Floralaan 54
6708 EK Wageningen
Email: consultants@bananahill.net
+31 (0)6 41667461
+31 (0)84 7304043

Trip Report

Mozambique, April 2009

Flemming Nielsen


This trip report discusses the main findings of my visit to Mozambique from 7 to 27 April 2009.

Thank you to the many people met, including:

Erik Schurmann Hansen (erik.schurmann@adpp-mozambique.org), Jacob Menyani Zulu (zulumenyani@yahoo.com), Henderson Maposa (hend_maposa@yahoo.co.uk), Bachir Afonso (bachirafon83@yahoo.com.br), Francis Atsimbon Vitung (FVitung@yahoo.fr), Marc Schut (marc.schut@wur.nl), Sandra, Luisa Alcantra Santos (luisa@zebra.uem.mz), Anna D. Lerner (anna.lerner@gtz.de), Per Hartmann (phartmann@map.gov.mz), Jose Carlos A. Monteiro (cessemonteiro@gmail.com), Elisabeth Johanna Klara Specht (especht@chimoio.mocambique.net), Calisto Bias (cbias@map.gov.mz), karen.adipsa@yahoo.com (karen.adipsa@yahoo.com)

Minimum Farm Gate Price

A main objective of this trip was to gain insights into the economics of Jatropha production f sml scale farmers.

Two farm gate prices that are particularly interesting for the project are:

Where nothing else is mentioned "farm gate prices" are implied.

In farming systems where farmers can shift their resources between various alternative cash crops these prices can be assessed relatively easy. However, in the Bilibiza project area it was found that there is very little in terms of cash economy.

Most farmers tell that sesame is the only crop they can sell. This year they get 35 MZN/kg (0.97 EUR/kg) but that is unusally high and not likely to be reached again soon. In a group interview with 20+ farmers as well as individual interviews we probed "what will happen if the price for sesame drops to 20 MZN/kg". The answer was that it would not make any difference. They would continue to produce and sell the same amount. We next halved the suggested price and received the same answer again. Farmers insisted that even at 1 MZN/kg they would continue to produce as they presently do because they have no alternative and they need income to buy some necessities they can't produce themselves.

We probed for the response to increased prices on sesame and here the reaction was that production would continue as usual. Food crops for own consumption is the first priority and farmers are not tempted to change this even if cash crop prices increases.

That this is not just empty talk was illustrated by an incidence that took place during my stay at Bilibiza. Two farmers brought sacks of Jatropha to the project and asked for a price of 75 MZN/kg. They were offered 2 MZN/kg which they accepted. The project staff who offered the 2 MZN insist that the farmers would have accepted 1MZN/kg if he had offered that. Notice that this is not even farm gate prices as the farmers transported the seeds themselves.

Based on the current evidence the minimum farm gate price that farmers will accept is 2 MZN/kg or lower and the price elasticity of supply is very low.

During the period that farmers have been paid 5 MZN/kg by the project they have expanded the area under Jatropha. However, the expansion is likely driven to a large extent by the extension effort of the project and the attention the project gets from the presidents wife, ministers and other dignitaries. Under these circumstances it is not possible to assess the minimum farm gate price required for continuous expansion of Jatropha production under more normal circumstances.

Provided the high priority given to food production for own consumption any significant expansion of Jatropha production is likely to come through increasing the number of farmers cultivating Jatropha instead of increased Jatropha production by the farmers already producing. This is positive seen from a food security perspective. However, it will keep transport costs high as long as oil is processed at one central location and thus have negative impact on the carbon and energy balance of the project.

Labour use for harvesting Jatropha

Three tests of harvesting rates were carried out; Two at Primero de Maio and one at Metambo.

  1. The first tests consisted of two farmers of about 25 and 45 years of age harvesting Jatropha for fifteen minutes from a two year old hedge with moderate fruit bearing. The fruits were decorticated on the go and a total of 1.5 kg wet seeds were ready in 15 minutes. It is equivalent to 3 kg/h per person. The farmers said that the speed of picking was slightly higher than what they usually do.

