Weeding and Pruning: Farmers reported weeds as a big problem, and commented on the lack of knowhow on combating it. Weeding was found to constitute 35% of total labour during the rainy season, though the shade from Jatropha trees is expected to suppress weeds after two or three years. While pruning was recommended for encouraging branching and higher yields, it was reported that farmers were originally suspicious of it, finding it wasteful to cut into healthy plants. This led to lower yields and more weeding due to a relatively open canopy. Tall plants also made harvesting harder. However, pruning was adopted by many FC s after observing its positive effect on plant strength and development. Of the farmers interviewed, a few had been trained on pruning. Some reported having been advised to cut down to the stump at the end of the first year (when the plants were approximately knee high), to trim ramifications in the second to enable better branching, and thereafter every two years. While the literature recommends pruning during the rainy season (because that enables the wounds to heal better), trials 17 indicated that wet season pruning led to fungus growth and dry season pruning gave better results. So the recommendation was for an annual pruning (trimming) in June (middle to end of the dry season). It was not clear how many farmers were aware of this, and of those how many were actually applying this knowledge in the fields. Fertiliser: No fertility enhancing techniques or inputs were used on Jatropha. The concept of maintaining soil fertility is in general somewhat alien to the area, since farmers are used to shifting agriculture, clearing new land when the soil in their current plot is exhausted. However, compost from leaves and manure is now being used, but only for high value crops, not Jatropha. Farmers were not aware of the use of Jatropha shells and husks as organic fertilizer. Most of them had thrown them away whereas just scattering them in the fields could improve fertility and making 9

11 compost or briquettes out of them would be still more effective. They have (so far) no access to Jatropha presscake as fertiliser. Water: Water availability is a serious problem in the area. Only vegetables are being irrigated at the moment, and some seedlings in nurseries. There is not enough water even for vegetables (many farmers reported losses to their vegetable harvest due to water scarcity), so there is little potential for irrigating other food crops, leave alone Jatropha. The project provided training on constructing stabilised wells complete with concrete rings for lining, a rim and cover and a hand rope pump. These wells can be used to water the nurseries so sowing can be done at the end of the dry season and seedlings can be planted at the beginning of the rainy season. They will be increasingly useful during prolonged dry spells and increasingly irregular rainfall patterns associated with climate change. Pests and Diseases: In terms of susceptibility to pests and diseases, preliminary research done in collaboration with UEM showederror! Bookmark not defined. that Cabo Delgado was the best place to grow Jatropha in Mozambique. However, termites have been a widespread problem in the project area. They attack the roots and the plant rots as a result. In addition beetles attack the leaves. And rats have been attacking the seed in storage. No pesticide or herbicide/fungicide has been used. Termites are, however, not perceived to be the primary problem 17. They tend to infest plants that are already weak, due to low soil fertility or exhausted soils or water logged conditions (note that the intense rainfall during the rainy season leads to water logging despite the sandy soil). Though no pesticide has been used so far, a kind of lime sediment (dichotomous earth) seems to be effective 17, and could possibly be brought from nearby coastal areas. The yellow flea beetle (Aphthona dilutipes) was a serious problem for Jatropha in Manica province (observed during the Caritas ADPP project, and again in plantations established originally under this project that were subsequently abandoned), reducing plant growth and often killing the plants. Many farmers in Manica had lost all their Jatropha plants. Though Cabo Delgado has so far not experienced this problem on a large scale, it may be too early to make a definitive conclusion since often, when a species is introduced on a large scale (even though it may have been grown before on a small scale, like Jatropha in this case), it can take a while for the pests to set in. Typically there can be an intermediate period of susceptibility to pests, after which natural enemies have time to evolve/develop and pest infestation again recedes. Chemical pesticides (containing Chlorpyrifos or Cyphenothrin) worked well against these beetles in trial plots in Manica, but are not a feasible option for the poor small scale farmers targeted by the project. Other species observed to infest Jatropha plants are the less devastating golden flee beetle, the rainbow shield bug (Calidea dregii) and the green leaf mining caterpillar (Stomphastis thraustica). In general though, pest damage was most severe in areas that showed signs of low soil fertility. Thus managing soil fertility could be an important means of controlling pest damage. Though a research collaboration on plagues, pest control and bio pesticide had been set up with UEM, this was later 17 Flemming Nielsen (Banana Hill Consultants) and Jan de Jongh (Arrakis), informal communication 10

12 discontinued, and FACT pulled out their funding. The results of the research were also not shared with the project. These included results from trials on three biopesticide extracts, made from Tephrosia, Tobacco and Neem respectively. However some preliminary findings indicated that the time of planting affected insect density years after the Jatropha had been established. When planting is done during periods of high pest infestation (February and March), the pest pressure is also high in subsequent years. II.1.5 Harvest and Yield At all localities the main Jatropha harvest takes place in March April. At some localities there is a second small peak when the rains start, around October. In areas with enough water fruits can be harvested throughout the year, but with peaks. So far, it has been hard to get reliable data from farmers on the amounts they harvested and the typical yields achieved. Among the factors that contributed to this: (i) The output has been meagre so far and therefore not worth it for farmers to keep records of how much they harvested from how many trees. (ii) During the evaluation it was observed that a lot of dried out Jatropha was left on trees or had fallen on the ground. This can be attributed to an overlap with activities connected to other crops and lack of sufficient labour (see II.2 on labour input). The fact that it was believed (and farmers had been told) that Jatropha fruit could be harvested later when it was dried out and almost black, may also have contributed. They were told this based on analyses indicating that acid content is highest in oil extracted from yellow fruit. However, analyses from elsewhere indicate that the oil from black dried out seeds has the highest acidity. This issue is still not resolved; please see II.4.2 on PPO quality control for details. (iii) Farmers got disgruntled with the price offered by the project for Jatropha seed (5 MZN per kg) after having sold seed at a higher price of $1 (or 27 MZN) per kg to external buyers from Tanzania. They were therefore not open about how much they produced and harvested, and had also lost interest in harvesting for selling to the project. When questioned regarding Jatropha yield, there was a tendency on the part of farmers to exaggerate and cite figures from hearsay or to cite the highest yield someone might have got from a single tree. For example, farmers in Primeiro de Maio village mentioned yields of 5 kg per plant per year, an impossible figure. According to projections (based on calibration of data from elsewhere 18 with local conditions), an optimal yield of about 0.8 kg per tree can be achieved in the area by the seventh or eighth year 8. This is a low figure, certainly compared to the claims made earlier in the literature. However, high yielding varieties tend to exhaust the soil and may not be the best choice for an area where the poor quality and not the amount of available land is the main constraining factor (see II.3.1). Project information for 2010 (the third year) on the other hand indicates that 240 kg was produced from 300 plants on average per family 19, implying an average yield of 18 Only two long term data series are known globally; these are from Nicaragua and Paraguay 19 Source: Bachir Afonso, e mail communication 11

