1 Introduction

1.1 Importance of the Banana Crop

Bananas and plantains (Musa spp.), hereafter called banana, are an economically important food and income security crop (FAO 2017). The crop is a principal source of carbohydrates for millions of people worldwide and fetches a large revenue share in domestic and international markets (Siddhesh and Thumballi 2017). It is currently cultivated in over 130 countries, on over 11 million hectares with a global production of about 155 million tons (FAO 2019). About one-third of the global banana production comes from Africa of which more than 50% is produced in the Great Lakes Region (GLR) including Burundi, Rwanda, the Democratic Republic of the Congo (DR Congo), Uganda, Kenya, and Tanzania (FAO 2019). Bananas of four use types are grown in the region, including cooking, brewing, dessert, and roasting bananas. Choice of a specific banana type varies by location, farmers’ preferences, market demand, and climate. Cooking bananas dominate in most areas, followed by dessert, brewing, and roasting types (Lusty and Smale 2003). The GLR is the highest banana-consuming region of the world with annual per capita consumption almost 15 times the world’s average and triple that of Africa (FAO 2019) (Fig. 10.1).

Fig. 10.1
figure 1

Banana consumption trend (2014–2018)

Bananas are also processed into puree, used for making banana drinks (Adeniji et al. 2010). Ripe bananas with a sweet taste and fine flavor and texture are also made into jam (Aurore et al. 2009). Leaves are used for thatching, for weaving baskets and mats, as a food wrapper for marketing and cooking, coverings over food, tablecloths, and plates, while extracted fibers are used as raw material for making specialty papers (Muraleedharan and Perumal 2010). The banana plant also holds medicinal potential and offers feed for animals. Starch extracted from the fruit has industrial uses, while the banana field confers high social status on the owners (Lusty and Smale 2003) who are predominantly men (Ajambo et al. 2018). The crop also controls soil erosion, sequesters carbon, and recycles nutrients (Lufafa et al. 2003; Ocimati et al. 2018; Kamusingize et al. 2017). Despite the great contribution of bananas to the GLR, average farm productivity has been declining since 2014 as opposed to African and global trends over the same period (FAO 2019) (Fig. 10.2).

Fig. 10.2
figure 2

Banana productivity trends (2014–2018)

This progressive decline in banana productivity in the GLR is attributed to sub-optimal management, declining soil fertility, increased moisture stress, postharvest losses, gender inequalities in access to resources, and biotic constraints. The burden of pests and diseases has especially been on the rise in the past three decades (Jones 2000; Tushemereirwe et al. 2004; Nyombi et al. 2013; Tinzaara et al. 2018). In the late 1990s, the fungal diseases black leaf streak or black sigatoka and Fusarium wilt had been considered the most important banana diseases in the region (Blomme et al. 2017a). However, the introduction and spread of a bacterial disease called Xanthomonas wilt, with significant impacts on banana yields and productivity, has become a force to reckon with in the past two decades.

1.2 Xanthomonas Wilt of Banana and Enset Emerges as a Serious Problem

Xanthomonas wilt (XW) of banana and enset, caused by the bacterium Xanthomonas vasicola pv. musacearum (Xvm), is currently one of the biggest threats to banana production in the GLR (Tripathi et al. 2009; Nkuba et al. 2015). First reported in Ethiopia (Castellani 1939; Yirgou and Bradbury 1968, 1974), XW was later observed in Uganda and the Eastern DR Congo in 2001 (Tushemereirwe et al. 2004; Ndungo et al. 2006). Since then, the disease has spread to the entire GLR, compromising production and food and income security (Reeder et al. 2007; Ndungo et al., 2008; FAO 2012). Xvm is mainly spread by insect vectors, contaminated cutting tools, and infected planting materials (Eden-Green, 2004; Tinzaara et al. 2006; Shimwela et al. 2016a, b, 2017). XW also spreads through long-distance trade (bunches, leaves used as wrapping material); other vectors such as birds, bats, and browsing animals are also common (Smith et al. 2008; Buregyeya et al. 2008; Blomme et al. 2014; Kikulwe and Asindu 2020).

Symptoms during the vegetative stage of a plant are yellowing and wilting of leaves, while floral infections start with wilting of the male buds and rachis, followed by premature ripening and rotting of the fruits, and finally the death of the plant (Ocimati et al. 2013, 2015; Nakato et al. 2014; Blomme et al. 2017a; Kikulwe and Asindu 2020). XW causes acute infections that can lead to the complete loss of a banana garden. It infects all edible Musa cultivars in the GLR and causes up to 100% yield loss, severely compromising livelihoods and food security for banana farming households (Kalyebara et al. 2006; Abele and Pillay 2007). Farmers respond to XW by abandoning bananas in favor of other crops (Desire et al. 2016; Ocimati et al. 2018, 2020). There are currently no chemicals, biocontrol agents, or resistant varieties available to control XW (Tripathi et al. 2009). XW management has relied on cultural controls that have evolved or been improved through iterations of co-innovation with farmers over the past two decades. For instance, it was the farmers who first pointed out that SDSR was effective (Andrade-Piedra et al., 2016). Some of these cultural techniques, their evolution, and the approaches adopted for their wider scaling are discussed in this chapter.

