Response surfaces of semidwarf wheat for seeding rate, and levels and times of application of nitrogen fertiliser

Abstract A three factor central composite second order design in incomplete blocks was used to quantify the responses of semidwarf ‘Karamu’ wheat established at five seeding rates (170–580 seeds/m 2 ), and fertilised at five rates of N at the onset of tillering (0–124 kg/ha) and during spikelet growth (0–38 kg/ha). N applied at tillering raised grain yields by increasing spike populations and the mean number of spikelets per spike, but generally depressed mean grain dry weights and, at low rates, reduced grain content. Persistence of N in the soil after higher rates of application may have accounted for its linear responses in that it influenced more of the yield components. N fertiliser applied during spikelet growth tended to reduce spike populations, and its effects on other yield components depended on the level of N applied at early tillering. Thus it depressed spikelets per spike and mean grain weight when no N was applied at tillering and increased the values of these components when high rates were applied. Applications of N at tillering generally increased grain N content and grain N yield but had little influence on grain yield. Seeding rate increments modified components of yield but had little effect on grain yield or grain N content. Spike populations increased linearly with seeding rate but mean numbers of spikelets per spike declined especially at lower levels of N nutrition. At low seeding rates N appeared to be limiting grain set whereas at higher seeding rates assimilates appeared to be limiting. Thus the effects of seeding rate and N fertilisers at tillering were related to their common effect on spike populations.