Choosing the right wheat variety for the system

Choosing the right wheat variety for the system (part A)

Christine Zaicou-Kunesch, Mario D’Antuono, Karyn Reeves. Department of Agriculture and Food WA

Key messages

Mace dominates the area sown to wheat in WA due to its consistent high yield over a large range of environments. 

Avoid sowing Corack early in environments where staining is a risk.  

Magenta and Cobra perform better than Wyalkatchem in high rainfall environments BUT staining and sprouting susceptibility are risks.

Background

The challenge for WA growers is choosing the right wheat variety for a system.   Managing the risk factors and productivity within systems such as wheat on wheat and early sowing options is important.   Mace has become a dominant variety across the state contributing to 53% of the area sown in 2013 (source: CBH and published in Wheat Variety Guide for WA, 2014). Productivity and grain quality have driven the adoption of the variety.

DAFWA agronomy trials and national variety trials (NVT) funded by GRDC are able to provide information to support growers with their decisions on variety choice and management for their farming system.  The individual trial results provide an indication of a varieties performance at one location and one year.   Collation of this information over a number of years provides another level of information.  For example, the interrogation of the GRDC’s NVT database by Young (2013) enabled direct comparisons of varieties in an Agzone.  Comparisons of the yield of Mace to Wyalkatchem shows a consistent and stable advantage to Mace in a wide range of growing seasons from 2008 to 2012.   Kelly (2013) described cluster analysis as the most accurate prediction of relative yield performance of varieties for an environment.  Reporting on the outputs of clusters is expected to be available on the NVT online website.

A new method of analysis is proposed which will assist growers in an Agzone to choose a variety based on its performance relative to Wyalkatchem (and Mace with future analysis) if sown before or after the end of May.   A second level of analysis to assess performance based on rainfall, rotation and soil pH is also provided.  

Aims

The aim is to support growers with variety selection and management by providing:

  • An understanding of whether the performance of current and new varieties (in terms of predicted yield) in an Agzone could be explained by a set of factors comprising sowing day (# in a year), soil pH, rainfall, soil texture or rotation. 

Analysis will be based on National Variety Trial data and wheat agronomy research from WA.  The focus of this analysis will be Mace, Magenta, Corack and Cobra and recently released varieties. 

Method

Influence of factors (sowing date, soil pH, rainfall, soil texture or rotation) on variety performance

The GRDC national variety trials database (2007 – 2012) for wheat trials in WA and the site descriptions were provided by NVT online <http://www.nvtonline.com.au>.   The yield data and trial site details included location, sowing day of the year, previous crop and soil pH in the surface soil (0-10 cm) and at depth (10 - 60 cm).  Rainfall data was provided and modelled by Scott Chapman (CSIRO) and computed as the growing season rainfall (sowing to heading) for a mid-range cultivar for each trial.

Predicted yields for each NVT of each variety within each Agzone were described according to whether they were above (better) or below (worse) the predicted yield for Wyalkatchem.  Three analytical tests were undertaken. 

Logistic regression analysis was undertaken to test yield response of the variety compared to Wyalkatchem for the following factors

  1. day of sowing before or after 31st May (d150);
  2. subsoil soil pH (greater or less than 4.5); and
  3. rainfall low (less than 160, 188, 242, 139, 189 and 265mm in Agzones 1-6 respectively)
  4. rotation (Canola, Cereal and Legume - Lupin/pasture).

The regression coefficients were tested and provided an indication if the varieties predicted yields were superior or poorer compared to Wyalkatchem. Note: the factors (1-4) were considered separately.

The Fisher Exact Test (FET) was used to investigate if the varieties predicted yields in the NVT’s relative to Wyalkatchem were higher (better) or lower (worse) when sown early (<d150) or when sown late (>d150).   The ‘FET p-value’ in Table 1 is an indication of the significance of the responses in each Agzone.  If the p-value was less than 0.05, then the variety performed differently (better or worse) when sowing date was before or after the end of May (d150).

The binomial test was used to identify if a variety’s predicted yield response was better or worse than Wyalkatchem irrespective of sowing day.  The test was assessed in terms of odds of superior to inferior predicted yields compared to Wyalkatchem, and a binomial p-value was derived. If the binomial p-value was less than 0.05, then the variety performed better or worse than Wyalkatchem in each Agzone.  Refer to the columns ‘Compared to Wyalkatchem sown (≤d150)’ and ‘Compared to Wyalkatchem sown >d150’ to identify if the response was better or worse.