1.3 Technology Development for Xanthomonas Wilt Management

Promoting complete disease mat uprooting was only partially successful. During the initial years of the XW epidemic in the GLR, management recommendations were based on practices used in Asia for controlling Moko and blood disease (caused by the bacteria Ralstonia solanacearum and Ralstonia syzygii subsp. celebesensis, respectively) that have similar spread mechanisms (Blomme et al. 2014). These practices included (1) rogueing entire mats or fields when a single plant was infected within a mat, followed by disposal of rogued plants (by burying or burning), referred to here as the complete uprooting of diseased mats (CMU), (2) sterilizing garden tools between farm operations using fire or a 3.5% sodium hypochlorite solution, (3) timely removal of male buds using a forked stick to prevent transmission of bacteria by insects, and (4) the use of clean planting materials (Karamura et al. 2006; Blomme et al. 2014, 2017b, 2019).

These control strategies were enforced through massive awareness creation campaigns and bylaws forcing the use of the practices. However, XW control failed and the disease continued to spread to new areas. It was later realized that smallholders greatly resisted this impractical, costly strategy (Tushemereirwe et al. 2006; Blomme et al. 2014; Kikulwe et al. 2018). Uprooting and destroying whole banana mats (i.e., the mother plant and attached lateral shoots) was time-consuming, labor-intensive, and expensive, especially in fields with many diseased mats. This made it impractical, especially for elderly and female farmers who were not up to this backbreaking task or could not afford to hire labor. Besides CMU, debudding was also labor-intensive and unattractive for women who reported they got tired and their chests hurt after debudding. Women also voiced concerns of being stung by bees and having male buds fall on their heads (FGD with female farmers in central Uganda). Farmers did not see any benefits from using the practices, and the disease continued to spread across farms over the years (Mwangi and Nakato 2009).

1.3.1 Learning Leads to Single Diseased Stem Removal

The drawbacks of CMU stimulated further research to gain insights into the epidemiology of the disease and varietal responses in order to develop farmer-friendly management options (Blomme et al. 2014). These studies found that Xvm bacteria do not colonize all lateral shoots in a physically interconnected mat (Ocimati et al. 2013, 2015). Healthy-looking plants were still observed and banana bunches were still harvested even in heavily infected fields. These observations formed the basis of a new XW control package that comprised the cutting of only the diseased banana plants in a mat, a practice referred to as the single diseased stem removal (SDSR) technique (Blomme et al. 2017b).

The SDSR package comprises (1) regularly (at least weekly at onset) cutting, at soil level, of all symptomatic stems; (2) removing the apical meristems of cut plants in the vegetative stage to prevent regrowth; (3) adding soil onto the cut surface to minimize vector contact with bacterial ooze; (4) disinfecting the cutting tools with fire after cutting all visibly diseased stems; and (5) disposing of cut stems at the edge of the garden or in a compost heap (leaving the removed stems intact, as additional cuts would enhance the oozing out of bacteria). Farm tool use for de-leafing and de-suckering excess asymptomatic plants is avoided or minimized during the first 3–6 months of SDSR application while using a forked wooden stick to remove male buds after the formation of the last hand of the bunch (Blomme et al. 2017b). Where necessary, e.g., to let in more light for intercrops, excess leaves are only bent at the petiole level and aligned along the pseudostem (Blomme et al. 2017b). The regular (weekly) use of SDSR in controlled trials and on farm reduced and maintained XW incidence to below 1% within 10 months (Blomme et al. 2017b, 2019).

Single diseased stem removal is based on the idea that the continued removal of only the diseased plants in a mat will reduce the amount of Xvm inoculum, reduce the risk of Xvm spread to other plants in a mat, and ultimately bring down disease incidence to an acceptable level. The method also demands less time and labor than removing a complete mat, so it is more acceptable to farmers (Blomme et al. 2021). The total time for SDSR application on a single plant has been observed to be 88% less, averaging 4.3 minutes compared with 36.5 minutes for CMU (Blomme et al. 2021). Consequently, the cost of SDSR was 96% lower than that for CMU (Blomme et al. 2021). Moreover, farmers who use CMU lose several rounds of harvests from a mat, whereas with SDSR they continue harvesting healthy bunches as the field gradually recovers from the disease. Compared to CMU, the labor-saving SDSR increased the potential of women to use management practices in Burundi (Iradukunda et al. 2019). In Uganda, female-headed households were more likely to adopt SDSR, with an average of 15% non-adopters compared to 28% low adopters and 24% full adopters (Gotor et al. 2020).

1.3.2 Outcome of SDSR Adoption

SDSR was observed to reduce and maintain XW incidence on farm to below 1% within 10 months, resulting in positive and significant impacts on banana production and sales (Blomme et al. 2017b, 2019). An on-farm study in Rwanda showed that SDSR reduced disease as effectively as CMU, but offered a 96% recovery in control costs (Blomme et al. 2021). Success on farm has also been shown to vary with the level of adoption (Ocimati et al. 2019; Kikulwe et al. 2019). For instance, maximum benefits have been shown to occur when farmers adopt the full SDSR package, with the value of banana production increasing by US$462/ha/year (Kikulwe et al. 2019).

1.3.3 Potential Hindrances to the Success of SDSR

Despite the huge ray of hope for XW management through SDSR, latent infections have been shown to occur in infected lateral shoots which are attached to an infected mother plant, and it can take up to 24 months for symptoms to resurface in some mats, even though most latently infected plants continue to grow vigorously and produce edible bunches (Ocimati et al. 2013, 2015). Thus, follow-up over several years is recommended when practicing SDSR, as symptoms may still appear later.

Tool sterilization, a core component of SDSR and CMU, was infrequently used due to the expense and low availability of household bleach (3.5% sodium hypochlorite solution) in remote rural areas and because tools were damaged by repeated heating on fire (Blomme et al. 2014, 2019; Kikulwe et al. 2019). A later study showed that washing tools with cold water and soap or detergent (more readily available and cheaper) and immersing tools in boiling water for a minute are as effective as bleach (Ocimati et al. 2021). Using soap or detergent and boiling water will increase the options for tool sterilization and encourage more farmers to sterilize their tools.

