Previous PageTable Of ContentsNext Page

Can large kernel size increase grain yield in sorghum?

Zongjian Yang 1, Graeme Hammer 1, Erik van Oosterom 1, David Jordan2 and Alistair Doherty3

1 Agricultural Production Systems Research Unit, School of Land, Crop and Food Sciences, The University of Queensland, Brisbane, Qld 4072, Australia. Email:
Hermitage Research Station, Department of Primary Industries and Fisheries, Warwick, Qld 4370, Australia.
Queensland Department of Primary Industries and Fisheries, Toowoomba, Qld 4350, Australia.


The large seeded sorghum line KS115, despite its low kernel number, provides a promising source for increased grain yield of elite sorghum germplasm, as it has a long kernel filling duration (KFD). The aim of this research was to determine whether the promising grain yield of KS115-derived germplasm was associated with a longer KFD or a kernel growth rate greater than expected for the reduced kernel number. A hybrid of KS115 (ATx642/KS115) and a normal-seeded hybrid (ATx642/RQL36) were grown in two field experiments at Gatton, Queensland. Panicles were sampled regularly to measure kernel volume, water content and dry mass during the kernel filling period. Results showed that the hybrid based on KS115 produced significantly larger kernels and out-yielded the normal-seeded hybrid ATx642/RQL36 in both experiments. High kernel mass of the large-seeded hybrid was associated with both an extended duration of kernel filling and a higher rate of dry mass accumulation. As the higher rate of kernel filling merely compensated for a reduced kernel number, prolonged kernel filling duration, rather than the high filling rate of individual kernels, is a major cause for the promising grain yield of the large-seeded hybrid.

Key Words

Kernel filling, kernel growth rate.


Kernel size is an important factor determining final grain yield and nutritional quality in sorghum (Kriegshauser et al. 2006). A large-seeded sorghum line KS115, which was developed at Kansas State University (Tuinstra et al. 2001b), has been used as a gene source for improving kernel size and quality in sorghum breeding programmes (Kriegshauser et al. 2006). Although high kernel mass is offset by low kernel number, hybrids based on this material seem to have higher grain yield (Tuinstra et al. 2001a; D. Jordan, pers. comm.). High kernel mass of KS115 was associated with both a high kernel filling rate and a long kernel filling duration (Tuinstra et al. 2001b). The physiological mechanism that causes the genotypic variation in kernel filling rate and duration is still not known and uncertainty remains as to the relative importance of these kernel filling characteristics in the determination of final grain yield. In this study, the kernel filling performance of a hybrid based on KS115 was compared with a normal-seeded hybrid with common female parent. The aim was to identify the kernel-filling characteristics associated with kernel size that could improve grain yield in sorghum.

Materials and Methods

As part of a series of field experiments to understand kernel size determination in sorghum, two hybrids (ATx642/KS115 and ATx642/RQL36) were sown on 17 January 2006 and 14 November 2006 at the Gatton field station, Queensland, Australia. Male parent KS115 produces exceptionally large kernels whereas RQL36 is a normal-seeded line. A randomised block design with three replicates was used in both experiments. Each plot consisted of four rows, 1 m. apart and 15 m. long. Plots were over-sown and thinned to a density of 5 plants m-2 (50,000 plants ha-1) two weeks after sowing. Approximately 30 main shoot panicles in each plot were tagged and date of apical and basal anthesis was recorded for each labelled plant. Anthesis date was recorded when 50% of spikelets on the panicle had extruded anthers. Labelled panicles were harvested regularly throughout the kernel filling period. Fifty kernels were randomly sampled from each of the apical and basal portions of the harvested panicles to determine their volume, fresh weight and dry weight. Kernel volume was measured by displacement of water in a pipette.

Using the R2LINES procedure of the Genstat Release 9.1 statistical package, kernel dry weight data were fitted to a broken linear model as a function of growing degree days after anthesis. This model comprises an initial positive linear increase followed by a phase of constant dry weight. Physiological maturity was defined as the time when the two lines intersected, and KFD was calculated as the thermal time from anthesis to physiological maturity. As linear kernel filling starts a few days after anthesis, care was taken to fit the model to observations after the linear phase had started, i.e. kernel weight had reached at least 10% of its final value (approximately 130Cd after anthesis). The mean rate of kernel growth during the effective filling period was estimated from the slope of the linear relationship between kernel dry weight and thermal time. Degree-days during the period were calculated by summing daily degree-days. Daily degree-days were calculated from hourly temperatures by adjusting for a base temperature of 5.7C and an optimal temperature of 23.5C (Hammer and Muchow 1994). Analysis of variance was performed using Genstat 9th Edition. Main effects of genotypes, treatments and their interactions were tested using Fisher’s LSD method.

Results and Discussion

Final kernel dry weights estimated from the broken linear model and measured in the bulk harvest were both characterized by significant difference between genotypes (Table 1, Figure 1). There was a trade-off between kernel size and number. Hybrid ATx642/KS115 produced larger kernels, but kernel number per plant was significantly reduced compared to the normal-seeded contemporaries (Table 1). The decrease in kernel number in ATx642/KS115 was associated with reduced panicle branching.

Table 1. Kernel dry weight, number and yield for the two contrasting hybrids in the two field experiments.



Kernel dry weight
(mg kernel-1)

Kernel number

Grain Yield
(g plant-1)










LSD (5%)













LSD (5%)




*Within column of each experiment, means followed by a different letter are significantly different (P < 0.05).

