Introduction

Timor-Leste (Fig. 1) is a small country with a population 0.94 million (DNE 2004), predominantly (85%) engaged in agriculture, covering c. 15,000 km2 and administratively divided into thirteen districts (Fig. 2). The terrain comprises narrow coastal plain and dissected uplands. The climate is tropical dominated by summer rainfall varying from annual means of 600 mm in the North-West to 3,000 mm in the highlands with an extended dry season particularly in the North of the country (Barnett et al. 2007). The diet is dominated by the cereals maize and rice and by the tubers of sweet potato and cassava. From 2005 to 2008 the average annual production (t) of maize was 83,400 t, of rice 59,527 t, of sweet potato 21,739 t and cassava 48,407 t (FAO 2009). In 2002 81% of households grew maize, 68% cassava and 44% sweet potato (Tilman de sa Benevides 2003). A typical farming operation in Timor-Leste is unable to derive a high level of personal income from 1 to 2 hectares of land. In fact, during 2008 more than half of the nation’s rural population lived below the poverty line of US$0.88 per day (World Bank 2008). Most of these families suffer from food insecurity (WFP 2006) producing insufficient cereal staples of maize or rice to last a full 12 months. Seven out of ten families surveyed during 2007 (SoL 2008) went without maize for four or more months. All were forced to practise food rationing from one to 6 months of the year; many gathered wild food regularly, and the worst affected families consumed the seed they had set aside for planting in the next season. Food insecurity is partly the result of low levels of food production—a consequence of poor land and the generally low levels of technology, combined with high crop losses, both pre- and post-harvest.

Fig. 1
figure 1

Location of East Timor (Timor-Leste) in SE Asia (map from Timornet)

Fig. 2
figure 2

Districts of East Timor (Timor-Leste) (Map from Wikipedia)

Timor-Leste is emerging from conflict and there is tremendous potential for the country to not just rehabilitate old systems but to design and implement new ones that are market-driven and provide holistic rural development. Several approaches may be taken but first there is a need to establish a high degree of food security in rural and metropolitan areas. Longley et al. (2006) suggest that rural livelihoods can be protected and improved by the following measures:

  • Ensure that vulnerable farmers have access to agricultural inputs and services (seed, irrigation, pest control, animal health, micro-finance, extension), either through direct distribution or interventions designed to stimulate choice such as vouchers

  • Increase agricultural production through access to appropriate technology options

  • Increase rural incomes through the promotion of agricultural product and labour markets

  • Establish the capacity, structures and institutions necessary for the sustainable delivery of inputs and services

  • Promote reforms necessary to address the structural causes of vulnerability.

All these interventions rely heavily on direct Government involvement, at least initially. Government direction in formulating a market-oriented rural economy is particularly relevant in Timor-Leste where the Government purchases a large amount of grain, particularly rice, for use as food subsidies to supplement national production shortfalls. The Ministry of Agriculture and Fisheries (MAF) also manages agricultural extension programs. It is essential for farmers to access appropriate technology options to improve food security on the farm and to eventually generate surpluses for sale.

Improved crop cultivars for food security in Timor-Leste

Ensuring availability and continuing supply of improved seed of food crops has proved successful in catalysing the process of improving food security in post conflict situations in other parts of the world (Sperling et al. 2008; Erskine and Nesbitt 2009). After the Khmer Rouge regime in Cambodia, for example, when food production was dramatically reduced, improved rice seed was imported from the International Rice Research Institute (IRRI) and other research programs for testing in the nation’s various rice growing ecosystems (Nesbitt 1997). Rice cultivars lost during the Khmer Rouge regime were also repatriated from the international rice genebank at IRRI. The number of cultivars was depleted when seed was consumed by farmers during food shortages, after seed was mixed when communal seed banks were established, or as a result of the enforced migration of farmers to unfamiliar rice ecosystems (CIAP 1994). A thirteen year research/extension program, conducted between 1988 and 2001 resulted in rice yields increasing from 1.45 t/ha to 2.1 t/ha. Varieties sourced from IRRI, Cambodia, Thailand and Nigeria were released and grain production increased by more than 50% as it became more economically viable for extra area to be planted to the crop (Young et al. 2001). The extra production was valued at US$1.3 billion at 2001 rice prices. A similar increase of 45% in grain production of food crops in Timor-Leste would substantially reduce food insecurity problems and provide surpluses for export.

The small size of Timor-Leste, its recent birth as a nation and associated conflict history, together with the lack of earlier sustained and systematic research on key staple crops make it a unique crucible to test the effect of a major post-conflict initiative of agriculture research and development on national food security.

