1 Introduction

Implementing agricultural innovations is key for coping strategies in the contexts of climate change and food security (Senyolo et al. 2018; Bommarco et al. 2018). The agricultural sector is still the central focus for rural development, especially in remote areas of Sub-Saharan Africa. Links to Small and Medium-size Enterprises (SMEs), among others, are often lacking, due to comparatively high costs for logistics, transport, and communication (Letiche 2010; Meyfroidt 2018; Stephens et al. 2018). Peri-urban and rural areas need specific and tailored livelihood strategies (Fraval et al. 2018). Thus, enabling environments for business models and alternative ecosystem services (Bommarco et al. 2018) are difficult to establish. Low economies of size and scale hinder the establishment of profitable economies (Tomich et al. 2018; Letiche 2010). Nevertheless, manifold implementation models for upgrading agricultural activities do exist and are continuously being tested and adapted in international research projects (Candel 2017). Specific implementation models disseminate innovations despite various structural problems of research and development in Sub-Saharan Africa (Lipton 1988).

Among these theoretical models, the main challenge remaining is how agricultural innovations can be disseminated efficiently and effectively through outscaling and upscaling, given varying site conditions and diverse target groups (Senyolo et al. 2018). While pro-poor approaches focus mostly on small-scale farmers, the question of the right setting for agricultural innovations persists. In less favorable areas, typically low-cost innovations are more suitable due to limited capacities (capital), while higher income farmers in favorable production areas might seek higher investments and more revenue through market integration (Tomich et al. 2018). Additionally, at local levels, some farmers are more innovative than others, seeking different agro-ecological transformation strategies (Tittonell 2014). These superior performing farmers are more likely to adopt new techniques, even if they are riskier (Steinke and van Etten 2018). Innovators and catalyzers are key to more efficient and reliable adoption of agricultural innovations (Steinke and van Etten 2018; Below et al. 2015; Uckert et al. 2015). The issue of innovation adoption is an important research topic bridging the gap between “maximum yield potential” and actual yields harvested in farmers’ fields (Foley et al. 2011). Such research should be long term and monitored over time in order to arrive at sustainable improvements. Agricultural innovations should be continually optimized in response to changing conditions (Tomich et al. 2018; Mutabazi et al. 2015, Senyolo et al. 2018, Below et al. 2015). Furthermore, an adequate incentive structure is a necessity for the long-term adoption of successful techniques (Nhantumbo et al. 2016).

These applied adoption theories are closely linked and indispensably coupled with out- and upscaling methods, which seek efficient and effective horizontal and vertical dissemination of innovations. While horizontal out-scaling is the dissemination from farmer-to-farmer or means-to-farmer-approaches (i.e. radio, field exhibitions, leaflets, farmer field schools, or extension), the vertical way uses efficiencies through upscaling via policy or education programs. These have the highest dissemination rates with the lowest inputs (i.e. regulations/laws, curricula, education programs). Hence, out- and upscaling activities are key for effecting innovative agriculture. Good examples include, first, bio-fortification through introducing new seeds with high vitamin content and, second, low-cost irrigation systems to overcome water scarcity (Laurie et al. 2015, 2018). Household or individual levels of access to food, dietary quality, and nutritional impacts within these systems are often underrepresented in research (Stephens et al. 2018). Hence, linking food security and nutrition-sensitive agriculture is an important research activity. One outcome should be the mitigation of stunting.

In the first part of this special section (Sieber et al. 2017) the large research project Trans-SEC was outlined as having the following methodological components: (1) the Trans-SEC conceptual framework (TCF); (2) the basic research principles for inter- and trans-disciplinary research; (3) the upgrading strategies (UPS) as agricultural innovations; as well as (4) the Trans-SEC research model (TRM). The overall Trans-SEC goal is to test innovations in order to improve the food security situation at the community level. In this regard the question remains whether the innovations can be disseminated effectively, such that small-scale farmers can cost-efficiently adopt these new practices.

