A systematic literature review identifies pertinent and reliable literature using a predetermined protocol to achieve a certain objective or answer a research question. These reviews enable the identification and exploration of the best evidence, contradictory findings, and gaps in the existing literature (Gupta et al., 2018). For a review to be “systematic” it should aim at answering a specific research question, follow a pre-defined protocol, and comprehensively review all the available information (Uman, 2011; Gupta et al., 2018). This systematic literature review aims to identify the factors influencing the adoption of CSA in Africa.

To investigate the research question raised by this study, this systematic review has been conducted in accordance with the reporting standards outlined in the Systematic Evidence Synthesis (ROSES) framework. The ROSES framework aims to improve the quality and transparency of reporting in the environmental field by providing guidelines and a review protocol (ROSES, 2023). The ROSES review protocol (see Supplementary Table 1) was followed when preparing for this paper, screening articles to include or exclude, and writing this systematic review.

The literature search was conducted using the Web of Science (WoS) database in November 2023. The search queries were: “climate smart agriculture AND adoption AND Africa,” “climate-smart agriculture AND adoption AND Africa,” and “climate-smart agriculture AND uptake AND Africa”. The search queries were applied on the title, topic, and abstract. The search was further refined by only including articles written in English, specifying the publication period (2019–2023), and focusing on peer-reviewed original articles. This recent time frame was chosen to ensure that the information presented in this review is current and reflective of the most recent developments in the field. These search queries generated 207 hits. After removing the duplicates 170 articles remained.

The WoS articles that appeared following the search query were exported into Excel where they were screened following different steps. The flow of this process is demonstrated in Figure 1. First, the title of each article was screened systematically. The title needed to refer to the (dis)adoption of CSA, when it was not immediately clear if the article could be relevant for this review inclusion was ensured if the titles featured terms such as perception, adoption, obstacles, challenges, engagement, lessons, uptake, barriers, constraints, opportunities, or preferences, or synonyms of these words. Moreover, since this article is not about upscaling CSA, the adoption of other sustainable practices related to CSA, or only about one sustainable practice (e.g., conservation agriculture), these articles were removed from the review. In addition, review articles or studies solely using qualitative methods were also removed. Following these criteria resulted in the exclusion of 122 articles.

The second round of screening of the remaining 48 articles was done based on the abstract of each article. After screening all abstracts, 18 articles were excluded. The main reason articles were excluded in this stage was the lack of data or information about factors influencing the adoption of CSA.

The third step entailed gathering important information from each article and reading them thoroughly. For each article the following information was gathered: author names, title, journal, publication year, abstract, where the research was conducted, which method was used, size of the sample, and the main results related to the (dis)adoption of CSA.

The 23 articles that remained after this process were then subject to a critical appraisal to guarantee good quality articles. It was important that the sample size was large enough (>99 respondents) and that the sample strategy was clearly described. Similarly, the data collection process must be clearly described and standardized. Next, the statistical method used needs to align with the study objectives; solely qualitative studies were removed. Lastly, the study limitations need to be acknowledged and discussed. Articles that did not comply with two or more of these requirements were removed from the final list of articles. Two articles that did not describe the data collection process nor the study limitations clearly were removed. In both articles it was unclear how the data was collected (e.g., which type of interview) and how it was standardized. In addition, it was unclear who collected the data (e.g., enumerators, researchers, etc.). The limitations of the research were also not discussed.

Each included article was carefully reviewed, focusing on sections pertaining to the drivers and barriers influencing CSA adoption. The practices and factors influencing CSA adoption, as delineated in the articles, were systematically summarized in Excel to facilitate a comprehensive understanding of patterns and distinctions. Notably, an observation emerged that many studies concentrated on five to ten specific CSA practices (e.g., conservation agriculture, crop rotation, intercropping, etc.), while some addressed CSA adoption holistically. Methodological variations, such as the use of percentages vs. narratives detailing positive or negative effects, prompted the need for a structured presentation to enhance clarity and coherence across diverse studies.

After going through the steps denoted in Figure 1, 21 articles from 2019 to 2023 researched and discussed the factors influencing the adoption of CSA amongst African farmers and met the predefined criteria. A full article list can be found in Supplementary Table 2. All studies were concentrated in one of 11 African countries, mostly on the South-East side of the continent, denoted in Figure 2. In general, large data samples were employed, with only three out of the 21 studies having sample sizes of fewer than 300 farmers.

