High yielding sorghum and closing the yield gap

Author: Loretta Serafin and Guy McMullen, NSW DPI, Tamworth | Date: 21 Jul 2016

Take home message

  • Time of sowing can have large impacts on final crop yield, however being able to identify when to plant depending on the season is not possible.
  • Nitrogen nutrition had the largest impact on crop yield across the three seasons.
  • Hybrid performance varied between seasons and sowing times, however there was no consistent pattern of which hybrid yielded the best.

Introduction

The Liverpool Plains is one of the highest yielding environments for the production of grain sorghum in Australia, making it an extremely important crop in the rotation due to the profit margins which result. The high yields result from in the environmental conditions which the crop is exposed to, in combination with the soils and agronomic management.

While, average yields are high compared to other sorghum production zones, there is a question over whether current yields are fully meeting their potential. There is also a need to determine the contributions that each of the agronomic decisions have on sorghum yields so that their relative impact on the gross margin can be attributed.  

In this research six key agronomic practices have been investigated; row spacing, plant population, hybrid selection, nitrogen (N) and phosphorus (P) nutrition and time of sowing at two locations on the Liverpool Plains during the growing seasons of 2013-14, 2014-15 and 2015-16.

In each season, trials were conducted at two locations, the NSW DPI Research Station at Breeza and in commercial sorghum paddocks at Pine Ridge (2013-14), Willow Tree (2014-15) or Premer (2015-16) in the respective seasons.

In this paper, results from the three years of trials from the Breeza site will be discussed as this provides a data set over three consecutive years at the same location. The 24 treatment combinations used in these trials were partially factorial (Table 1). The data from a subset of ten of these treatments is presented here (Table 2) as final analysis is ongoing.

The trial treatments were designed around a stepped approach of segmenting the key factors starting with a high input or “Rolls Royce” treatment option (shaded in Table 2; treatment 24) which used the top rates of nitrogen (200 kg /ha) and phosphorus (20 kg/ha) in combination with a recent released hybrid; MR Scorpio. This was then compared to a range of contrasting treatments where only one factor was changed each time; for example less nitrogen, down to a “basic” version (Table 2; treatment 3) which used an old hybrid, MR Buster, with no additional crop nutrition.  

The data generated from this complex set of interactions is invaluable in partitioning yield and ultimately economic return from crop inputs.

Table 1. Treatment combinations

Agronomic Factor

Options included in trial

TOS 1

TOS 2

Time of sowing (TOS)

2013-14

2014-15

2015-16

31st October

12th November

14th October

10th December

20th December

3rd December

Row spacing

Normal row: 90 cm (2013-14 only) then 100 cm (2014-15 and 2015-16) or

Twin rows (two rows 7.5cm apart)

Plant population

50, 75 and 100,000 plants/ha

Hybrid

MR Buster, MR Scorpio and 85G33

Nitrogen

0, 100 and 200 kg N/ha or a 100:100 split at 6-8 leaf stage applied as Urea, side banded at sowing or spread on the surface for the split application.

Phosphorus

0, 10 and 20 kg P/ha with seed at planting applied as Triple Superphosphate

Breeza 2013-14 Results

In the 2013-14 season, two trials were run at Breeza on 1.8 m wide raised beds. Both trials were pre-irrigated to ensure a full profile at the start of the season and then one trial was treated as dryland and the second trial was irrigated; receiving two in crop irrigations.

In this season, across all treatments there were only small differences as a result of varying the time of sowing in both the irrigated and the dryland trial, with TOS1 yielding between 0.24 (dryland) and 0.35 (irrigated) t/ha more than TOS2. Differences for the individual treatment combinations varied more.

The largest impact on crop yield was from reducing the nutrition inputs (Table 2), when comparing the high input treatment number 24, using MR Scorpio at a plant population of 50,000 plants/ha in combination with 200 kg of N/ha and 20 kg P/ha with treatment 1 which had MR Scorpio at the same population but no applied nitrogen or phosphorus. The reduction in yield was 2.62 and 2.73 t/ha for TOS 1 and 2 respectively in the irrigated trial and 1.62 and 1.11 t/ha for TOS 1 and 2 respectively in the dryland trial.

If we compare these results to treatment 2 with zero nitrogen and 10 kg P/ha we can see that the majority of the yield is coming from the application of the N as opposed to the P. This is reinforced by comparing treatment 23 with 24, where only the P has been omitted and the N rate has remained at 200 kg N/ha. In both trials and sowing times we see a much smaller variation in the yield of less than 0.4 t/ha, further reinforcing the benefit of the nitrogen and the minimal impact of phosphorus at this site.

