Kernel set in a maize hybrid and its two parental inbred lines was examined across a range of water-stress levels during the period bracketing silking. Kernel set was greater in the hybrid than in the inbred lines under all levels of water stress. Differences in kernel set between the hybrid and its parental inbred lines were attributable to a greater kernel number per unit plant growth rate and not to differences in plant growth rate. Greater kernel set per unit plant growth rate was attributable mainly to kernel set per unit ear growth rate (i.e., kernel set efficiency) and to a lesser extent to dry matter partitioning to the ear.
Greater kernel set of a hybrid compared to its parental inbred lines is attributable to kernel set per unit plant growth rate around silking.
Zea mays L.
Heterosis for grain yield in maize has been associated with four physiological processes: (i) Heterosis for leaf area index (LAI) due to increased leaf size, resulting in increased light interception and dry matter accumulation of hybrids. (ii) Heterosis for ‘stay green’, which in addition to heterosis for maximum LAI results in increased light interception and dry matter accumulation during the grain-filling period. (iii) Heterosis for sustaining photosynthesis of green leaf area during the grain-filling period, resulting in increased canopy photosynthesis during this period. (iv) Heterosis for harvest index due, in part, to a heterosis for kernel number (Tollenaar et al., 2004). The physiological mechanisms underlying the heterosis of kernel set and its response to stress are not known.
Kernel number per plant (KNP) is associated with plant growth rate during a critical period bracketing silking (PGRs) (Tollenaar et al., 1992). In maize, the KNP-PGRS relationship has been described by two successive curves to account for the first and second ear in prolific, or a single curve in non-prolific plants (Tollenaar et al., 1992). According to this model, differences in kernel set in hybrids compared to the parental inbred lines could be related to PGRS, kernel set per unit PGRS (KNP/PGRS), or both.
In addition, differences between hybrids and their parental inbred lines in kernel set per unit PGRS may be attributable to either or both dry matter partitioning to the ear and kernel set efficiency (i.e., kernel set per unit of ear growth rate) during the period bracketing silking (Andrade et al., 1999).
In this work we studied physiological processes underlying kernel set in a hybrid and its two parental inbred lines.
An experiment was conducted in 2003 at the Arkell Research Station near Guelph, Ontario (43º39’ N, 80º25’ W and 375 m above sea level), using a hydroponic system in the field (Tollenaar and Migus, 1984). During the whole life cycle, except for ~20 days around silking, pails were watered four times a day using a nutrient solution as described by Tollenaar (1989). Three seeds were planted on 21 May 2003 and seedlings were thinned at the 3-leaf stage to two plants per pail. Plant density was 60,000 plants ha-1. Treatments consisted of one hybrid (CG60 x CG102) and its two parental inbred lines and of three water regimes (100 (=control), 75 and 60% of water availability relative to the control). The experimental design was a split plot with three replications. The main plot was genotypes and sub plot was water regime.
Control pails were maintained water-replete at all times. Irrigation to pails assigned to water stress treatments was stopped at 3 to 8 days before silking. At that moment, all pails were covered with plastic bags at the bottom of the plants to avoid the entrance of water from rain. Watering regimes started when wilting of the ear leaf was noticed at noon. Full irrigation to all the pails was reestablished at silking+13 days of each genotype. To determine the pail water holding capacity (Pc), 30% of the control pails of each genotype were watered to excess and were weighed immediately after the drainage from the bottom of the pails finished. The Pc was determined at the beginning and every 5 to 7 days during the period of water stress treatments. To determine the water used by control plants, control pails were watered to excess in the morning and were weighed after 24 h. (= P+1) every day during the period in which the water stress treatment was performed. The water used by control plants during one day was estimated as the difference between Pc and P+1. Treatment pails were supplied with 75 or 60% of the water used by control plants
Shoot biomass of tagged plants was quantified at approximately 3-8 d before and 13 d after silking through a combination of destructive and non-destructive sampling (see below). KNP was determined at maturity. Silking was recorded for each treatment, as the date when 50% of the plants presented at least one emerged silk from the husks.
Destructive sampling. Morphometric variables, i.e., basal stem diameter and diameter and length of the ear, were measured on 4 to 6 plants per replicate for each treatment. Immediately after measurements, plants were harvested. Plants were separated into stem plus leaves and tassel, and ears and were oven dried at 65º C until constant weight. Allometric relationships were established between morphometric variables and dry weights of shoot and ears. Ears included kernels and rachis. All equations were significant at P<0.05, and average R2 was 0.82.
Non-destructive sampling. Before silking, 12 to 18 plants per replicate for each treatment were tagged. Shoot and ear biomass were assessed for each tagged plant using the allometric relationships determined by destructive sampling.
Growth rate during the critical period for kernel set was estimated as the ratio between accumulated biomass in shoots or ear and the duration of the period, assuming ear biomass to be negligible at 10 d before silking. An estimate of the threshold PGRS for kernel set was obtained from the mean plant growth rate of plants that had a kernel set between 1 to 15% of the maximum KNP; maximum KNP was defined as mean KNP for plants with PGRS ≥ 4 g d-1. A linear model was fitted to the relationship between ear growth rate during a period bracketing silking (EGRS) and PGRS (EGRS = a + b * PGRS) in order to determine the minimum PGRS for dry matter partitioning to the ear (PGRSP = a/b). T-tests were used to assess differences between genotypes for intervals of PGRS and EGRS.