As previously noted in the introduction of this review paper, CSA comprises a compilation of existing practices. Across the 21 WoS articles analyzed, over 50 distinct CSA practices were identified, with most articles addressing between five to ten practices. While some studies assessed the individual impact of each practice, others examined the combined effects. However, there may be some discussion about how or why certain practices are categorized as CSA, such as the spiritual invocation of rain (Moutouama et al., 2022) and fodder and forage production (Zampaligré and Fuchs, 2019).

To get a better overview of the practices, it was opted to group them into 11 categories (see Table 1). Some practices such as cover cropping, mixed cropping, and intercropping are discussed in 11 articles, whereas other practices are only discussed in one article.

In the 21 WoS articles used for this review, over 40 factors or independent variables, influencing the adoption of CSA amongst African farmers were discussed. This paper focused on recurrent factors mentioned consistently across multiple studies, ensuring a robust and widely applicable analysis. This approach enables a nuanced understanding of the main influences driving CSA adoption in Africa while maintaining clarity and relevance in the discussion. Therefore, all factors that were only mentioned in one study (e.g., the influence of receiving remittances, being a member of a faith-based group, etc.), were left out. In total, 25 factors were researched in multiple studies and will be further discussed in the next sub-sections. The factors were subdivided into five groups: personal, farm-related, financial, environmental, and informational factors. Figure 3 denotes the general effect of each factor on CSA adoption whereas Supplementary Table 3 gives an overview of how the factors influence the adoption of specific CSA practices. The following subsections give a generalized summary of the effects of each factor. To enhance readability, consistent references to individual articles throughout the text have been omitted. Instead, a consolidated summary is presented in Supplementary Table 4. This table provides an overview of which articles investigate specific factors and whether they report positive, negative, or mixed effects.

The educational level of farmers emerges as a key determinant influencing the adoption of CSA, with a range of explanations provided by researchers. Higher-educated farmers exhibit a greater propensity to adopt CSA practices, attributed to farmers' capacity to obtain, process, and utilize information relevant to agricultural practices (Makate et al., 2019a; Abegunde et al., 2020; Obi and Maya, 2021; Oyawole et al., 2021; Sisay et al., 2023). Belay et al. (2022) found that a 1-year increase in education results in a 21.40% increase in the likelihood of adopting CSA practices. Additionally, CSA often requires substantial financial investments, which is something that more educated farmers can afford since they are more likely to possess diverse income sources (Kangogo et al., 2021). However, it is noteworthy that while higher education increases awareness and adoption of certain practices, it can also decrease adoption in other CSA practices, suggesting nuanced relationships between education and adoption outcomes (Ouédraogo et al., 2019). This is also concluded by Mthethwa et al. (2022), who suggest that educational levels may hinder CSA adoption, but they offer limited clarity or explanation regarding their observed outcomes.

Farmers' risk attitude significantly influences the adoption of CSA practices, as revealed Kangogo et al. (2021) and Musyoki et al. (2022). Notably, risk-averse farmers exhibit a lower likelihood of adopting CSA technologies such as terraces, ridges, and bunds, which may involve investments in terms of labor, time, and tools. This reluctance stems from their aversion to financial risks and the unwillingness to deplete limited cash reserves on these practices (Musyoki et al., 2022). On the other hand, risk-taking farmers are more likely to adopt various CSA practices, indicating their propensity to engage in new and innovative strategies. For instance, Kangogo et al. (2021) suggest that risk-taking plays a crucial role in decisions related to irrigation and changes in cultivation calendars, highlighting the importance of farmers' risk attitudes in shaping the landscape of CSA adoption.

Farming experience has been found to positively influence CSA adoption. Research by Belay et al. (2022) revealed that a 1-year increase in farming experience corresponds to a 3.90% rise in the likelihood of adopting CSA practices. More experienced farmers exhibit a higher likelihood of adopting CSA practices due to heightened awareness and a better understanding of changes in farming practices. This experiential advantage enables them to identify and comprehend the changes associated with CSA more easily (Ouédraogo et al., 2019). Additionally, the positive relationship between farming experience and CSA adoption is attributed to more efficient engagement with extension services and stronger social networks, providing these farmers with enhanced support systems (Abegunde et al., 2020). However, contrasting findings indicate that less experienced farmers may be more inclined to adopt specific technologies, such as weather-smart technology, possibly driven by a desire to mitigate risks in the absence of extensive experience (Antwi-Agyei and Amanor, 2023).