13 around 0.8 kg per plant already in the third year likely to be a substantial overestimate. There was also some confusion about weights of whole or dry and decorticated seed. Currently it seems farmers are selling the seed to the BBC without drying it 20 (it is dried at the BBC), while earlier 17 it seems the farmers were drying the seed themselves before selling. This could make a significant difference to calculating yields (and prices), since the moisture content of freshly harvested fruit can be around 20 25%. II.2 LABOUR INPUT Main Question: How much labour is going into Jatropha seed production? The main problem farmers reported on Jatropha cultivation was that the labour input required was untenable at the price they currently get for Jatropha seed from the project (5 MZN/kg for participating farmers and 2.5 MZN/kg for other farmers). Especially shelling by hand was mentioned by all farmers as a cumbersome and time consuming task where mechanisation is badly needed. Weeding was also cited in Primeiro de Maio village (but not in Koko) as a big problem for the larger plots (like 4 ha), taking up to 2 months in a year to carry out. A major complication is that the harvesting times for Jatropha coincide with those for food and other crops, with weed outbreaks, and with land preparation for the next cycle. In most localities in the project area, Jatropha is harvested between February and April (though in some places there is a second, smaller, harvest in October). The main Jatropha harvest (Feb Apr) coincides with the peak demand for harvesting of food crops 5. During the evaluation visit, we saw Jatropha fruit fallen all over the ground since farmers priority was to harvest food crops (mostly to protect them from baboons that don t attack Jatropha). While project reports mentioned that Jatropha does not shatter much, so the ripe seeds can be left on the plants for several weeks (if labour is not available), it is possible that this leads to higher acidity in the extracted oil, though results are not yet conclusive (see II.4.2 for details). Originally, while the project was being set up, Jatropha harvests were expected to take place after the other major crops. However, in reality the fruit is ripening earlier than expected, and harvesting takes place between January and April. Such occurrences are to be expected due to the experimental nature of the crop in general, and the lack of pre existing knowledge about its particular growth characteristics in the area. Project reports also mention labour scarcity as the main factor limiting agricultural production 5. The report states that the labour peak was caused by weeding, which took up 35% of total labour hours during the rainy season. Another 30% was spent on chasing wild animals. In the case of weeding, the shade from Jatropha trees is expected to suppress weeds after two or three years, and most of the labour would be required for harvesting. 20 Source: Bachir Afonso, during evaluation visit 12

14 Thus labour shortage is a constraint within the current farming system, and a major potential obstacle to the take off of Jatropha production in the area. Hiring external labour is not an option since farmers don t have the resources. In this context, the availability of mechanical shelling machines would relieve the labour pressure considerably. The need for mechanised ploughing and tractors was expressed by most farmers, to enable more efficient land preparation for the next cycle of planting while crops are being harvested. With mechanisation many felt they would like to expand Jatropha production to more land since plenty is available in the area, and production needs to be augmented to enable the viability of processing. Farmers anecdotal figures combined with other data from Southern Africa indicates they pick between 2 and 3 kg Jatropha fruit per hour 21. When shelling is included in the process, the average rate reduces to around 1.5 kg/hr. Thus on average farmers are able to generate a value of around 30 MZN per day (if paid 2.5 MZN/kg for seed) and around 60 MZN per day (if paid 5 MZN/kg). The comparison with the minimum wage for hired agricultural labour (80 to 100 MZN per day) is therefore not too favourable. Increased yields can improve the returns to labour a bit, and some improvements can also be made on harvesting efficiency and picking methods. Manual shelling is very tedious and not feasible for a single person to do continuously. It also comprises a major part of the total time spent on seed production, thus reemphasising the need for mechanical shelling. With a hand operated shelling machine 22, the shelling time can be cut down to two minutes per kg; thus 18.5 kg of seed can be produced per day, generating a value of 46 MZN per day (at 2.5 MZN/kg) and 92 MZN per day (at 5 MZN/kg). This compares considerably with minimum wages. However, the combined effect of improved yields and mechanised shelling can be substantial 23, leading to a 3x improvement on returns to labour when yields are tripled and mechanical shelling introduced. Most of all, larger fruits (from breeding and selection) could substantially increase the picking rate and, when combined with shelling machines, can make a really big difference in labour productivity. II.3 PRODUCTION, INCOME & OTHER BENEFITS Main Question: What did the producer gain from the crop? And how is this likely to evolve with increased production? II.3.1 Improved Accessions The main seed source for the project was a local naturalised accession of Jatropha Curcas. The advantage of using local seed sources is that they had been observed to perform well (and had few pest problems) at the project site. There is said to be little genetic variation between African accessions, so trials using different locally available accessions were not deemed to be useful. Many say that African accessions are all closely related to the Cabo Verde accession, having been introduced in Africa from Cape Verde but research has not yet come to a definite conclusion on this issue. 21 Data from elsewhere in Africa (Zambia, Tanzania) indicates 2 kg per hour without shelling. 22 Of the kind produced by BYSA (see Gota Verde project, Chapter 2 of this meta evaluation), costing around $ see Table 5 in Chapter 4 (CEDISA, Peru) of this meta evaluation 13

15 Seed from elsewhere may not be as well adapted to the agro climatic conditions in the project area and would therefore introduce more uncertainty. However, they may yield better, and for this reason, provenances from Guatemala and Tanzania are currently being compared in trials in Cabo Delgado. However research towards identifying locally suitable accessions of Jatropha Curcas should probably not be focused on achieving high yields. In an area like Cabo Delgado where poor quality (and not availability) of land, water scarcity and labour shortage are the main factors limiting production, high yielding varieties (that tend to exhaust the soil) would probably not be the optimal choice. Accessions with high tolerance to drought and dry conditions (see also II.6.4), and relatively large seeds (reducing the labour to income ratio) would be most appropriate. II.3.2 Income, Payment and Profit The farmers of the area have relied on subsistence farming and shifting cultivation of food crops via slash and burn clearing methods. Before the project began, maize had been the main income source, and they had to sell it despite not having enough for their own domestic consumption. They are not selling maize anymore, and the bulk of maize, sorghum and beans are now being grown for self consumption. At around the same time as this project started, ADPP started another project supporting food crop production. Sesame production started around then, because the market price for sesame went up. It has now become the farmers main source of income. There seems to be a fairly secure market for sesame seed from people coming from Nampula (Mozambique) or Tanzania. However these are mostly middlemen who get their cut and pay the farmers 30 to 35 MZN per kg. Price negotiation is not helpful since there is no coordination between different FCs to fix a reasonable price (see II.5.3). Hopefully the project can facilitate this in the near future. Also, the project has been in touch with sesame processing companies for making BBC their collection point, since companies don t want to visit individual villages for small quantities. For example, the Fundacao Contra Fome has offered to buy sesame seed at 90 MZN/kg, almost three times the price that farmers are currently getting. The BBC might itself start processing sesame for edible oil later this year, and might provide an additional market. The project has been buying Jatropha seeds at 5 MZN/kg from participating farmers (and at 2.5 MZN/kg from other farmers). Seven of the 17 farmers interviewed had already sold some Jatropha seed to the project. Amounts sold varied among the farmers interviewed from a few kg to 280 kg in 2010 (at 5.0 MZN/kg) and lower amounts in As mentioned before, sesame is the main source of income for most farmers. For some this was followed by groundnuts and beans while for others Jatropha was already the second highest source of income. The annual earnings in 2010 from selling sesame varied from 2000 MZN to 4500 MZN among farmers interviewed, with annual earnings from Jatropha going up to a maximum of 1400 MZN for the same year. In 2009, many reported the same two crops as being the main income sources but in lower quantities. In 2008, maize was the main income source (at 3 MZN/kg) yielding a total income of 2400 MZN (maize being the sole source of income that year) for a farmer holding 7.5 ha of land, thus at the higher end of the income range. 14