Control of XW with clean planting materials (such as tissue culture plantlets) was also poorly adopted because of its high cost. There is little private-sector investment in banana seed production, because the crop is perennial and farmers rely on naturally regenerated planting materials (suckers) that are readily available in villages. However, informally sourced suckers are associated with a higher risk of disease spread. And farmers do not have the technical means to verify whether healthy-looking planting materials are XW-free or not (Jogo et al. 2013; Rutikanga et al. 2013; Kikulwe and Asindu 2020), especially given latent infections and incubation periods as long as 24 months (Ocimati et al. 2013, 2015). So, there is still a high risk of transmitting the disease through infected planting materials, even when farmers apply the other control practices effectively. The realization that XW causes less damage to varieties with persistent male bracts and flowers has led to the identification of varieties like KK (Musa ABB genome) as alternatives to the farmer preferred but highly susceptible Pisang Awak (Musa ABB genome) variety, which is called Kayinja in Uganda.

XW control may be uneven if farmers and traders fail to fully implement cultural practices. Women farmers often face constraints such as limited access to information, resources, and limited decision-making power (Kikulwe et al., 2018; Ajambo et al. 2020). The need to understand gender relations in terms of norms and who does what in banana production is crucial for informed targeting of SDSR. Men and women may be responsible for different production activities and fields; for instance, in Burundi, men do most of the production work for the commercial fields, while the women manage plantains grown for home consumption. Men are also more likely to attend trainings on bananas (Iradukunda et al. 2019). SDSR scaling approaches ought to integrate these dynamics and ensure access to information by those who do the actual work.

2 General Conceptual Framework Adopted for Scaling

For successful scaling to occur, stakeholders must understand the development problem at hand before exploring the appropriate innovations that can address the problem (IDIA 2017). The next steps are to foster dynamic relationships among multiple actors who enable scaling, facilitate learning interactions, and invest in local capacity and leadership to support sustainability (Hartmann and Linn 2008; Wigboldus et al. 2016; IDIA 2017). Scaling XW management in the GLR used this framework highlighting the disease as the baseline scenario (problem); its incidence was assessed and full impact on banana production was estimated (Fig. 10.3). The base scenario (XW) then informs on the need to innovate and generate technologies to combat XW considering the broader socioeconomic needs and benefits. Developed technologies and innovations are then piloted among target groups, also known as the diagnostic phase of scaling innovation (commonly done through trials, demonstration plots, farmer field schools, exchange visits, and lead farmers).

Fig. 10.3
figure 3

Conceptual framework for XW scaling in East and Central Africa

After testing the innovations, extra effort is allocated to creating an enabling environment for wider scaling, by identifying external and internal factors deemed influential to scaling the technologies. For instance, institutions are identified and examined for their capacity, resource availability, possible roles in scaling, and other elements necessary for collaborative scaling. Existing technologies and scaling channels are also critically examined for their readiness to be used in the fight against XW. This is possible through scaling readiness assessments in which scores are allocated to different technologies and scaling channels with the final choice based on overall scores for each component (Sartas et al. 2020).

Once the enabling environment is laid out and rules are defined (policies, legislative frameworks, etc.), strategic plans to take innovations to scale are drawn up, and the implementations are rolled out. The plans are drafted by a team comprising the developers of the innovation, farmer representatives, scaling partners, and scaling champions. Once the innovations are implemented, a monitoring and evaluation (M&E) system is set in place to document outcomes and impacts. This is also called the navigation stage (Sartas et al. 2020). The M&E is facilitated by the scaling champions and includes the primary scaling partners, farmer representatives, and the secondary scaling partners who also double as extension agents. Based on lessons learned from continuous M&E, systems are set in place to ensure that the scaling strategy is sustained. Depending on performance, innovations are improved or replaced with others that perform better. The overall process is generally overseen by the M&E framework, which draws lessons at every stage to inform the next steps.

2.1 XW Scaling Interventions in GLR

In the last two decades since the emergence of XW, several management interventions have been designed and scaled through collaborations of national and international institutions including National Agricultural Research Organization (NARO) and the Ministry of Agriculture, Animal Industry and Fisheries (MAAIF) in Uganda, Institut des Sciences Agronomiques du Burundi (ISABU) and Bureau Provinciale de L’Environnement, de l’Agriculture et de l’Élevage (BPEAE) in Burundi, Institut National pour l’Étude et la Recherche Agronomique (INERA) and Inspection Provinciale de l’Agriculture, Pêche et Élevage (IPAPEL) in DR Congo, Rwanda Agriculture Board (RAB), universities, Bioversity International, the International Institute of Tropical Agriculture (IITA), Catholic Relief Services (CRS), Réseau Burundi 2000Plus (RBU), and various universities and community-based farmer organizations (CBFOs). Some of these interventions have always been crosscutting, while others are country specific (Table 10.1). However, the level of scaling of these technologies differed in each country. To bridge this gap and make the XW management innovations accessible, several interventions were implemented over the past decade, mainly by NARIs and CGIAR centers. The scaling approaches evolved over time, as later efforts tried to address the limitations identified in the first ones.