These considerations asked for an adequate out- and upscaling dissemination strategy. Hence, each innovation had to be tailored to the specific site conditions of the implementing region. This challenge of developing an adequate method for local characteristics is a tightrope walk: on the one hand, there is the goal to (1) maximize generalizability and reproducibility of food security strategies, considering guiding principles for applying innovations and, on the other hand, each implementing method needs (2) to be specific enough to tailor agricultural innovations to site-specific conditions and target groups. In this regard, the Trans-SEC project part II reported here targets the conceptualization of the general Trans-SEC food and nutrition security model. It specifically analyses potentials for out- and upscaling of respective agricultural innovations with regard to:

  • Constraints preventing the efficient and effective dissemination of agricultural innovations (so called upgrading strategies (UPS)) – these constraints lead to mal-adoption;

  • Description of the Food and Nutrition Security Model (FNSM) for interventions to build development options for small-scale farmers; and

  • The UPS-specific out- and upscaling levels that are known to build dissemination strategies.

2 Implementation constraints, system complexity and scalability

In order to explain the general findings of Trans-SEC, we systemize system-immanent problems and, from there, derive solution-based strategies. In this regard the three steps from constraints (see Section 2.1) toward system-analysis using the model FNSM (see Section 2.2), up to levels for effective out- and upscaling (see Section 2.3) are illustrated in the following with examples of agricultural innovations.

2.1 Implementation constraints for agricultural innovations

Successful interventions are complex to implement and, thus, need to be considered in an agricultural system approach (Graef et al. 2014, 2015). The bottlenecks must be carefully discussed in order to overcome the related challenges. In this regard, Fig. 1 illustrates the constraints of agricultural innovations (UPS) along Food Value Chain components (FVC), which are generally not visible before implementation. In summary, these bottlenecks are either the high labour input for short implementation periods (ridging, weeding, see 1a – Fig. 1) or, if investments are needed (costs for oxen, ridging tool, cost sharing of fertilizer, see 1b – Fig. 1), the complex underlying social organisation, including business models and cooperative systems. Liquidity for financial input is often lacking and, therefore, needs business models among involved farmers. The formation of a business group for making economic investments is key for success (business groups, see 6, 7, 8 or 1b – Fig. 1). The motivations, incentive structure, and composition of the group, as well as power relations related to opinion holders, innovators, and excluded farmers must be carefully and simultaneously considered. In many cases, market-related technical hurdles, such as the availability of only large-sized fertilizer bags, cannot be easily overcome without political pressure: convincing market participants to provide smaller, more affordable, units of fertilizer will help poor farmers afford to use them (see 1b, Fig. 1). Moreover, the logistics and infrastructure of market systems often need high investments or low-cost decentralized solutions. Both need long-term planning and often are only realizable at the small-scale level. Soft skills, such as technical aptitude of innovators, and characteristics of actors, including high intrinsic motivation, endurance, and cleverness, are sometimes unpredictable or not easy to identify (see 7 – Fig. 1). Across all UPS, one important factor is ‘knowledge’ through education and training measures (see 8 – Fig. 1). All these constraints determine the ‘enabling environment’ of agricultural innovations for UPS.

Fig. 1
figure 1

The tested agricultural innovations as upgrading-strategies (UPS) along the Food Value Chain (FVC) and implementation constraints

Full size image

We found two major types of interventions, which can be defined as (a) coping strategies in enabling environments with high pressure on scarce favourable conditions such as water and good soil quality; and (b) linking to markets to upgrade subsistence farming systems by generating additional income from markets and by stabilizing livelihoods though division of labour. These two types of interventions are illustrated in Fig. 1 (see legend), which differ in (a) stabilizing farm systems without direct market linkage (coping strategy) shown in UPS numbers 1, 2, 4, 5, 8, and 9, while (b) the remaining ones, 3, 6, and 7, are for income generation seeking (market strategies).

All interventions described in Fig. 1 are not singular, but rather complex, interlinked, and system-relevant. This means that integrated approaches are needed not just for the simultaneous consideration of the implementation processes of innovations, but also the related constraints. These constraints limit the available implementation capacities of the stakeholders involved, such as farmers and extension officers (Sieber et al. 2015a, b). Due to this high complexity, careful priority-setting in financially scarce environments is needed. This maximization of impacts and simultaneous cost-minimization is one of the greatest challenges for implementing organizations (Akullo et al. 2018). Therefore, detailed knowledge of the system complexity and its specific challenges is key. Finding site-explicit tailored implementation models for research, such as the exemplified FNSM in Fig. 2, is a challenge.