16 Note that the above is anecdotal information, obtained via informal discussions with farmers. According to project information 19, for 2010 (the third year), 240 kg was produced from 300 plants on average per family giving an income of 1200 MZN from Jatropha per family in But this seems rather unlikely 24 given the yields possible (see II.1.5) and the fact that a lot of trees from 2008 did not survive to 2010 (see Table 1). In any case, Jatropha already seems to be contributing up to a third of the incomes of some farmers. However, the value of this annual income is still abysmally low, hovering around subsistence level. In Table 2 the returns from land use from growing Jatropha, maize and sesame have been given (maize is the staple food crop in the region, and sesame is the main cash crop). Note that this is just to give an idea of incomes generated, and a direct comparison of Jatropha with these other crops may not be justified in this case, for various reasons explained in the next paragraph. Data in Table 2 have been obtained from varied sources as cited, since data on food crop yields were not available from the project. For Jatropha, the optimal yield of 0.8 kg per plant (expected in the 7 th or 8 th year) has been used, and an equivalent of 1000 plants per ha even though it has been planted as hedges (see I.5). Table 2: Returns on land use from Jatropha and other crops Crop Yield (kg/ha) Price (MZN/kg) Income (MZN/ha) Jatropha Maize (rain fed) Sesame Such a comparison would indicate that other crops give far better returns for land use, especially since sesame prices offered can go a lot higher (to 90 MZN/kg; see text above) than what farmers are currently getting. However it must be said that this comparison is not too relevant in this context where land availability is not a constraint, and Jatropha has been planted as hedges, and is therefore not interfering much with the use of land for other crops. The estimates made in section II.2 of returns to labour are more valid for estimating Jatropha viability, because competition for labour and not for land is the determining factor. Participant farmers interviewed during the evaluation felt the price (5.0 MZN/kg) was too low, and indeed returns for labour (see II.2) do not compare well with minimum wages at current yields and prices, and without shelling machines. However, this price is on the high side for Jatropha PPO to be competitive with fossil diesel (see II.4.5). This implies that a substantial price increase is not feasible in the near future. 24 Project data was not being monitored at this time since the person in charge was no longer around 25 Maize yield in Cabo Delgado and maize price (6 to 9 MZN/kg) for North Mozambique in , both from Special Report FAO/WFP Crop and Food Security assessment Mission to Mozambique; see 26 Smallholder sesame yields are around 250 to 400 kg/ha in Mozambique; see 15

17 II.3.3 Flexibility and diversification Jatropha is grown as hedges and has not interfered with diversified food crop cultivation. The main crops farmers had been cultivating before the project began were maize, beans, sorghum and groundnuts, and to a lesser extent cassava, sweet potato and pigeon peas. Since the start of the project in 2007, they are additionally cultivating Jatropha, sesame and in some cases other vegetables like garlic, cabbage, green pepper, carrots and cauliflower. These new supplementary food crops were facilitated by a parallel ADPP project offering support on food crops, that involved many of the same farmers. While Jatropha plots and intercropping with food crops are not encouraged in the area, the fact that the project ran parallel with another ADPP project focusing on improving the viability of food crop cultivation enabled diversification of income and the sharing of inputs and technical expertise. Equipment and training on well construction (as part of the project being evaluated) has facilitated food crop cultivation. Jatropha cultivation itself represents a diversification opportunity for farmers in terms of cash crops, from the one or two food cash crops (mainly sesame) that they currently produce. However, there does appear to be competition for labour between Jatropha production and that of other crops (see II.2). II.3.4 Investment & Credit Access Sixteen of the project FCs elsewhere (in Pemba Metuge, Ancuabe, Macomia) are now officially registered, which makes them eligible for some nominal credit from the government. However this is only for food crops and not for Jatropha. Some farmers have already applied for and received some credit from the government for inputs (water pump, seeds) for food crops, and for selling vegetables. No interest is charged on these loans. In case of loan default, the farmer in question cannot benefit from further loans or credit. However the farmers interviewed in Primeiro de Maio and Koko villages said they received no credit at all. The procedure for registration is tedious and time consuming, and too complicated for farmers. The project has been helping out, but support is needed for transport and tax payments: farmers need to go first to the Chief of Post and then to the District Office in order to register, and to pay a tax of 2500 MZN. II.3.5 Other benefits for producers (i) Training and Capacity Building: Extension workers usually come from ADPP Bilibiza, and stay in their area of work for a week or so to reduce transport costs and increase the number of effective working hours. It also makes them more directly accountable to the communities. Demonstration fields in the villages are visited roughly once a week. However no extension workers were available for interview at the time of the evaluation. Typically the extension worker trains the farmers (on everything, all crops) that happen to be at the demo fields, the idea being that these will in turn train others. But it seems that such secondary training did not take place too often. However, several farmers in Primeiro de Maio and Koko villages had been trained directly by extension workers on pruning, increasing productivity, soil conservation, selling and 16

18 negotiation. Some of the women were trained on making clay stoves, planting, cleaning and cooking. As far as training on processing methods is concerned, only some club presidents have been trained. The president of Primeiro de Maio for example said he trained others, but very generally, since there was no equipment. The other farmers present in P de M had not been trained on processing and were not aware of the potential of Jatropha PPO as fuel and energy source. In any case, training for farmers on processing was not part of the project, only some demonstration of the facilities. However many producers (50 women and 1 man) from the Bilibiza FC (which is officially not part of the project being evaluated, since they were mobilised in 2010) attended a workshop on soap making (probably due to their proximity to the BBC). They were provided with plastic gloves and a mask to use while making the soap (mixing caustic soda). Some in this FC were also trained on food nutrition, planting and running a business. (ii) Transformational effect of local energy access: The project has had no impact on energy access for producers. (iii) Participation in cooperatives and joint action societies: 36 Farmers Clubs 27 were included and are growing Jatropha. Farmers in Primeiro de Maio said they learnt a bit about capacity for negotiation from the extension worker. Farmers from Koko said they were trained on selling and negotiation. However it was observed during the evaluation that farmers are not organised for effective negotiation of prices and conditions. In the case of sesame for example (see Section II.5.3) external buyers are able to set their price. (iv) Increase in physical capital: Training and materials were provided for the construction of wells and hand rope pumps, thus leading to increase of physical capital. The wells were lined with concrete rings of 6 cm thickness and covered with a concrete plate. The water from these wells is potable. It is being used to irrigate the nurseries and some food crops. Vehicles (two small trucks and five motorbikes) were bought for the project. The trucks were used for collection (and for construction of wells etc) and the motorbikes for transport of the extension workers. However the motorbikes are the property of ADPP and will probably be taken away for other projects. II.4 PROCESSING Main Question: What can be gained from processing (M/F)? How is quality control of PPO/biodiesel organised? II.4.1 Oil Extraction Oil pressing finally started in late 2009, due to delays in obtaining equipment. The central oil pressing unit, the BBC was set up then, with a skilled operator. However, during the evaluation visits to the processing centre, oil was only being extracted for testing purposes, and no skilled operators were present. The engines were running 27 These clubs, previously created by ADPP, are groups of a maximum of 50 small farmers who work individually and support each other. Cooperatives are apparently not popular in Mozambique. 17