Table 10.1 XW disease management approaches used in SDSR by country

The eradication and containment strategy of XW spread was deployed in the early years of the epidemic (2001–2007) by NARIs and their partners in Uganda, Rwanda, and DR Congo where localized outbreaks would be identified; then teams would be dispatched and paid to cut and bury the diseased mats (Mwangi and Nakato 2009). This costly approach was later replaced by the community sensitization for action strategy, practiced mainly in Uganda, involving printing and distribution of information leaflets, factsheets, brochures, and posters about XW. It was assumed that this information would trigger action by farmers, but it was later observed that although over 85% of farmers were aware, fewer than 30% were practicing (Bagamba et al. 2006; Tushemereirwe et al. 2006). As a countermeasure, the participatory development communication (PDC) approach was developed, tested, and promoted by NARO. This approach mobilized community stakeholders to explore solutions to XW, which they constituted into an action plan that detailed what needed to be done, when, where, and how to do it, who would take what responsibility and mechanisms for M&E. The PDC led to a drop in XW prevalence by 70% (Kubiriba and Tushemereirwe 2014). Unfortunately, this approach was limited by its dependence on an external PDC resource person and thus was unsustainable.

The community action intervention was later introduced by NARO in Uganda. The community action intervention was an improvement on the PDC that integrated mechanisms for enforcing the actions generated through the PDC. Its implementation led to the development and enforcement of bylaws at community level that led to a drop in XW prevalence by 68% and a banana yield recovery of 22% (Kubiriba et al. 2012).

The farmer field school (FFS) was introduced to reinforce adoption of XW strategies. The FFS mobilized community members for participatory discovery, decision-making, problem-solving, and stimulating local innovation while using the field as their classroom under the guidance of a technical facilitator (Ochola et al. 2015). The implementation of FFS in Uganda in 2006–2008 was accompanied by a drop in XW prevalence of 43% (Kubiriba et al. 2012; Kubiriba and Tushemereirwe 2014). A shortcoming of the FFS was its tendency to encourage the one-way flow of information from facilitators to farmers.

To address the limitations associated with the earlier approaches, Learning and Experimentation Approaches For Farmers (LEAFF) was developed by Bioversity International in collaboration with NARO in Uganda, with support from the McKnight Foundation. LEAFF aimed to foster sustainable and inclusive community-based management of XW (Tinzaara et al. 2019). LEAFF is similar to FFS: both are group-based, adult learning approaches that teach farmers how to experiment and solve problems independently. LEAFF, however, is more inclusive and involves farmers in the development of control measures through farmer action research networks which support experimentation and learning, value chain strengthening, enterprise development, and knowledge sharing. The community-level networks act like knowledge-sharing channels through which adoption challenges are explored.

LEAFF farmers were connected through mobile phones and the Internet to increase interactions within and between the groups, which are often lacking in FFS. This increased farmer-to-farmer interactions to enhance the sharing of experiences and skills in the quest to increase the effectiveness of the control measures. LEAFF exposed farmers to experimentation skills, including constraint identification, prioritization, hypothesis setting and testing, data collection and analysis, participatory M&E and reporting with emphasis on XW management, while tapping into the farmers’ own experiences. LEAFF was found to be a powerful approach against XW; when applied in the study areas of Bushenyi and Kiboga districts, it contributed to a reduction in disease prevalence by 27% and then to its eradication in 82% of the farms vs 64% where the approach was not applied (Tinzaara et al. 2019). LEAFF worked directly with 220 farmers, but through trainings, meetings, and radio talk shows, the program reached an estimated 22,000 farmers.

2.2 Scaling Readiness Across Three Countries in GLR

Supported by the RTB scaling project, scaling readiness (see Chap. 3) was implemented in four countries in GLR: Uganda, Rwanda, DR Congo, and Burundi. The 2-year, RTB-funded scaling readiness project (January 2018 to December 2019) aimed to come up with a strategy to promote the implementation of previously identified XW management practices. The project kicked off with a series of workshops, including one in Kampala, Uganda, in April 2018, where scaling champions and primary scaling partners from the four project countries met to conduct scaling readiness assessment exercises. During this participatory effort, three teams from Burundi, Uganda, and DR Congo devised country-specific “rich pictures” and “visions of success.” The need for gender-responsive scaling using action research and social learning approaches was highlighted during the workshop (Rietveld 2021).

The rich picture from the April 2018 workshop created a shared vision of how different institutions would work together to see that SDSR went to scale in each country. Figure 10.4 depicts the rich picture for Uganda showing how NARO would conduct research and disseminate the results through multiple channels to men and women farmers. The channels included local government institutions and secondary scaling partners who would help to train and pass on the information to other banana farmers through their extension staffs. Information on SDSR was also expected to reach the farmer through radio and TV shows, schools, churches, and local NGOs that promoted banana production. It was thought that religious and community leaders could be equipped with the SDSR messages and empowered to teach other farmers and that the traders in banana markets could share SDSR messages with farmers as they brought their fruit to sell.

Fig. 10.4
figure 4

A picture drawn by community members showing actions to scale SDSR in Uganda

The vision of success (also created during the March workshop) depicted the elements that needed to mesh together to enable SDSR interventions to go to full scale. It was drawn on two joined flipcharts, pointing out the institutions to be engaged and the activities they were to carry out to take SDSR to scale. The vision map also showed the situation before the SDSR interventions and the desired situation that would result from them. During this exercise, several public- and private-sector actors were identified to take part in SDSR scaling. There were calls for training and building the capacity of partners to design and implement gender actions to improve the impact of scaling (Fig. 10.5). The vision of success was later used to design the strategic plan for implementing XW management in each country. An assessment was also done of the institutions that were identified as scaling partners; only those with a shared vision to scale XW management practices and with strong community ties and skilled staff were picked to help implement the practices. The scaling champions from the CGIAR centers also used the vision map to develop their scaling action plans and select scaling channels.