  2. The second test included the same two persons as the first test. They spent 14 minutes picking Jatropha from one hedge. Decortification was done afterwards and took 26 minutes. The total harvesting time of 40 minutes resulted in 2.46 kg wet seeds. It is equivalent to 1.8 kg/h per person. Without the decortication 5.3 kg/h per person were harvested.

  3. The third test included two women of around 25 years of age. They harvested Jatropha for 12 minutes and spent 46 minutes decorticating the 2 kg. About 70% of the fruits were black while 30% were yellow. The picking rate including decortication is 1 kg/h per person. Without decortication it is 5 kg/h per person.

The farmers said that they usually pick the fruits first and decortify when they are anyhow sitting together.

To summarise the findings:

These figures are at the low end of figures from the literature. A reason may be that the hedges are young and have relatively few fruits and that the poor pruning practise makes it difficult to reach fruits on long branches.

Considering that the major time spend on harvesting the seeds is for decortication we should look into the feasibility of simple mechanical decortifiers. It could make Jatropha more attractive/beneficial to the farmers. This is particularly important because of the time of the labour peaks as discussed later in this document.

Relative economic advantages to farmers

With the harvest tests it is possible to give some indication of how attractive Jatropha production is to farmers.

First of all the price of 5 MZN/kg that is currently paid to farmers is too high for oil extraction to be profitable. A price of 2.5 MZN/kg is considered realistic.

Using the figures from the harvest trials the income from eight hours of harvesting is 20 to 60 MZN/day.

Through interview with a number of farmers it was found that manual labour is paid 30 or 35 MZN/day depending on whether meals are included. Work agreements are often based on payment for a given task so a fast and strong person may earn 30 MZN for three hours work while others spend ten hours doing the same task. For calculations an eight hours working day is assumed.

If it is assumed that the only time farmers spend on Jatropha is for harvesting, it appears that Jatropha is on average a better income earner than employment.

To assume that only harvesting takes time is approximately true after the first year when the raising of seedlings and planting takes significant time. Apart from harvesting weeding and pruning is required for maintenance. Usually the Jatropha hedges are so dense after two years that they shade out most weeds so weeding is minimal. Pruning is quick and is done during the dry season when the labour demand is low so the opportunity cost is close to zero.

A comparison with the only cash crop in the area, namely sesame, would be interesting but no data are available. To collect useful data on sesame production data collection for one full season would be required. That is outside the scope and means of this project.

Another way to look at the economic benefits to the farmers is to assume the Jatropha oil is used locally and substitutes kerosene or diesel. Locally kerosene prices were found to vary from 22 MZN/l by the main tarmac road to 50 MZN/l at isolated areas.

For the calculations the following assumptions are made:

The value of one persons work will range from to 22 to 240 MZN per eight-hour working day. Again this compares favourably with salaries paid for manual labour.

Ideally the transport time for carrying the seeds to the press as well as the purchasing cost of the press should be included in the calculation. Data on transport time is not available and I am not aware of any data set they can be derived from.

(Jan, do you have data on cost of presses and expected lifespan?)

Economics of Jatropha for lamps

It was found that the sale of kerosene is decreasing because farmers are switching from kerosene lamps to LED lamps.

The lamps are home made from cheap parts made in China and consist of a bright blue LED (10 MZN) connected to 3 size A batteries (5 MZN a piece). The total cost of the lamp is 25 MZN and it provides light for two month. Because the LED works till it is physically damaged and because batteries that have become too weak for use in radios are reused for LED lamps the real price is significantly lower. Farmers will continue to add an increasing number of weak batteries to maintain the voltage.

The LED lamps gives less light than kerosene lamps and are most popular with young people. Their use is increasing fast.

If Jatropha oil lamps are introduced they will have to compete with the LED lamps. The last mentioned provides light for a maybe 0,1 MZN/h (assuming the long term cost is half of the price of a new lamp and lifetime of 120 hours). If one liter of Jatropha oil cost 20 MZN it has to give light for 200 hours (assuming zero cost of the lamp). With a kerosene price of 50 MZN/l one liter of Jatropha oil must give light for 500 hours.

These figures are indicative only and do not take into account that the amount of light varies between the different technologies.

In Negue farmers make candles by stuffing bamboo sticks with Jatropha kernels. This may be the cheapest way to make light from Jatropha.