The Research Program

The MAF in Timor-Leste is addressing food security through a focus on problems of poor germplasm and seed quality in a targeted research, seed production and distribution, and capacity building program (within a new Research Department in MAF). These activities commenced soon after the vote for independence in 1999 under the Seeds of Life (SoL)—Timor-Leste program (da Costa et al. 2003). The key focus of research was maize, sweet potato, rice, cassava and peanuts. Cultivars of these crops from similar agro-ecological environments in other countries were identified by research centres within the Consultative Group for International Agricultural Research (CGIAR). These included IRRI which offered assistance with the evaluation of rice cultivars, the International Maize and Wheat Centre (CIMMYT) with maize, the International Centre for Tropical Agriculture (CIAT) with cassava and beans, the International Potato Centre (CIP) with sweet potato and the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) with peanuts. Several Non-Governmental Organisations (NGOs) including World Vision International (WVI) and Catholic Relief Services (CRS) offered assistance with establishing and managing trials. The program was funded by the Australian Centre for International Agricultural Research (ACIAR). Cultivars that were higher yielding than local cultivars were identified in replicated trials conducted on research stations to 2006. Meanwhile the emphasis in SoL shifted to an extensive system of the participatory evaluation of promising lines for release with growers under their conditions on-farm. In the 2007–2008 season, for example, on farm trials included 243 maize, 190 sweet potato, 87 cassava, 190 peanut and 77 rice trials. The percentage distribution by District was 16% in Alieu, 31% in Baucau, 16% in Manufahi, 25% in Liquica, 5% in Manatutu and 7% in Ainairo. A greater number and distribution of trials is underway in the 2008–2009 season (SoL 2009c).

Sufficient data, varying across cultivars from 175 to 668 on-farm trials, were gathered on four of the key food crops for the Varietal Release Committee of MAF to release seven new cultivars in March, 2007 (Table 1):

  • Maize—Suwan 5 and Sele

  • Peanut—Utamua

  • Rice—Nakroma

  • Sweet Potato—Hohrae 1, Hohrae 2 and Hohrae 3

Table 1 Yield advantage of MAF/Seeds of Life cultivars over traditional cultivars

After reviewing the results of trials conducted during 2007–2008 and a summary of results over the entire program period, two cassava cultivars will be recommended for release later in 2009. It may also be possible to release another rice cultivar at the end of 2009 and new sweet potato cultivars in 2010.

Yield advantages of the released cultivars over those traditionally grown are highly significant (Table 1) and farmers prefer their size (sweet potatoes, maize, peanuts), and eating qualities (sweet potatoes, maize, rice). Farmers’ opinions of the released sweet potato cultivars are presented in Table 2 and the preferences for Nakroma rice are presented in Table 3 expressing what the farmers look for in new cultivars. Higher yield is not always the highest priority for farmers. Household surveys conducted in more than 1500 farm households over 4 years recorded and analysed livelihood information to examine these preferences (SoL 2007; SoL 2008; SoL 2009c). Extra surveys specifically targeted adoption constraints by gender (SoL 2009a) and various socio-economic groups (SoL 2009b). The SoL team was aware of these consumption trends and is keen to measure the level of adoption and the impact of the new cultivars on farm households.

Table 2 Farmers’ opinions (% of 115 respondents) of new sweet potato cultivars
Table 3 Farmers’ preferences for Nakroma rice

Seed production

The distribution and spread of new crop cultivars relies on the farmers receiving high quality seed that they can depend on to possess the traits they require. A cost-effective and sustainable seed multiplication and dissemination system should be established to meet the farmers’ needs. There are a number of methodologies to meet this requirement ranging from Government controlled systems to village-based seed enterprises (Bishaw and van Gastel 2009). The latter involves stimulating community-based seed businesses as local rural enterprises which in turn contributes to both market and rural development as well as food security. At present, the MAF in Timor-Leste takes responsibility for seed production and dissemination. Breeder and foundation seed are produced under researcher control both on Government research stations and, for foundation seed, on farmers’ fields. Small amounts of seed are then given to farmers to multiply for their own use. Another methodology may be adopted in later years as the demand for seed of recently released cultivars increases.

Adoption of newly-released cultivars in Timor-Leste

A year after the cultivars were released, an early adoption study was undertaken to seek preliminary evaluation of the releases. The number of farmers growing at least one of the improved cultivars of their own accord in the following season was very high with their adoption by 396 of the 544 interviewees (73%). There was no difference in adoption levels between men- and women-headed households (73 and 72% respectively). High levels of adoption were recorded for one or more the new rice, peanut and maize cultivars (81, 74 and 81% respectively), whereas sweet potatoes had a lower adoption rate (35–69%) possibly due to runner production problems in the dry season (Table 4). Anecdotally, the new cultivars are viewed as adding extra diversity to farmers’ fields, not diminishing it. The adoption of new cultivars does not result in the immediate genetic erosion of traditional cultivars as the new varieties are just added to the mix of 3–5 local varieties of each species grown every year. Further research will clarify the biodiversity implications of the new cultivars over time. In the meantime, MAF/SoL has mounted a collection of locally grown varieties for posterity and for use in the national evaluation program.

