Results & Discussion
Summaries of the results are presented in the following Tables and Figures.
1. Overall level of contamination
Results in Table 1 show that AWTA Ltd testing detected greater numbers of fibres of levels 5 and 6, and fewer fibres of levels 7 and 8 than detected by either of the SARDI tests. The average differences per 10g of sample were that, compared to SARDI, AWTA Ltd detected an average of:
- 9 (approx.) more fibres equal to or greater than level 5;
- 7 (approx.) more fibres equal to or greater than level 6; and,
|
Table 1. Average numbers of Dark Fibres/10g for all samples tested |
|
Dark Fibre Level |
AWTA Ltd |
SARDI
( AWTA Ltd Prep) |
SARDI
(SARDI Prep) |
|
5,6,7or 8 |
21 |
11.8 |
12.8 |
|
6,7 or 8 |
18 |
11 |
11 |
|
7 or 8 |
3.7 |
7.4 |
8 |
The inclusion of the AWTA Ltd Shirley Analyser into the SARDI procedure does not introduce a major change to the SARDI results, indicating that any extra contamination introduced by the AWTA Shirley Analysing procedure was minimal.
Figure 1. Relationship between Dark Fibre detection results for AWTA Ltd and SARDI (AWTA-prep). Click on the image to see a larger version.
2. Test variation
The numbers in Table 2 show that there is considerable variation in Dark Fibre counts between the replicate 10g samples. As expected, the variability between the labs is generally higher than the within lab variability. For Darkness levels 5 and 6 AWTA Ltd has greater variability than does SARDI, though the opposite is the case for levels 7 and 8.
Importantly for determination of very low levels of contamination, the between replicate variation is level dependent, being much lower at low contamination levels, as shown in Figure 2.
The AWTA Ltd Shirley Analysing step adds slightly to the variability between replicates of the SARDI testing procedure.
It is worth noting that the between replicate variability for the level 7 and 8 fibres of approximately 6df/10g to 8df/10g is high relative to the average level of contamination (4df/10g to 8df/10g).
Figure 2 . The increase in variation between replicate subsamples as Dark Fibre count increases for both AWTA Ltd and SARDI (AWTA preparation). Click on the image to see a larger version.
The medullation counts of the two laboratories are very similar as shown in Table 3.
|
Table 3. Average numbers of Dark Fibres/10g for all samples tested |
| |
AWTA Ltd |
SARDI
( AWTA Ltd Prep) |
SARDI
(SARDI Prep) |
|
Average values |
2.5 |
2.4 |
2.3 |
|
Differences between replicates |
3.9 |
2.4 |
2.7 |
3. Inspection time
Table 4 shows the average inspection times and the ranges of times from the trial. AWTA Ltd operators used an extra 40 minutes of time/10g sample on average. The average time to test 2x10g subsamples at AWTA Ltd is 4.5 hours.
|
Table 4. Inspection Times (hours:minutes) for 1x 10g Samples.
Note that 2 x 10g is the recommended rate of testing. |
|
|
AWTA |
*SARDI(AWTA-Prep) |
|
Average |
2:15 |
1:35 |
|
Minimum |
1:10 |
0:45 |
|
Maximum |
3:15 |
3:15 |
*Note: SARDI inspection time was for detection and classification only; the AWTA Ltd time also includes the verification phase.
4. Comparison with expected levels of contamination
Tables 5 and 6 show the levels of contamination for the three sets of samples with different levels of expected contamination, viz. 'control' (no contamination),' mild' and 'heavy'.
|
Table 5.
Average Dark Fibre Counts for the Three Sets of Samples with Different Levels of Expected Contamination. |
| |
Control |
Mild |
Heavy |
|
AWTA Ltd: |
|
|
|
|
5,6,7or 8 |
3.9 |
<16.0 |
29.4 |
|
6,7 or 8 |
2.6 |
13.1 |
24.7 |
|
7 or 8 |
0.6 |
3.0 |
<5.1 |
|
Medullation |
0.6 |
2.8 |
5.5 |
|
|
|
|
|
SARDI (AWTA-Prep): |
|
|
|
|
5,6,7or 8 |
1.4 |
7.7 |
17.4 |
|
6,7 or 8 |
1.4 |
7.6 |
16.0 |
|
7 or 8 |
0.9 |
4.8 |
10.9 |
|
Medullation |
0.7 |
1.3 |
3.5 |
|
Table 6.
Ranges of Dark Fibre Counts from the Three Sets of Samples with Different Expected Level of Contamination. Average of 2 x 10g Samples (Dark Fibres/10g rounded to nearest whole number). |
|
Classification |
Control |
Mild |
Heavy |
|
AWTA Ltd: |
|
|
|
|
5,6,7or 8 |
2 – 5 |
8–28 |
17 – 39 |
|
6,7 or 8 |
0 – 4 |
5–26 |
17 – 32 |
|
7 or 8 |
0 – 2 |
0 – 8 |
1 – 11 |
|
Medullation |
0 - 2 |
1 - 5 |
2 - 13 |
|
|
|
|
|
SARDI (AWTA-prep): |
|
|
|
|
5,6,7or 8 |
1 – 2 |
1– 21 |
9 – 29 |
|
6,7 or 8 |
1 – 2 |
1 - 21 |
8 – 26 |
|
7 or 8 |
0 – 2 |
0– 14 |
2 – 26 |
|
Medullation |
0 - 4 |
0 – 3 |
0 - 9 |
Table 5 shows the very clear distinctions between the three sets of samples based on their average values. Table 6, on the other hand, shows the level of overlap between the ranges of values for each set.
This overlap is shown clearly in Figures 3, 4 and 5 which show the Dark Fibre counts for the AWTA Ltd tests of the 1 x 10g sample, and the AWTA Ltd and SARDI (AWTA-prepared) lots as the average of 2 x 10g subsamples, respectively. In each case there was no overlap in Dark Fibre counts between the control and the heavy contamination sets, indicating that the procedures were providing the expected level of separation.
Click to view a larger version of the above image.
Click to view a larger version of the above image.
Click to view a larger version of the above image.
5. Operational issues for Dark Fibre testing
The Dark Fibre test is unlike any other test currently performed by AWTA Ltd. The technique incorporates a large degree of subjectivity, requires concentration and attention to detail by the operator, is prone to cross contamination at a level far more demanding than other tests conducted, and, in its current form, takes an average of 4.5 hours per test. The long time per test means that turnaround time for a Dark Fibre guidance test would be very slow if more than a few tests were received simultaneously.
These problems with the test appear insurmountable in the short term. Consequently the test will be expensive to perform routinely.
In this trial, 2 x operators worked an average of 2.25 hours per 10g sample, i.e. 4.5 hours per test and 9 x operator hours/test.
The cost of a Dark Fibre Detector is approximately $8,000 to $10,000 from CSIRO. The cost of a microscope is $3,000 to $6000.
The 2 x 10g Shirley Analysed subsamples could be obtained from the workflow associated with the existing Yield & Micron test with minimal extra effort.
Quality control of any Dark Fibre test would need to be strict and between lab monitoring would be required if the service were to be provided from more than 1 laboratory.
Download a PDF version of this FactSheet
FURTHER INFORMATION
Ian Ashman
General Manager Customer Relations
AWTA Ltd
Ph: 03 9371 4103
Fax: 03 9371 4191
Email: ian.ashman@awta.com.au