Compatibility between local farming systems and Jatropha production

An ideal bio-fuel crop has labour peaks that do not coincide with the labour peaks for other crops. To test this assumption monthly data collection from farmers fields has been initiated. For various reasons this research is delayed and it was therefore decided to use a fast approach to get rough estimates. Labour calendars were made with two communities.

In Ngeue markers were used to draw on big sheet and in Primero do Meio Jatropha seeds were used to indicate work load by month.

Labour calendars were made separately for Jatropha and all other crops.

In Ngeue there are two harvest seasons for Jatropha;

Outside these main seasons smaller amounts of fruits appear.

The main labour peak on the farm coincides with the first Jatropha harvest. This is the time of weeding. The second harvest coincide with the land preparation which is less intensive than the time of weeding but can be difficult because under nourishment is common at the beginning of the rainy season.

At Primero do Meio there is only one peak harvest season for Jatropha which coincides with the first season at Negue. This difference was corroborated by the extensionists and project staff.

The main harvest of Jatropha is problematic because it severely limits the Jatropha production that a household can undertake because they will not reduce on labour for crop production.

Ripe Jatropha seeds can be left on the plant and harvested later. From various sources there are indications that this leads to slightly reduced oil content and the seeds may harden so they are less ideal to press. Do we have any data on the following questions or do we need to consider researching these issues to be able to give recommendations;

Manilla envelopes for collecting seeds from trials

It has not been possible to have special net bags produced for monthly seed collection from the trials so instead it was decided to test standard manilla paper envelopes. The main concern was that the seeds would mould because air passage through the envelope is limited. To test this an enveloped was filled with fresh seeds and left to air dry in a well ventilated partly shaded place. The moisture from the seeds soaked the envelope for a few days after which it dried. After ten days the envelope was opened and no indication of mould was found. One advantage of using sealed envelopes is that the risk of loosing or mixing seeds during weighing is reduced.

640 manilla envelopes with individual water proof labels has been left with Henderson. Each envelope has information on trial type, location, month and year. Jatropha plants in all trials with several treatments have been colour coded with strings and the envelope has the same colour code to reduce risk of error during seed collection.

The seeds in the sealed envelopes will be stored at the office and weighing will be done by Flemming or Jan during their visits.

Seed Sorting by Shaking

Large seeds should be selected for propagation because large seeds are generally associated with higher yield and higher oil content. Small differences in seed size are difficult to detect by eye so it was tested if the selection can be made by shaking seeds in a container and then skim the top seeds. Ten batches of 300 gr of seeds were shaken in a 1.5 liter container for 45 seconds. The container was prepared so seeds could easily be extracted from the top and bottom.

Table 1: Weight of 20 seeds (in grams) taken from the top and bottom of a container with 300 gr of seeds shaken for 45 seconds.























The seeds from the top had an average weight of 13,5gr and the bottom seeds averaged 12,9 gr. A two-tailed t-test gave p=14% so the differences are not significant.

It was difficult to move the container only vertically. The result was that the seeds in the container tended to describe a circular movement where the seeds would rise on one side of the container and sink on the other side. A device or technique that ensures vertical shaking only may work better.

Commercial machinery exist but the cost is only justifiable if sufficient amounts of seeds is processed. It may be economically viable to have seed processing equipment at BBC, particularly at the present time when the demand for and the price of quality seeds is high.

Trials in Cabo Delgado

More trials had been established since my last visit and during my visit we established a number of new trials and labelled the existing ones with coloured plastic strings.

Table 2. Status of trials, April 2009

Type of Trial



Provenance trial



Hedge improvement trial



Yield measurements in farmers' plots



Yield measurements in farmers' hedges



The low number of sample sites in farmers' plots is related to the fact that most farmers plant hedges and not plots.

The main challenge is to get the seed collection undertaken timely and with few errors. As mentioned above, the tools are in place, trees are marked etc. and the main challenge is therefore purely a management issue.

Trials in Manica

The trial at Mandonge, Sussundenga, Manica Province is in poor shape after having been left without weeding for two seasons. Most plants survive but have hardly grown since they were planted. All plants are marked by severe pest attacks from flea beetles and none of the plants have started bearing fruits.

