The Precision of OFDA2000 and FLEECESCAN for Estimating the Diameter Characteristics of Fleeces: A Case Study

Abstract

A number of authors have reported that the precision of the measurement of a fleece will have an impact on the decisions taken on-farm with respect to sheep selection for on-going breeding or culling and fleece selection for preparing classed lines of wool of different Mean Fibre Diameter (MFD) for sale. Determination of the precision of MFD for a fleece test has been the subject of many research studies All of these relate to the precision of a single test conducted at one site on the animal, whereas the precision that is important is the precision of the measured value as it represents the entire fleece. The precision of any value based on a calculation for one site will underestimate the precision relevant to the entire fleece, as it inevitably does not include the variation over the fleece. This paper summarises the first attempt to validate a model, which aims to represent the precision of the entire fleece. Any estimate of precision will be dependent on the procedures used to determine the average value. As any change in the procedures is likely to have an impact the precision, quoting a 95% confidence limit without the detailed procedures that accompany it is meaningless.

The 95% Confidence Limits relevant to the entire fleece were determined for three different measurement systems, as they are used commercially in Australia, as follows:

Notes:
^a The Flank sampling is in very close proximity to the commonly used
   pinbone site and is provided here for the purposes of comparison.
^# The Between-Sheep Ranges of SDD and CVD for the OFDA2000 measured
   fleeces were lower than the Between-Sheep Ranges measured by
   FLEECESCAN and Laboratory LASERSCAN.

These Confidence Limits are relevant to the sampling procedures and laboratory practices of the participating testers. The Confidence Limit will be widened if the number of measured fibres is less than that used in this study. Some FLEECESCAN operators are known to measure 600 fibres. This will result in a small increase in the reported Confidence Limits (i.e. ± 1.02µm to ± 1.04µm for MFD). Unlike the traditional midside sampling and Laboratory testing no Industry agreed Standards exist for either the OFDA2000 or FLEECESCAN. Hence, wool producers need to be aware that individual operators can determine their own sampling and testing procedures.

The commercial implications of the results presented above relate to sheep selection and fleece classing. The lower the Confidence Limit the better will be the ability to differentiate between individual fleeces and individual sheep. In past times, many wool producers have used Laboratory testing of midside samples to improve their flock. As the uniformity of the flock increases the further economic gains will be limited by the Confidence Limit. Wool producers will need to take this into account when deciding on what sampling and testing procedures should be used on their flock. For example, it would be commercially beneficial to test the breeding rams at a higher sampling and testing frequency (i.e. measure more specimens with FLEECESCAN or more sites with OFDA2000) than the breeding ewes.

Citation

"The Precision of OFDA2000 and FLEECESCAN for Estimating the Diameter Characteristics of Fleeces: A Case Study", K.A. Hansford , J.W. Marler & I.M. McLachlan, Paper 36, Wool Industry Science & Technology Conference, October 2002, Hamilton, Victoria, Australia