Genetic diversity of potato for nitrogen use efficiency under low input conditions in Ethiopia

Potato is a prime food security crop for smallholder farmers in the highland part of North western Ethiopia. In this region, nutrient availability, especially nitrogen (N) is a major constraint for crop productivity. To obtain insight in the possibility of improving potato for growth under low N input conditions in Ethiopia, we evaluated CxE diploid back cross population, modern European and Ethiopian potato cultivars and local Ethiopian cultivars for their ability to grow and produce tubers under low and high N input conditions. The experiments were conducted under rainfed and irrigation conditions. Eighty-eight Dutch cultivars and 9 Ethiopian cultivars were evaluated in three locations in North-western Ethiopia, in 2013 and in 2015. The two years represent two different growth seasons: rain-fed (June-October 2013) and irrigated cultivation (February-June 2015). Similarly 100 CxE diploid back cross potato genotypes were evaluated in both rainfed and irrigation production seasons in 2014. The Growth of the plants was monitored throughout the growth cycle using canopy cover measurements, with modelled canopy characteristics, and other agronomic traits were measured as per the description. The effect of season and location was further investigated by a GGE Biplot genotype-by-environment interaction analysis, and genetic factors determining phenotypic traits and yield were identified through QTL mapping and association mapping. Ethiopian cultivars showed a remarkable, environment-dependent difference in utilisation of the canopy for tuber production. While total photosynthetic capacity was higher in Ethiopian cultivars than in Dutch cultivars in rainfed production season at Injibara, tuber production was higher in Dutch cultivars. This low radiation use efficiency was not observed in the other rain-fed location (Debre-Tabor). A Genotype by Environment analysis using GGE biplots demonstrates that, Irrespective of the N levels and locations, rainfed production season test environments were grouped as one mega environment and irrigation production season test environments as the other mega environment, indicating most of the variation for yield and nitrogen use efficiency (NUE) in the dataset may be caused by the effect of rain-fed vs irrigation season. Further trials are needed to confirm this result. The QTL mapping with the CxE diploid population and GWAS analysis with the Dutch cultivars discovered both season-environment and N-specific QTL as well as constitutive QTLs. Overall, N availability affects Dutch and Ethiopian cultivars differentially, with strong environmental interaction on canopy and yield traits. Rainfed and irrigated seasons in Ethiopia may require different breeding programs for improved yield under varying fertilizer levels. Both constitutive and environment-specific QTLs were identified that may be targets for breeding prorgams towards improved yield under Ethiopian cultivation conditions.