Toxic or synergistic relationship: The interdependence and determinants of dairy climate smart agriculture practices in the North Rift and western parts of Kenya

Dairy farming underpins rural livelihoods in Kenya but is increasingly vulnerable to climate variability while contributing to greenhouse gas emissions. Climate-Smart Agriculture (CSA) practices offer pathways for resilience and mitigation, yet most previous studies assess single practices in isolation, overlooking complementarities, trade-offs, and regional differences. This study examines the joint adoption of five CSA practices—feed management, improved breeding, manure management, animal health, and milk value addition—using survey data from 447 dairy farmers in Kisii and Uasin Gishu counties, complemented with high-resolution rainfall and temperature records. A Multivariate Probit (MVP) model revealed high adoption of breeding (87 ​%), animal health (85 ​%), and feed management (81 ​%), while manure management (39 ​%) and milk value addition (22 ​%) lagged. Regional contrasts were evident: Kisii farmers favored manure management (44 ​%), while Uasin Gishu led in feed management (89 ​%) and value addition (39 ​%). Strong complementarities emerged between feed management and breeding, animal health and feed management, and manure management and value addition, with no evidence of substitutability. Adoption was shaped by climate factors (rainfall, temperature), socio-economic characteristics (gender, age, household and herd size, farm size), and institutional access (cooperatives, credit, extension, early warning systems). Findings underscore that bundled CSA strategies are more effective than isolated interventions, providing comparative, policy-relevant insights for promoting climate-resilient dairy farming in Kenya and similar smallholder systems. • Geospatial analysis informs climate smart agriculture (CSA) studies. • CSA practices show synergy in simultaneous adoption. • Gridded meteorological data outperforms farmer perceptions on climate change. • CSA adoption shaped by climate, socio-economic, and institutional factors. • Multivariate Probit model captures simultaneous CSA adoption decisions.