19 on sunflower oil and not Jatropha PPO, due to the high acidity levels of the latter, as well as low production volumes. But soap making activities, which are still viable with acidic oil, had also been suspended. Farmers do the de husking by hand before the seed is sold to the BBC. It was reported during the evaluation visit however that farmers are not drying the seed themselves. Apart from causing confusion in determining the price (since the moisture content affects the weight significantly), this may be a problem depending on how long the seed stays with the farmers before it is collected (and the possibility of fungus developing if the seed is not dried properly). After the seed is brought to the BBC from the field, it is dried for a week in the sun, and then for another 2 weeks in sun/shade before oil extraction. The BBC has two motorised oil expellers: a Sayari Press (7.5 kw) from Tanzania, and a Chinese Double Elephant (DE) Press (5.5 kw) bought in Maputo. Both can extract 1 litre of oil from 5 kg of seed in about 5 minutes, but the DE press is cheaper. 28 Though the DE got stuck repeatedly and technicians could not be easily brought from Maputo, Niels Anso (Dajolka) finally succeeded in getting it to run smoothly, and project staff now have more faith in it than in the Sayari. After extraction, the oil is taken through a 3 stage filtration system: first filtered by sedimentation, then filtered again with cloth (residues remain in cloth), and finally passed through a cartridge micro filter, after which it collects in the cartridges of the apparatus. While the scale of the project demanded large motorised screw presses for extraction, hand operated expellers were also considered due to their affordability and easy operation and maintenance. But testing on Bielenberg and Piteba hand presses showed that heavy clogging made their use unviable. Local community members also found the production of less than a litre of oil per hour to generate very little value (however hand presses can now extract up to 2.7 litres per hour). Another argument mentioned against hand expellers was that though they would enable community members to extract oil locally and use it themselves, the oil would still have to be transported to the BBC for quality control. However, quality control is not needed for soap production. II.4.2 Quality of PPO Quality Control of the first batches of PPO produced from locally grown Jatropha has revealed that it is quite strongly acidic while it should be close to neutral for use in engines. While the acidity level was reported to be to around 7.5 (reduced from still higher values) during the evaluation, it should be at most 2.0 to be viable for use in engines. Krishna Raghavan and Jan de Jongh (Arrakis, Netherlands) had measured acidity and other properties earlier and found that acidity could be lowered from 17 to 3 via a neutralisation process using caustic soda. However a lot of oil was lost in sediments in the process. They had also trained others earlier at the BBC on testing for acidity using titration methods, but these people (Henderson Maposa and Itai 28 DE cost $3000 in Mozambique, including plate filter; $2000 if bought in China with local costs depending on the local taxation regime. Sayari was bought in Tanzania for $4200 including a diesel engine, with total costs (transport, spare parts etc.) coming to $6300, but import taxes probably doubled the price. 18

20 Chikanya) have since left and expertise for testing is currently not available. Moreover, calculations show (see II.4.5 below on processing costs) that it is not at the moment economically viable to reduce the acidity and make the PPO suitable for use in engines. However, oil quality can improve with quality management along the entire seed and oil production chains. This will require monitoring of all activities for their potential impact on oil quality: from soil preparation to training farmers at what stage to pick the seeds, to drying and storage techniques and distribution. Once the oil meets fuel quality standards, regular on site testing may not be necessary, and samples can occasionally be sent to labs elsewhere 29. Research from different sources indicates that the state of ripeness of fruit when it is picked (as indicated by the colour of the shells) can influence the acidity of the oil extracted from it. However mutually contradictory results have been reported on this from tests carried out in different places. At INIA in Peru it was found that oil from blackened and dried out fruit is the most acidic 30. On the other hand, Niels Anso (Dajolka) got samples of green yellow and brown black seeds from Mali tested at ASG Labs (Germany), and his results showed that the green plus brown/black are within acceptable acidity levels, but the yellow ones were too acidic and had too much phosphorus. In the absence of conclusive evidence, it seems safest to pick seeds when they are light brown, but not black. Apart from the right stage for harvesting, storage in galvanised tanks leads to the creation of polymers that can block fuel filters. Also during pressing the quality can change (too high pressing temperatures increase the phosphorus content of the oil). The moisture content may also be a problem, depending on how long the seed stays with the farmers before it is collected (and the possibility of fungus developing if it is not dried properly). In general there is some correlation between the acidity level and other characteristics like phosphorus content and particulate matter content. If the acidity level is low enough, chances are that other aspects (phosphorus and particulate matter) also are. Laboratory testing and quality control are hard to achieve in Bilibiza (especially for phosphorus and particulate content) samples need to be sent to labs in Germany or South Africa for testing. In December 2009 a sample was sent to ASG(Germany) for testing and gave negative results on nearly all the critical variables. There may now be a possibility for getting the testing done at government laboratories (Agriculture Department) in Pemba. Testing at the BBC would in any case only be viable for acidity control and titration, in case of improved availability of equipment, chemicals and expertise. For example, the chemicals needed for titration like methyl alcohol, phenol red etc. are not available in Mozambique, and had to be smuggled in from abroad (it would not pass through customs in Mozambique). For checking the phosphorus content (which needs to be checked if the oil is to be used in cars), a gas chromograph is needed, and is not available in Mozambique. A centrifuge would be needed for removing particulate matter and water (to avoid engine corrosion). But the main problem at the moment is the lack of reliable local expertise; equipment can eventually be brought from elsewhere if trained personnel would be available locally for operating it. Furthermore (as 29 For example processing facilities being built in Beira are likely to have all lab facilities for quality testing, making this easier and cheaper 30 See Annex 2 in Chapter 4 of this meta evaluation (on the CEDISA project). 19

