Effects of different water regimes on field-grown determinate and indeterminate faba bean (Vicia faba L) .2. Yield, yield components and harvest index

Faba bean yields are highly sensitive to variations of water availability. For indeterminate faba bean, high levels of water could promote excessive vegetative growth at the expense of pod growth. Therefore, the present study was undertaken with the objective of comparing the yield and harvest index of determinate and indeterminate faba bean under different water regimes and finding how the balance between vegetative and reproductive growth responds to variations in water availability. Two field experiments were carried out in 1989 and 1990 at Reading, United Kingdom, using the determinate cultivar 'Tina' and the indeterminate cultivar 'Gobo'. Rainshelters and trickle irrigation were used to establish four different water regimes: unsheltered and well-irrigated (W-1); sheltered at flowering and subjected to post-flowering water stress (W-2); sheltered throughout the season and subjected to continuously increasing water stress (W-3); open rainfed (W-4). High water availability (W-1) promoted vegetative growth and decreased the harvest index (HI). The decreases in HI were more pronounced in the indeterminates with reductions up to 48% as compared to 25% in the determinates. Except for the indeterminate W-1 in 1989, final pod yields were higher in W-1 (4.6-7.6 t/ha) than the rest (1.4-5.4 t/ha) in both cultivars. Under irrigation, the indeterminates required a longer post-flowering duration (Di) to produce high pod yields because irrigation promoted post-flowering vegetative growth and delayed the onset of pod-filling. Consequently, in the shorter 1989 season, the indeterminate W-1 produced a lower pod yield than the rest whereas the opposite occurred in the longer 1990 season. In the water-stressed treatments, final pod yield was determined by the LAI at the onset of pod-filling, with higher values resulting in higher pod yields. Pod growth rate was maximal around a critical LAI of 4. Except in the 1989 indeterminate W-1, there was a positive linear relationship between pod yield and post-flowering leaf area duration. Seed yield was correlated with the numbers of pod-bearing nodes and pods per m(2) in both cultivars. No significant correlations were found between yield components, thus ruling out the presence of yield compensatory mechanisms in these cultivars. It is recommended that for the environment where this experiment was done, the determinates are preferable for late-spring plantings under irrigation. For early-spring plantings under irrigation and for both early- and late-plantings under rainfed conditions, indeterminates are more suitable. (C) Elsevier Science B.V.