Fig. 10.5
figure 5

Shared vision for SDSR scaling in Uganda

After the workshop in Kampala, teams from DR Congo, Uganda, and Burundi devised country-specific scaling strategies. The three core innovations received above-average scores in Uganda possibly because these techniques had been refined over time and promoted in prior projects. However, these projects were not on a large scale, with full packages like it was planned for the SDSR scaling project. As such, they were deemed ready for scaling. Basing on the scoring of complementary components for Uganda (Table 10.2), the strategy focused on developing radio messages and factsheets with gender-sensitive illustrations of the XW management practices and on mobilizing agents to scale out the innovations to banana farmers. During the workshop in Uganda, tool sterilization was ranked the lowest of the core components, because fire damages metal tools. Farmers did not know in advance how long the tools should be kept in the fire, but this was later addressed by scaling agents who promoted sodium hypochlorite (bleach) for sterilizing knives and machetes.

Table 10.2 Scaling readiness score for SDSR innovations (Uganda, May 2018)

In a similar exercise held in DR Congo, tool sterilization with fire also received a low readiness score, for the same reasons, resulting in the recommendation to use cutting brigades to inspect and cut infected banana plants together with the community members instead of relying only on individual farmers. It was hoped that the cutting brigades would ensure regular monitoring and ensure the sustainability of the XW management activities.

Scaling readiness assessments were repeated twice for each country at an interval of 6 months under the supervision of the scaling project leader (principal investigator) in collaboration with the scaling champions of the respective countries, the primary scaling partners, farmer representatives, and secondary scaling agents (extension agents). During the reviews, core components were retained because their scores did not drop. But those complementary components with declining scores were replaced with others that earned low scores during the first review but gained by the second assessment. Bottlenecks were identified during implementation and alternatives were proposed. For instance, in Uganda, following the second phase of scaling readiness scoring, short videos were added, using drama centered on XW management. For the overall scaling readiness score for Uganda at the first readiness assessment, see Table 10.2.

In Burundi, the scaling readiness assessment exercise was used to refine and validate (1) the SDSR core components discussed during the RTB Scaling Fund kickoff workshop in March 2018 and (2) the additional complementary component shared during the scaling readiness meeting held in Uganda during May 2018. Both core and complementary components of SDSR were scored high and ready to go to scale. Although video and radio messages scored highly, participants in the scaling readiness assessment realized that these approaches would be too expensive given the lack of community-based communication infrastructure in Burundi. Therefore, media were abandoned in favor of a comprehensive factsheet (Fig. 10.6). To increase the innovation readiness of the factsheet, it was validated for comprehension and efficacy in the field (Table 10.3).

Fig. 10.6
figure 6

Factsheet used in Burundi to scale SDSR

Table 10.3 Scaling readiness assessment score for SDSR innovations (Burundi, May 2018)

2.3 Enabling Policy and Institutional Factors for Success of XW Management Scaling

The institutional and policy environment was critical for the success of scaling SDSR innovations in all four countries. At the start of the project, Rwanda had no policy framework which would allow implementing the proposed innovations, so the 2-year project was used for trials in farmers’ fields comparing the effectiveness, labor cost, and time demands of CMU and SDSR to influence policy decisions and enable the innovations to sail through to the implementation stage (Blomme et al. 2021). There was also an initiative termed information and communication technologies for banana Xanthomonas wilt (ICT4BXW) to collect digital data from extension providers and farmers for developing an early warning system to curb the spread of XW and provide decision support on cost-effective XW control.

In Uganda, there was already a government buy-in by MAAIF for the SDSR package, since it replaced the mat uprooting technique originally promoted by NARO, its partners, and the government. More importantly, there was an established development strategy and investment plan (by MAAIF), policies (bylaws), and a partnership capable of taking XW innovations to scale. The success of the scaling project in Uganda was enhanced by integrating it into the Bill & Melinda Gates Foundation (BMGF)-funded Banana Agronomy project, which had a component for linking banana farmers to markets by selling collectively through the banana cooperative union. The linkage with markets was a motivation for adopting XW control, as farmers envisioned higher yields and a surplus they could sell.

In Burundi, farmers embraced the SDSR innovations as cost-effective and feasible. This played a pivotal role toward renewed trust in banana as a preferred food and cash crop. BPEAE (a government agricultural extension agency) staff quickly acknowledged and supported SDSR innovations for XW control. Even though SDSR had not been officially integrated into national policy, the many staff trained in the technique resulted in wider scaling. The SDSR team also received requests for advice from colleagues in other provinces facing XW. The head of the banana program of ISABU, the Burundi NARI, also supported the SDSR innovations. Much thought went into the scaling activities in the DR Congo, since design of interventions had to take into account the post-conflict situation, poor coordination of field activities, and the limited involvement of youth and women in managing banana plantations. The lack of an enabling environment was exacerbated by limited government commitment and the poor skills of government extension services, even though there was an existing legal framework for XW detection and control. This left most of the coordination to be done by international NGOs, with some assistance from INERA.

2.4 Case Studies of Scaling Outcomes

During the implementation of the SDSR package, case stories on XW management were collected from farmers and documented for wider scaling. Some of the inspiring stories documented from the field in Uganda are highlighted below.