Table 4 Adoption of all tested cultivars across all Districts in 2006/2007

Impact of adoption of new cultivars: household level

Releasing improved cultivars that are higher yielding has the potential of not only improving food security by increasing grain production for on-farm consumption, but also providing the opportunity to sell the surplus for cash. Some farmers who were able to take the opportunity of sowing the new cultivars to increase their rice, sweet potato, maize and peanut yields sold part of their crop at the market. With their income they were able to purchase other types of food, and pay for educational expenses and clothing. Examples of the uses of cash income from rice grain sales are presented in Table 5. Although the amounts involved are small, there is an increasing number of farmers realising this advantage and the overall impact is expanding. SoL personnel are currently in the process of measuring the area occupied by new cultivars in order to estimate the economic impacts they have made to date.

Table 5 Use of cash earned from selling Nakroma rice

Mrs Freitas is one adopter who had taken the opportunity of growing Hohrae sweet potato cultivars and selling part of her produce to buy household goods. She is quoted as saying:

My name is Juvita Da Costa Freitas from Ostico village, Ostico hamlet, Vemasse Sub-district. Sweet potato that I planted in 2008 was sold and provided me with US$ 58.00.Footnote 1 I sold it in the market at US$7.50/sack. After selling, myself and my husband decided to buy a set of plastic chairs that can be considered as souvenir for my hard work in planting sweet potatoes. Sometimes, I also sell sweet potato from home and it is purchased by school teachers and students. I also share the cuttings with the neighbours who wish to plant, and some of the yield is also shared with my brothers. Life experience that we had with sweet potato is a story that can be remembered by our children when I pass away.

Potential impact of new cultivars at the national level

The potential impact of the new cultivars at the national level was estimated ex ante. Data on national production and harvested area are available for maize, peanut, rice and sweet potato from FAO (FAO 2009) (Table 6). The yield data of the new cultivars are from cultivation by farmers under their own conditions in six Districts and hence realistic for use in a national estimate. Conservatism was brought into the national estimate by using the average yield advantage of all new cultivars of a crop rather than that of the best cultivar. For potential national impact estimation, two scenarios were used. In the Linear Scenario an average adoption rate across the new cultivars of a crop was used and extended nationally with farmers adopting new cultivars on 75% of their land. In the Conservative Scenario an arbitrary lower adoption rate across farmers was assumed at 50% for maize, peanut and rice, again estimating that adopting farmers only grow the new cultivar on 75% of their land. For both Scenarios the mean sweet potato adoption rate of 53.7% was assumed because this lower rate in comparison to the cereals was partly due to the low availability during the dry season of sweet potato runners and this will be addressed in a MAF runner multiplication campaign. The estimates of increased production from the new cultivars are high in all the crops. In maize the adoption of new cultivars nationally is estimated to produce an extra 23,511 t annually representing a boost of 28% over current total national production (Table 6). In the conservative estimate the extra national production was estimated at 15,168 t or an extra 18% more national production. For rice the figures are lower but still substantial: the adoption of the new rice cultivar nationally is estimated to produce an extra 8,317 t annually representing an additional 14% over current national production. In the conservative estimate the extra percentage production was estimated at 9% rice grain. In sweet potato the new clones have dramatically improved yields compared to the local material but the national cropped area is small; the new cultivars will result in an extra national production of 30% compared to the current total national production. In peanut the new cultivar is anticipated to increase national production by 16% in the linear model and 12% in the conservative model of adoption. The total increase in food production in these crops would contribute considerably to the current 60,000 t of rice import equivalent currently required for food sufficiency in Timor-Leste (Heringa 2009).

Table 6 Average area harvested and production for maize, peanut, rice and sweet potato from 2005–2008 (FAOSTAT) with the respective yield advantage and adoption averaged over cultivars. Also anticipated extra production from the new cultivars in t/annum and % extra production compared to the current national total production with two scenarios: Linear and Conservative (assuming only 50% adoption for maize, peanut and rice)

Equity

It is important that national agricultural research programs serve all farmers equitably. The MAF/SoL research program has taken specific measures to ensure that this is the case in Timor-Leste. There has been no research, for example, on hybrids which would require the annual purchase of seed, thereby depleting short cash reserves in the subsistence farmer community. All released cultivars are equally productive when sown on similar soils by wealthy or poor farmers. None of the released cultivars place a burden on farmers to purchase extra inputs or require extra labour. All data collected on adoption is gender-disaggregated to determine whether male and female headed households react differently to each cultivar. Although there are some differences between adoption rates by men and women headed households, none has so far been significantly different. In fact many of the farmers who have successfully sold product off-farm are women.

