The interplay between dairy farming, climate change, and farmers can be effectively analyzed through the lens of socio-ecological systems. This approach provides a comprehensive understanding of how these elements interact and adapt within a dynamic, interconnected system. A socio-ecological system is a combination of biophysical and ecological units that interact with social systems shaped by stakeholders and institutions (Flórez et al., 2024b; Pallero et al., 2018; Martín-López et al., 2012; Glaser et al., 2008; Ostrom, 2009). Socio-ecological systems are defined by the dynamic interaction between their ecological and social components. The ecological system provides essential ecosystem services that sustain human life, economic activities, and production. In turn, the social system generates both positive (environmental benefits) and negative (externalities) impacts that influence the ecological system’s functionality (Martín-López et al., 2012). Managing these systems effectively requires an ecosystem-based approach and strategies that are aligned with social, economic, and ecological contexts (Pallero et al., 2018).

Social valuation is essential for assessing socio-ecological systems, offering insights into how communities perceive, value, and interact with their environment. It examines cultural, social, and personal factors to better understand people’s relationship with nature and its resources. According to Scholte et al. (2015), a complete social valuation of ecosystem services involves three core elements: knowledge (what we know), perception (how we perceive it), and use (how we engage with it). Integrating these elements provides a comprehensive understanding of human-environment interactions (see Figure 1).

Social valuation bridges individual attitudes with actionable strategies, ensuring a holistic approach to valuing goods and services and informing policies for sustainable outcomes. While studies on social valuation in the livestock sector are emerging, existing literature often treats knowledge, perception, and use of mitigation strategies separately, lacking an integrated framework for understanding their social value. This fragmented view limits understanding of how these components intersect to influence farmers’ engagement with mitigation, particularly in terms of their values, beliefs, and decision-making. This gap is especially significant in regions like Africa, where socio-ecological vulnerabilities and resource constraints are more pronounced. The following subsections provide an overview of existing literature on these topics.

In Africa, livestock farmers’ knowledge of climate change is shaped by education, media exposure, and local experience, with better-informed communities showing higher awareness (Mahl et al., 2020; Silvestri et al., 2012; Wetende et al., 2018). While farmers broadly recognize climate impacts such as feed scarcity, heat stress, and shifting rainfall, understanding of GHG emissions and specific mitigation options remains limited (Tadesse and Dereje, 2018; Feliciano et al., 2022). Perceptions and adaptive responses vary by socio-demographic factors including age and gender, often differing from meteorological records (Abazinab et al., 2022; Kasulo et al., 2012; Kumar et al., 2023; Bryan et al., 2023). Evidence from dairy systems in Northern Ethiopia highlights ongoing adoption challenges in climate-smart agriculture and the increasing risk of heat stress in cattle (Balcha et al., 2023; Balcha et al., 2024). Common adaptation strategies include livestock and crop diversification, seasonal mobility, feed conservation, improved pastures, crop–livestock integration, and drought-tolerant forages, combining traditional and modern knowledge (Descheemaeker et al., 2016; Popoola et al., 2019; Silvestri et al., 2012; Tadesse and Dereje, 2018). Although interest in climate-smart forages is increasing (Flórez et al., 2024c; Junca Paredes et al., 2023), adoption remains constrained by limited resources, weak institutional support, and competing socio-economic priorities (Feliciano et al., 2022; Wetende et al., 2018; Mahl et al., 2020).

To estimate the social value of climate change mitigation strategies, we developed a social value indicator structured around three conceptually grounded components commonly used in social valuation frameworks: knowledge (understanding), perception (perceived relevance and benefits), and use (current or intended application of the practice). A total of 46 dairy farmers participated in the social valuation survey, selected through purposive sampling from CIAT ¹ -supported projects to ensure prior exposure to forage-based climate mitigation practices. While participants were not tested through a formal knowledge assessment, their involvement in training activities and demonstration farms provided baseline familiarity with the practices evaluated. This study is a case study conducted in Nandi and Uasin Gishu counties, two dairy-producing regions in Kenya engaged in CIAT forage and livestock initiatives. As such, results reflect local conditions and are not intended to be statistically representative of other regions in Kenya.

Two data collection workshops were held in July 2024 – one with 20 women dairy farmers and the other with 26 men. Each workshop was divided into two segments: a field visit to demonstration farms and a conference room session. During these sessions, participants were engaged in presentations, discussions, and exercises addressing the social and technical aspects of adopting GHG mitigation strategies. The workshop was structured in two segments to balance experiential learning with standardized data collection. The field visit allowed participants to observe mitigation practices in a real farm setting, ensuring a shared reference point, while the conference session provided a controlled environment for discussion and survey implementation. To minimize bias, no performance evaluation or persuasive messaging was delivered during the field visit, and the social valuation survey was administered individually before group discussions or technical presentations to reduce social desirability and peer-influence effects. This research was developed as a case study involving farmers participating in CIAT-led projects in Kenya, and it is not intended to provide statistically representative results. Instead, the study aims to demonstrate the application of the social valuation methodology and its potential for future research with larger sample sizes and expanded sampling designs. The data collection process followed these sequential steps:

Participant recruitment through purposive sampling from CIAT-supported dairy projects.

