Environmental control of tillering and yield of wheat

Abstract

Four experiments were conducted in controlled and field environments from 1982 to 1984 to establish relationships between dry matter growth rate and tillering, yield components and grain yield of wheat.

In the first experiment two semi-dwarf cultivars, Oroua and Karamu, were grown in pots in a glasshouse and subjected to drought during four phases of development: emergence to the six leaf stage (6L); 6L to flag leaf emergence (FL); FL to ear emergence (EE), and EE to 10 days after anthesis. The number of stress days (SD) accumulated during each drought period was calculated from measurements of evapotranspiration of the stressed and the well-watered treatments and the duration of the drought in days.

Drought at any phase of development decreased green surface area of mainstems to 60-80%, absolute growth rate per plant to about 60-80% and consequently dry matter growth to 75% of the well-watered control. Grain and straw yield decreased to about 75% of the well-watered control, although the early droughts, which accumulated slightly more SDs, depressed yield slightly more than the late droughts. No moisture sensitive periods were evident. Drought decreased yield by decreasing ears per plant, grains per ear and weight per grain. Karamu and well-watered control plants used water slightly more efficiently than Oroua and drought-stressed plants. There was no interaction between the cultivar and drought treatments.

In the second experiment the growth, development and tillering of Oroua and Karamu wheats were studied in the field. Plant growth rates were varied by using low (150 m⁻²) and high (600 m⁻²) plant populations, and by applying either 0 or 100 kg N ha⁻¹. At the start of the tiller death phase (Zadoks(Z)45), plant growth rates were further varied by thinning alternate rows and by spraying a photosynthesis suppressant (DCMU) at 875 g ha⁻¹ in 15 l water.

Oroua produced more tillers per plant, and hence 7% more ears, than Karamu. However, this effect was counteracted by 23% more grains per ear in Karamu and as a result there was no significant difference in grain yield between the cultivars. Plants growing at the low population produced 140% more ears per plant and 25% more grains per ear thus increasing the grain yield per plant by 190% over the plants at the high population. Nitrogen increased ears per plant by 18%, but grains per ear and weight per grain remained unchanged. Consequently, grain yield per plant increased by 15%. Tiller appearance at Z30 was correlated with available tillering sites at Z20, and number of shoots per plant was correlated with dry matter per plant at Z30. Thinning alternate rows increased both ears and grain yield per plant by 11%. DCMU decreased ears per plant by 24%, grains per ear by 26% and weight per grain by 19%. As a result grain yield per plant was 44% less than the control.

In the third experiment Oroua was grown in small pots in growth cabinets with 0, 50, 100 and 150 mg N applied per pot. Every additional increment of nitrogen caused an increase in the nunber of tillers, green surface area and dry matter per plant, and delayed anthesis and maturity by 3-7 days. Every additional 50 mg N up to 150 mg caused an increase of two ears per plant. Grains per ear showed a quadratic response to N, reaching a maximum at 100 mg N per pot. Weight per grain increased linearly with applied N, although it failed to reach levels typically achieved in the field. The response of grain yield to N was both linear and quadratic.

In the fourth experiment Oroua was raised in pots in growth cabinets. The plants were exposed to either dim (120 µmol m⁻² s⁻¹) or bright (450 µmol m⁻² s⁻¹) light, and groups of plants were subjected to opposite bright or dim light exposures during specific developmental phases. Exposing brightly-lit plants to dim light decreased yield by 11-30%. The decrease was associated with 17-25% fewer ears per plant and 6-23% fewer grains per ear. Conversely, grain yield was increased when dimly-lit plants were exposed to bright light. There was a 32-55% increase in ear numbers in plants exposed from emergence to 3L and 3L to 6L stages. Grains per ear increased by 11-39% when dimly-lit plants were exposed to bright light from the flag leaf stage onwards. Weight per grain remained unchanged.

In the third and fourth experiments in the growth cabinets many observations were similar. Tiller appearance followed a Fibonacci series of leaf appearance of MS, the first tiller being appeared in all the treatments at leaf number interval (LNI) 3. Tiller appearance started to decline from the potential rate at the seven leaf stage in most of the treatments. At higher N levels and in dim light tiller appearance continued until ear emergence. In all treatments, mainstems (MS) and older tiller cohorts formed type I (Deevey, 1947) curves and the younger cohorts type III. Nitrogen and light intensity treatments were important in forming type II curves. Mainstems and older cohorts also showed a significant hierarchy in accumulating assimilate from a common source with MS> cohort 1> cohort 2. At the start of the stem extension phase, at Z30, tillers with less than 3 leaves or 75-100 mg dry weight senesced without producing ears.

In all four experiments the number of shoots per plant at Z30 was highly correlated with dry matter per plant. The number of ears per plant was highly correlated with the increase in dry matter per plant from Z30 to Z55, and the number of grains per ear was highly correlated with the increase in dry matter per culm from 20 to 10 days before anthesis. Weight per grain was mainly determined by dry matter growth per grain from anthesis to maturity, and any shortage of assimilate required to reach maximum grain size in any specific environment was compensated by the translocation of pre-anthesis stem reserves.