Results and discussion

Sowing day influence on performance

Sowing day did not influence the performance of Mace relative to Wyalkatchem.  Mace has performed better than Wyalkatchem when sown before or after d150 (Table1, Binomial test p-value <0.05) in all Agzones except Agzone 6.  In contrast, Cobra has performed better before d150 and worst than Wyalkatchem after d150 in Agzones 1 and 2 (FET test, p<0.05).  In Agzone 3 Cobra performed better than Wyalkatchem both before and after d150 (however there was a small sample size) (Binomial test, p<0.05). 

Corack yielded higher than Wyalkatchem when sown before or after d150 in all Agzones (Binomial test, p<0.05) except Agzone 6.  However the risk factors for Corack are staining susceptibility in high rainfall environments (Table 2).   Early sowing of this variety will increase the risk of reduced grain quality.    

Trial number (sample size) will influence the results of the FET and Binomial analysis.   In particular Agzone 4, 5 and 6 had small sample sizes.  The analysis was based on research up to and including 2012.   Inclusion of the 2013 data in further analysis will be important for analysis of the newer varieties in these zones.  

The influence of rainfall, soil texture, soil pH and rotation on each variety’s performance relative to Wyalkatchem was assessed using logistic regression analysis.  There were a limited number of factors which significantly influenced the performance of these varieties.   In Agzone 2, both Magenta and Cobra performed better in the high rainfall and poorer in the low rainfall compared to Wyalkatchem (analysis not provided). However staining and low falling number with pre-harvest rain are risks in high rainfall environments but influenced by sowing time and season (Table 2).  The analysis did not indicate if Corack, Cobra, Mace or Magenta performed better or worse than Wyalkatchem when subsoil pH was greater or less than 4.5. 

When dry sowing and with no indication of when the season will break choosing a variety such as Mace, which performed better than Wyalkatchem when it was sown before and after d150 may be a risk management strategy.   This is because if the variety was sown before d150, but emerged after d150, its performance is more likely to be better than Wyalkatchem at both sowing times. Other varieties which performed better than Wyalkatchem when sown before d150 could be targeted if sowing into moist soil or if there is an indication that the season will break close to seeding and before d150 (Table 1). 

Table 1 The performance of a variety (better or worse) relative to Wyalkatchem, sown before (≤) or after (>) d150 and based on the number of NVT’s in each Agzone.

AgZone

Variety

N of trials sown

≤ d150 >d150

Compared to Wyalkatchem

Sown ≤ d150

Compared to Wyalkatchem

Sown > d150

FET

p-value

Binomial

p-value

1

Cobra

7

5

better

worse

0.02

1.00

 

Corack

7

5

better

better

1.00

0.01

 

Mace

16

7

better

better

1.00

0.01

 

Magenta

17

10

worse

worse

0.19

0.05

2

Cobra

13

16

better

worse

0.03

0.71

 

Corack

13

16

better

better

0.30

<0.01

 

Mace

27

29

better

better

0.54

<0.01

 

Magenta

34

48

better

worse

<0.01

0.32

3

Cobra

4

4

better

better

1.00

0.07

 

Corack

4

4

better

better

1.00

0.01

 

Mace

8

8

better

better

1.00

<0.01

 

Magenta

14

11

better

better

0.41

0.04

4

Cobra

8

3

better

worse

0.55

1.00

 

Corack

8

3

better

better

1.00

<0.01

 

Mace

10

13

better

better

0.60

<0.01

 

Magenta

13

20

worse

better

0.15

0.49

5

Cobra

9

1

same

worse

1.00

0.75

 

Corack

9

1

better

better

1.00

0.02

 

Mace

13

6

better

better

1.00

<0.01

 

Magenta

22

8

worse

worse

1.00

0.02

6

Cobra

5

0

better

n/c

1.00

0.06

 

Corack

5

0

better

n/c

1.00

0.38

 

Mace

6

4

same

better

0.20

0.34

 

Magenta

11

5

better

worse

0.28

1.00

Table 2 Comparison of the staining (%) of grain of wheat varieties sown at Binnu and Eradu in 2013

Location

Eradu

Binnu

Sowing date

9th May 2013

8th May2013

Bonnie Rock

4.5

Cobra

8.2

7.1

Corack

10.9

9.0

Emu Rock

8.9

Mace

3.5

1.4

Magenta

10.9

Wyalkatchem

6.3

4.0

LSD

2.0

0.8

(Source: Zaicou and Reynolds, DAFWA Trial and Demo reports 2014– on line in February)

Figure 1 The difference (deviations) of predicted yield (Cobra - Wyalkachem) for each NVT in Agzone 2 relative to sowing day of the year.   Symbols indicate the growing season rainfall (mm) intervals. For example, the circle symbol indicates the interval >75.3 and ≤ 188mm.

