Australian Crop Accreditation System
National Variety Trials
NVT 2009 Trials Report
Alan Bedggood
Summary
The 2009 NVT trials were yet again affected late breaks and a dry season with associated spring frosts, with Central Queensland being the exception where very early trial plantings were achieved. Continual improvements and interactions between ACAS and the Service Providers continue to deliver a relatively smooth operating and successful trials program with all Service Providers meeting their GRDC milestones.
Few trials were lost due to operational errors, with most abandoned trials being caused by drought condition during the cropping season. This mostly affects non-cereal crops. There are more canola trials than pulse trials sown in the marginal areas and thus canola proportionally suffer more trial failures.
Turnover and retraining of field staff is a continual battle for the Western Australian service providers.
2009Sowing
The sowing of NVT trials is obviously determined by the timing of rainfall and preparedness of sowing teams. As with 2008, the 2009 season saw very different situations in each of the GRDC regions.
Central Queensland in the north received rains in early April that enabled all wheat trials to be seeded in the one week – 21-24 April. This was followed by an extensive dry period that meant any unsown trials would be very late. Agrisearch did well to implement the program so early.
Southern Queensland received the same early April rains but it was considered too early for seeding trials. Consequently, all trials were sown after the subsequent rains which didn’t come until the week of May 26 (and continued to rain for days, holding up all field operations. Agrisearch punted on the impending rains and sowed the Roma trials the day before the rains arrived. All southern Queensland trials were sown within the week of 17-21 June.
Seeding in northern NSW began at the end of April with faba beans and lupins utilising the mid April rains that were too early for most other crops. A few trials were sown early on adequate moisture (Wellington canola, Coolah wheat). However, most trials were sown towards the end of May and early June after the rains that started in northern NSW just after mid May.
The rainfall experienced across the southern GRDC region during the week of April 29 set up the whole area for an ideal early sowing.Many southern NSW NVT trials were sown on this rain over the following couple of weeks. However, subsequent weeks were very sunny and dry. Initially, this was good for cropping but it eventually became too dry to continue. The dry period also caused poor germinations in trials sown later, a couple of weeks after the rains.
Follow up rains eventually came in the week of May 26 when seeding of the last of the trials was enabled.
DPI NSW again ensured that the majority of all wheat trial locations had their early and main season trials sown several weeks apart, with even the southern regions Early trials nearly all sown by the first week of May.
The Victorian NVT trial managers wasted no time after the late April rains to get all trial seeders into operation. Many trials were sown in May with the most notable exceptions being the Wimmera trials (Kaniva, Tarranyurk, Horsham, etc). These trails were all sown up until June 18. Although in the historical sowing window for the Wimmera, the decision to hold off seeding of cereal trials for so long drew many criticisms from breeders.
In a repeat of 2008, South Australia experienced good sowing rains in 2009 allowing practically all trials to be sown in May or the first week of June. The SARDI field staff are very adept at getting trials sown quickly as the earlier trials give the best opportunity for results.
Western Australia experienced very poor opening season rains that did not allow a good seeding program to occur. Some trials were relocated to sites where moisture might be found. More and more trials were seeded dry to match the farmers who were doing the same.
The rains that did eventually come in the fourth week of May validated the earlier dry sowings and allowed the last of the trials to be sown. Quite a number of the southern wheat trials were not sown until the third week of June.
DAFWA completed sowing their trial component for NVT at the end of June.
The 2009 Trials
ACAS has again sought trial results from the pulse and oat breeding programs to enhance the trial data set for growers and to enhance the METs for those crops.
Trial designs were imported into the NVT database mid-season to enhance the process at harvest. Although good in concept, this did not overcome the lateness of obtaining the last of the pulse trial results from WA.
Better communications with DAFWA staff is all that can be aimed for. There is no financial leverage available for enhance access to trial results.
Trials sown
The number of trials sown in 2009 is now the norm for the NVT program. The previously used total of 580 trials was an approximation and did not account for the number of trials results now provided by DAFWA.
Added to the NVT trials are the 116 advanced trials results from the pulse and oat breeding programs. These additional results are included to enhance the robustness of the METs for those crops.