The social valuation survey was conducted incorporating an individual exercise structured into the following sections:

Section Introduction - Demographic, socioeconomic, and dairy production information: Participants completed a detailed survey to collect essential demographic and socioeconomic information, including farm size, number of cattle, milk production levels, and income sources.

Section Materials and methods - Social valuation of GHG mitigation strategies: This section assessed participants’ knowledge and perceptions of climate change and GHG emissions, their willingness to adopt mitigation strategies, and their readiness to invest in such strategies. Urochloa hybrids were selected as mitigation strategy because they are among the most familiar and contextually relevant mitigation practice for farmers involved in CIAT projects in Kenya. Urochloa hybrids are a particularly attractive climate change mitigation option because their deep root systems build soil carbon and they perform well even under low soil fertility, achieving high yields and nutritional quality. This allows for the intensification of livestock production, increasing carrying capacity on a smaller land area, freeing land traditionally occupied by native or naturalized pastures, and ultimately reducing greenhouse gas emissions per unit of meat or milk (Enciso et al., 2019). As this is a case study focused on applying the social valuation methodology, the research does not compare multiple mitigation options. Future studies could expand the analysis to other alternative strategies.

The survey questions were inspired by the National Climate Change Perception Survey conducted in Costa Rica (MINAE, 2021) (see Table 3 in the results section for its components).

The survey included numerous questions related to social value; to simplify the analysis, we generated composite indicators that summarize this information into a few key variables. We used Principal Component Analysis (PCA), a statistical method that reduces the dimensionality of datasets with interrelated variables by transforming them into a smaller set of uncorrelated principal components that retain most of the original variance (Jolliffe and Cadima, 2016). In a sample of n individuals described by p variables (X₁, X₂, …, X_p), PCA identifies z components (z < p) that explain most of the variance. Each of these z variables is called a principal component (Greenacre et al., 2022; Amat-Rodrigo, 2017).

Indicators were calculated as weighted averages of the principal components, with weights based on the proportion of variance explained. Because survey items were measured on different scales, all values were standardized to a 0–1 scale using each variable’s maximum possible value. PCA was performed using R software (R Core Team, 2023). The number of principal components was determined using scree plot inspection to retain only components explaining meaningful variance. The survey items used different response scales (binary, Likert, and ordinal) because they capture distinct dimensions (knowledge, perception, and use), which require different measurement approaches. To ensure comparability across variables in the PCA, all indicators were standardized prior to analysis to mitigate scale effects.

The PCA results enabled the construction of a consolidated indicator by integrating the original variables, the most relevant principal components, and their respective variance contributions. The proportion of variance explained was calculated from the eigenvalues of each principal component relative to the total variance, as is standard in PCA; no statistical significance testing was applied, as component retention was based on explained variance and interpretability rather than hypothesis testing. Following the generation of components, we applied the methodology of Peña Méndez and Gutiérrez Sánchez (2014), Saldarriaga-Isaza et al. (2025), and Flórez (2022). This approach involves calculating a weighted sum of key variables within each significant principal component, proportional to the variance each explains. For example, if PCA identifies three principal components as relevant, with PC₁ defined by X₁ and X₂, PC₂ by X₃ and X₄, and PC₃ by X₅, then the indicator (I) is calculated as: I = P r o p V a r P C 1 ∑ i = 1 n P C i * X 1 + X 2 Max X 1 + X 2 * P r o p V a r P C 2 ∑ i = 1 n P C i * X 3 + X 4 Max X 3 + X 4 * P r o p V a r P C 3 ∑ i = 1 n P C i * X 5 Max X 5

The formula is adapted based on the number of relevant principal components identified and the dominant variables within each. It enables the calculation of indicators on a standardized scale: within the range [0, 1] when all variable values are positive, or [-1, 1] when the dataset includes both positive and negative values.

The social value indicator comprises three sub-indicators aligned with the core dimensions of social value:

Knowledge: Measures participants’ awareness and understanding of climate change and its impacts.

These sub-indicators are derived from specific survey questions, and their interpretations are detailed in Table 1. To support comparability and interpretation, indicator values were categorized into five equally sized ranges using the equal interval classification method, which is recommended for analyzing continuous data. An equal-interval classification was applied, as it is widely used for emerging research topics where empirical evidence is still limited and no standardized thresholds exist to support more specific distributions (Meyer, 2024). This approach offers a clear and nuanced understanding of the social factors influencing the adoption of GHG mitigation strategies in dairy systems.

Table 2 presents the basic characteristics of the dairy farmer sample surveyed in Nandi and Uasin Gishu counties (46 respondents). The data shows a slightly higher proportion of respondents from Uasin Gishu. The gender distribution was balanced, and the average age of participants was just over 42 years. The average education level was 2.2, which likely corresponds to completion of secondary school. Farmers reported large household sizes and high numbers of dependents, indicative of the elevated dependency ratios commonly observed in rural agricultural communities. Dairy farming emerged as the principal livelihood for nearly all respondents, who had an average of 13 years of experience in the sector. Despite their experience, household incomes remained low, with a monthly per capita income of just under USD 37, reflecting the economic challenges many smallholders face.