21 mentioned above) it is hoped that streamlining of quality control can reduce acidity and other contaminant levels (thus eliminating the need for frequent testing) once definitive information is available on how acidity levels evolve in ripening fruit II.4.3 Producing Biodiesel The project is not producing any biodiesel, and given production cost as well as the technical expertise and imported chemicals needed it is not expected to do so in the near future. II.4.4 Other Products Soap: Given the problems with quality control of PPO, soap making would appear to be the most viable use for the oil at the moment. In the long term also the plan would be to use oil of high acidity for making soap. While we were told during the evaluation that Jatropha PPO is currently only being used for making soap, there were no indications of soap being made at the BBC or elsewhere at the time. However, workshops on soap making had been held. The female farmers of the Bilibiza FC (an FC with almost all female members that was mobilised later, after the project) were very enthusiastic about the workshop they had attended where they were allowed to keep the soap they made. However they had only started planting in 2010, and had thus not produced the seed themselves. An earlier plan to invite a trainer from Arusha (Tanzania) for advanced training on making specific types of soap and soap based products (washing soap, toilet and luxury soap, shampoo, shower gel, hand wash, disinfectant, mosquito repellent) had to be shelved because the project ended before such a stage was reached, and there was no money left. A continuation of work under a new EEP funded project might enable this to happen. Peanut butter and sesame oil: Given that groundnuts and sesame are common crops in the area, these were identified as other value added products with potential for relatively easy production at the BBC and a wider external market. The idea would be to use one of the extractors for edible oil extraction and the other for Jatropha. Some trials on peanut butter were reasonably successful; it was even distributed locally for evaluation and got positive feedback but there has been little activity since then. II.4.5 Investment and processing costs; Profit distribution PPO: At the moment PPO production is just barely viable economically, even apart from its acid levels and potential additional costs that would be incurred from quality control and acidity reduction in the laboratory. Table 3 below shows PPO production cost using the extractors at the BBC at two different seed prices. As shown, PPO production costs about 30 MZN/litre (36 MZN/litre including VAT) at a seed price of 2.5 MZN/kg, and 44 MZN/litre (51 MZN/litre including VAT) at a seed price of 5.0 MZN/kg. Fossil diesel costs about 50 MZN/litre locally, so additional laboratory costs for reducing the acidity of PPO would make the whole process unviable. 20

22 Table 3: Two cases for cost price calculation of Jatropha PPO 8 (This version incorporates the reduction in travel costs communicated later by project staff) Case 1: Seed price 2.5 MZN/kg Case 2: Seed price 5.0 MZN/kg Component Assumptions Cost/litre PPO (MZN) Cost/litre PPO (MZN) Purchase of seeds 4.5 kg seeds needed to press 1 litre of PPO Transport Seeds collected by own 6 7 truck Salaries Only part time 8 8 Processing PPO used to drive 3 4 engines for press Other Costs 2 2 Subtotal Tax 17% VAT Total Costs Originally, the plan was to reduce the seed price to 2.5 MZN/kg after 2010 to keep PPO competitive with mineral diesel. However, diesel prices have been rising, and this may not be necessary. Given the level of farmer discontent (during the evaluation visit) with the current price of 5 MZN/kg, it may also not be feasible. Farmers can also produce their own oil to substitute diesel or kerosene in maize mills, water pumps etc. Using a hand press (5.5 kg seed for one litre of oil at about 1 lit/h) with an 8 hour working day and including time spent on picking and shelling seeds (about 1.5 kg/h) generates a value of around 62 MZN/day 31 if a price of 50 MZN/lit is assumed for the fuel substituted (diesel or kerosene). This is comparable to the value of around 60 MZN/day generated by selling seed (at a price of 5 MZN/kg), and the minimum wage for hired labour of 80 to 100 MZN/day. Hand presses operating at double the speed (extracting 2 lit/h) have however been reported in other projects, but would increase the value generated by oil extraction for local fuel substitution only slightly, to around 67 MZN/day. Oil extraction via a hand press for replacing fossil fuel in engines is therefore (at the moment) not profitable, even if additional issues of quality control could be resolved. Soap: In the end, it was concluded that good soap could be made for a price that could compete with other soaps on the market. PPO that does not pass quality standards for engine use can still be used for soap. Most of the soap made so far has been in training workshops for producers. Members of the Bilibiza FC (comprising of 46 women and 4 men, some of whom were interviewed for the evaluation) were among those trained. The soap was not sold, but each could take home the bar they made. The women were very enthusiastic about setting up a small soap making business, and demonstrated the process to us during the visit, including the use of plastic gloves and a mask while mixing the caustic soda. A few bars were sold in Pemba (20 bars at 100 MZN per 1 kg bar). Soap bars were made of 1 kg (100 MZN) and 380 gms (40 MZN). The cheaper soaps on the market cost around 15 MZN for 100 gms, so Jatropha soap is competitive at the cited prices. Though it was more expensive than the cheapest brown soap bars on the market, local communities are 31 Own calculation. 21

23 generally prepared to pay a bit more for its health properties and better appearance. However no more soap making activities were held. In order to maximise benefits for producers and their self sufficiency, they need to be able to extract their own oil, or at least to have easy access to it. A barter arrangement whereby farmers are paid for Jatropha seeds with soap production kits consisting of oil, caustic soda, gloves etc. might be one way to proceed for now. In this way the high quality oil can still go to the engine fuel market. Calculations of the production cost of Jatropha soap, and of profit margins were not available from the project, so an estimate has been made in Table 4 of the value that can be generated from soap production. This assumes producers buy the PPO for 45 MZN per litre (around the production cost at seed price 5 MZN/kg, as estimated in Table 3). Since they are also producing the seed themselves (and can possibly produce the oil) they can make an even wider profit margin. Labour data and amounts of inputs have been taken from Kakute (Tanzania) 32. Table 4: Value that can be created by soap production Inputs Amount Price Cost (MZN) Jatropha PPO 20 litres 45 MZN/litre 900 Caustic Soda 3 kg 34 MZN/kg 102 Wrapping 50 TOTAL COST (25 bars of 1 kg) 1052 TOTAL 100 MZN per bar 2500 PROFIT (26 hours labour) 1448 Thus a value of 1448 MZN can be generated in 26 hours, or 446 MZN per day. This is a significant improvement on returns to labour from seed sale (around 60 MZN per day), as well as on the minimum wage for hired labour ( MZN per day). This calculation uses a sale price of 100 MZN per 1kg bar, considered viable in local markets in Cabo Delgado. Better quality production for wider markets can enable wider profit margins and even better returns for labour. Thus soap production is by far the most economically viable activity for producers at the moment. II.5 MARKETS Main Question: Are prices covering the expenses (including transport) and what are the prospects? II.5.1 Protection against volatile oil markets The seed price was set to allow maximum benefits for producers while keeping PPO price competitive with that of fuels (like mineral diesel and kerosene) that it could substitute. Estimates made by project staff 17 indicated that a price of 2.5 MZN/kg would be above break even for farmers, and was acceptable for farmers (based on interviews they did earlier). At the time of the evaluation however farmers were very dissatisfied even with the price of 5 MZN/kg they were getting from the project. Part of the reason could be that these estimates do not take labour shortage (and competition for labour with other crops at peak harvest time) into account. The fact that external buyers from Tanzania were willing to pay a lot more would also 32 See Table 2 in Chapter 5 of this meta evaluation (TaTEDO, Tanzania) 22