Trials harvested
Losing 9% of the trials in 2009 again reflected the dry conditions experienced in some regions.
Two sites in Central Qld (7 trials)
were sown very early but did not have sound plant establishments – poor operator judgement at Springsure. These sites were subsequently abandoned, making the Qld total lowest as a percentage.
South Australia experienced the best growing season of all which is reflected in the high yields and higher percentage of trials harvested.
Trials Published
Victoria experienced a heat wave during grain filling period that decimated many crops, including the NVT. This made it harder to estimate yield potential from the relatively good plant growth experienced during the year. Of the Victorian trials harvested, 27% were not released to the public.
The individual trial results are not flagged as released as there is little to no site characterisation or trial management information available. Resulting trial reports for these trials will contain yield values with statistical results from the analysis but little else.
Again, the canola trials are those that suffered the most by the drought. This is due the number of the Early maturing trials sown in marginal areas which are prone to drought. There are fewer pulse trials sown in these areas to fail.
The breeder’s trials do not have the accompanying data for the trials – sowing date, fertiliser and chemicals, covariates, paddock history, rainfall, etc. The NVT database generated reports for these breeder’s trials will be practically empty but they were perceived to still allow growers to compare varietal yields in areas additional and complimentary to NVT. Subsequently, the trials were published if they met the same criteria of achieving lower than 15% CV for the analysed results and that the trial mean yields were over 0.30 t/ha.
These breeding trial results are not reflected in the tables above.
Analyses
The NVT database program is now familiar with biometricians and trial manager alike and ACAS has responded to these clients by having incorporated many tweaks to the processes to facilitate the data entry, providing comments and streamlining the data extraction – analysis – data import process.
Again, the biometricians as a group responded well to individual trial analyses with most trials being analysed within a week of harvest data entry. In a repeat of previous years, obtaining the last of the pulse trial results from DAFWA was the major holdup in generating the METs from the 2009 harvest.
The METs were provided for all crops within a week, when they were eventually begun. The delaying of METs until all crops might have been convenient for Brian Cullis but much less so for those that were awaiting the results for publications. Getting MET results over a week or two is much more preferable than getting them all in several days.
Hopefully, more coordination can overcome the hurdles faced over the 2009 harvest.
As for other years, the 10-day dispute resolution period was applied to all barley, wheat and canola results. The breeders of the other crops waived this option after agreements with the NVT Manager in 2005 that trials with a CV greater than 15% would not be made public
Goodness of Trials
ACAS has continued to place a higher priority on the need for site audits to give confidence to the results of the NVT trials program. It also validates the comments provided by the wheat and canola breeders. Few other breeders provide comments on the status of the NVT trials that they have visited.
Table 4 indicates the number of trial locations (where the majority of locations have multiple trials) that had been visited by ACAS staff and or reported upon by visiting plant breeders. ACAS made a much more concerted effort to visit most NVT trial locations in the 2009 season, with in excess of 90% of all locations visited and reported upon by either Neale Sutton or Alan Bedggood
The value of site audits remains high. There are a number of instances where operational mishaps are not reported and may impact upon the trial results. This includes a wide variety of seeding mishaps, very weedy trials and misguided attempts to clean up weeds within trials.
Poor site-maintenance (no signs, no site plans, no cut-out access track, weedy headlands) was also evident at some sites but this was more restricted in 2009 than in previous years. Although not affecting trial results, this does not set a good image for GRDC and the NVT program in terms of acceptable site management.
Tiny Tags
The 2009 season saw the best utilisation of the Tiny Tag data in the support of the release of the NVT trial results. The minimum temperature graphs were generated in sufficient time to support the release decisions on the trial results. Comments regarding the potential influence of frosts on the trial results can now be made to help explain those sites that show widely divergent varietal comparisons.
The Tiny Tag data was also used to class some 2009 trials as invalid and not to be used in the 2009 METs.
A concerted effort was made to generate the minimum temperature graphs for all available trial sites for 2009. As an interim, these were put into an Excel table format to allow access to them. A better way is needed to capture and display these graphs – perhaps in html format on a website.
