The NVT Long Term Yield Reports provide Australian growers and advisors with the best available tool for making variety selection decisions based on crop yield. Information is available for all cropping regions in Australia for ten crops including wheat, barley, canola, oat, triticale, chickpea, field pea, faba bean, lentil and lupin.
To generate the data, a multi-environment trial (MET) analysis of NVT data is conducted using multiple years of individual trial data. For most crops, a five year period is used, but this can vary up to ten years for barley and down to three for canola (likely case for the 2015 analysis). The most appropriate duration is selected based on correlation between varieties across years.
A factor analytic (FA) approach to the analysis is undertaken using the national dataset for most crops. However, for wheat and barley, the FA method was used on mega-region based datasets, splitting the dataset into agro-ecological regions.
The long-term analysis produces Best Linear Unbiased Predictions (BLUP) values for each variety in each trial using the MET dataset. These values are also referred to as “Production Values” or “PVs” and are the best prediction of a variety’s performance in a particular trial. They are much more reliable than the “single-site” analysis that produces variety predictions using the information from a single trial in a single year.
Once the Production Values have been determined, they can be used as a very powerful tool for predicting yield performance. PVs not only give the best prediction of yield performance at a single trial level, but datasets from across groups of trials, specific regions or different years can be combined to show the most likely variety performance outcomes. This gives great confidence in the dataset and growers and advisors can customise information to suit their specific needs.
Historically, the Long Term Yield analysis combined data from all trials across a region to produce a single yield prediction per variety for that region. This method did not account for seasonal yield variation, such as the affect of a dry spring or a late autumn break. Years where faster maturing varieties yielded well were combined with years where longer season varieties yielded well, and the end results didn’t separate, identify or account for this variation.
Now, the Production Value system uses the complex MET analysis to provide the most accurate prediction of variety performance down to the individual trial level. The downside to this level of accuracy is that the wheat dataset, for example, can have over 11,000 rows of data for users interpret. This application helps users interpret the data by grouping trials together with similar mean yields. Data is presented at 0.5 tonnes/ha yield groupings and predictions are averaged across the group of trials selected. This might be national (not recommended), state or region based, dynamically selected across different sites, or at the individual site level.
The data can be viewed in the default tabular format or in a line graph where trends and patterns of specific varieties can be compared. The datasets also provide the number of trial that each variety has been tested in, and the number of trial constituting each yield group providing users with a data reliability index. A higher number of trials ensures more confidence in the data.
The NVT Long Term Yield Reports App – Providing users with leading edge variety performance data in an easily interpretable format. The data is generated through the GRDC funded National Variety Trials program.
Close localities have been grouped into a single town.
Max Error % is no longer shown in the App.
Selecting a Town shows results at a trial level.
View Details - When a count of trials is clicked on in any cell (at state, region or dynamic region level) a popup windows is displayed and shows the trial details that make up that count, showing Town Name, and the trial year of each trial in that town.
Tapping the View Details button when a Town is selected displays a popup window showing information about all the trials at that town and a link to download the Trial Report from the NVT Online communications website (nvtonline.com.au) is provided for released trials.
Tapping the Info icon in the column heading when a Town is selected displays a popup window showing Mean Yield t/ha, Trial Code and a link to download the Trial Report from the NVT Online communications website (link provided for released trials only).
Tapping the Info icon next to a variety (shown at any level other than national or dynamic region crossing state boundaries) displays a popup window showing the “Variety Passport” information for the appropriate state
When a Town is selected
•The y-axis value is the BLUP of the Variety for each Trial in that Town
•The x-axis legend shows Trial details: Mean Yield t/ha, Trial Year + Alpha code, Site
•Error Bars display the PEV values for the individual variety
A Sort button allows the graph to be sorted by either
•Trial Year (ascending) and then Mean Yield (descending), or
•Mean Yield (descending)
A “Hide Error Bars” checkbox allows error bars to be hidden. By default, the checkbox is not checked and Error Bars are displayed.
A filter button in the top menu bar allows the display of results, and data included in calculations to be refined. All filter options apply to the Graph as well as the Table Report.
Filters affecting display of data only: (data used in calculations is not affected, only Variety Sort Order is affected)
1)“Hide Untested Predictions”
If “Hide Untested Predictions” Include All tickbox is ticked, then only the values for varieties that were tested in at least one trial within the selected area are displayed.
2)“Years Tested in NVT”
Filters affecting data included in calculations:
1)VAF - If “Variance Accounted For” Include All tickbox is unticked, then the data records used for calculations only include those records with a VAF value greater than or equal to the value selected by the slider control.
2)PEV - If “Predicted Error Variance” Include All tickbox is unticked, then the data records used for calculations only include those records for Trials with an absolute average PEV value greater than or equal to the value selected by the slider control.
b)Accuracy - If “Accuracy” Include All tickbox is unticked, then the data records used for calculations only include those records for Trials with an average Accuracy value greater than or equal to the value selected by the slider control.
c)Sowing Date - If “Sowing Date” Include All tickbox is unticked, then the data records used for calculations only include those records for trials with sowing dates within the selected Month/Day range.
When a link in the navigation breadcrumb above the table or graph is clicked
•The table/graph results are displayed for the selected region, state or national level.
•The map zooms out to the appropriate level
A town marker can be tapped to select it and show the report just for the trials(s) relevant to that location
Known Issues: Print report menu item is not operational.
Ratings and Reviews
- Australian Crop Accreditation System Limited
- 40.3 MB
Requires iOS 6.0 or later. Compatible with iPad.
- Age Rating
- Rated 4+
- Copyright © 2013 Australian Crop Accreditation System Limited
Up to six family members will be able to use this app with Family Sharing enabled.