2.4.1 Case Story 1: SDSR Package for XW Management

Federesi Kenema, an amiable, 38-year-old mother of five, says her banana farming fortunes have gotten better since she was trained by NARO. Before the training, Kenema was in the dark about how to deal with XW, which had plagued her garden. Her crops in Kakinga village in Rwimi sub-county in Bunyangabu district were always devastated by the wilt, and as a control measure, she would ill-advisedly uproot most of her bananas. “I would notice that many of my bananas had gotten ripe whilst still young. Their leaves would turn yellow and wilt and when I would cut the pseudostem, it would ooze a yellowish liquid. It seemed as if the wilt was spreading from one sick plant to another, but much as I racked my brain, I could not come up with any plausible clues on how it was happening.”

“The experts who were conducting the training came to my farm and saw firsthand how it had been damaged by XW. Then, I was trained in SDSR and other techniques like sterilizing cutting tools and bending dried leaves in heavily infected banana gardens,” Kenema explains. Before, a farmer friend had talked her into cutting all of her diseased banana plants, rationalizing that it was the only cure. Unknowingly, however, the practice damaged the soil structure and soil fertility and it was a lot of work to boot. After seeing how counterproductive the practice of complete mat removal had been, Kenema adopted SDSR.

“I wanted to control the spread of XW and I was told that I would have to regularly cut each infected stem at ground level, instead of uprooting the whole mat,” Kenema recounts. “Because bees frequently spread the wilt by moving back and forth between the male buds of banana plants, I was advised to remove male buds from my healthy banana plants with a forked stick, instead of using knives or pangas, which easily spread the infection.” Kenema adds, “The experts also advised me to sterilize farm cutting tools like knives and pangas with fire or bleach whenever I used them to prune. There was less disease after I started using those practices. My banana garden is flourishing. I now harvest bigger banana bunches, meaning I can now sell at more competitive prices to the ever increasing banana buyers from across the country and from Rwanda.”

In the past, at the peak of XW, Kenema could only harvest 7–12 small bunches every fortnight and each bunch went for the paltry price of 7000 Uganda shillings ($1.86). Now, thanks to the new practices she employs, she harvests 20–40 bunches every fortnight. She sells each bunch, depending on the size and season, between 10,000 and 15,000 Uganda shillings ($2.65 to $3.97). Kenema has been able to improve her family’s standards of living, with the better monthly profits, worth 650,000 Uganda shillings ($172).

2.4.2 Case Story 2: Rising from the Ashes of XW

Mzee Stanley Rwabuchye, a banana, bean, and coffee farmer in Ntaza II, Ishaka municipality, regained his normal life after an intervention of NARO, the district, and Bioversity International. He grew up in a banana-growing family, and in his own words, “bananas have always been our source of food and income. The disease started in 2008 when the banana leaves started yellowing, but we continued with our work. Alarm bells went off when the banana bunches started yellowing and getting rotten while immature. During this time, a team from Bioversity International, the district and NARO came to our village. They told us to cut and bury the infected plants. The more we buried the infected plants, the more the disease spread. The team gave us posters to guide in controlling the disease.”

He still has the banana bacterial wilt (XW) poster he was given in 2008 at the height of the outbreak (Fig. 10.7). The paper is brown with age, but the way he holds onto it says something about the importance of the information on it. “The team came back and advised us to stop burying the infected plants, but instead to chop them up and leave them on the ground to dry. We were also told to stop pruning the dry banana leaves and to use a forked stick to remove the male bud. By this time, my bananas were almost wiped out so I decided to cut down everything. After all the few remaining plants were also going to get infected. Shortly, NARO gave me new, disease-free suckers, which I planted and started my banana garden afresh. We were encouraged to disinfect our tools using bleach or fire.”

Fig. 10.7
figure 7

Poster describing the spread and control of XW

Rwabuchye says, “I want to tell you about the time when I cut down my banana plants. Life was difficult. We lost our source of food and income, and started depending on donations. After our bananas were wiped out, we started growing maize, cassava and sweetpotatoes. These helped with food, but not for money. It was a desperate time, I tell you.”

“Life is now back to normal. The plantations are thriving and trucks have come back to our village to buy bananas. Before the XW outbreak I used to harvest 80 bunches every month, but the price was as low as 2000 shillings ($0.53), even 500 per bunch and I’d earn 250,000 shillings per month ($66). Today, I harvest half of that, 40 bunches, but the price is better, 15,000 to 20,000 shillings per bunch, so I earn about 500,000 shillings per month.

“I thank Bioversity International and NARO for teaching us this technology and saving our livelihood. If they had not come to our rescue, we would have lost everything and become destitute. Matooke (plantain) is our food and livelihood; my parents used to make beer from bananas and pay school fees for us. I have also educated my children with income from bananas.”

2.5 Project Learning Outcomes

The rate of spread and the severity of bacterial wilt disease caught farmers and research scientists off guard. Nobody quite knew how to control this disease. Some said diseased banana fields should be destroyed. Others said that you only had to remove the male bud on an infected plant, while others insisted that sterilizing farm tools was the key. One thing was certain, the disease disrupted many countries’ food security and created social unrest among the farmers. Farmers and scientists alike were alarmed by its virulence, for it could destroy the entire bunch as well as up to 100% of the plants in a field within a matter of weeks. Among the most affected varieties was Kayinja (Pisang Awak), which is used to produce beer in Uganda. As XW spread, researchers realized that bees spread the disease from plant to plant as they collected nectar, pollen, and sap from the male floral bud. Farmers have always removed male buds from young bunches of cooking varieties to allow the bunches to fill up and grow bigger. Beer bananas would retain their buds as farmers believed that their removal would affect the quality of juice and beer (Bagamba et al. 2006). This explains why the disease was rife in areas where Kayinja beer bananas predominated, which hampered efforts of scaling XW control packages. It was also realized that the disease caused less damage to varieties with persistent male and female floral bracts and neuter flowers. Although farmers were taught and adopted practices for managing the disease, developing resistant varieties offered the only effective long-term solution in terms of both time and cost.

