A New Way to Assess Wheat Variety Performance in Western Australia

A new way to assess wheat variety performance in Western Australia

Summary

To support you with variety decisions, it is valuable to review how the varieties have performed in a range of years and in similar environments to your farm. A new approach (using production values) has been developed using the information available through the Grain Research Development Corporation’s (GRDC) National Variety Trials project.

Production values (PVs) will help you select the best variety for a particular growing environment. The PVs are based on a multi-environment trial (MET) data-set comprising NVT wheat trials grown in WA over a five year period.

Introduction

In 2014, the Department of Agriculture and Food Western Australia’s (DAFWA) wheat agronomy project collaborated with GRDC’s Statistics for the Australian Grains Industry (SAGI) to develop a new approach to support growers with variety information and selection. The information provides more detail on how the varieties have performed in a range of years and in similar environments

The new approach of presenting this information has been developed through GRDC’s National Variety Trials project and a new statistical analysis, production value – plus system. The analysis results in a unique production value (PV) for a variety at any given year and site combination. The PVs currently referred to wheat performance in Western Australia (WA) based on a multi-environment trial (MET) data-set comprising NVT wheat trials grown in WA for the five year period of 2009—13.

In previous years, trial data has been provided as single site data and as long term summaries from the last five years of data. The ‘predicted yield of varieties’ was provided for the six agzones in WA, averaging the response of varieties over an agzone and years. This single figure can mask any variety by environmental response.

By assessing the production value of a variety over time and in locations relevant to your business, you get an indication of:

1. how much more a variety is likely to yield (tonnes per hectare (t/ha)) compared to other varieties in that environment.

2. the stability of the variety over a number of locations and years within an agzone or region.

It is important to note that PVs are obtained for all varieties (commercial and advanced breeding lines) evaluated in NVT wheat trials for the five year period.

How to use PVs

Interpreting production values

PVs for each environment are expressed in kilograms per hectare (kg/ha). They are shown as positive or negative differences relative to a base-line (PV=0) which reflects the expected average yield of all the varieties in the current NVT data set, if grown in that particular environment.

Thus, varieties may be viewed as having expected yields that are above or below the baseline for a particular environment.

Mace had a production value of +450kg/ha in 2011. This is an indication that it is expected to yield 0.45t/ha higher than expected average in this type of environment. It had a stable PV over seasons indicating consistent performance over a range of environments.

Since selection decisions are based on the comparative performance of varieties, it is the difference between PVs that is critical. For example at Mingenew:

Magenta with a PV of 600kg/ha in 2011 would be expected to yield 150kg/ha more than Mace (PV = 450kg/ha) when the varieties are grown under these conditions. Magenta’s PV’s change over seasons which indicates a big influence of environment on performance.

PV’s across a region/agzone

When using PVs for variety selection decisions it is recommended that you consider not just a single environment, but a number of environments, both in terms of location (trial site) and season (trial year). This gives an indication of the stability of a varieties performance over a range of conditions

Whilst viewing a single graph location additional information has also been provided in a table in order to identify the type of environment experience at that site and year (Table 1). At present the trial sowing date, growing season rainfall, previous crop history and soil pH are reported. Site correlation is a relativity measure that is included in the reporting. It indicates how the variety ranking of the trial resembles neighbouring trial result rankings within the current Agzone and year.

Table 1. Location details for NVT’s at Carnamah in 2009, 2010, 2011 and 2013

Location details / Carnamah North in 2009 / Carnamah North in 2010 / Carnamah in 2011 / Carnamah in 2013 /
Sowing Date / 5/06/2009 / 7/06/2010 / 22/05/2011 / 22/05/2013
GSR (May – Oct) mm / 271 / 159 / 327 / 216
pH..CaCl2at 10cm / 5.9 / 5.2 / 5.3 / 5.4
pH..CaCl2at60 cm / 5.1 / 5.6 / 5 / 6.1
Prev.Crop / Canola / Wheat / Canola / Canola
Site Correlation / 0.50 / 0.76 / 0.59 / 0.36

Table 2. Production value (kg/ha) of wheat varieties and their classification at Carnamah between 2009 , 2010, 2011 and 2013