Fig. 2
figure 2

Food and Nutrition Security Model (FNSM)

Full size image

2.2 Trans-SEC’s food and nutrition security model (FNSM)

The integrated and generic Trans-SEC - Food and Nutrition Security Model (FNSM) (see Fig. 2) was developed specifically to classify Trans-SEC’s agricultural innovations (Sieber et al. 2017).

Trans-SEC implemented a bundle of interventions that was selected and developed through participative involvements by stakeholders, mostly small-scale farmers. The assumption is that this tacit-knowledge within the community (social capital) is the best proxy for the actual implementation needs. In this case, the stakeholders themselves select the agricultural innovations to be implemented. The possible up-grading steps of farmers are achieved by following the pathway described in Fig. 2:

The FNSM in Fig. 2 classifies the innovations into (A) coping activities; (B) small scale business in agriculture; (C) seeking opportunities for income generation; and (D) down- and upstream business with/without agricultural integration (SMEs). From the short and mid-term perspective, no sustainable solutions are possible. Food aid is a short-term solution. From the mid-term perspective, implemented innovations can be tested and, if successful, implemented for long-term use.

Against this background, these are the following general pathways: (A) fostering coping strategies to stabilize poor livelihoods and food insecure situations; (B) starting off from a more stable situation and establishing a small-scale business with slightly increased inputs. Once the food and nutrition situation are stabilized or the situation is per se secure, further economic development can be achieved through (C) increasing inputs (land, labour, capital) to establish market integration and also connect FVC components across different sectors; or (D) generate income in other sectors such as tourism, energy, and agriculture-related services such as consulting and trading. Exclusive business outside the agricultural sector as alternative income source (e.g. credits, insurances, and infrastructure) is favourable if labour can be utilized more efficiently.

The dynamics within the FNSM are driven by the actors involved, who identify and commence the pathway from A to B and, eventually, C and D of how food and nutrition securing strategies are set up and applied. Generally, a consensus among decision makers and additional stakeholders is key for success; however, sometimes a single innovator, as a positive deviant, may achieve a similar or even better positive result. Free riders and strategic behaviour by opinion holders hinder positive impacts within the communal society. Thus, the question on the adequate implementation model of how agricultural innovations can be implemented in a sustainable setting arises. Here, Trans-SEC tested only one implementation model out of a wide range of possible variations of models (inter alia Graef et al. 2014, Tittonell et al. 2012, Schut et al. 2015, Notenbaert et al. 2017), each with different components and implementation procedures. A promising method is positive deviance (Steinke and van Etten 2018), which first contacts innovators to establish a small upscaling centre at community level. An adequate enabling environment for any intervention on agricultural innovations is also important. The major challenge is to overcome the constraints within the agricultural system that hinder stakeholders from properly implementing changes. Based on our research, reasons for this (see Fig. 1) are ranked in order of importance: (1) the opportunity costs of labour; (2) social organization of farmer implementation groups; (3) access to credit; (4) technological, process-oriented knowledge; and (5) cultural, traditional reasons and gender. The ranking changes depending on the individual type of agricultural innovation; therefore it is only generally representative across all tested innovations.

2.3 Levels for effective out- and upscaling

Out- and upscaling of successfully tested agricultural innovations need a tailored concept for dissemination. The characteristics of this concept depend on site conditions as well as the specifics of the target group. Twofold dissemination procedures are possible: (1) Up-scaling is most efficient since, through implementation of findings with policy programmes in other regions, target groups can be easily addressed, presuming that actual policies are effective. (2) Out-scaling, involving farmer-to-farmer outreach for knowledge transfer, either to other regions or simply to other target groups within the same region, is equally important as a complementary dissemination measure. Based on interviews within the Trans-SEC consortium, the following levels were identified for dissemination of agricultural innovations. They relate mostly to up-scaling activities, although (3) and (4) can also involve out- and upscaling activities simultaneously (see Fig. 3).

