Social media and drone adoption in mountainous fruit production: the mediating role of technology cognition

Social media platforms have been reshaping the dissemination of agricultural technology information, yet the mechanisms through which these platforms influence advanced machinery adoption among smallholder farmers in mountainous regions remain insufficiently understood. Based on the Technology Acceptance Model (TAM), this study constructed an analytical framework of “social media use → technology cognition → adoption behavior” and introduced terrain conditions as a contextual moderating variable. We collected survey data from 650 mountainous fruit farmers across Shaanxi, Yunnan, and Hunan provinces in China, and employed Logistic regression models, Bootstrap mediation tests, and instrumental variable methods to systematically examine the effects of social media use on plant protection drone adoption decisions. The results demonstrate that social media use intensity exerts a significant positive effect on drone adoption (OR = 2.68, p < 0.001), with users showing a 34 percentage point higher adoption probability than non-users. Mechanism analysis reveals that technology cognition—combining perceived usefulness and perceived ease of use—partially mediates the relationship between social media and adoption behavior, accounting for 40.8% of the total effect (95% CI: 0.28–0.54). Heterogeneity analysis indicates that terrain conditions significantly moderate this relationship: the marginal effect of social media among mountainous farmers is 2.74 times that of plain farmers (OR = 4.14 vs. OR = 1.51). Regional analysis reveals social media effects range from OR = 1.93 (Hunan) to OR = 3.77 (Shaanxi), supporting the terrain moderation mechanism. This study extends TAM in mountainous agricultural contexts and identifying terrain as a critical moderator, providing empirical evidence for optimizing digital agricultural extension strategies and bridging the technology adoption gap in mountainous areas.