24 have contributed to the general sense of dissatisfaction. As in many cases over the world, these short lived external markets offer higher prices (in this case the Tanzanian buyers were offering $1 or 27 MZN per kg, compared to 5 MZN/kg offered by the project) for seed to be used for planting. Some farmers have indeed sold to these external buyers, but it is not clear how many and how much. Farmers in the region have a history of being exploited by the government and by companies, and saw this project as being different. Project coordinators feel it will take time to win back their trust after the Tanzanian experience. II.5.2 Local market development It is too soon to say about this, since hardly any PPO has been processed. At the moment, Jatropha soap is the main product that can generate substantial income (see Table 4), and has a viable local market. But there has been little activity on soap production, apart from a trial workshop or two and the sale of 20 bars in Pemba (see II.4.4 and II.4.5). Though the price is higher than that of the cheapest brown soaps, local farming communities elsewhere in the region (Zambia, Zimbabwe) have been prepared to pay a bit more for its medicinal qualities and its cleaner brighter look. With more advanced training and better quality soap, there is of course a potential for entry into wider specialised national markets for natural and health products. However, catering to local demand should be prioritised in the early stages. Peanut butter and sesame oil would also appear to be products that could be easily produced at the BBC with currently existing project equipment, and that have good local marketing potential already, as well as potential for entry into wider national markets once quality can be improved. For the longer term, farmers expressed a strong desire for locally situated generators running on PPO for electricity. The project plans include an electric fence to protect food crops from elephants and baboons, which would be immensely useful but PPO quality is at the moment not suitable for use in generators. Use of PPO for stoves and lamps was another alternative farmers looked forward to. Women were especially interested in stoves running on PPO that would absolve them of the tedious, time consuming and often dangerous business of collecting firewood. However, at the moment PPO for stoves and lamps may not be viable; also stoves and lamps that can operate efficiently on Jatropha PPO are still in the developmental stages (see II.8.4). II.5.3 Innovative & Equitable Business Models Farmers have no interaction with large scale processing companies in this project. Some external buyers do come around to buy sesame seed, food crops, and Jatropha seed for planting. So far there was no evidence of farmers having negotiated with them, though there seems to be significant potential for gains from higher prices especially for sesame. The fact that the various FC s do not meet each other is a factor that contributes to the lack of negotiation capacity. If farmers of one village decline to sell at the price offered, the buyer can just proceed to another village. This was discussed during the evaluation and the farmers expressed significant interest in coming together regularly with FCs from other villages to enable coordinated and effective price negotiation if the project would provide some support. 23

25 The question of negotiating prices of Jatropha seed is not relevant at the moment since the arrangement right now is for the project to buy all the seed for processing at the BBC. However farmers have sold outside the project since external buyers do come round and offer higher prices, especially for seed for planting, and this could be a potential hurdle for the project. Alternative models for distribution and marketing of seed might alleviate this problem (see III.2). II.5.4 Flexibility and Diversification On the processing side, not much has been achieved in this domain, despite the significant potential for soap production for local markets, and for manufacturing peanut butter and sesame oil at the BBC. These could supplement farmer incomes the year round due to their value added nature and storage possibilities and are already viable (unlike PP for fuel). On the agricultural side, a parallel project supporting food crop cultivation has enabled differentiated sources of income to some extent. There seems to be a secure market for sesame for example (see II.3.3) and the possibility for negotiating higher prices. Other food crops are sold locally to officials from District HQ who come by periodically to buy them. Jatropha also adds flexibility because it can be left unattended for many years if other crops are more profitable, and quickly be brought back into production when needed. II.6 FARMING SYSTEM RELATED ISSUES Main Question: Does the crop fit in the farming system or produce synergies, or does it distort system relations? Does it affect food security? II.6.1 Risk for producers Changing weather patterns constitute a substantial risk for farmers. For example, several farmers suffered losses recently due to the late arrival of rainfall, but had no recourse to support or compensation. There are no risk sharing measures in place, and agricultural insurance is unknown in the region though perhaps larger scale private enterprises have access to it. However ADPP has introduced measures for minimising the damage caused by extreme weather events, for example by providing wells and hand rope pumps (to see farmers through dry periods) and building small dams, earth boundaries, etc. (to prevent water runoff and erosion during periods of heavy rainfall). II.6.2 Availability and Quality of Land As mentioned earlier, availability of land is not a constraining factor in the area. In fact, there is such an abundance of land that famers have traditionally not spent any time on improving soil quality and fertility in their existing fields. Instead, when the soil is exhausted they just move on, clearing more forested land via traditional slash and burn shifting agriculture that has been practiced for a long time. Consequently, deforestation is becoming a big problem in Quirimbas National Park. By motivating farmers to concentrate on soil conservation in their existing fields, the project could in principle contribute to slowing down the deforestation though it is too soon to give a verdict. The project is promoting conservation farming (crop rotation, intercropping, use of mulch, agroforestry and jatropha hedges against erosion) to 24

26 make it more viable for farmers to stay on their plots instead of moving on to clear new land (i.e. slash and burn). As part of this conservation farming initiative, the project has also helped establish nurseries for planting various trees. EnviroTrade is paying farmers for planting trees. They pay 10,000 MZN 5000 MZN and 2500 MZN respectively in years 1, 2 and 3 for planting 50 trees. Among the trees being promoted is Falderbia Abida (a good nitrogen fixer). Another factor that might motivate the farmers settling down and dissuade them from further slash and burn activities is the facilitated access to water and other provisions to which the project has contributed. Jatropha hedges delineating territory may also promote a sense of ownership of land. The project s planned provision of an electric fence running on PPO, to keep wild animals away from the fields, may if implemented also serve as an added incentive for settling down. Though deforestation rates have not been monitored directly, farmers have been observed to stay longer on the same plot. However, many farmers have fields in remote areas, far from their homesteads, and it is hard to keep tabs on the extent to which they might be continuing to expand there by cutting down more forest. As to whether Jatropha cultivation can improve the quality and productiveness of degraded land, also for other crops, there is no verdict here since Jatropha was not planted on degraded land, only as hedges around agricultural fields planted with food crops. II.6.3 Food Security Food shortage is a problem, though the availability of land is not a constraint in the area and the Jatropha was anyway planted as hedges, leading to little interference with land used for food crops. So the competition for land between Jatropha and food crops has not been a factor here. However, competition for labour (see II.2) is a problem. Though farmers seem so far to have been prioritizing food crops over Jatropha, its cultivation and harvest has probably taken some time away from tending to food crops. While farmers groups consulted by the project responded that time and effort is only spent on Jatropha and other cash crops when their food needs are secured, only 23% reported producing enough food for their household needs. Subsistence agriculture with no external input prevails in the area. There is little competition for inputs between food and cash crops since food shortages lead to the automatic prioritisation of food crops. For example, only vegetables are irrigated due to the water scarcity in the area. Some other reasons contributing to farmers prioritization of food crops are: timely harvest protects food crops from wild animals that don t affect Jatropha, producers find the prices offered for Jatropha not attractive enough, and the fact that they were told that Jatropha seed could be left on the trees without much damage. However the last may not be true (see II.4.2), and if a Jatropha based economy becomes viable and profitable, farmers may give it higher priority. It is possible that that occurs at the expense of food crops, due to the severe shortage of labour. II.6.4 Traditional Jatropha Cultivation See II.1.1 II.6.5 Water Availability 25