Farmers who grow bananas as a business are more responsive to adopting XW management technologies in Uganda. Through the scaling readiness scoring approach, communication materials for scaling XW management measures were harmonized for all stakeholders. At the end, implementing teams were in a much better position to transfer timely, coherent, and effective messages on XW management to banana farmers. To improve efficiency of scaling through radios, fan clubs were created. This led to closer interaction between the radio DJs and fans. The closer ties fostered a favorable environment for hands-on practice sessions on some of the SDSR practices that had previously been aired on radio talk shows. The continued DJ-fan interaction saw XW management interventions move quickly beyond the initial project sites, i.e., wider scaling of the SDSR interventions.

Working with partners already implementing agricultural projects enhanced the sustainability of the project as they did not request additional funding to add XW management components to their existing activities. This is one reason why the implementation of the scaling project in Uganda was integrated into the BMGF-funded Banana Agronomy project led by NARO in collaboration with IITA, Bioversity International, Makerere University, and CABI. Lastly, a quick and sustainable model for wider scaling was realized through the use of “Star performing model farmers.” These farmers were identified at the start of project implementation, trained on SDSR through demonstration plots in the banana farming communities. Each demo host was tasked to train 15 other banana farmers on a voluntary basis. This not only made it easier to communicate SDSR messages to other banana farmers but also quickly increased the number of beneficiaries of the scaling project.

We learned from experience that XW control requires commitment of all stakeholders along the banana value chain as there is a need for continued sensitization. In sites with land shortages, intercropping affected adoption of some of the XW technologies. Suspension of sterilized and unsterilized cutting tool use in heavily infected fields is not appropriate for farmers and for increased banana productivity; XW management should include all other recommended banana agronomic packages which do require the use of these cutting tools.

In Rwanda, approval by the government to scale the innovations to the farmers was still pending. This induced the scaling team to undertake a unique approach centered on doing trials to convince the government that SDSR controlled XW much better than CMU (see Blomme et al. 2021).

In DR Congo, the SDSR scaling project brought together many institutions to complement each other and to share the same message on XW management. These included farmers and their groups, community leaders, Bioversity International, IITA, INERA Mulungu (the National Agriculture Research body of DRC), IPAPEL (the DRC extension body), and local, national, and international NGOs. The NGOs that participated included Food for the Hungry International, World Vision, and Projet Agricole de Buhengere (PABU). Different communication methods were used, including music, field demonstrations, radio programs, training workshops, and posters. Despite this, women’s participation and access to extension services remain low. For instance, only 34% of women participated in the trainings on SDSR. Where bananas are intercropped, the banana mats are predominantly managed by male farmers, while women manage annual crops (Blomme et al. 2017b). Conflicts may arise if men try to prune bananas without taking into account the women’s intercropped bean plants, so it is important to integrate gender when designing XW scaling projects to ensure that both genders benefit equally.

Interactions with farmers revealed that they rarely sterilized tools with household bleach or fire in any of the scaling sites. This led to the exploration of cheaper, readily accessible, and easy-to-use alternatives. Washing metals in cold water with laundry soap or detergent or immersing tools in boiling water for a minute were found to be as effective as household bleach (Ocimati et al. 2021). Soap and detergents are readily available in shops even in remote communities. Washing tools to sterilize them is easy to use even where other techniques are not applicable, e.g., it may be hard to light a fire on rainy days or fire be a hazard, especially where farms are far from the homestead. Washing tools for a minute in soap and water was long enough to sterilize them (Ocimati et al. 2021). Repeated heating in fire will certainly weaken steel tools, and farmers are loathe to buying new ones any more frequently than is necessary.

Failure to take gender into account can hamper uptake of management recommendations. In Burundi, for instance, women were not motivated to use XW strategies because men sell the bananas and keep the money. In addition, access to information was unequal, with men participating in more trainings while women had limited access to radios and other sources of new ideas (Iradukunda et al. 2019). Ugandan women had limited access to extension, the main channel used to share information (Kikulwe et al. 2018). The labor requirements of the XW control strategies also influenced uptake (Gotor et al. 2020). Women were less likely than men to adopt backbreaking techniques like CMU. However, most of the approaches developed to scale out XW strategies integrated gender issues.

2.6 Impact of XW Scaling Interventions

In all SDSR scaling efforts, the focus has been on recovering banana productivity, improving food security, income, and well-being of banana farming households in the GLR. Quantification of such impacts saw various impact assessment studies, focus group discussions, key informant interviews, field visits, and observations in the project countries. SDSR has been extended to over 100,000 farmers in Burundi, DR Congo, and Uganda, with a recovery in banana production area  of about 50,000 ha and an estimated annual value of total production ranging from US$10–30 million or US$100 to 300 per farmer/annum was attributed to these efforts (Rietveld 2021). Additional and detailed country-specific impacts are documented as follows.

Burundi: In the province of Muyinga, the household survey conducted in November 2019 showed that the number of households having XW on-farm fell from 57% to 10% and the number of households with an XW incidence of 10% declined from medium (level 3) to low (level 2) on a four-point scale: 1 = very low (one or two plants), 2 = low (only a few plants), 3 = medium (many plants), and 4 = high (more than one-quarter of plants are symptomatic).