Farmers' land size is a critical factor influencing the adoption of CSA practices, as revealed by various studies. Makate et al. (2019a) and Sisay et al. (2023) explain that farmers with more extensive landholdings are more willing to apply CSA technologies since they have more space and flexibility to experiment and implement practices like agroforestry, integrated soil fertility management, and conservation agriculture. Moreover, wealthier households often have larger land sizes and higher asset indices, making it more feasible to adopt CSA practices such as stress-tolerant livestock, terraces, integrated soil fertility management, composting, stone bunds, farmer-managed natural regeneration, improved crop varieties, and fodder and forage production, as they possess the resources to manage associated risks (Zampaligré and Fuchs, 2019; Musyoki et al., 2022; Sisay et al., 2023). However, some researchers found mixed results, implying that the relationship between farmers' land size and CSA adoption depends on the practices. For instance, the positive association between land size and soil testing aligns with the scale advantage larger landowners have. In contrast, the negative association with intercropping suggests that farmers with smaller land sizes are more involved in crop diversification, probably to meet household food needs (Kangogo et al., 2021).

The impact of farmers' on-farm income on the adoption of CSA is mostly positive. Studies consistently show a positive and significant effect of (annual) on-farm income on the use of various CSA technologies, including conservation agriculture, agroforestry, climate information services, and more. This relationship is attributed to farmers with higher incomes being able to afford inputs and equipment required for implementing these technologies (Sisay et al., 2023). Moreover, more farm income reduces risk aversion and enhances exposure to information (Abegunde et al., 2020). In addition, this relationship between on-farm income and CSA adoption also has to do with selling agricultural produce for income generation, indicating that farmers engaged in commercial farming are more likely to adopt CSA practices. In contrast, those primarily cultivating for household consumption or with alternative income sources outside farming exhibit lower CSA adoption rates (Mthethwa et al., 2022).

The impact of farmers' off-farm income on the adoption of CSA practices reveals multifaceted dynamics. Abegunde et al. (2020), Djido et al. (2022), and Musyoki et al. (2022) found that average monthly off-farm income had a statistically significant and negative influence on CSA adoption, suggesting that reliance on off-farm activities as major income sources diminishes the commitment to agricultural production. This weak commitment is reflected in the reduced likelihood of adopting various CSA practices, as diversifying income sources may lead farmers to perceive less need for enhancing agricultural resilience through CSA adoption. Furthermore, cultivating a diversified income portfolio empowers smallholder farmers who are heavily reliant on agriculture for their livelihoods to mitigate the impacts of climate-related risks. This strategy is recognized as a crucial approach to enhancing farmers' resilience against climatic shocks (Djido et al., 2022). In contrast, Kurgat et al. (2020) noted a positive association between off-farm income and livestock diversity, emphasizing the role of financial resources in acquiring inputs for livestock production and the potential for larger land areas to cultivate diverse feeds, reducing reliance on purchased feeds.

Access to credit emerges as an important factor influencing the adoption of CSA practices in the reviewed studies. The positive and significant impact of credit access is evident in the adoption of multiple CSA practices such as conservation agriculture and integrated soil fertility management (Zampaligré and Fuchs, 2019; Kangogo et al., 2021; Sisay et al., 2023). According to Kangogo et al. (2021) and Sisay et al. (2023) this can be explained because having access to credit enhances farmers' financial capabilities, empowering them to cover transaction, equipment, and input costs linked to the diverse CSA technologies they may choose to adopt. Moreover, more access to cash empowers households to withstand and recover from losses resulting from climate change (Zampaligré and Fuchs, 2019). For instance, farmers with access to credit show a 20% higher likelihood of adopting climate information services, aligning with the understanding that CSA practices often involve substantial initial investments, and credit accessibility facilitates acquiring inputs (Ouédraogo et al., 2019).

The changing patterns of rainfall also seem to impact the adoption of CSA among smallholder farmers. On the one hand, in case of less rainfall or droughts, empirical results find a positive relationship with CSA adoption. When there is variability in rainfall amounts, farmers are compelled to actively seek diverse practices to conserve and harvest water, particularly during water-shortage seasons, thereby relying on CSA practices to ensure water access and management in response to shortages (Mthethwa et al., 2022). For instance, drought conditions positively influence the decision to implement ridges and bunds, as these structures mitigate surface run-off. The reduced run-off enhances soil moisture retention, providing farmers with a valuable resource during dry spells, and facilitating crop cultivation (Musyoki et al., 2022). On the other hand, an abundance of rainfall resulting in floods also impacts CSA adoption. Research shows that farmers who experience floods, are more likely to adopt stress-tolerant livestock and fertilizer application. The faith in stress-tolerant livestock's ability to cope with flooding, coupled with the desire to enhance farmland productivity after floods, motivates farmers to adopt these practices (Musyoki et al., 2022).