The value of utilising new hybrids has been compared in these trials through the use of MR Buster as an old hybrid, and MR Scorpio and 85G33 as new hybrids. If we compare similar treatments, with the exception of changing the hybrid such as treatments 10, 5 and 8 we find small differences of up to 0.5 t/ha between hybrids but no consistent pattern.

The comparison of the “Rolls Royce” treatment 24 with the “basic” treatment 1 showed a 0.98 – 2.89 t/ha difference in yield, validating the major improvements in yield which can be obtained from getting all of the crop agronomics right.

Breeza 2014-15 Results

In the 2014-15 season, only a dryland trial was conducted at Breeza, on 2 m raised beds. The beds were pre-irrigated to provide a full profile and then remained dryland. The two times of sowing were planted 12th November for TOS 1 and 20th December for TOS 2 and resulted in a large difference in yield, with TOS 1 on average yielding 6.46 t/ha versus 2.57 t/ha for TOS 2. The late planting suffered under cold conditions during grain fill and high midge pressure. As a result of these combined stresses on the crop in TOS 2 there were very few differences of note between the different treatments.

Large differences in yield between treatments were observed in TOS 1. The largest impact on yield was obtained from reducing the nitrogen nutrition to the crop. The “Rolls Royce” treatment 24 provided the highest yields in TOS1 with 200 kg N/ha and 20 kg P/ha. There was a negligible impact on yield when the P was omitted in treatment 23 but the nitrogen rate was maintained. In contrast, when the N rate was reduced to 0 but 10 kg of P was retained as in treatment 2 the yield was reduced by 1.70 t/ha. Reducing both the P and N to zero (treatment 1) resulted in a larger reduction in yield of 2.72 t/ha.

The yield performance of the different hybrids, comparing treatments, 5, 8 and 10 was much larger in the 2014-15 season, than in the 2013-14 season. Differences of 1.3-1.4 t/ha were found between hybrids when all other factors were kept the same but again the pattern of which hybrid performed the best was not consistent across the sowing times, with MR Buster being the highest yielding in TOS 1 and MR Scorpio in TOS 2.

Comparing the “Rolls Royce” treatment 24 with the “basic” treatment 1 resulted in a 0.84 t/ha yield difference in TOS 1. The differences in TOS 2 were negligible for reasons discussed earlier.

Breeza 2015-16 Results

In the 2015-16 season a dryland trial was conducted at Breeza, on 2 m raised beds. The beds were pre-irrigated to provide a full profile and then remained dryland. The two times of sowing were planted on the  14th October for TOS 1 and 3rd December for TOS 2 and resulted in a large difference in yield, with TOS 1 on average yielding 3.21 t/ha compared to 6.20 t/ha for TOS 2.

Differences between treatments at both times of sowing were much smaller in the 2015-16 season than in previous seasons (Table 2). However, similar to previous seasons the largest impacts were obtained from reducing both the nitrogen and phosphorus nutrition. The “Rolls Royce” treatment (24) yielded 0.44 and 0.69 t/ha more than the base MR Scorpio treatment (1).

Varying the hybrid selection (treatments 5, 8 and 10) resulted in yield differences of 0.30 – 0.65 t/ha however in line with the other seasons, there was no consistent pattern.

Incorporating the use of twin rows, as in treatment 15 compared to treatment 5, had a larger positive impact on yield with the later sowing time, than it did in TOS 1; 0.54 compared to 0.22 t/ha respectively.

Table 2. Grain sorghum yields from Breeza 2013-2016
 Breeza Sorghum Yields 2013-2016 Irrigated Dryland Dryland Dryland
Trt No.
Hybrid 
 
