A genetic evaluation system for tree breeding (TREEPLAN®)

The effective deployment of improved genetic material in plantations depends on the ability to accurately choose elite genotypes. This involves managing large amounts of information on the attributes for a very large number of potentially valuable parental trees from different generations of breeding. Rapid improvement of genetic stock requires constant trial assessment, predictions of genetic worth and decisions about which tree parents will produce the best offspring when mated. Traditionally, this process has been sub optimal, due to resource constraints and inadequate genetic evaluation methods delaying decision-making and hence genetic progress in plantations.

To improve this situation, a new software system has been developed by the Southern Tree Breeding Association (STBA) in collaboration with scientists from the Animal Genetics and Breeding Unit, University of New England, and from the Breeding Strategies Project of the CRC-SPF. The CRC-SPF has incorporated theoretical developments in model building and trait mapping. The integrated system consists of two components: a database for maintaining all information from the STBA's national tree improvement programs (STBA-DMS) and a genetic evaluation program (TREEPLAN®).

TREEPLAN® uses the most sophisticated genetic evaluation algorithms available to cope with the diversity of information contained in the national tree improvement programs for Eucalyptus globulus and Pinus radiata. The TREEPLAN®evaluation system is flexible, allowing for new sources of information and analysis options.

Major obstacles for national evaluations are the sheer number of trees included in the process and site heterogeneity. In addition, evaluations need to be updated regularly as the STBA collects new information. TREEPLAN®overcomes these problems by:

• using reduced individual tree models;
  
• mapping assessed trials to a small number of key traits that have the biggest impact on profit, thus focusing on the financial impact of decisions;
  
• allowing for multiple site qualities, ages of measurement and measurement systems;
  
• using models that are tailored to the measurements and experimental design of each site; and
  
• grouping similar genetic material to generate comparable information for each different population being assessed.

There are plans for further development of the system to include molecular information, non additive genetic effects and new statistical methodologies. Equally important will be improvements in usability for tree breeders. The system is currently being tested using the latest national evaluations for Blue gum and Radiata pine.