Soil quality is another factor influencing the adoption of CSA practices. According to Pangapanga-Phiri and Mungatana (2021) farmers with poor soil quality are more likely to adopt CSA practices such as soil and water conservation. Similarly, Teklu et al. (2023) found that poor soil fertility positively influences the adoption of agroforestry, row planting, and improved crop varieties with 4.7, 4.3, and 5% respectively. Neither of the researchers further explains this relationship.

Access to training emerges as a crucial determinant positively influencing the adoption of CSA practices among farmers. Various studies consistently demonstrate the significant and favorable impact of training on the utilization of diverse CSA technologies, such as conservation agriculture, agroforestry, integrated soil fertility management, terraces, integrated pest management, crop diversification, and more (Musyoki et al., 2022; Sisay et al., 2023). The findings underscore that farmers with access to training are more adept at employing CSA practices compared to those without such access. Specific training interventions are highlighted as influential factors in the adoption of particular technologies. For instance, training on variety selection positively influences the adoption of intercropping, and drought-tolerant or improved crop varieties whereas training on climate information services significantly promotes the adoption of such services (Ouédraogo et al., 2019).

Training is intertwined with farmers' access to extension services, which is also a pivotal factor influencing the adoption of CSA. Studies consistently affirm the positive and significant impact of extension contacts on various CSA technologies. Contact with extension services is associated with a heightened awareness of specific practices and technical support, which results in farmers being better informed and ready to implement drought-tolerant or improved crop varieties, integrated soil fertility management, micro-dosing practices, agroforestry, and organic manure adoption (Ouédraogo et al., 2019; Zakaria et al., 2020; Oyawole et al., 2021; Sisay et al., 2023). The multifaceted role of extension services encompasses providing information on climate change, adaptation, resilience, and agricultural management practices, contributing to increased awareness and adoption of CSA practices (Abegunde et al., 2020).

Belonging to agricultural-related associations or groups emerges as a key determinant positively influencing the adoption of CSA practices among farmers. The evidence across various studies consistently underscores the significant and positive impact of group membership on CSA adoption such as changing the cultivation calendar, certified seed adoption, crop rotation, and soil testing (Kangogo et al., 2021). Membership in these associations helps build strong social networks, fosters social capital, provides farmers with access to public spheres and creates platforms for the exchange of experiences and challenges. Such interactions within agricultural groups not only contribute to enlightenment but also offer practical advice on coping with agricultural problems, thereby enhancing CSA adoption (Abegunde et al., 2020; Belay et al., 2022). Furthermore, these groups enable farmers to exchange information about accessing agricultural-related innovations, services, training, and knowledge on new technologies, while also providing opportunities to obtain key inputs necessary for implementing CSA practices (Zampaligré and Fuchs, 2019).

Access to climate information is another crucial factor significantly influencing the adoption of CSA. Farmers equipped with reliable information on current and future temperature, rainfall, and climate conditions demonstrate a higher propensity to adopt CSA practices, including conservation agriculture, agroforestry, and crop diversification (Moutouama et al., 2022; Sisay et al., 2023). Information can be shared and received in various ways such as farm groups, extension services, mobile-based platforms, and community displays (Obi and Maya, 2021; Mujeyi et al., 2022). Moreover, access to weather forecasts can come from scientific sources or indigenous knowledge systems (Mujeyi et al., 2022). Access to climate information can be linked to a 20.20% increase in the likelihood of adopting selected CSA practices (Belay et al., 2022). This information enables farmers to adjust their farming activities and strengthen their resilience in response to climate change (Zampaligré and Fuchs, 2019).

The last factor to be discussed is closely related to the previous factor, climate information. Research shows that also awareness of CSA practices is a major determinant of CSA adoption. Therefore, a primary obstacle to the adoption of CSA is a lack of awareness about the practice. In the context of CSA, having sufficient awareness and appropriate information is crucial for the decision to embrace a new technology. A substantial portion of the population is not familiar with these practices (Moutouama et al., 2022). Teklu et al. (2023) found that farmers who are aware of CSA are between 6.70 and 33% more likely to adopt CSA practices such as improved crop varieties, agroforestry, crop residue management, compost, row planting, soil and water conservation. Only for crop rotation they did not find a significant effect, this might be because crop rotation is also used by farmers who are not so familiar with CSA. Both articles do not provide further information explaining their findings.