A significant majority of respondents reported receiving support from external sources. Cooperatives were the most common providers of support, followed by NGOs, research institutions, and government agencies. Private sector involvement was minimal, suggesting a potential gap in public–private collaboration in these areas. The types of support most frequently received included training, milk marketing assistance, production inputs, and rural extension. Less commonly, farmers accessed credit, technology transfer, or monetary subsidies. Notably, around one-fifth of respondents reported receiving no external support at all.

Cooperative affiliation was nearly universal among participants, indicating that cooperatives play a crucial role in connecting dairy farmers with essential services, training opportunities, and access to markets. Overall, the data reflects a population of experienced dairy farmers operating under modest economic conditions, with cooperatives serving as key channels for institutional support and development.

Table 3 provides an overview of the basic results of the social valuation survey. The surveyed dairy farmers demonstrated a high level of awareness (knowledge) of climate change and GHG emissions, with nearly all respondents acknowledging their existence and anthropogenic origins. This reflects a strong foundational understanding across the sample. Notably, women were equally – or in some cases more – likely than men to recognize that climate change is occurring, though a slightly lower proportion attributed it primarily to human activity. These findings underscore the importance of designing gender-sensitive awareness campaigns that reinforce scientific explanations and actively engage women.

In terms of perception, both male and female respondents view climate change as a current and pressing issue rather than a distant threat. However, women reported slightly lower levels of discussion on the topic in social settings and a lower perceived personal impact. This may reflect differentiated roles in household and community decision-making, but also gender-differentiated levels of education, and points to the need for inclusive communication strategies that foster dialogue across genders.

Regarding the willingness to act, a large majority of farmers expressed belief in the feasibility of reducing GHG emissions and indicated readiness to adopt mitigation practices. Women expressed slightly greater optimism about the potential to reduce emissions (100% vs. 96% among men), yet they also reported lower levels of preparedness (65% vs. 81%) and no perceived access to necessary resources (0% vs. 12%). These disparities suggest that women face systemic barriers, including limited access to technical training, financial capital, and decision-making authority on the farm. Despite these constraints, women expressed full willingness to participate in training (100%), highlighting a valuable opportunity for targeted capacity-building efforts. Addressing these gender-specific barriers will be essential to ensure the equitable and effective implementation of climate change mitigation strategies in Kenya’s dairy sector.

Table 4 provides an overview of the indicators derived from our analysis. Our analysis estimated a knowledge indicator of 0.962 and a perception indicator of 0.807, both classified as very high. These results suggest that dairy farmers in the sample are well-informed about climate change and GHG emissions and perceive them as pressing concerns affecting Kenya, their farms, and their families. Farmers broadly acknowledge the reality of climate change and recognize its risks and impacts on their livelihoods. Despite this high awareness, the willingness to act indicator is low (0.380), largely due to knowledge gaps in implementation strategies and, more critically, financial resource constraints. Notably, 93% of respondents reported lacking sufficient resources to adopt GHG mitigation measures. These limitations significantly reduce their capacity to translate concern into action.

Combining the three sub-indicators yields a social value indicator of 0.723, categorized as high. The values were consistent across the two counties surveyed, with Nandi scoring 0.723 and Uasin Gishu 0.721, indicating no significant regional variation. However, meaningful differences emerged by gender and age. Men had a higher social value indicator (0.735) than women (0.707), driven primarily by the latter’s lower willingness to act score (0.345). This gap may reflect gender-specific barriers such as limited access to training, financial capital, or decision-making authority on the farm and in the household. Similarly, youth scored lower (0.646) than adults (0.734) and older adults (0.737), likely due to a lower perception score (0.688) and less farming experience.

While these patterns are observable in the data, further research is needed to understand the root causes of gender and age disparities–whether they stem from resource access, information availability, decision-making roles, or broader sociocultural dynamics.

Finally, we asked dairy farmers to identify their preferred GHG mitigation strategies for implementation on their farms (Figure 2). The most preferred strategy, chosen by the majority of respondents, was to replace half of their natural pastures with Urochloa hybrids and remove male animals from the herd – an approach that simulates the use of artificial insemination. The second most favored option involved combining Urochloa hybrids with the replacement of local cows by purebred Friesians. Additionally, still one-third of respondents were interested in adopting Urochloa hybrids alone, without any complementary herd management practices. In contrast, improved manure management was the least attractive option, with few farmers indicating a willingness to adopt it. These preferences appear to be driven primarily by the perceived productivity and income benefits associated with improved feeding and genetic upgrading, while the GHG mitigation impact is seen as an added co-benefit rather than the main motivation. This highlights the importance of aligning climate-smart interventions with farmers’ economic incentives, ensuring that mitigation strategies are both technically viable and perceived as beneficial to farm productivity and livelihoods.