Jose Monterio from IIAM who is in charge of the Forestry Research office at Sussundenga wants to terminate the trial. This was discussed with Teresa Alvy (IIAM Maputo) and Luisa Santos (UEM).

In favour of the trials is that they will show how different provenances perform under difficult conditions. However, considering that all plantings of Jatropha in the area have shown very poor performance it is unlikely that Jatropha production will ever become important at sites with conditions similar to Mandonge.

Due to the severe flea beetle attacks common in the area it provides a good location for the study of the biology of flea beetles that is planned in collaboration with UEM. Logistics would be easy due to the many researchers based in the area. However, other smaller plantings of Jatropha exist in the area and they can be used for such a study.

The decision about the future of the trial rests ultimately with IIAM. FACT has money for maintaining the trial but it is carried out on IIAM land and under its authority.

If the trial is abandoned it is recommended to move ten or more plants of each provenance to a location more suited for Jatropha, e.g. at IIAM Nampula. There they should be part of a national reference collection of Jatropha provenances which can be used to provide planting material for testing and multiplication by research organisations, private companies and farmers. Experience has shown that it is very difficult to obtain provenances internationally and particularly in a timely fashion. The seeds are only viable for about a year and we have therefore always run into the problem that when the last provenances for a trial were received the seeds from the first provenances were no longer viable. Even big companies like Sun Biofuel experience this problem.

If IIAM decides to abandon the trial and use the plants for a reference collection at another location FACT cannot automatically support that financially. However, it may be possible, so Monterio and Teresa please inform us about how you want to proceed.

Sun Bio-fuels

Sun Biofuels has started their big plantation outside Chimoio since my last visit. It is the largest Jatropha project in Manica Province so far. More than 900 ha have been planted using direct seeding in a 4x1.5 m spacing to facilitate the use of machinery from the fruit industry. Seeds were bought from Diligent, Tanzania. Fertiliser, pesticide and mechanical power is used so the approach and problems they face are different from Jatropha production by small scale farmers. Three seeds were planted together and later thinned. Only few plant stations have are empty.

Pest problems are minimal so far and the plants grow well. All plants appears to branch readily at ground level which is different from the experience we have from Cabo Delgado where branching is minimal unless insects damage the terminal bud.

96 accessions are being tested. If superior material is identified the current plants will be substituted over a number of years.

Research with UEM

Three studies were planned for last season in collaboration with UEM:

All studies should ideally have been undertaken during the second part of the rainy season. For various reasons this did not happen. Luisa Santos (UEM) has now included the planned studies in a proposal to GALP which may fund full scholarships whereas the FACT funding is limited to paying for specific research projects. GALP is currently establishing a large Jatropha plantation in Manica Province. Due to the uncertainty about the GALP funding it was agreed with Luisa Santos that FACT/UEM will continue to work together on the agreed topics.

Course on Jatropha Production, September 2009

Henderson and I started making a more detailed outline of the course with content specified per day as well as resources required. Since this has already been communicated it will not be repeated here.

Organise BBC

The set-up of oil production plant in Bilibiza (BBC) was discussed with project staff and ADIPSA which had some interesting experience. They suggest that we organise Farmer Clubs in a Forum which will negotiate collective contracts with the oil factory. The factory will run as a purely commercial entity to ensure efficiency. The Forum will prevent exploitation and that desperate farmers undercutting each other. To optimise transport efficiency Farmers Clubs should receive a higher price when they can fill a lorry.

ADIPSA will put me in contact with a person who is expert in these issues and who has recently relocated to Pemba.

ADIPSA recommends supporting the outline setup for several years or we run a high risk of it collapsing soon.

Pruning Time

The issue of correct pruning time remains unresolved. Two commercial Jatropha producers, namely D1 and Sun Biofuels argue that the rainy season is best because the wounds heal quickly.

Our recommendation has been the opposite, namely to prune during the dry season to avoid fungi entering the wound.

During this visit when we encountered farmers who had pruned we inquired about their timing. They had all pruned during the dry season and no adverse effect had been observed.

Previously rainy season pruning has been been observed to work well in many cases too. However, the few severe cases of die-back after pruning have all been related to pruning done during the rainy season. We therefore find it safest to continue to recommend dry season pruning only.

Flemming Nielsen April 2009