Conclusion

A methodology to assess the performance of varieties using the NVT’s has been undertaken to support growers with variety selection.  The NVT data set from 2007- 2012 was analysed to assess how factors such as rainfall, soil pH, and rotation affect the performance of a variety.  This information can be used to support growers in choice of variety given the break of the season and rainfall. The analysis did not provide a clear indication of which variety to choose given soil pH <4.5 or rotation.

The analysis validates the rapid adoption of Mace across the state and across sowing days. However the information must be used in collaboration with the knowledge of the agronomic traits.  Mace dominates the area sown to wheat in WA due to its consistent high yield over a large range of environments.  Avoid sowing Corack early in environments where staining is a risk.  Magenta and Cobra perform better than Wyalkatchem in high rainfall environments BUT staining and sprouting susceptibility are risks.

Key words

National Variety trials, agronomy, wheat, varieties

Acknowledgments

The authors thank the GRDC and DAFWA for funding the project “Wheat Agronomy – Building the Systems Profitability”.  Appreciation to Alan Bedggood and Neale Sutton for provision of the NVT dataset generated from the National Variety Trials project funded by GRDC.

GRDC Project No.:     DAW218

Paper reviewed by:    Ben Curtis and Brenda Shackley

Choosing the right wheat variety for the system (part B)

Christine Zaicou-Kunesch, Department of Agriculture and Food WA

Key messages

Trojan and Harper may be long season wheat options to replace Yitpi.

Monitor for disease expression in a wheat on wheat system but choosing a high yielding variety adapted to your environment is still the most important consideration.

Recheck the leaf rust resistance of your variety because a new leaf rust pathotype was discovered last year.  It is important to budget for fungicide application in rust prone areas. 

Background

DAFWA agronomy trials and national variety trials (NVT) funded by GRDC are able to provide information to support growers with their decisions on variety choice and management for their farming system.  A new method of analysis of the National Variety Trial database was outlined in the companion paper ‘Choosing the right wheat variety for the system- Part A’ by Zaicou-Kunesch, D’Antuono and Reeves, Crop Updates 2014. This paper, Part B, reports on the findings for DAFWA agronomy trials conducted in 2012 and 2013.  

Aims

The aim is to support growers with variety selection and management in a continuous wheat environment and long season environments. 

Method

Very early sowing systems - long season wheats

To investigate the effect of fungicide on the performance of wheat varieties sown early in May, two trials were conducted by Kevin Young, DAFWA.  They were sown on 1st May at Gibson and 2nd May at Grass Patch in 2013 onto canola stubble.   Fungicide treatments were nil and plus.  At Gibson the ‘plus’ fungicide (0.5 L/ha Tilt® Xtra), was applied at 22nd July and 10th September.  At Grass Patch the ‘plus’ fungicide treatments was a single application of Tilt® Xtra (0.5 L/ha) on 20th August (disease incidence at this site was low).

Continuous wheat environment

Four agronomy trials that were sown on wheat stubble investigated the interaction between nitrogen and fungicide application on wheat varieties.  The trials were located in 2012 at Mingenew (Zaicou-Kunesch and Beard, 2013, MIG Trials Report 2012) and Corrigin (Amjad, Coutts and Thomas, 2013, Facey Group Crop Update Book 2013.  Trials in 2013 were located at Binnu (Zaicou-Kunesch, Beard and Reynolds 2014, DAFWA trials and demos Report 2014) and Cunderdin (Amjad and Thomas, 2014, Crop Updates 2014).  Refer to the published papers for methodology.

Results and discussion

Very early sowing systems - long season wheats

If Yitpi is a benchmark variety for longer season environments because of its maturity and moderate susceptibility to frost damage, are there varieties which growers can consider as alternatives for very early sowing in the southern districts?