Variety / Classification / Production value at Carnamah North 2009 / Production value at Carnamah North
2010 / Production value at Carnamah 2011 / Production value at Carnamah 2013 /
Carnamah / AH / 9 / -24 / -39
Cascades / AH / -270 / -362
Cobra / AH / -52 / 366 / 155
EGA Bonnie Rock / AH / 131 / 113 / 299 / -5
Emu Rock / AH / 357 / 201 / 38
King Rock / AH / 131 / 108 / 354
Mace / AH / 261 / 300 / 244 / 91
Tammarin Rock / AH / 136 / 193
Yitpi / AH / -133 / -87 / -52
AGT Katana / APW / 142 / 202 / 65
Corack / APW / 439 / 642 / 175
Correll / APW / -139
Envoy / APW / 9 / 342
Espada / APW / 73 / 80 / 25 / 52
Estoc / APW / -77 / -198
Fang / APW / -319
Gladius / APW / 20 / 85
Magenta / APW / -37 / -96 / -122 / 110
Scout / APW / -166 / -62 / -32
Trojan / APW
Westonia / APW / 173 / 166 / 213 / 91
Wyalkatchem / APW / 169 / 182 / 183 / 36
Young / APW / 124 / 190
Zippy / APW / 329 / 451 / 632
Clearfield STL / APW-imi / -281 / -402
Grenade CL Plus / APW-imi / -503 / -111
Impose CL Plus / APW-imi / 204 / 227 / 166
Justica CL Plus / APW-imi / -112 / -331 / -63
Arrino / ASWN / 238 / 241 / 371
Binnu / ASWN / 223 / 162
Calingiri / ASWN / -118 / -149 / -253 / 13
Fortune / ASWN / -84 / -45 / -439 / 23
Yandanooka / ASWN / -75 / -57 / -424 / -52

Acknowledgements

We acknowledge SAGI’s contribution in the presentation of results in this report. In addition we acknowledge the GRDC for their financial support of DAFWA’s wheat agronomy project DAW00218 which has enabled the collaboration with SAGI.

Further information

Production value-plus

Production value (PV)-plus was developed by biometricians within the GRDC funded “Statistics for the Australian Grains Industry” (SAGI) project. SAGI is responsible for the design and analysis of all data collected within the NVT program and for the continued improvement of the associated statistical methodology. A recent research paper by SAGI (Smith et al. 2014) reviewed the provision of grower information in the NVT and resulted in the development of the PV-plus system.

The variety information provided by the PV-plus system is a major step forward. Previously, the information reported to growers comprised long-term regional (agzone) means. These were obtained from a statistically valid MET analysis, as described in Smith et al. (2001), but were not ideal as they often failed to identify important local variety by environment interaction. Growers and advisers therefore resorted to the practice of using results from the analyses of individual trials to make varietal selection decisions. Such results are merely a snap-shot of what occurred on a specific area of land in a specific time-frame and are based on a very small amount of data. If used for variety comparisons, the risk of errors can be unacceptably high. SAGI acknowledged both the need for the provision of variety information at this local rather than regional level and the potential for using a more up-to-date and superior statistical approach for the analysis of MET data. PV-plus was developed on the basis of these fundamental principles. PV-plus therefore delivers more accurate variety information at an individual environment level. This information is not the same as the individual trial results and is more reliable since it is based on more data and an appropriate statistical model for variety by environment interaction.

SAGI

Statistics for the Australian Grains Industry (SAGI) is a GRDC supported project, which is responsible for the analysis of all data, collected within the NVT program and the continued improvement of NVT multi-environment-trial (MET) analysis methodologies. A recent research paper by SAGI (Smith et al. 2014) reviewed the provision of grower information in the NVT. A main recommendation of this paper was that variety performance should be presented for each given location and year combination based on the long-term analysis of all trials in WA over the past five years. This analysis is referred to as the production value - plus system, which results in a unique production value (PV) for a variety at any given year and site combination.

References

Smith A (2001) The analysis of crop variety evaluation data in Australia. “Australian and New Zealand Journal of Statistics 43:129-145

Smith AB, Ganesalingam A, Kuchel, H, Cullis BR (2014) Factor analytic mixed models for the provision of grower information from national crop variety testing programme. Theoretical and Applied Genetics (accepted).

Important disclaimer
The Chief Executive Officer of the Department of Agriculture and Food and the State of Western Australia accept no liability whatsoever by reason of negligence or otherwise arising from the use or release of this information or any part of it.