TREE BREEDING AUSTRALIA

Blue Gum (Eucalyptus globulus) Breeding Values

To assist growers in understanding what breeding and index values represent, TBA provides a brief summary and a more detailed explanation to assist. 

Brief summary

A TBA seedlot or genotype will have Estimated Breeding Values (EBV's) provided to describe its genetic worth. The following table indicates the information provided and what it means:

TREEPLAN run name EGLOB2020-02-27 TREEPLAN analysis (EGLOB=E. globulus, 2020=year, 02=month, 27=day).
The later the year the better the predictions (improved models and more data) – only compare EBVs in the same TREEPLAN analysis.
EBV heading EBV value EBV comments
National (Aus) Estimated Breeding Values (EBV) – Primarily used for breeding decisions but can also be used for deployment. A set of EBV's includes an $NPV value (INDEX), and the 3 genetic traits impacting performance (growth=Volume, wood density=Density and Kraft Pulp Yield=KPY).
INDEX ($NPV) $1425 NPV=Net Present Value, 7% = discount rate.
This represents the additional $NPV ($1425) per ha at harvest, this genotype/seedlot will provide using a national production system and an 7% discount rate. A single INDEX for each seedlot/genotype allows ranking of genetic material (higher INDEX=more profit) based on a set production system.
Volume ebv 27.34 Volume=volume production in cubic metres per hectare.
Volume EBV's are produced for the national breeding program. This represents the estimated breeding value of 27.34 m3/ha above baseline production of 229 m3/ha. Overall Volume for this seedlot/genotype = 256.34 m3/ha (229 + 27.34). Used for comparative purposes against other seedlots/genotypes and also primarily for breeding decisions.
Density ebv 13.53 Density=core basic density in kg/m3 from baseline production of 537 kg/m3. This seedlot/genotype has an average Density of 550.53 kg/m3 (537 + 13.53).
KPY ebv -0.22 KPY=Kraft Pulp Yield in % from baseline production of 55.7%. This seedlot/genotype has an average KPY of 55.48 % (55.7 - 0.22).
Regional INDEX (NPV) values and Volume EBV's – Primarily used for deployment decisions. A set of EBV's includes an Index ($NPV) value, and 3 genetic traits. As Density and Kraft Pulp Yield do not change across regions, the national ebv's for these traits apply, therefore only Index (NPV$) and Volume are different.
GIPP = Gippsland, GTR = Green Triangle Region, Tas = Tasmania, WA = Western Australia.
Index Gipp $1656 Regional $NPV are described similarly to the above national Index, but using regional Volume ebv's to calculate the Index.
Generally the NPV will be higher for regions, as Australia wide results are primarily for breeding decisions whereas regional results are more useful for deployment. It’s the same seedlot/genotype but re-valued for use in a specific region.
Index GTR $1604
Index Tas $1563
Index WA $1529




Volume Gipp ebv 29.01 Regional Volume EBV's are similar to the national Volume ebv's but reflect the Volume of each regional area when using the same seedlot/genotype. The same baseline of 229 m3/ha applies. More appropriately used for deployment decisions as an additional indicator to the regional Index to understand predicted growth levels.
Volume GTR ebv 28.35
Volume Tas ebv 30.1
Volume WA ebv 27.53





Detailed explanation

The genetic merit of a seedlot or genotype (tree) allows decisions when selecting material for use. TBA uses Breeding Values and Index Values to provide this comparative performance information. This lay person explanation is to quickly assist a decision maker to understand what the numbers represent and how they may be interpreted. There is significant science and statistical data behind the values produced and TBA can assist with a more detailed understanding. 

TBA is seeking to improve performance over time by identifying genotypes (trees) with better characteristics (traits) for future use. TBA breeds new varieties and tests the genetic performance of individual trees (genotypes) over time and uses their performance as well as the performance of relatives to assist in predicting genetic worth. 

An annual TREEPLAN genetic evaluation is undertaken to allow us to report the genetic merit of genotypes (or seedlots) for each trait we are seeking to improve. We use early age assessments to estimate the genetic worth of each trait at harvest. Environmental conditions and silviculture regimes will also impact actual performance achieved. We are interested in describing the genetic worth after using replicated experimental design to remove performance noise caused by environmental or silvicultural impacts.

Each seedlot or genotype (tree) is allocated an Estimated Breeding Values (EBV) for each trait of importance. 

