Previous PageTable Of Contents

G E interactions for high yield potential of winter wheat in Tasmania

Tina Botwright Acua1, Geoff Dean2 and David McNeil1

1 Tasmanian Institute of Agricultural Research, The University of Tasmania, Private Bag 98, Hobart, Tas 7001. www.tiar.tas.edu.au Email Tina.Acuna@utas.edu.au
2
Tasmanian Institute of Agricultural Research, Mt Pleasant Research Laboratories, PO Box 46, Kings Meadows, Tas 7249

Abstract

Wheat production in Tasmania has achieved yields in excess of 12 t/ha in some field trials, which indicates that the state may be ideally located to study the physiological basis of high yield potential in the HRZ. Pattern analysis of 10 wheat cultivars and breeding lines in 9 Tasmanian environments was used to identify high yielding genotypes and to establish the potential impact of genotype x environment (G E) interactions on grain yield. Results of the analysis revealed very large E main effects. G E was similar in magnitude for G, which indicated that genotype ranking was largely unchanged across environments. Longford was the highest yielding environment, where two CSIRO winter wheat breeding lines, 95102.1 and K89.44, had grain yields greater than 11 t/ha. In contrast, two environments were affected by frost at flowering and yielded only 2 t/ha. A field trial is planned this year to evaluate the contribution of phenology, biomass and yield components to high yield potential, compared with the wheat ideotype proposed for the HRZ of south-eastern Australia.

Key Words

Wheat, yield potential, G x E

Introduction

The high rainfall zone (HRZ) for agricultural production of cereal crops in southern Australia includes areas with an annual rainfall of between 450 to 900 mm. Yield of wheat is a function of crop management, such as planting density and nutrient inputs, and yield components such as tillering and the number and weight of grain in spikelets and ears. In the HRZ of Western Australia and Victoria, between 450 – 550 ears/m2 are required to achieve a high yield potential of wheat (Riffkin et al., 2003; Zhang et al., 2007). However, average grain yield of wheat across these environments is only half the potential yield of 6.3 t/ha due to waterlogging, physical and chemical constraints to root growth and a lack of adapted cultivars (Zhang et al., 2006). Tasmania faces similar abiotic constraints, yet the long growing season, high rainfall (600 to 1200 mm) plus access to low-cost irrigation water has resulted in grain yield of winter wheat varieties approaching 12 t/ha in some managed field trials. Such high yields, which are comparable to that being achieved in the high-input production environments in the UK (Foulkes et al., 2007), indicate that the state is ideally located to study the physiological basis of high yield potential of the wheat ideotype proposed for the HRZ (G. O’Leary, pers. comm.). Firstly, however, there is a need for a systematic analysis and identification of superior germplasm and environments for realising high yield potential of wheat in Tasmania. Genotype x environment-type analyses (e.g. Cooper and Hammer, 1996) are one approach to achieving this aim. Typically such analyses involve large numbers of genotypes, but there are some examples in the literature of G E interactions with relatively few entries. Here we report on the results of a G E analysis for 10 wheat cultivars and breeding lines grown in nine environments in Tasmania from 2004 to 2006.

Methods

Ten long-season (predominately) winter wheat cultivars and breeding lines (Alberic, Brennan, Kellalac, Mackellar, Teesdale, Tennant, 95102.1, H123.1, K37.18 and K89.44) were evaluated for agronomic performance and grain yield in replicated field trials grown at nine managed environments in northern Tasmania from 2004 to 2007. All wheat cultivars and breeding lines had red grain, with the exception of Brennan and Kellalac, which have white grain. Genotypes selected for the analysis were either commonly grown in Tasmania, being considered for release or have been bred for high yield in the HRZ. The number of genotypes included the analysis was limited to those with relatively consistent representation in field trials over the time period. Trial abbreviations are shown in Table 1. All field trials had a randomised complete block design with four replications, with the exception of WB04 and CR05, which had three. Plots were 8 m long and 1.2 m wide, with 0.2 m between rows. Two of the nine field trials were sown in March as dual purpose crops for grazing and grain production. All grain-only trials were sown in May and harvested in January the following year. Field trials were amended with 150 to 250 kg/ha of fertiliser (typically 9:13:17 N:P:K) at seeding and top-dressed once with 75 kg N/ha (WB04) or twice with 50 kg N/ha (N05, PER06 and CR05 sites). Low input sites at SP04L and LF06L were top-dressed once with 50 kg N/ha with nil fungicide application; high input SP04H and LF06H sites received twice this amount of N and two applications of the fungicide, ‘Bumper’. Climate data, including irrigation, are shown in Table 1. The PER06 site experienced severe frost on 16 October and 16 November.