Kernel number per plant was greater in the hybrid compared to its parental inbred lines at all water availability levels (P<0.05, Fig. 1A), but PGRS of the hybrid was not consistently greater than that of its parental inbred lines (Fig. 1B). Therefore, the greater KNP of the hybrid was not related to greater PGRS.
Figure 1: (A) Kernel number per plant and (B) plant growth rate during a period bracketing silking as a function of water availability relative to the control, for a hybrid (CG60 x CG102) and its parental inbred lines (CG60 and CG102). Bars indicate the standard errors.
The relationship between KNP and PGRS was curvilinear with a significant (P<0.05) threshold of PGRS for kernel set. Compared at equal PGRS, KNP was greater in the hybrid than in the inbred lines at all PGRS (P<0.05, Fig. 2). Therefore, a greater kernel set per unit PGRS underlies the greater kernel number of the hybrid. The threshold PGRS for kernel set was similar for the hybrid and the parental inbred line CG60 (2.19 and 2.16 g d-1, respectively; P>0.05) and the threshold PGRS of the inbred line CG102 was lower than the former two genotypes (3.12 g d-1, P<0.05). At high resource availability per plant (i.e., PGRS ≥ 4 g d-1), the maximum KNP was higher for the hybrid (506±8.0) compared to its parental inbred lines (182±26.6 and 286±11.4 for CG102 and CG60, respectively). Hence, a greater kernel set per unit PGRS of the hybrid was indicated by a low threshold of PGRS for kernel set along with a high maximum KNP compared to its parental inbred lines.
Figure 2: Mean kernel number per plant at 1-g d-1 intervals of plant growth rate during a period bracketing silking for a hybrid and its parental inbred lines. Bars indicate the standard errors.
Ear growth rate during a period bracketing silking (EGRS) was greater in the hybrid than in its parental inbred lines at PGRS < 4 g d-1 (P<0.05) and EGRS was greater in the hybrid than in CG60 (P<0.1) when compared at PGRS <2 (Fig. 3A). The minimum PGRS required to partition dry matter to the ear was 55% higher for CG102 (PGRSP = 1.97 g d-1) than for CG60 and the hybrid (PGRSP =1.01 and 1.16 g d-1, respectively), showing the greater relative importance of the ear as a metabolic sink within the internal hierarchy of the plant in the two latter genotypes. This phenomenon could be related to a lower dominance of the tassel over the ear (Echarte et al., 2004). In addition, the minimum PGRS required to partition dry matter to the ear was highly correlated with the PGRS threshold for kernel set (r=0.99, P=0.01). Therefore, a low minimum PGRS required to partition dry matter to the ear could be underlying a low PGRS threshold for kernel set. At high PGRS (i.e., PGRs ≥ 4 g d-1), greater EGRS was not associated with greater KNP across genotypes (Fig. 2 and Fig. 3A). The weak relationship between EGRs and KNP at high resource availability per plant could be related to morphogenetic limitations. The excess of assimilates in the ear that was not used to set more kernels was probably assigned to the cob or used to increase the potential kernel size.
Compared at equal EGRS, KNP was greater in the hybrid than in its both parental inbreds at all ranges of EGRS (P<0.05) (Fig. 3B). A greater kernel set efficiency in the hybrid than in the inbred lines could be associated with a lower minimum kernel growth required for kernel fixation, a greater synchrony of silk pollination, which in turn diminishes the competition among kernels in different position of the ear, a lower dry matter assigned to the cob and/or a greater potential kernel number.
Fig. 3: (A) Mean ear growth rate at 1-g d-1 intervals of plant growth rate during a period bracketing silking and (B) mean kernel number per plant at 0.25-g d-1 intervals of ear growth rate during a period bracketing silking, for a hybrid (CG60 x CG102) and its parental inbred lines (CG60 and CG102). Bars indicate the standard errors.
Kernel number per plant was greater in the hybrid than in its parental inbred lines under all levels of water stress. This was attributable to a greater kernel set per unit plant growth rate rather than to a greater plant growth rate during a period bracketing silking. The greater kernel set per unit plant growth rate in the hybrid compared to its parental inbred lines was attributable mainly to kernel set efficiency (i.e., kernel number per unit ear growth rate) and to a lesser extent to dry matter partitioning to the ear (i.e., ear growth rate per unit plant growth rate).
Andrade FH, Vega C, Uhart S, Cirilo A, Cantarero M and Valentinuz O (1999) Kernel number determination in maize. Crop Science 39, 453-459.
Echarte L, Andrade FH, Vega CRC and Tollenaar M (2004) Kernel number determination in Argentinean maize hybrids released between 1965 and 1993. Crop Science 44, (5) In press.
Tollenaar, M (1989) Response of dry matter accumulation in maize to temperature. I. Dry matter partitioning. Crop Science 29, 1239-1246.
Tollenaar M, Ahmadzadeh A and Lee EA (2004). Physiological basis of heterosis for grain yield in maize. Crop Science. 44, In press.
Tollenaar M, Dwyer LM and Stewart DW (1992). Ear and kernel formation in maize hybrids representing three decades of grain yield improvement in Ontario. Crop Science 32, 432-438.
Tollenaar M and Migus W (1984) Dry matter accumulation of maize grown hydroponically under controlled-environment and field conditions. Canadian Journal of Plant Science 64, 475-485.