27 Water availability is a serious problem in the area. Currently only vegetables are irrigated when water is available. As described in II.1.4 and II.3.5, wells and pumps were constructed with support from the project to make this possible. Some rainwater harvesting measures were reportedly initiated by ADPP (like the construction of ridges and gullies) but not observed during the evaluation visit. Though high yielding Jatropha plants generally need quite a lot of water, this is not feasible in the area. In fact it does not seem feasible to irrigate Jatropha at all in the area, or to cultivate high yielding varieties that can further exhaust the soil (see II.3.1). Though Jatropha is not being irrigated, accessions that can provide good yields under water stress would be very useful. However, it should be ensured that such improved accessions don t absorb more water and nutrients from the soil, to the disadvantage of other crops in the vicinity. Farmers in the area have observed that other plants in the vicinity of Jatropha plants don t do so well. II.7 WIDER SOCIAL AND ENVIRONMENTAL ISSUES Main Question: Is the project socially or environmentally unacceptable? II.7.1 Land Ownership Land tenure rights are dictated by customary law. In general, a family or person is entitled to stay on their land if they ve occupied it for 10 years or more. According to local tradition, when a boy grows up he inherits his father s axe ( ) so he can find, clear (slash and burn) and occupy new land to start a family. It is thus not common for children to contend for their parents land. In case of death, the spouse usually inherits but in the end such decisions are taken by the chief. Land ownership is not registered in Mozambique, and in any case the tradition has been to move on to a new piece of land after a couple of years. Also as far as external private investors are concerned, land ownership is not possible in Mozambique; only a 99 year lease is possible. II.7.2 Wasteland Wasteland or degraded/marginal land was not used for growing Jatropha. It was grown as hedges around plots planted with food crops. Thus the issue of the project possibly taking over land that is unofficially used by the landless poor is not relevant. II.7.3 Environmental Risks Jatropha has been present in the region for some time without spreading beyond where it was planted, so the risk of it being invasive locally is quite minimal. Biodiesel is not being produced, and its production is not foreseen for the near future, so there is little risk of chemical leakage (for e.g. methanol). There is little evidence of forests being destroyed to plant Jatropha (see II.6.2) but in case this should happen, a study undertaken with the University of Copenhagen revealed that it could take 1900 years to pay back the carbon debt. Finally, there could be positive environmental effects if rampant battery usage (and the ensuing risk of leakage of toxic chemicals ) can be reduced by using Jatropha PPO. II.8 LOCAL AND REGIONAL ECONOMY ISSUES 26

28 Main Question: Does the project fulfil a need that was not previously satisfied? II.8.1 Appropriate Technology The project is situated within Quirimbas National Park in a remote part of Cabo Delgado, 35 km from the main road, reachable only via a very poor dirt road that can become quite impossible to navigate during the rainy season. However, part of the reason for this location was precisely to enable supplementary income and energy access to remotely located communities. Thus the proximity of technical expertise was not a criteria that could be fulfilled. This has indeed been a hurdle to the development of the project. While the project provided access to technical training of high quality via regularly visiting foreign experts, the locally based personnel whom they trained have since all left the project probably for the same reasons of location and isolation and bad (or non existent) basic facilities. At the time of the evaluation visit, one technician (Itai Chikanya) was still present, but he also left soon thereafter. A good example of innovative appropriate technology catering to specific local constraints and needs would be the use of PPO in generators that power electric fences to keep wild animals (a major threat to food crops) away from the fields. Another would be the substitution of diesel, kerosene etc. in maize mills and water pumps. These are all examples of uses that boost production efficiency while increasing demand for PPO. However, quality control issues need to be addressed first (see II.4.2). On the other hand, the substitution of fuelwood by PPO (in stoves etc.) may not yet be viable for local circumstances (see II.8.4). II.8.2 PPO and Local Energy Generation Biofuel is not yet being used so there has been no effect on keeping money circulating in the local economy by replacing fossil fuel or other products transported from elsewhere. At this point it is also difficult to make well defined estimates of the potential for this happening in the near future. Only some observations can be made regarding the possibilities for substitution by biofuel, or of product substitution from its processed products and by products. Of course once PPO quality meets fuel standards, there are many more potential uses that would enable local energy generation and reduce dependence on national grids while reducing costs. While the BBC and EPF Bilibiza have generators (running on diesel or PPO) and the EPF was to get connected to grid electricity (soon after the evaluation, in anticipation of a visit from the President), the villages don t have electricity access, and given their remote locations and sparse populations most probably will not have it any time soon. Originally, the import of commercial kits for adaptation was not considered suitable for the project due to their high complexity and price. Research was done on developing low cost conversion kits from locally available materials, resulting in a first prototype that was successfully tested in the Netherlands. A second prototype was tested in Mozambique 33 in 2009 but failed for several reasons. 33 At Evretz, the farm of a Zimbabwean farmer experimenting with biofuels 27

29 Therefore two Chinese Feidong (Lister type) engines were purchased and modified for the BBC. These engines are easily available and commonly used for maize grinders. However, contrary to claims in the literature that Lister type engines can be run on both diesel and PPO without modification (except for adding an extra filter), it was found that modifications were necessary for dealing with the high viscosity of the Jatropha PPO. The engines were modified to a two tank system (a tank was added, among other things) by Niels Anso (of Dajolka) using imported parts from Denmark, at a total cost of around 200. The principle was to use the diesel tank to start up the engines (since Jatropha PPO is too viscous in a cold state), and thereafter to use the water circulating in the pipes (used for cooling the engines) to pre heat the Jatropha PPO to reduce its viscosity. One of these Feidong (15 kw) engines functions as the drive engine of the BBC generator set. It was used for a while (on cooking oil) to supply power to the workshop (water pump, oil expellers etc.) and as a back up for the school centre. The two tank engine and stabiliser cost about $13,000 all together. The other Feidong engine functions as the drive engine of the maize mill of the EPF school. There is also a plan to use one of these engines/generators for powering electric fences against elephants and other animals. In addition, one car (a Nissan 4Wd) was modified by Niels Anso to run on PPO (via single tank fuel injection, an adaptation using parts imported from Denmark, costing around 700). It was driven to Maputo (over 2000 km) on sunflower oil as part of a presentation of project results to the public. None of these engines are currently running on Jatropha PPO not enough has been produced and moreover, the quality (see II.4.2 on quality control) is not yet acceptable for use in engines. All the engines have been running since last year on edible (sunflower) oil bought from outside. Before last year, fossil diesel was being used. Sunflower oil is being bought to run the engines at 45 MZN/litre. Diesel costs 50 MZN/litre locally, possibly a bit more in remote areas due to transport costs. II.8.3 Multifunctional Platforms The engine set up at the BBC could be considered as a kind of MFP, since it can provide backup support for electricity generation for the EPF Bilibiza, for grain milling and so on. However it is not based in producer communities and so far the replication of such units in producer villages remains a distant prospect. II.8.4 Other Direct Local Uses of Jatropha Oil (PPO) Soap: At the present time soap seems to be the most viable product for local markets, given that its production is not affected by the high acidity level of the PPO. If production could be stepped up to cater to the entire market for soap in Cabo Delgado (or at least within the communities participating in the project) a sizeable amount of money, spent on soap manufactured outside the area or brought from Maputo, could be kept within the local economy and generate income for the producers (see II.4.5) 28

