Cost considerations

Cost components

Email från Ola November 07: Vi skall leverera en lista på åtgärder som våra tekniker skall kostnadsberäkna. Målet är att komma igång med det under mötet. Gunnar kan lämpligen förse Curt med de material han och Björn hade i sin analys.  Dag har dragit ur oss uppgifter under åren och har kanske en sammanställning.

One of Dag's pets (which at least Ola bought), is that cost components are more important for optimal decisions than e.g. genetic parameters. Cost components are in principle easy to construct, but in practice not that easy. The meaning and philosophy should be understood. For optimisation generally the relations among costs is the important matter. The total cost of forest tree breeding is important only for lobbying for increased funding, and that is better done with rougher means. The lobbying done since 1995 has evidently not worked well as the resources for operative tree breeding has been reduced the last decade. But to get reliable relative cost components and make the exercise instructive and be able to use these cost components operative. Cost components are suggested to be used to give a rough idea "do we do right or should we change in some direction", the instrument may be to dull and complicated for every use. 

The sum of the costs should add to the budget. E.g.. if the annual budget for long term breeding is (200000 SEK)/ (breeding population), when this figure should be possible to obtain from the cost components. Say that where are two cost components: per family and per tree. When the total cost is obtained by adding. But there are other costs and it may be traps in calculating overheads and hidden costs. What costs are included in per tree cost, is it plant production, or all costs from seed to final measurement? What is relevant is marginal costs, thus what is the added cost for adding something. E.g. what is the marginal cost for a field site more, thus to have the same number of plants on two different sites instead of one. The genetic gain can be predicted as a function of among other things the number of sites and plants and families. If marginal costs of adding sites and plants and families are known, the allocation of resources on number of sites, families and test plants which maximizes gain can be found. If no such cost components are known, no efficient reply on questions like how many sites, test plants and breeding population members should we use.  

Of course cost estimates are useful in themselves, and for that reason the break down on marginal cost for an added unit are not that needed.

Cost components which have been used by Lindgren and collegues in studies:
    - Additional cost per experimental plant
    - Additional cost per field test site
    - Additional cost per pollen parent
    - Additional cost per seed parent
    - Additional cost for a graft in a seed orchard
    - Additional cost for a cutting in a seed orchard

All cost component are built up by smaller units and it could be practical to have a more detailed list.

The cost components I remember I have used in studies are as follows:

Note some cutting cost estimates by Sonesson (2002) at

Results and experiences from the Central Swedish Clonal Forestry Program

Default values for "breeding cycler"

Now we start to discuss the use of markers in operative breeding. Estimates of costs are needed to see if it is meaningful to go on with discussions. I suggest as heuristic rules that the cost is set to 100 000 SEK as fixed cost to initiate analyses, and 200 SEK as additional cost for labwork including labor per genotype.