RESPONSE OF RAIN-FED BREAD AND DURUM-WHEAT TO SOURCE, LEVEL AND TIMING OF NITROGEN-FERTILIZER ON 2 ETHIOPIAN VERTISOLS .2. N UPTAKE, RECOVERY AND EFFICIENCY

In trials conducted at 2 highland Vertisol sites in Ethiopia in 1990 a nd 1991, 2 locally popular wheat cultivars, 1 spring bread wheat (Trit icum aestivum L.) and 1 durum wheat (T. durum Desf.), were supplied wi th nitrogen (N) fertilizer at 0, 60 and 120 kg N ha(-1) in the form of large granular urea (LGU), standard urea prills or ammonium sulfate. N was applied all at sowing, all at mid-tillering of split-applied bet ween these two stages (1/3:2/3). While durum wheat exhibited the highe st N concentration in grain and straw, bread wheat, because of its hig her productivity, resulted in a greater grain and total N uptake. In g eneral, split application of N and use of LGU as N source enhanced gra in and total N uptake, apparent N recovery and agronomic efficiency of N, particularly under severe water-logging stress. Where significant, the interactions among the experimental factors substantiated the sup erior responsiveness of the bread wheat cultivar to fertilizer N, and the beneficial effects of split N application and LGU as an N source. Split application of N tended to nullify the positive effects of LGU, presumably by approximating the delayed release of N achieved with LGU . Considering the potential benefits to Ethiopian peasant farmers and consumers, split application of N should be advocated, particularly on wafer-logged Vertisols; LGU could be an advantageous N source assumin g a cost comparable to the conventional N source urea.