In this section, factors exhibiting nuanced effects on the adoption of CSA practices are explored, showcasing how these relationships can be influenced by various contextual factors. The decision to narrow the focus to three factors with mixed results was driven by their recurring presence in the reviewed studies, indicating their prominence in the literature on CSA adoption.

First, “gender” emerges as a factor with mixed results even though the conventional assumption suggests a higher readiness for CSA adoption among male farmers. In some publication this is confirmed. For instance, Pangapanga-Phiri and Mungatana (2021) discovered that female farmers, often hindered by limited access to agricultural inputs, exhibit lower technical efficiency, posing challenges to their adoption of CSA practices. This finding was corroborated by studies from Kangogo et al. (2021) and Musyoki et al. (2022), which highlighted gendered disparities in resource access and information, contributing to lower CSA adoption rates among female farmers. On the other hand, Musyoki et al. (2022) found that female farmers tend to adopt more individual CSA practices, while male farmers are more inclined to combine multiple practices, potentially due to their superior resource access. Interestingly, Oyawole et al. (2021) reported contrasting results, indicating that female farmers adopt more CSA practices than their male counterparts, particularly in agroforestry, green manure, organic/compost use, and zero/minimum tillage. This diversity underscores the complex interplay between gender dynamics, resource availability, and CSA practices.

Secondly, the influence of “age” on the adoption of CSA practices presents a nuanced picture with varied outcomes. Older farmers, presumed to possess greater experience and access to resources, are expected to exhibit higher CSA adoption rates. Studies by Zampaligré and Fuchs (2019), Kurgat et al. (2020), and Sisay et al. (2023) support this notion, highlighting the propensity of older farmers to adopt certain CSA practices such as stone bunds and crop diversification. This inclination may stem from their economic stability and larger land holdings, facilitating access to improved crop varieties and enabling the adoption of labor-intensive techniques. However, the age of farmers also exerts a negative impact on the adoption of innovative technologies. Younger farmers, characterized by greater adaptability and access to information, demonstrate higher adoption rates for practices such as pest-resistant crops, conservation agriculture, and drought-tolerant varieties. Obi and Maya (2021) note a diminishing likelihood of older farmers embracing new farming practices because they are more risk-adverse, further highlighting the generational disparity in CSA adoption. Consequently, while older farmers may leverage their experience and resources to adopt certain CSA practices, younger farmers exhibit a greater propensity for embracing novel and technology-driven approaches to agricultural sustainability.

Thirdly, the variable “distance to market” presents a nuanced relationship with CSA adoption, contrary to the initial expectation of many researchers of a negative impact. Makate et al. (2019a) highlighted the higher transaction costs associated with increased distance to markets, potentially discouraging CSA adoption due to difficulties in accessing input and output markets. Similarly, Abegunde et al. (2020) emphasized how distance from towns or markets affects farmers' access to crucial resources such as information, technologies, and credit facilities, while also increasing transportation costs and complicating monitoring efforts for distant farmland. Zakaria et al. (2020) reported both negative and positive effects of distance on CSA adoption, with some practices like adopting drought-tolerance varieties and adjusting planting periods positively influenced by distance to market or town, yet others, such as adopting early-maturing varieties, bunding, and irrigation, are negatively impacted. This inconsistency could stem from the challenges smallholder farmers face in accessing extension services when located far from markets. Surprisingly, Makate et al. (2019b) found a positive association between distance to market and the adoption of improved legume varieties, potentially attributed to extensive advertising and community seedbanks mitigating the negative impact of distance. Additionally, Musyoki et al. (2022) discovered a positive influence of distance to market on household decisions to adopt fertilizer, possibly due to collaboration with local NGOs facilitating doorstep delivery of farm inputs, bypassing the need for market visits.

Although there are other factors with nuanced effects (e.g., farmers' marital status, household size, livestock ownership, hired labor, access to a phone and media, and more), these three factors already demonstrate the complexity of the relationship between context and CSA adoption, highlighting the multifaceted considerations that influence farmers' decisions in adopting sustainable agricultural practices in SSA.