Contribution of morphological traits for prediction of sweet potato weevil resistance in Uganda

Abstract

Sweetpotato is an important source of nutritional and bioactive substances especially to resource poor households in sub-Saharan Africa (SSA). However, several pests affect it with the sweetpotato weevils (SPW) being the most destructive. These sweetpotato weevils can cause 67-100% yield loss in sub-Saharan Africa especially in drought prone areas. Hitherto, there are no weevil resistant sweetpotato varieties that have been released by breeding programs in SSA. This has been attributed predominantly to large genome and complex genetics of sweetpotato characterized by an auto-allopolyploid genome (2n=6x=90). As a result, the genetic basis for the inheritance of many traits including resistance to SPW in sweetpotato has not been understood. Previous studies identified several traits that could be responsible for resistance against the sweetpotato weevil. The aim of this study was to isolate the most efficient resistance predicting parameters against sweetpotato weevil and to determine genotype by environment (G*E) effects and stability of the key resistance parameters across locations. Thirty genotypes including local collections, released varieties and breeding lines were evaluated at three locations (Namulonge, Serere and AbiZARDI) and two seasons (2017A and 2017B) in Uganda. Data were collected on storage root yield, sweetpotato weevil damage (infestation) and weevil resistance factors including vine vigor, ground cover, vine pubescence, vine weight, root neck length, root latex content, root cortex thickness and storage root dry matter content. Genotype means across seasons in the three locations for weevil damage, storage root yield and other measured attributes varied significantly. The mean genotype weevil damage ranged from 1 in genotypes MSD 380, ADJ 190 and SPK004 to 5 in genotype NKB3 with overall grand mean of 2.6. While the mean genotype storage root yield ranged from 5.3 tons per hectare in genotypes NASPOT1 and ADJ 190 to 29.1tons in genotype NKB 3 with grand mean of 15.2 tons. G*E was significant across seasons and locations for weevil damage severity. Significant negative correlation coefficients were observed between weevil damage and morphological traits (vine vigor, ground cover, root neck length, root latex content, root cortex thickness and storage root dry matter content). The path coefficient analysis was able to point out ground cover and root neck length as the most important resistance predicting morphological traits. Additive Main effects and Multiplicative Interaction (AMMI) analysis indicated the test environments to be very diverse. The relative performance of the genotypes was significantly different for SPW severity and ground cover in the test locations. Therefore, Root neck length was stable across environments while ground cover was unstable across the locations. It is therefore possible to indirectly select for weevil resistant sweetpotato varieties for Uganda using the identified resistance sweetpotato attributes