The temperature data from every 2009 Tiny Tag unit was collected and combined into a MS Access data file to allow data interrogation to occur. It was combined with a table of unit identification numbers and related trial locations so that the full NVT data set can be combined with the temperature data if required.
Efforts will be made to get all of the 2008 and 2007 full data sets into this same format before the 2010 data collection begins.
SERVICE PROVIDERS
There are six Service Providers employed by GRDC to conduct the NVT trials across the country. The NVT database and the trial management process has now become more routine and accepted by the Service Providers. As for previous years, there are ongoing operational issues arising this season that needto be addressed with continual improvement being the goal. These will be discussed after the comments on individual Service Providers.
Kalyx Pty Ltd
Kalyx employ a dedicated NVT data manager who has developed many skills in data management and the operations of the NVT database program (as well as the Kalyx commercial trialling operations). This has helped ACAS staff through many implementation processes of operational improvements to the NVT system.
Kalyx, through Peter Burgess, has developed a very good rapport with many farming groups that facilitate finding trial sites and the availability of the trials for field days and farm walks.
Staff turnover continues to plague Kalyx, with new staff continuing to create teething operational issues at some trials sites. Mostly, these can be adjusted for in the analyses of the results.
Kalyx staff continue to make a concerted effort to provide all information as part of their Services Agreement in a timely manner.
DAFWA
As per the previous year, there was a concerted effort made to capture results from the 2009 DAFWA trials into the NVT database, with a high success rate. The next challenge for the 2010 season is to capture more site characterisation data and to obtain the harvest data earlier to enable the METs to be generated.
Few 2009 DAFWA trials were inspected by ACAS staff as the lat longs are not provided to ACAS staff. Latitude and longitude values for many of the barley trials that were provided turned out to be incorrect, with no trial being at the designated point. This will hopefully be remedied for the 2010 season by generating maps using the lat longs provided and getting DAFWA staff to validate them before ACAS staff make site visits.
Leigh Smith, DAFWA, South Perth, continues to visit all DAFWA CVT trials for data integrity checks. However, there are known examples of disastrous trials still being harvested and provided to ACAS, resulting in their flagging as invalid for METs. More site audits of DAFWA trials are required in future.
SARDI
As for previous years, the 2009 South Australian NVT trials were again managed successfully by SARDI, with their thoroughness in trial management (mostly positive trial reports from breeders) and commitment to the NVT process. The results were loaded in a timely manner and Service Agreement milestones for GRDC were met.
There are still remaining concerns about the Kimba site being with a long term farmer cooperator (good) but on a farm that is very poorly managed and not reflective of the district (not so good). Complaints for the past three seasons and from two breeders raise this as an issue. Certainly, the new NVT project in 2010 will be an ideal time to “relocate” some trials.
The Bool Lagoon sites fall into this category as well. Trials on lunettes amongst the Bool Lagoon swamps may be well drained and produce reasonable trials but how reflective are they for the district practice is questionable. This was not addressed for 2009 but will hopefully be addressed in the New Project.
DPI Vic
NVT trials were conducted across the Mallee, Wimmera and south west of Victoria with Agrisearch conducting trials in the north central and north east.
As for 2008, the 2009 late break saw many DPI Vic trials sown quite late, albeit within the long term sowing windows for the Wimmera. A number of cereal breeders were critical of the lateness of the trials in a run of short seasons previously experienced.
The reasonably good winter rains set up a good trialling program regardless of the seeding times.However, the hot dry spell in October decimated many crops and trials across Victoria.
Instructions for DPI Vic staff to sow earlier were not adhered to for 2009. The 2010 season needs more stringent attention to seeding times.
Not undertaken in 2009 due to the fixed mind-set of the Departmental staff, moving canola trials further south when sowing is delayed because of the dry start to the season is a requirement that will need enforcing in the New Project in 2010.
Agrisearch
Agrisearch has centralised computer management at Toowoomba and seed packaging at Wagga Wagga. David Leah is the coordinator for Agrisearch. As well as trials in Victoria (managed from Shepparton), NSW (managed from Wagga Wagga and Narrabri) and Qld (managed from Toowoomba), Agrisearch also conduct a winter seed increase of the control varieties for the cereals, with wheat in WA and wheat, barley and triticale at Wagga Wagga.