The large-seeded hybrid (ATx642/KS115) out-yielded the normal-seeded hybrid (ATx642/RQL36) in both experiments (Table 1). In the first experiment, grain yield of ATx642/KS115 was 10.4% higher than that of ATx642/RQL36. In the second experiment, the yield advantage of ATx642/KS115 was even greater than in the first experiment, but that was partly due to infection by sugarcane mosaic virus (SCMV) which caused premature leaf senescence especially for ATx642/RQL36.

Figure 1. Kernel dry weight versus thermal time after anthesis for the two hybrids studied in Exp 1. Vertical lines indicated the time of physiological maturity for ATx642/RQL36 (472 oCd) and ATx642/KS115 (532 oCd).

Mean kernel growth rates during the effective filling period were estimated from the slopes of the linear growth phase (Figure 1). The rate of kernel growth for the large-seeded hybrid ATx642/KS115 was significantly greater than that of ATx642/RQL36 in both experiments. Kernel growth rates were higher in the first experiment than in the second experiment for both hybrids, possibly because of the occurrence of SCMV in the second experiment. Genotypic variation for KFD was present (Table 2). The large-seeded hybrid (ATx642/KS115) consistently had a longer period of kernel filling than ATx642/RQL36 in both experiments. This indicated that large kernels of ATx642/KS115 were achieved by both an extended duration of kernel filling and a higher rate of dry mass accumulation.

The greater filling rate of per kernel for ATx642/KS115 was compensated by a lower kernel number. As a consequence, kernel filling rates at the whole plant level were similar between hybrids (Table 2). The yield advantage of the large-seeded hybrid (ATx642/KS115) was mainly caused by an extended kernel filling period. In the first experiment, the KFD for ATx642/KS115 was 60.3Cd longer than that of ATx642/RQL36, which would result in 21.8 g of extra kernel mass per plant for ATx642/KS115 (17.7% increase). In the second experiment, 32.4 g per plant of extra kernel mass would be produced in ATx642/KS115 (35.1% increase) due to an extension of kernel filling by 117.4Cd compared to ATx642/RQL36. The results suggested that KS115 could be a good source of genes to increase kernel size and improve grain yield of sorghum, predominantly through the longer KFD. The association between kernel-filling duration and grain yield has been studied for a variety of crops. In several experiments, genetic variation in kernel-filling duration has been found to be positively related to yield in different cereals including maize (Daynard and Kannenberg 1976), wheat (Gebeyehou et al. 1982) and barley (Leon and Geisler 1994). Increase in grain yield in maize and soybean was achieved by selection for longer kernel-filling period (Cavalieri and Smith 1985; Smith and Nelson 1986).

Table 2. Kernel filling rate and duration ( standard error) determined from the broken linear model for the apical kernel of the two contrasting hybrids in the two experiments.



Kernel Filling Rate
(mg oCd-1 kernel-1)

Kernel Filling Duration

Kernel Filling Rate/PlantA (mg oCd-1 plant-1)



















AEstimated by multiplying total kernel number per plant by the filling rate of apical kernels of the main shoot panicle.


Hybrids examined in this experiment showed significant variation in final kernel dry weight. Greater kernel mass of a large-seeded hybrid was associated with both a high kernel filling rate and an extended duration of kernel filling. As there was a trade-off between kernel filling rate and kernel number, prolonged kernel filling duration, rather than increased filling rate of individual kernels, was the main cause for the higher grain yield of the large-seeded hybrid.


We would like to thank Kurt Deifel and Ian Broad for excellent technical assistance. Seeds used in this study were provided by Hermitage Research Station, DPI&F, Australia.


Cavalieri AJ, Smith OS (1985) Grain filling and field drying of a set of maize hybrids released from 1930 to 1982. Crop Science 25, 856-860.

Daynard TB, Kannenberg LW (1976) Relationships between length of actual and effective grain filling periods and grain yield of corn. Canadian Journal of Plant Science 56, 237-242.

Gebeyehou G, Knott DR, Baker RJ (1982) Rate and duration of grain filling in durum-wheat cultivars. Crop Science 22, 337-340.

Hammer GL, Muchow RC (1994) Assessing climatic risk to sorghum production in water-limited subtropical environments. I. Development and testing of a simulation model. Field Crops Research 36, 221-234.

Kriegshauser TD, Tuinstra MR, Hancock JD (2006) Variation in nutritional value of sorghum hybrids with contrasting seed weight characteristics and comparisons with maize in broiler chicks. Crop Science 46, 695-699.

Leon J, Geisler G (1994) Variation in rate and duration of growth among spring barley cultivars. Plant Breeding 112, 199-208.

Smith JR, Nelson RL (1986) Selection for seed-filling period in soybean. Crop Science 26, 466-469.

Tuinstra MR, Kriegshauser TD, Vanderlip RL, Kofoid KD, Hancock JD (2001a) Can long gain-filling duration improve yield potential and grain quality of sorghum? In Proceedings of the 56th Corn and Sorghum Research Conference. American Seed Trade Association, Chicago. pp. 185-195.

Tuinstra MR, Liang GL, Hicks C, Kofoid KD, Vanderlip RL (2001b) Registration of KS 115 sorghum. Crop Science 41, 932-933.

Previous PageTop Of PageNext Page