Spatial analysis and factors influencing the spread of verticillium wilt in cocoa plantations in the agroecological zones of Beni and Mambasa (North Kivu and Ituri, DRC)

Abstract

Cocoa verticillium wilt, caused by Verticillium dahliae, poses an emerging threat to the cocoa sector in the Democratic Republic of the Congo (DRC). This study assessed the incidence and severity of the disease in 90 plantations spread across six regions and 18 villages in the Beni and Mambasa territories. Statistical analyses were adapted to the nature of the data: the Shapiro-Wilk test for normality, the Levene test for homogeneity of variances, followed by the nonparametric Kruskal-Wallis test. Multiple comparisons were performed using the Dunn test with Benjamini-Hochberg correction. The results show significant spatial heterogeneity (Kruskal-Wallis, χ² = 38.473, df = 17, p = 0.0021), with major clusters in the village of Mikwata (1.85 ± 0.58): Makumo axis, Makumbise village (1.77 ± 0.71): Makumo axis, and in Kamirihi village (1.78 ± 0.95): Mbau axis, while minor clusters were recorded in Kisanga village (0.41 ± 0.36): Mutwanga axis, Kazebere village (0.45 ± 0.14): Kazebere axis, and in Vusivwameso village (0.52 ± 0.20): Kazebere axis. A significant negative correlation between altitude and incidence (ρ = −0.39; p < 0.001) indicates a modulating effect of topographic factors. These results underscore the need for an integrated and spatialized approach to plant health management, prioritizing low-altitude areas.