The trials on the south coast at Gibson and Grass Patch investigated the performance of longer maturing varieties and the value of fungicide on production as alternatives to Yitpi.   At Grass patch, disease incidence was low and fungicide did not improve productivity of varieties significantly.   At Gibson, Staganospora nodorum was the dominant leaf disease with S.tritici developing later in the season (pers comm. K Young).  Estoc and Envoy were higher yielding than Yitpi but similar to Mace in the absence of fungicide.   Necrosis on the top 3 leaves was reduced by between 20 and 57% (Figure 1).  Disease control improved yields of all varieties (Figure 1) and improved Trojan’s yield to make it the second highest ranked variety.   On the south coast, monitoring and managing disease will be key management strategies for long season wheats.

Figure 1 Effect of fungicide application on the grain yield (t/ha) and leaf necrosis (top 3 leaves) of varieties sown on the 24th April 2013 in Gibson (Source Young, DAFWA.  LSD depicted by error bars)

Harper and Trojan are recently released varieties.   The sprouting tolerance of Trojan and Harper is similar to Yitpi, however frost susceptibility is not known.   Their development was similar or 5 days earlier than Yitpi in 2013 (Figure 2). 

Figure 2 Days of flowering from Mace of varieties sown in DAFWA’s phenology trials which were sown on 26th April at Katanning and Esperance in 2013. (Mace flowering date – 28th and 17th August at Katanning and Esperance respectively) (Source Shackley and Young, 2014).

A comparison of Trojan to Yitpi in the NVT’s indicates Trojan has a yield advantage over Yitpi but not Mace (Figure 3).   However, when compared to Mace, it was significantly lower yielding than Mace in 2012 but not in 2013.  Harper has only been included in the 2013 NVT’s.  It did not perform better than Mace in Agzones 5 and 6, but yielded similar or better than Mace in Agzone 3.

Figure 3  Grain yield (t/ha) of Trojan compared to a) Mace and b) Yipti in NVT’s in Agzones 5 and 6 in 2012 and 2013.  Figure 3  Grain yield (t/ha) of Trojan compared to a) Mace and b) Yipti in NVT’s in Agzones 5 and 6 in 2012 and 2013.

Figure 3  Grain yield (t/ha) of Trojan compared to a) Mace and b) Yipti in NVT’s in Agzones 5 and 6 in 2012 and 2013.

Figure 4 Grain yield (t/ha) of Harper compared to a) Mace and b) Yitpi in NVT’s in Agzones 2,3,5 and 6 in 2013.  Figure 4 Grain yield (t/ha) of Harper compared to a) Mace and b) Yitpi in NVT’s in Agzones 2,3,5 and 6 in 2013.

Figure 4 Grain yield (t/ha) of Harper compared to a) Mace and b) Yitpi in NVT’s in Agzones 2,3,5 and 6 in 2013.

Disease assessments were undertaken on NVT’s at Wongan Hills, Mt Madden and Hyden in 2013. Harper and Trojan had better disease ratings than Yipti (Table 3).  Trojan and Harper may be long season wheat options to replace Yitpi.

Table 3 Leaf disease ratings on wheat varieties, courtesy of NVT trials in 2013. (0 - no infection/necrosis, 9 - full infection/necrosis) (Source: Young and Zaicou-Kunesch, DAFWA).

Wongan Hills

Mt Madden

Hyden

Hyden

Name

YLS/SN

YLS/SN

YLS

LR

Cobra

2.8

3.3

3.8

0.7

Corack

6.1

4.9

3.4

2.3

Envoy

 

7.2

6.0

0.0

Estoc

 

7.0

6.4

0.0

Harper

5.9

6.0

6.6

0.0

Mace

5.7

5.5

5.0

1.0

Magenta

2.0

2.6

2.4

0.0

Scout

7.4

7.5

8.2

0.0

Trojan

3.9

5.2

5.8

0.7

Wyalkatchem

5.2

4.5

4.4

1.9

Yitpi

6.2

7.3

9.0

2.3

mean

4.9

5.7

5.3

0.7

F pr

 <0.001

 <0.001

 <0.001

 <0.001

LSD(5%)

1.0

1.2

1.2

0.8

Where : YLS = yellow leaf spot and SN = S. nodorum; LR = leaf rust

Wheat on wheat – variety management

Wheat on wheat is an important rotation in the wheat belt.  The analysis of the NVT database does not indicate any varietal differences in grain yield based on rotation however there are limited number of wheat NVT’s sown on wheat. (Zaicou etal -part A. Crop Update 2014).

