Correlate the cyanogenic potential and dry matter content of cassava roots and leaves grown in different environments

Open Access JournalCassava (Manihot esculenta Crantz) is an essential stable food crop in Sub-Saharan Africa commonly consumed amongst the low-income communities in Africa. Though cassava roots and leaf have vast economic and commercial benefits, it produces cyanogenic glycosides, which are toxic and most often responsible for the bitter taste of some cassava cultivars. The study evaluates the cassava roots and leaves’ cyanogenic potential and dry matter content of the Genetic Gain Assessment trial grown in a different environment. It establishes the association between the cyanogenic potential (CNP) and the roots and leaves dry matter (DM). Genetic Gain Assessment (GGA) cassava genotypes (N = 400) selected for the Uniform Yield Trial (UYT) breeding stage were planted under IVS (Dry season in Inland Valley Hydromorphic area) and Upland (rain-fed conditions) in two locations of IITA Research Farms, namely; Ibadan (IVS and Upland) and Mokwa (Upland) in Nigeria. The CNP content of cassava leaves in IVS, Mokwa, and Upland ranged from 3.39 to 272.16 mg/100 g, 4.28 to 228.72 mg/100 g, and 13.13 to 127.39 mg/100 g, respectively. However, the respective CNP range in root samples across IVS, Mokwa, and Upland was 0.76–76.31 mg/100 g, 0.94–136.53 mg/100 g, and 2.37–47.11 mg/100 g. Also, the mean ± SD of DM content of leaves were 27.97 ± 3.01%, 28.81 ± 4.01%, and 13.65 ± 3.69%, respectively, in IVS, Mokwa, and Upland, while the root samples had mean ± SD of DM content of 38.09 ± 4.80%, 32.69 ± ,5.93% and 24.63 ± 5.07% respectively. Furthermore, location and genotype had a highly significant effect (p < 0.001) on the CNP and DM of roots and leaves. Also, linear regressions were established between CNP and DM of root and leaf with regression equation; DM-Root = 1.1999*DM-Leaf (r = 0.956) and CNP-Root = 0.29006*CNP-Leaf (r = 0.54). The relationship between the DM (root and leaf) and CNP (root and leaf) could serve as a valuable “inter-prediction” tool for these parameters