Yield data for 10 wheat genotypes and 9 environments were extracted from appropriate single-site analyses for each site. G E interactions were analysed using the pattern analysis tool in IRRISTAT (IRRI, 2000). Data for row and column means were transformed with location standardised. The transformed data was clustered using an agglomerative hierarchical algorithm based on minimising incremental sum of squares (Table 2).

Results

Characterisation of environments

Northern Tasmania has a temperate climate with cool to cold winters and mild summers. Total rainfall in 2004 was close to average. In 2005, above average rainfall for the months August to December resulted in total rainfall exceeding the long-term mean by 125%. The 2006 season was particularly dry, with total rainfall only 60% of the long-term mean. The CT06 and PER06 sites were affected by frost at flowering.

Table 1. Site characterisation and abbreviations, including rainfall, irrigation and date of sowing and harvest.

Abbreviation

Year

Site

Growing season rainfall (mm)

Irrigation (mm)

Soil type

Sowing date

Harvest date

SP04L/SP04H

2004

Symmons Plains

386

-

Sodosol

27 May

19 Jan

WB04

2004

Westbury

479

-

Dermasol

26 May

26 Jan

CR05

2005

Cressy

541

-

Sodosol

26 May

25 Jan

N05

2005

Nile

512

65

Vertosol

18 Mar

13 Jan

CT06

2006

Campbell Town

208

-

Dermasol

22 May

18 Jan

LF06L/LF06H

2006

Longford

315

185

Dermasol

19 May

24 Jan

PER06

2006

Perth

385

-

Sodosol

14 Mar

24 Jan

Analysis of variance and pattern analysis of grain yield

Mean grain yields were computed for each environment using the appropriate analysis for each trial. These means were extracted for nine environments and ten genotypes. The main effect of environment was large and accounted for 91.1% of the total sums of squares (Table 2). Genotype and the G E interactions were only 4.7% and 4.2% of the total sum of squares. The analysis was repeated, omitting the two sites affected by frost. In this analysis, G, E and G E accounted for 20.9, 64.5 and 14.6, respectively, of the total sums of squares. The small size of G E interactions indicates that ranking of G was largely unchanged across environments, which makes it easier to select for genotypes with improved grain yield. G E analysis with a larger set of genotypes would be required to confirm these results. Cluster analysis is still reported to assist in the identification of potential environment and genotype groups. Three environment groups and four genotype groups were identified, which preserved 88.6% of the G x E-SS among groups (Table 2).

Main effects of environment on grain yield

The analysis identified three environment groups (Table 3). Environment group E1 included two sites at Perth (PER06) and Campbell Town (CT06), where severe frost at flowering resulted in exceptionally low average grain yields of 1.8 t/ha. Relatively poor growing season rainfall at Campbell Town in 2006 may have also contributed to a reduced grain yield at this site. Environment group E2 included five of the nine sites and had the largest average grain yield of 8.81 t/ha. Environment group E3 included a dual purpose crop at Nile (N05) and a grain crop at Cressy (CR05), with an average grain yield of 7.87 t/ha. The 2005 season, particularly in the latter half of the year, was particularly wet (Table 1) and waterlogged soils may have contributed to the slight reduction in grain yield at Cressy (CR05).

Table 2. Across site ANOVA for G x E interaction studies in the northern Midlands, Tasmania, 2004 – 07.

