The experiment was conducted in Uganda 2013, in two locations; National Crops Resources Research Institute (NaCRRI) in Namulonge representing the lowland at 1150 meters above sea level, and Buginyanya Zonal Agricultural Research Development Institute (BugiZARDI) in Buginyanya located in highlands sites in Eastern Uganda at >1750 m.a.s.l. A total of 4800 botanical seeds collected from 24 half-sib families were planted in the screen house in the Mukono district in central Uganda. Only 10 progenitors of 50 seedlings were selected to generate progeny for evaluation in the highlands. Twenty-five (25) seedlings (genotypes)/family were evaluated in each site. Data were collected in both sites on plant vigor (Pltv.) at three months after planting (MAP), incidence, the severity of CMD and CBSD at three, six MAP, and root yield traits at harvest 10 MAP: shoot weight (STW), storage root per plant (SRP), storage root weight (SRW), commercial storage root (CSR), storage root length (SRL), storage root girth (SRG), harvest index (HI), percentage dry matter content (DMC%) and cassava brown streak root incidence (CBSD-RI) express in percentage. The research findings reached were summarized; the possibility of the selection for traits of (STW, SRW, SRP, CSP, SRL, SRG, HI, DMC, and CBSD-RI in cassava at highland sites, prediction of cassava performance in highland could be achieved within limits through selection at lowland, and the most appropriate culling/selection of the disease is basically at the lowland where virus and disease pressures are high.

Keywords: Characterization, genetic variance, cassava seedlings, lowland, highland, root yield performance