Agronomy

Improvements to crop agronomy associated with the release of new cultivars may improve yields even further. Research is currently being conducted on plant spacing, weed control through the use of cover crops and soil improvement with legumes. There is also scope for investigating productivity increases with the use of fertilizers and pesticides. Possibilities for increased mechanization of crop production such as use of tractors, water pumps, threshers and milling equipment are also being considered.

Storage

Storage pests and microorganisms cause considerable losses in stored grain, tubers and roots in Timor-Leste. Such losses are estimated to be as high as 30% across the country. They are mainly due to damage from weevils, rats, mice and through fungal degradation. Fungal degradation is a problem in stored sweet potato and cassava while weevils are considered to cause large losses in maize. In studies completed by SoL (SoL 2007, 2008), weevil damage in maize was found to be as high as 63% in some cultivars when stored in the sheath for 9 months. These studies showed that there was significant variation between cultivars. Timor-Leste farmers are aware of this and use weevil tolerance as a key selection criterion when choosing new cultivars. Grain damage due to weevils can be almost eliminated by enclosing grain in airtight storage receptacles such as silos, plastic bags or oil drums (Guteres and Williams 2006). Grain plus the insects in the grain stored in air tight containers respires sufficiently to consume most of the contained oxygen thereby killing the insects. A program of providing 200 litre drums to Timorese farmers for their grain storage would, along with the higher yields of the new cultivars, largely eliminate food insecurity in Timor-Leste.

Diversified systems for sustainability

Rural households need to diversify their sources of food and income to overcome their vulnerability to the vagaries of weather and other unexpected disasters. Crop diversification is encouraged both in the crop species grown and by planting a number of cultivars of each crop. A wider range of legumes grown individually and intercropped with traditional sources of carbohydrate will also improve the protein status of rural diets. Small animal production may also support household protein diets and income. Other alternative sources of incomes should be encouraged such as the cultivation of vegetables and tree crops (coffee and fruit) or a household member finding off-farm work.

Capacity building

Human capacity improvement has long lasting effects on facilitating rural development. Training of research, extension and farming communities contributes to the formulation, transfer and adoption of food security improving technologies. The training may take the form of participatory research, field days, short courses, long-term training or on-the-job training. Farmers are more likely to adopt a technology if they understand the reasons behind its effectiveness and have the opportunity of observing it in action. By testing the technology themselves, they can best evaluate the benefits and costs of the new technologies. Research and extension personnel should be fully conversant with the science to possess the background knowledge for the technology transfer process.

Discussion

Food security exists when all people, at all times, have physical and economic access to sufficient safe and nutritious food to meet their dietary needs and food preferences for a healthy and active life (FAO 1996). In Timor-Leste widespread dry season food shortages are indicative of transitory food insecurity (Pinstrup-Andersen 2009). Successful rural development is dependent on initially achieving food security for the majority of Timor-Leste farmers. Initiatives to reach this goal are already in process with the development of appropriate technology options to increase agricultural production. Higher yielding crop cultivars of sweet potato, maize, groundnut and rice have been released, which are anticipated to improve productivity nationally. Early indications are of rapid adoption. Even at conservative estimates, the SoL program will contribute significantly towards resolving food deficits in Timor.

The use of the new cultivars increases grain production for on-farm consumption, and importantly also provides the opportunity of selling surplus for cash—some growers have sold at market for the first time. Notwithstanding, a major associated campaign to disseminate these cultivars more widely needs to be mounted.

Looking ahead the spectrum of improved cultivars needs to be increased with a larger range of crops and cultivars to improve diversity both nutritionally and across ecosystems. The development of further appropriate technology options will also expand this diversity. Maintaining a range of options of equitable and sustainable technologies that are easily extended to the farming community is of paramount importance. These technologies should include crop production, food storage and food processing. A balanced human capacity development program will facilitate this process.

Establishing an enduring agricultural research/development program in any country is a major long-term process, especially so in nations emerging from conflict. The early successes (and eventual sustainability) of the Seeds of Life program are recognized as being due to the fact that the program is firmly imbedded within the MAF—a Government body which has human capacity for both technology development and technology transfer. In addition to Government personnel, NGOs and farmers participate in the technology identification process and consider themselves to be part of the research program. All members of the evaluation team are included in program sponsored training courses and by the time the varieties are ready for release, both the technical personnel and farmers are fully aware of their attributes and potential. Good teamwork is recognized as a strength of the program. Maintaining good teamwork requires considerable effort from the management and support through a strong training program. Finally, in order to have maximum impact, the research program was simple in design and technically robust. Clearly capacity building within a national institution combined with a research focus on a few key technologies, relevant to a large audience using simple experimental designs are essential for success in developing countries emerging from conflict.