Résumé
From 2013 to 2015, a study was conducted on 29 coffee plots to assess the influence of landscape context and plot features on the epidemics of coffee leaf rust (CLR) descriptors in Gumay and Setema districts, Southwest Ethiopia. Maximum CLR severity and number of sporulating lesion density per leaf was significantly correlated with different land use types (tree density at 100 m, 200 m, pasture and crop lands at 200 m circular radius), and plot features (altitude, North-South aspect, East-West aspect and leaf area index) over different years of observation. High tree density in the surrounding coffee plots significantly increased the severity of the disease. There was a general decreasing trend of CLR severity, sporulating lesion density and proportion of leaf fall over the altitudinal gradient. The interaction of landscape context with CLR descriptors found were so complex and need further detail analysis.