30 Lamps: During the evaluation visit farmers identified LED lamps as their main source of light; many seem to have switched from using kerosene lamps. The LED lamps are self constructed makeshift torches using worn out radio batteries. The batteries are placed in LED frames extracted from cheap Chinese torches that are available locally. They provide a very dim light, dimmer than kerosene lamps. When the batteries wear out completely, they throw them in the rubbish or bury them. While there is potential for lamps based on Jatropha PPO in terms of better light and sustainability, they cannot compete with LED lamps in economic terms 34. A couple of models for lanterns have been developed that can run on Jatropha PPO. It remains to be seen however whether there is a local demand for these. That would depend on their production cost, efficiency in terms of PPO usage and on whether people are willing to invest in better quality light. One prototype lantern that can run on Jatropha PPO has been developed by JPTL (Tanzania). It can be constructed from materials available in most African villages for about Tsh 3500 (or 59 MZN). The LED lamps cost about 25 MZN (10 MZN for the LED frame and 5 MZN per battery), but since they use already discarded batteries and other parts, the effective cost is even lower. Stoves: Firewood is also used a bit for lighting, but mostly for stoves. It is collected mostly by women who spend 4 to 8 hours per week on this activity, often under difficult and dangerous conditions. Women (producers and their wives) we interviewed were particularly enthusiastic about the possibilities that the replacement of firewood by PPO based stoves would open up for them (see II.10). However a good prototype stove that can run efficiently on Jatropha PPO is not yet available. Moreover, some experts 17 believe that for the specific local circumstances of Cabo Delgado and given the abundance of firewood in the area alternatives cheaper than firewood would be hard to find, and therefore more efficient woodstoves might be most viable. Health problems might be avoided by locating woodstoves outside the huts. On the other hand, while stoves inside have increased the risk of respiratory diseases, they have been instrumental in warding off mosquitoes and thus malaria so there might be a trade off in the health benefits of placing stoves outside. Bio pesticide: There were some plans for testing for the use of Jatropha PPO as biopesticide, but they did not materialise. This is another potentially viable use for PPO that does not meet quality standards for use in engines. There is a considerable international market for bio pesticide and the prices are often more favourable than for PPO as fuel. Local staff in Bilibiza mentioned that a preliminary experiment by UEM on a mixture (Jatropha soap, normal soap, Tefrosia leaves and tobacco leaves) yielded good results but was not conclusive and the results were not made available. However the available reports only mention trials with Tefrosia, tobacco and Neem leaves or extracts. II.8.5 Local Application of By products / Organic Waste Recycling Shells: By products of processing are not being used currently. Farmers were not aware of the potential uses of waste products of Jatropha as fertiliser. They were throwing away the Jatropha shells as garbage, probably also because the shelling 34 For a comparison of costs, see activities/07 rapid economic assessment 1 29

31 takes place around the homesteads, far from the fields and it is cumbersome to transport them to the fields. While even just scattering the husks in the fields could contribute to improving soil fertility, training on making compost and fuel briquettes (the ash can still be used as fertiliser) out of the shells and husks would enable the replenishment of soil nutrients that are taken away by Jatropha cultivation. This is particularly important for maximising longer term returns from land use in the current context of poor farmers and impoverished soils. Seedcake and Biogas: Farmers do not have access to seedcake for fertiliser since hardly any oil is being extracted at the BBC. There had been a plan to invite experts from Diligent for training on biogas generation from organic waste, but there has been no progress, and no experiments have taken place yet. The eventual plan is to have one digester running on Jatropha seedcake (at the BBC) and one on kitchen waste (near the EPF school kitchens). If seedcake is used for biogas production, the remaining effluent should be returned to farmers for use as fertiliser. It is an effective and concentrated fertiliser, easier to transport. II.9 NATIONAL POLICY Main Question: What is the role played by national policy; is it conducive or obstructing? The Ministry of Energy recently issued a booklet establishing the policy and strategy of biofuels in Mozambique 3. The two main reasons cited for drafting this document were: (1) Promotion and use of agro energy resources for energy security and sustainable socioeconomic development, with an impact on reducing GHG emissions, and (2) the need to respond to the unpredictability and volatility of fuel prices on the international market, as well as to reduce the country s dependence on the import of fossil fuels. The objectives of government actions in the biofuel sector that were cited included engendering the sustainable production of biofuel from local energy resources to supplement imported fuel, reduce dependency and high oil import bills, diversify the energy resource matrix, promote rural development through biofuel investment and assistance to small scale producers, and so on. In this connection a preliminary study was done on the country s potential in terms of land, water and climate for producing bio fuels. A number of crops were compared in this context: Jatropha, coconut, sunflower, soya and peanut for biodiesel and sugarcane, sorghum, cassava, and maize for bio ethanol. Jatropha was deemed the preferred option for biodiesel, followed by coconut. Thus, while no concrete actions have yet been taken (like blending mandates or tax breaks) the government is clearly interested in activating the Jatropha sector. It remains to be seen, however, to what extent the interests of small scale production for local development will be prioritised over large scale national needs and export plans. II.10 OTHER KEY FACTORS FOR GENERAL EVALUATION 30

32 II.10.1 Diversification Though the project has significant potential in terms of diversification and branching of the Jatropha production chain, not much has been achieved yet. On the agricultural side, the simultaneous support given to food crop cultivation was a plus for diversifying livelihoods. However, the overlap of the Jatropha harvest cycle with that of other crops led to competition for labour and was not optimal for crop diversification with Jatropha as an element. II.10.2 Gender Land Ownership: As mentioned in section II.7.1 there are no land titles. While ownership is governed by customary law, the recognised owner is almost always the man (head of the household). Married women do not own land. In case of the death of the husband however, the wife usually inherits. Inheritance disputes between children are rare because land has always been plentiful. Some say local tradition (see II.7.1) may result in daughters being more likely to inherit because the sons have already cleared and occupied their own land elsewhere. The (unmarried) daughters stay back and might therefore inherit their parents land. However this most likely becomes the husband s land when they marry; usually the chief decides these matters. In case of divorce, sometimes (if they have kids) women can inherit some land. If there are no kids, the final verdict can depend on the grounds for divorce. Women of Bilibiza FC (This FC was added in 2010, so is a follow up of the project) not clear whether these preferences are recognised. Among the women of the Bilibiza FC (which consists of 46 women and 4 men) that we interviewed, no married women own any land. The unmarried women (there are 8 in this FC) do own land; they cleared the land themselves and said they were determined to keep it as their own even if they marry. But such decisions are made by the village chief, and it is Division of Labour: Agricultural labour (planting, maintenance, weeding, harvesting) is carried out together by men and women, but women have additional tasks and are putting in more hours per day on the whole. Additional tasks include looking after children and other household tasks, as well as the collection of firewood for stoves and light. The women also have to carry their babies on their backs while working on the fields. The collection of firewood seems to be a task exclusively carried out by women, and one on which they spend 4 to 8 hours per week. It is strenuous work (difficult on the hands, heavy to carry) and there are threats from wild elephants and other animals. All the women were very interested in the possibility of stoves that can run on Jatropha PPO instead of firewood. When told they d need a lot more Jatropha seed to produce enough oil for stoves and lamps, they said would like to expand production 31