Volume EBV (see note) Volume growth  growth in cubic metres per hectare
Density EBV Core basic density density in kg/m3
KPY EBV Kraft Pulp Yield KPY in percentage

Volume note - TBA reports a national Volume EBV for breeding decisions as well as other regionally based Volume performance for deployment decisions. These include Green Triangle Region (Volume GTR EBV), Tasmania (Volume Tas EBV), Western Australia (Volume WA EBV), Gippsland (Volume GIPP EBV). Growers should utilise the most appropriate MAI based on their end use

These EBV's provide a performance estimate for each trait, however when multiple traits are involved, decision making becomes more complex. Some traits can be negatively correlated, for example chasing growth can reduce density, so a balance is needed. 

TBA looks to maximise profit based on a defined production system (product end use). The production system identifies a single value in Net Present Dollars (NPV) for improvements to each particular trait for the defined end use. As each improved trait has a profit value for each unit of improvement, TBA can then combine the value of each trait into a single $NPV index value for each seedlot/genotype. This single NPV INDEX value estimates the genetic worth of each seedlot or genotype and the higher the NPV the more profitable the genotype for the specific end use modelled. It also allows all the genetic material in a TREEPLAN analysis to be assigned an INDEX NPV and rank the list from best to worst for a specific production system. 

Recommendation 1: the higher the estimated Index ($NPV) for a seedlot or genotype, the better genetic merit of that seedlot or genotype for the specific production system (product end use). 

It should be noted that TBA uses a national production system to allow the breeding program to identify new elite material for use, so a national INDEX describes the genetic material. TBA can also provide regional or customised production systems for use by growers depending on their end use and economic objectives. Therefore it remains important to understand the underlying production system being used to establish the INDEX value and if comparing seedlots or genotypes for use, you need to compare them using the same underlying production system.  

An annual TREEPLAN analysis is undertaken using all the historical data and any new performance data collected in the last year to re-describe the genetic worth of each genotype. 

Recommendation 2: identify the year of the TREEPLAN analysis and use the latest TREEPLAN run as more data and refined models may improve the predictions.

Recommendation 3: To avoid comparing apples and oranges, only compare breeding values if they are derived from the same TREEPLAN analysis.

For Australia (national) use, we currently predict breeding values against a national baseline productivity as provided in the following table. Some clients have developed customised baseline production models to suite their plantations and describe the genetic worth against their own models and the average baseline values may be different to the TBA nation baselines used. The same baseline productivity estimates are used for regional EBV's reported. 

Blue Gum - national baseline

Traits reported Baseline productivity Comments
Volume (m3/ha) 229 Average cubic metres of wood removed at rotation harvest
Density (kg/m3) 537 Average density of the wood in kilograms per cubic metre
Kraft Pulp Yield (%) 55.7 Average percentage of kraft pulp yield in the woodchips

The systems we use to predict EBV's are flexible to accommodate different users and different productivity systems. This allows us to rescale the EBV's depending upon user needs. 

To understand the EBV for each individual trait, we report the genetic difference of the genotype or seedlot and not the total productivity. For example, with growth (Volume), a Volume EBV of 25 means the genotype/seedlot will be 25 units better than the baseline level of production. TBA reports against an average baseline of 229 m3/ha meaning an average growth of 229 cubic metres per hectare and any reported Volume EBV variance will be above (or below = negative) this baseline. Therefore with a Volume EBV of 25, this genotype has a total predicted Volume at harvest of 254 (229 + 25). The baseline production can vary depending on the EBV's being reported and it remains important to understand the baseline being used for each trait.     

Recommendation 4: If you are looking to appreciate each traits worth at harvest (rather than just its genetic comparison with other seedlots), seek the baseline production values and their foundation (is it national, region, site type, etc) and if it is related to the area you intend to grow the material.

Traits can be negatively correlated, so chasing growth alone may reduce other traits and the resultant logs may not maximise profit. Each genotype/tree/seedlot has average EBV's for its production but generally, a grower is trying to improve profit, which is trying to identify the best combination of traits together for their production system. We provide an $NPV for each genotype or seedlot so users can compare the merit of material in the same analysis but it must be noted that the $NPV is directly related to the production system defined and end use. A grower may be selling his wood for fence posts, so his end use if different to grower selling wood for structural housing use. The production model and $NPV will be very different as the fence post producer may only chase growth (MAI) and not be interested in wood strength (STIFFNESS) so the economic model would be set up to recalculate the $NPV index and re-sort the rankings of genotypes with better characteristics for that specific use.