|dc.description.abstract||Four field experiments were conducted between 1986 and 1989 with a two-row spring barley cultivar 'Triumph' in Lincoln, Canterbury, New Zealand, to investigate the causes of screenings and the measures for minimizing screening rate in barley. The relationships between kernel dimensions, and between screening rate and plant variables such as kernel yield and yield components were examined under various conditions of screening rate (1 - 46 %), mean kernel weight (24.7 - 53.2 mg) and kernel yield (3.1 - 10.1 t ha⁻¹) using treatments of sowing density, N fertiliser, defoliation, plant thinning, sowing date and drought.
The shortest dimension of kernel structure, depth, showed a strong linear relationship with width (r = 0.97) and kernel weight (r = 0.99). Mean kernel weight for a crop was consistently correlated with screening rate. The effect of kernel and ear populations on screening rate was variable and only significant for extremely high populations (18,000 kernels m⁻²) and with unfavourable kernel growth conditions. Under favourable kernel filling conditions, screenings came mainly from terminal spikelets irrespective of ear type. Under poor kernel filling conditions, screenings came from the basal and terminal parts of ears and most kernels from higher order tiller ears became screenings. The hierarchy and relative advantage in DM accumulation (growth potential) of kernels within an ear or among ears determined around anthesis were not modifiable by post-anthesis treatments such as anthesis N, defoliation and thinning. Conversely, post-anthesis treatments greatly affected the level of kernel filling.
Variation in weight among kernels was mainly attributable to the rate of caryopsis linear growth. The terminal kernels, mostly screenings, were further disadvantaged by having a shorter growth duration. The rate of kernel growth was closely associated with light intensity, while final kernel weight was strongly correlated with gross light energy available for the whole period of linear kernel growth. Likewise, variation in developmental rate produced by different sowing times was explained by variation in temperature and temperature adjusted for photoperiod, thus enabling a simple model to be used to account for the direct effects of those variables on growth duration and final kernel weight.
Kernel weight and screening rate were not stable crop characteristics and greatly modifiable by post-anthesis environments. Treatments imposed at different crop developmental stages with various severity consistently indicated that the critical period in determining screening rate and kernel weight was the post-anthesis period and kernel growth potential established around anthesis had little effect on screening rate and kernel weight. Growth limitation imposed during linear kernel growth increased screening rate significantly, while treatments imposed before linear growth usually decreased screening rate, presumably by eliminating weak tillers, ears and kernels. As the most important determinant of screening rate was the environment and agronomic management during the kernel growth period, it is suggested that agronomic management for low screening rate should aim to eliminate undue stresses to the crop during this period.||en