Plant
population
('000/ha)
Applied
N
(kg/ha)
Applied
P
(kg/ha)
2014
Yield
(t/ha)
Yield
Gap
2014
Yield
(t/ha)
Yield
Gap
2014
Yield
(t/ha)
Yield
Gap
2014
Yield
(t/ha)
Yield
Gap
2015
Yield
(t/ha)
Yield
Gap
2015
Yield
(t/ha)
Yield
Gap
2016
Yield
(t/ha)
Yield
Gap
2016
Yield
(t/ha)
Yield
Gap
 TOS 1 TOS 2 TOS 1 TOS 2 TOS 1 TOS 2 TOS 1 TOS 2
 24 MR-Scorpio   50 200  20  6.56   6.69   3.83   3.99   7.42   2.62   3.19   6.22  
 23  MR-Scorpio
 50 200 0 6.83 0.27 6.30 -0.39 4.02 0.19 3.73 -0.26 7.29 -0.13 2.64 0.03 3.28 0.09 6.07 -0.15
 14  MR-Scorpio
50 100 0 4.52 -2.04 5.96 -0.73 2.97  -0.86 3.45 -0.54 6.75 -0.67 2.56 -0.06 3.19 0.00 6.52 0.30
 5  MR-Scorpio
50 100 10 5.64 -0.92 5.66 -1.03 3.01 -0.82 3.38 -0.61 6.10 -1.32 2.81 0.19 2.86 -0.33 5.72 -0.50
 10 MR-Buster 50 100 10 5.60 -0.96 5.36 -1.33 3.52 -0.31 3.24 -0.75 7.34 -0.08 2.32 -0.30 3.52 0.33  5.74 -0.48
 8 85G33 50 100 10 5.06 -1.50 5.35 -1.34 3.37 -0.46 3.57 -0.42 5.93  -1.49 1.54 -1.08 2.88 -0.31 6.01 -0.21
 2 MR-Scorpio 50 0 10 4.01 -2.55 3.80 -2.89 2.35 -1.48 2.87 -1.12 5.72  -1.70 2.44 -0.18 2.86 -0.33 5.72 -0.50
 1 MR-Scorpio 50 0 0  3.94 -2.62 3.96 -2.73 2.21 -1.62 2.88 -1.11 4.70 -2.72 2.12 -0.50  2.75 -0.44 5.53 -0.69
 3 MR-Buster 100 0 0 4.60 -1.96 3.80 -2.89 2.85 -0.98 2.51 -1.48 6.04  -1.38 2.49 -0.13 3.66 0.47 6.01 -0.21
 15 MR-Scorpio (Twin rows) 50 100 10 5.70 -0.86 5.29 -1.40 3.37 -0.46 2.91 -1.08 6.58  -0.84 2.56 -0.06 3.08 -0.11 6.26 0.04

Conclusion

Over the three years of research, the most important agronomic inputs for sorghum yield have been related to time of sowing and nitrogen nutrition. In dryland situations the most difficult aspect of selecting whether to sow early (as in TOS 1) or late (TOS 2) is picking when the seasonal conditions will be most favourable. As seen in this data set yield can vary dramatically between years and by a large or quite small amount depending on seasonal conditions.

Nitrogen nutrition has been consistently shown to be an important driver of final grain yield. Often the combination of nitrogen and phosphorus has proven beneficial even at this site, which is not considered to be P responsive.

While hybrid selection plays an important role around many plant characteristics such as maturity, midge resistance and lodging, in these trials there were variable and inconsistent responses to using newer released hybrids compared to the old genetics of MR Buster. This reinforces that while genetic gain is important, much larger benefits can be obtained from optimising crop agronomy.

The “Rolls Royce” treatment in this trial has demonstrated that further yield gains are likely if additional inputs, particularly in the form of nitrogen and phosphorus nutrition are applied to grain sorghum. The question which needs to be answered is how economic is it to chase these higher potential yields.

While the field trial component of this project has finished, the importance of finalising the interpretation of results from these trials and the economic impacts of these treatments cannot be overlooked. This paper has focused on obtaining the highest yields from grain sorghum; however the comparative gross margins from these trials will ultimately provide the most powerful insight into where investment should be placed for the best outcomes.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC and NSW DPI, the author would like to thank them for their continued support. In particular, Ian Carter “Connamarra” Pine Ridge, James Arnott “Berwicks” Willow Tree and Ed & Fiona Simson “The Plantation” Premer.

Thanks to NSW DPI staff;  Mark Hellyer, Peter Perfrement, Bronwyn Brennan and Delphi Ramsden for their technical assistance in the conduct of these trials and to Scott Goodworth and Steve Gengos, LPFS Breeza for their assistance. Thanks also to Pioneer and Pacific Seeds for their supply of trial seed.

Contact details

Loretta Serafin & Guy McMullen
NSW Department of Primary Industries, Tamworth
4 Marsden Park Rd, Calala
Ph: 02 67 63 1100
Fx: 02 67 63 1222
Email: Loretta.serafin@dpi.nsw.gov.au or guy.mcmullen@dpi.nsw.gov.au