About 65% of banana-growing households have been reached by the scaling project and about 70% of those are already applying the SDSR innovations. During a field visit by senior project scientists in the later stages of the project, no XW diseased plants nor male buds beyond the expected cutting stage were seen in the banana fields in the province of Muyinga. Previously abandoned banana fields were being quickly restored, showing that farmers were regaining their trust in bananas as a preferred food and cash crop. The BPEAE has become a committed supporter of the SDSR innovations, even though this has not yet been officially integrated into Burundi’s national policy. With this warm enthusiasm and high number of staff trained on SDSR, it is expected that future posting of such staff to new locations will help scale out the innovations to other provinces. There are already informal requests for advice from other provinces facing XW, a move likely to spark wider scaling in Burundi.

DR Congo: In one of the earlier projects in which single diseased stem removal was deployed, the results showed that 1 year after starting field trials in the North and South Kivu provinces of DR Congo, the proportion of diseased plants had fallen to less than 0.5% (InfoMus@'s News and Analysis 2014). For the latest interventions by scaling project (2018/2019), DR Congo took strides to further contain the spread of the disease by cutting sick plants and removing male buds, even though the impact of these interventions has not yet been assessed. However, indications are that most (75%) banana farmers now apply SDSR and some (10%) use both SDSR and CMU.

Uganda: Bioversity International conducted a quantitative survey with 1224 banana farmers to assess impact of adopting SDSR. The study found that adoption is constrained by limited knowledge of SDSR and influenced by the importance of the banana crop to the household (Kikulwe et al. 2019). The study further showed that training women farmers increased the chances of households adopting all three core practices under the SDSR package. Adoption of the full package (all three SDSR core practices) was more profitable than using one or two of them. Adopters obtained considerable increased earnings from their banana production compared to non-adopter, as adopting XW control measures reduced the disease incidence and increased banana yields and sales. Extrapolations from this survey suggested that of an estimated 800,000 banana-growing households in Uganda, about 600,000 had adopted some of the BXW control strategies.

Focus group discussions (FGDs) were held in six districts in the three major banana-producing regions, including Central, Midwestern, and Southwestern Uganda. The focus group discussions examined the contribution of the SDSR interventions to household food security and banana yields. The results showed that farmers could identify XW diseased plants, and they understood its spread mechanisms and how to use SDSR. Both men and women, young and old, had fully embraced male bud removal. Men and women were routinely sterilizing their tools, some on a daily basis.

Banana farming communities had started growing much less of the Kayinja variety, the most susceptible to XW. Those who had initially left banana production (about 25–50% in central and western Uganda) or who had reduced the area of land devoted to bananas were quickly replanting their fields after learning the SDSR package. Food security had improved and households that had adopted SDSR were eating more meals per day. The number of banana-based meals increased, and this was more appreciated in the central region where banana is expected to be on the table for every meal.

2.7 Future Interventions

Despite the lack of currently known genetic resistance against Xvm in cultivated banana in East and Central Africa, resistance has been reported in nonedible species such as Musa balbisiana, while varieties of the Musa acuminata subsp. zebrina (AA) set were also identified as potentially useful sources of Xvm resistance (Ssekiwoko et al. 2006, 2015; Nakato et al. 2019; Tripathi et al. 2019). It would take many years to incorporate these genes into edible varieties through conventional breeding. The difficulties of conventional breeding with this sterile crop call for the need to innovate transgenic approaches for XW control.

Genetically modified (GM) XW resistant bananas have already been developed and were field-tested in Uganda (Namukwaya et al. 2012; Tripathi et al. 2017). However, policies allowing the release of these materials are lacking. The public is skeptical of GM and consumer resistance is a major constraint to GM crop production (Panzarini et al. 2015). Given consumer rejection of GM crops, clustered regularly interspaced short palindromic repeats (CRISPR) may offer an alternative. Because CRISPR edits genes out instead of introducing new ones, it has brought about rapid positive changes in the attitude of many producers toward genetically engineered crops (Waltz 2018). Banana breeding using CRISPR is still in the early stages, but shows promise (Tripathi et al. 2019, 2021). Future scaling approaches need to address farmer and consumer attitudes about crops bred with CRISPR.

3 Lessons Learned and Policy Implications

SDSR effectively manages XW. Farmers using it report reduced banana yield losses. In much of East Africa, SDSR has contributed to a tremendous improvement in food security and incomes.

Men and women farmers accept SDSR because it saves money while effectively controlling XW without all the work of CMU. A single stem is cut while healthy bunches are still harvested, making SDSR more farmer-friendly. Nevertheless, it is important to further reduce the labor requirements of the SDSR package. Action plans for SDSR training and information dissemination have to consider gender from the beginning in order to guarantee ideal participation for women and youth. Barriers to overcome include work burdens, unequal allocation of cash and other benefits from the banana crop, and conflicts arising as men trample women’s beans while pruning bananas.

Tool sterilization with fire is still being promoted for XW management, even though fire weakens steel tools. To make sterilization cheaper, without ruining farmers’ tools, knives and pangas can be washed in cold water with soap or detergents.

The Rwanda case suggests the importance of engaging policymakers at all stages of technology innovation to ensure their buy-in and subsequent scaling to end users. Platforms are also essential for bringing stakeholders together and scaling innovations.

Producing banana as a business is key to the continued use of XW management practices. A sustainable banana market may enhance scaling of XW recommendations. In the GLR countries where banana is still considered a man’s crop, allowing women to take decisions and make money from bananas will boost the uptake of the SDSR package. XW management practices must continue to be used even when disease incidence drops.

Scaling agricultural technologies such as the XW control strategies require a clear linkage with food systems, e.g., by integrating with a systems-based project. In Uganda, for instance, linking with the ongoing Banana Agronomy project enabled contextualization of the strategies and improved the results.