Fig. 3
figure 3

Themes and levels of out- and upscaling activities for effective dissemination

Full size image

Legislation: Generally, it is favourable that simple regulatory measures, which are easy to communicate and implement, are transferred through laws and regulations. Inter alia, energy-efficient cooking stoves, which are relatively simple to construct, are ideal to implement via national or regional regulations. Especially helpful are all easy to implement innovations that are simple in construction, design, or guiding principles (e.g. educational objectives).

Curriculum within public authorities: All findings that are “standardisable” as educational tools or components within curriculum are adequate; either through excursion or practical training components as well as in university / school teaching programs. This measure is as efficient as legislation, but targets the students who, subsequently, are the “experts” and, thus, promoters of the specifics.

National action plans: e.g. the National Adaptation Programme of Action (NAPA) targets specific thematic areas, defining adaptation measures and relationships with development goals. These fields of actions outline country-specific programs to tackle identified challenges with regard to climate change, food security, and related risks. New findings in innovation research, such as guiding principles, can be region-explicitly involved in the NAPAs.

District level-specific actions plans: these plans are more detailed and represented through local government agencies at the district level, which tailor programs to site-conditions. Specific technology and promoter programs can be applied, although they are generally the same as in (3).

The following levels, 5 and 6, are identified for the dissemination of agricultural innovations, mostly for out-scaling activities, since region-specific concepts on agricultural innovations can be disseminated via farmer-to-farmer approaches, inter alia. These can be defined as follows (see Fig. 3):

At the sub-district local level, such as the Wared level or, even lower, farmer-to-farmer, the approach comprises dissemination events, including exhibitions, community radio, or local associations that transfer knowledge through leaflets, guides, or workshops. These can also incorporate focus group techniques. Specific local government agencies (LGAs) may be involved and also connected to up-scaling activities in singular cases.

At the local community level, extension officers can tailor community-specific concepts to out-scale and also to up-scale targeted activities, thus bringing them to the community. Via LGAs, extension officers can be trained in collaboration with Small and Medium-sized Enterprises (SMEs). The extension officers can accompany specific implementation processes throughout the outscaling process and require SME’s inputs at the request of involved stakeholders.

The last is especially important if agricultural inputs, such as fertilizer, are lacking. Feedback loops involving primary and farmer schools may be applied if they exist at the local level (e.g. school garden concepts). Further, if successful concepts are proven (e.g. kitchen garden concepts), then teaching programs can be disseminated via local authorities from one village to another.

In summary, on the one hand, the exemplified out- and upscaling levels in Trans-SEC show exceptional diversity across all tested innovations for the main characteristics, with the amount of time needed varying drastically. While on the other hand, national changes in legislation or a representative innovation program in NAPAs may need several years of lobbying and tailored policy action strategies from multiple actors. Nevertheless, specific examples of out-scaled agricultural innovations from one target group to another might be relatively easy to guarantee success in a short time. These challenges of out- and upscaling activities for effective dissemination differ across the four illustrated components: (a) country regulations and related sub-scales; (b) policy types; (c) target groups; and (d) supporting organizational settings from governmental (e.g. farmer school), private (e.g. SMEs) and non-public (e.g. NGOs) levels (Fig. 3).

3 Conclusions

This paper presents (1) the constraints to disseminate efficiently and effectively agricultural innovations; (2) a food and nutrition security model (FNSM) to conduct interventions successfully; and (3) the specific out- and upscaling levels for agricultural innovations (upgrading strategies UPS) as challenges to efficiently develop more detailed dissemination strategies. Social organization within specific business groups and micro credit systems, as well as technical and cultural limitations, are key challenges across all innovations. The FNSM classifies the innovation types applied in Trans-SEC and the pathway of a potential farmer development is described. Classification of the agricultural innovations in FNSM enables better planning and implementation in a sustainable way. This can enhance long-term adoption, including continuous adaptation to changing site / frame conditions over time. Efficient dissemination of successful agricultural innovations through out- and upscaling can be formulated and tailored at institutional, legislation, education and farm levels, from the local through (inter-)national levels. Five levels were identified, along with descriptions of the related challenges across national levels and related sub-scales of regulation sub-systems, policy types, target groups, as well as supporting organizational settings.