Adoption studies informing breeding market segments, Product Profiles and seed system interventions in Southeast Asia

by Jonathan Newby, Erik Delaquis, Luz Andrea Gomez, Adriana Bohorquez

Cassava production in Mainland Southeast Asia is a multi-billion-dollar sector supporting the livelihoods of millions of smallholder farmers, rural communities, and national trade balance sheets. Productivity is essential when competing in the global starch and carbohydrate market against alternative feedstock for a range of food and industrial applications. This is worsened by the introduction and spread of diseases, which continues to threaten the economic competitiveness of the sector. The Alliance cassava program has been leading a network of partners to tackle these threats – with the strategy requiring market economists and social scientists to work closely with breeders, agronomists, and seed system specialists to ensure that acceptable disease resistant varieties are accessible to a diversity of farmers as soon as possible.

Mainland Southeast Asia’s informal cassava seed system has been effective at driving the adoption of improved cassava varieties with both high fresh root yield (FRY) and starch content. Unfortunately, the same informal system efficiency also facilitated the rapid spread of cassava mosaic disease (CMD) across national borders following its introduction to the region. Following a lag in response and crossing, CMD-resistant varieties are now emerging from the breeding pipeline, this understanding is about to be harnessed to maximize the reach and speed of variety adoption.

Recent results from household surveys coupled with DNA fingerprinting of cassava varieties obtained from farmers’ fields are providing new insights into just how dynamic the seed system is, revealing how industry and farmers cope with the existential threat of crop disease to their livelihoods. This information is now being used to inform breeding strategies, refine seed product market segment definitions, adjust the essential traits within target product profiles, and inform seed system interventions through a range of public-private partnerships.

In 2023-2024, surveys were conducted throughout 8 provinces of Laos (300 samples collected); 9 provinces of Cambodia (706 samples collected); and 4 districts of Tay Ninh Province in Vietnam (221 samples collected).

Location of household surveys in Lao PDR, Cambodia, and Tay Ninh Province of Vietnam (red dots), as well as accessions donated from national partner gene banks in the Philippines and China (blue dots). 

DNA fingerprinting was also conducted on samples from the entire Philippines national gene bank collection (398 samples) and field collections in Guangxi Province of China (15 samples), adding more samples to the reference library. With the results of the DNA fingerprinting still undergoing analysis, there are already a number of interesting observations.

1. In Lao PDR, while there have been several small outbreaks of CMD (typically from informal introductions of planting material from outside the country by industry actors), lower adoption of recent releases from breeding programs in Vietnam and Thailand indicates that access to modern bred varieties is limited, with the country being largely reliant on domestic cassava seed (stems) during ongoing rapid expansion of the cassava industry.
2. There are some notable exceptions with pockets of unknown varieties appearing in Vientiane province, as well as variety 89 (a Thai clone never formally released in Thailand but popular due to its high FRY – despite other weaknesses). This shows that novel disease introductions through contaminated stems continue to pose a risk.
3. Large areas of older, highly CMD susceptible varieties (Rayong 11; Variety 89; Rayong72), demonstrate the continued vulnerability of Lao farmers and industry to CMD epidemic outbreaks.
4. Some sweet varieties were detected which match Latin American landraces. There is currently no CGIAR breeding pipeline serving this SPMS in Asia. CR63 is the reference name for a cluster of accessions in the genebank with multiple duplicates, including 16 accessions from Indonesia. This accession has shown some resistance to CMD in previous studies.

In Cambodia, on the other hand, the results clearly demonstrate the high connectivity of the Cambodian informal seed system to both the Vietnamese and Thai breeding programs.

