Abstract

Coffee production in Peru plays a crucial socio-economic role, supporting over 200,000 families and contributing significantly to export income. However, the spatial variation in coffee farming across ecological and socio-economic regions remains poorly understood. This study examines spatial patterns of coffee farm clustering in three Peruvian mountainous regions (Moyobamba, Tingo María, and Tocache) using descriptive statistics, geospatial visualization, and unsupervised clustering techniques. Farm-level reports and government geospatial records covering 2019–2023 were analyzed to evaluate cultivation area, altitude, and spatial distribution. Kernel density mapping, Moran’s I spatial autocorrelation, and Local Indicators of Spatial Association (LISA) were applied to identify statistically significant clustering patterns, while regression analysis and DBSCAN clustering were used to evaluate spatial trends and production hotspots. Moran’s I indicated moderate spatial clustering (0.34, p < 0.001), while regression analysis showed a weak negative association between altitude and cultivation area (β = −1.144 × 10−4, adjusted R2 = 0.023). Results suggest that measured environmental variables explain only a limited proportion of spatial variation in coffee production, indicating that additional unmeasured factors, potentially including socio-economic influences, may contribute to observed clustering patterns. These findings highlight the value of spatial analysis for understanding production heterogeneity and for supporting regionally adapted agricultural planning strategies.

IPC Classification

Keywords