Exploring the yield gap of orange-fleshed sweet potato varieties on smallholder farmers' fields in Malawi

Orange-fleshed sweet potato (OFSP) can contribute to combating vitamin A deficiency and establishing more resilient cropping systems in sub-Saharan Africa. There is limited understanding of the factors that affect yield and quality of OFSP on smallholder farmers' fields. This study aimed to assess the performance of six OFSP varieties, identify factors limiting productivity and explore options to close the gap between actual and attainable OFSP yields on fields of smallholder farmers. Data were collected in the 2015/16 growing season from 221 on-farm variety demonstrations in seven districts in Central and Southern Malawi. Dependent variables of interest included crop establishment, vine yields, storage root formation, root yields, percentage of marketable root yield, and weevil infestation. Using linear mixed models, a range of biophysical, climatic, management and socio-economic factors and variables was used to identify associations with these dependent variables. The root yield gap was explored using a multivariate boundary line model to identify the most yield limiting factors. Results show a large variability across farmers' fields and a wide range of interacting factors affecting the variables of interest. Varieties Chipika and Kadyaubwerere attained good yields and were preferred by farmers in terms of taste. Varieties Zondeni and Anaalcwanlre gave a poor root yield, but a good vine yield. Timely planting is crucial to attain good root yields by making better use of the available rainfall. There was a varietal effect on weevil infestation and Kaphulira was most affected. Weevil control is required for market-oriented producers to enhance the percentage of marketable roots. The average attainable fresh root yield ranged from 18 t ha(-1) for Zondeni to 32 t ha(-1) for Mathuthu, against actual yields of 5-9 t ha(-1). Elevation, planting date, rainfall and crop establishment could explain only 28 percent of the average yield gap, while 49 percent was explained for Mathuthu. Other factors that may explain the yield gap, but were not included in the model are: tillage methods and soil nutrient limitations. Male host farmers received better quality cuttings and planted in better soil moisture conditions, resulting in better establishment and vine yields. OFSP productivity can be enhanced through gender-sensitive extension, by ensuring male and female farmers can plant clean planting material of a suitable variety early in the rainy season. This requires additional efforts in vine multiplication of the required variety prior to the onset of the rains.