Source

df

MS

F

%TSS

% G x E

         

SS

All environments

         

Environment (E)

8

96.262

194.86 ***

91.1

-

Genotype (G)

9

4.407

8.92 ***

4.7

-

G x E

72

0.494

 

4.2

-

           

Stability regression

9

1.015

2.42 **

-

25.7

Regression deviations

63

0.420

 

-

74.3

           

AMMI component 1

16

2.98

9.61 ***

-

58.4

AMMI component 2

14

1.11

3.58 ***

-

19.0

AMMI component 3

12

0.76

2.45 *

-

11.2

AMMI component 4

10

0.63

2.03 n.s

-

7.7

AMMI residual

60

0.31

   

3.7

Main effects of genotype on grain yield and yield components

Four genotype groups were identified in the cluster analysis (Table 3). Group G1 included five of the ten genotypes, with an average grain yield of 6.66 t/ha. Genotype groups G2 and G4 had the largest average grain yields of 7.96 and 8.19 t/ha, respectively (Table 3). The cultivar Tennant was the only genotype in group G3. Although the G E interactions were small, there was a tendency for genotype groups G3 and G4, which had late flowering (data not shown) to have the greatest grain yield (3.1 and 2.5 t/ha, respectively) under frost conditions in environment E1. Genotype group G2, in contrast, tended to have the greatest grain yield of 10.4 t/ha in high-yielding environments such as E2.

Table 3. Main effect of grain yield of three environment and four genotype groups. Mean yield was 7.17 t/ha.

E group

Environment

Yield (t/ha)

G
group

Genotype

Yield (t/ha)

           

E1

PER06

1.72

G1

H123.1

6.66

 

CT06

1.96

 

Brennan

6.05

       

Mackellar

6.79

E2

LF06L

9.92

 

Teesdale

6.99

 

LF06H

10.61

 

Kellalac

6.79

 

SP04L

7.94

     
 

SP04H

8.18

G2

95102.1

8.21

 

WB04

7.44

 

K89.44

7.71

           

E3

N05

8.52

G3

Tennant

6.97

 

CR05

7.21

     
     

G4

Alberic

8.16

   

 

K37.18

8.21

           
 

l.s.d

0.70

l.s.d

 

0.66

Conclusion

The analysis revealed that G E interactions for the 9 environments and 10 genotypes to be relatively small, which indicated that genotype ranking was largely unchanged across the managed environments assessed in this study. The implication for wheat breeders is a relatively high level of confidence that genetic material with high yield potential should perform well across Tasmanian environments. CSIRO breeding lines, 95102.1 and K89.44 had the largest grain yield across sites, which slightly exceeded the simulated potential yield of 11.1 t/ha for wheat grown in northern Tasmania (G. O’Leary, pers. comm.). Superior genotypes identified through this analysis will be used in a field trial planned to study the physiological basis of yield potential (Fischer, 2007) in Tasmania. A major constraint to yield potential identified in this analysis was frost at anthesis, the impact of which depended on flowering date (Boer et al., 1993). Genotype ranking did tend to change in environments that experienced out of season frost, although this effect is perhaps outweighed by the overall large impact of these frost events on grain yield. Field trials are planned over the next four years over a wider geographical area, including regions in the northwest, north, southern Midlands and southern Tasmania. It will be worthwhile to reassess whether or not there exists significant G E interaction for grain yield in response to the greater diversity of climate and soil types.

References

Boer, R., Campbell, L.C., and Fletcher, D.J. 1993. Characteristics of frost in a major wheat-growing region of Australia. Australian Journal of Agricultural Research 44,1731-1743.

Cooper, M., and Hammer, G.L. 1996. Plant Adaptation and Crop Improvement. CAB International, Wallingford UK.

Fischer, R.A. 2007. Understanding the physiological basis of yield potential in wheat. Journal of Agricultural Science 145,99-113.

Foulkes, M., Snape, J., Shearman, V., Reynolds, M.P., Gaju, O., and Sylvester-Bradley, R. 2007. Genetic progress in yield potential in wheat: recent advances and future prospects. Journal of Agricultural Science 145,17-29.

IRRI. 2000. IRRISTAT for windows IRRI, Los Baos, Philippines.

Riffkin, P., Evans, P.M., Chin, J., and Kearney, G. 2003. Early-maturing spring wheat outperforms late-maturing winter wheat in the high rainfall environment of south-western Victoria. Australian Journal of Agricultural Research 54,193-202.

Zhang, H., Turner, N.C., Poole, M.L., and Simpson, N. 2006. Crop production in the high rainfall zones of southern Australia - potential, constraints and opportunities. Australian Journal of Experimental Agriculture 46,1035-1049.

Zhang, H.P., Turner, N.C., Poole, M.L., and Asseng, S. 2007. High ear number is key to achieving high wheat yields in the high-rainfall zone of south-western Australia. Australian Journal of Agricultural Research 58,21-27.

Previous PageTop Of Page