Leaf disease risk in a continuous wheat rotation is high for stubble borne diseases such as yellow spot and Staganospora nodorum (Glume Blotch), but severity will be influenced by the environment.   At Corrigin in 2012, yellow seeding was observed at the two leaf stage (10%) but the very dry conditions limited the spread of the disease throughout the rest of the season (Amjad etal 2013).  At Mingenew in 2012 and Binnu in 2013, despite the low rainfall during the growing season, fungicide application was profitable when leaf spot diseases were present (on average there was a 0.3 t/ha and 0.2 t/ha yield response across all varieties respectively).    At Cunderdin in 2013, the dry early season conditions limited disease development prior to flag leaf emergence.  Although fungicide applied at Z32/33 significantly reduced disease severity the subsequent yield response (0.35 t/ha across all varieties) was not statistically significant (Amjad and Thomas 2014).   Fungicide application is more likely to reduce disease levels and result in yield increases in years when there is good spring rainfall that promotes disease. 

The interaction between nitrogen and fungicide on disease incidence and yield has been investigated in the agronomy trials.   The results indicate that sub-optimal levels of nitrogen nutrition resulted in small increases in disease severity and the primary benefit of optimal nitrogen is reflected in plant growth rather than disease management (Amjad and Thomas 2014; Zaicou-Kunesch, Beard and Reynolds 2014. Table 1). Adequate nitrogen is not an alternative to fungicide but an additional measure to make plants less vulnerable to disease. 

In both 2012 and 2013, the season influenced the profitability of additional nitrogen and fungicide application in the wheat on wheat system in the northern and central agricultural regions.   Mace was consistently the highest yielding variety at each trial.   Choosing high yielding varieties adapted to the environment, ensuring optimal crop nutrition, and monitoring and managing disease, continue to be important management strategies for the wheat on wheat system.

Table 1 Percentage leaf area diseased (LAD) of 4 varieties with nitrogen and fungicide treatments sown at Binnu in 2013. Figures are average disease on top three leaves at Z71 (early grain development). Disease was a mixture of SNB and YS.

Variety

Nil Fungicide

Plus Fungicide

N0

N30

N60

N90

N0

N30

N50

N90

Cobra

28

27

23

21

15

13

11

14

Corack

29

34

26

33

18

18

14

13

Mace

28

28

26

21

16

18

13

14

Wyalkatchem

28

30

22

19

15

10

11

11

Source: Zaicou-Kunesch, Beard and Reynolds 2014.

LSD(5%)

5.59

New leaf rust pathotype – what does it mean?

The finding of a new leaf rust pathotype in WA  (the first occurrence of virulence for the resistance genes Lr13, Lr17a, Lr17b, and Lr26) will reduce the resistance rating of some wheat varieties such as Mace, Corack, Emu Rock, King Rock and Wyalkatchem. Refer to the Wheat Variety Guide for WA, 2014 (available on line).

Growers in rust prone areas should be prepared for the chance that Mace (and other varieties) may respond as an MS type and budget for a fungicide spray after flag leaf emergence.   If spraying is likely to be carried out for the control of yellow spot or S. nodorum (Glume Blotch) then that spray is also likely to help control the leaf rust.  

Conclusion

Growers on the south coast capitalise on early sowing options through adoption of Yitpi because of its longer maturity and reduced risk of low falling number following pre-harvest rain.  Trojan and Harper may be long season wheat options to replace Yitpi.  They have a similar falling number risk as Yitpi, similar maturity or 5 days earlier, depending on the environment, and yield better than Yitpi.

The environment has a huge role in disease expression in the wheat on wheat system and fungicide application can be effective in improving grain yield.  However in the wheat on wheat system, choosing high yielding varieties adapted to the environment, ensuring optimal crop nutrition, and monitoring and managing disease continue to be important management strategies.  In rust prone areas budget for the application of fungicide, particularly  since the discovery of a new leaf rust pathotype which will affect the disease resistance ratings of some wheat varieties.  

Key words

National Variety trials, agronomy, wheat, varieties

Acknowledgments

The authors thank the GRDC and DAFWA for funding the project “Wheat Agronomy – Building the Systems Profitability”.  Acknowledge DAFWA’s wheat agronomy team (B Curtis, K Young, B Shackley, M Amjad, B French, A Smith, B Haig, R Bowey and P Barlett) for their contribution to the research.

GRDC Project No.:     DAW218
Paper reviewed by:    Ciara Beard, Ben Curtis