1. Significant diversity was detected, including many varieties not currently in the CIAT DNA fingerprint library to enable variety matching, meaning they have never been collected.
2. There are more recent releases from both the Thai and Vietnamese breeding programs in farmers’ fields. There is a wide diversity of farmer names for the same varieties, and several varieties identified with the same name.
3. The variety “Purple-Shoot” has gained traction in several provinces. This variety has been seen in Cambodia for several years during ad hoc variety identification; but also appears in a SPIA study in the Gai Lai Province in the Central Highlands Region. This variety may have been developed in a Vietnamese breeding program.
4. KM140 has been detected over the Vietnamese border from Tay Ninh into eastern Cambodian provinces.
5. TMEB419, a CMD-resistant release, has seen some adoption in Kratie Province, with activities of a GIZ development project linked to the variety evaluation

Tay Ninh Province is the engine of varietal turnover in Vietnam. The province has over 50 large starch mills that not only process fresh roots from the surrounding provinces within Vietnam, but also process a large percentage of roots produced in Cambodia. This connectivity in the root trade also creates linkages in informal stem transport networks.

The adoption study in Tay Ninh shows several interesting points to be considered by breeding programs and seed system interventions.

1. The province was extremely efficient in replacing the popular but highly CMD-susceptible KM419. This variety was estimated to be present in 50% of farmers’ fields in Tay Ninh during the last adoption study in 2015, just prior to the introduction of CMD. This demonstrates how existing seed systems can replace highly susceptible varieties very quickly.
2. Industry and seed value chain actors sought out an alternative (non-resistant) variety from elsewhere in Vietnam, which covered large areas within a few years.
3. The resistant variety TMEB419 is present in all districts but in small areas, with farmers reporting its susceptibility to root rot and Cassava Witches Broom Disease (CWBD).
4. The Thai variety KU50 (known in Vietnam as KM94) has almost disappeared, despite being some tolerance to CMD and high starch yield. Farmers prefer more erect plant types, and modern improved varieties may provide a yield advantage if mildly afflicted by CMD.
5. The high adoption of the non-resistant variety KM140 again shows that farmers are prioritizing other traits. In this case, farmers often report the ability to harvest earlier (less than 10 months) while still achieving acceptable FRY and starch content.

Implications for breeding and seed system development

There is a Seed Product Market Segment (SPMS) for early maturing varieties, with farmers willing to trade-off higher potential yield for the ability to harvest earlier to escape flooding and obtain cash-flower earlier. This has previously been flagged as a “Future SPMS” in MIPPI. However, several questions remain:

1. How large is this market segment within Southeast Asia, and how do partners prioritize it?
2. Does this SPMS require its own pipeline, or is it served by the existing pipeline?
3. When should trials be harvested to align with most farmers’ expectations?

A staggered-harvest trial using the advanced clones is currently being conducted to determine whether any of the advanced CMD resistant clones could replace KM140 as an early-harvest, CMD resistant variety. Staggered-harvesting trials are difficult to conduct, particularly early in the pipeline with more clones. The development of high-throughput phenotyping methods to evaluate root formation and starch accumulation would greatly add to the development of this SPMS.

While the informal seed system has been effective at spreading improved varieties, large differences between provinces, countries, and farmers’ socioeconomic characteristics remain. Without a network of researcher-managed or farmer-managed trials, it is difficult to conclude the extent to which these differences are driven by environmental factors (soil, climate etc.) versus exposure and access to new varieties through various project and industry actors. Previous Alliance research in the product target areas has demonstrated that both geography and farmer characteristics affect demand and WTP for cassava stems, as well as a measurable effect of quality guarantees related to phytosanitary status.

Without a formal exchange of breeding products between countries, industry actors will continue to move varieties across national and continental borders with little heed to the risks. Strengthening networks for official germplasm exchange and evaluation will not only help the spread of the new CMD resistant varieties, but will help protect the region from introductions of untested materials potentially carrying new pathogens. High risk examples include Cassava Brown Streak Disease (Africa) or Cassava Frogskin Disease (LAC) that have not yet been detected in Asia. The Australian Center for International Agricultural Research (ACIAR) has recently added support to the Asia Cassava Breeders Network to strengthen the definition and prioritization of current and future market segments, identify and priories discovery activities, and enhance collaborative germplasm exchange and evaluation.