29
SEED TESTING INTERNATIONAL APRIL 2026
β€’ RULES DEVELOPMENT
Analysis of Data for Three Seed Lots 
Tested by 11 Laboratories 
Comparison of box plots revealed that the data 
from one laboratory in Europe was markedly 
lower than for other laboratories. This did not 
appear to be related to the test conditions used 
and suggested that there was a problem in 
evaluation of RE. This laboratory was therefore 
excluded from further analysis.
RE counts after 72 h at 20 Β°C revealed small 
differences in the mean data for the three 
seed lots, as was seen for the data from the five 
laboratories in India (Table 3).
Tolerances: Only three tests out of 30 (ten 
laboratories Γ— three lots) were out of tolerance, 
with outliers in laboratories 3 (lot 1), 5 (lot 2) and 
9 (lot 3).
Repeatability (β¨π‘Ÿ  ) and reproducibility (⨍𝑅 ) 
values are presented in Tables 7 and 8. The value 
for β¨π‘Ÿ  (1.04, Table 7) is close to 1.0 and is therefore 
acceptable. The value of 2.14 for ⨍𝑅 (Table 8) is 
close to that theoretically derived using the 
over-dispersion factor between laboratories 
established by Miles (1963).
Conclusion
Seven seed lots tested within five laboratories 
in India were ranked similarly and achieved 
acceptable repeatability and reproducibility. 
Analysis of the data for three of these lots 
when tested in five Indian and five European 
laboratories also revealed acceptable 
repeatability and reproducibility. This provides 
support for the inclusion of Allium cepa as a 
species to which the RE test can be applied.
Acknowledgements
We are grateful to Jean-Louis Laffont for 
completing the statistical analysis and his advice 
on interpretation of the data. We thank the 
participants in the comparative test in both India 
and Europe, particularly for their patience as we 
encountered problems with the availability of 
seeds. 
3. ISTA (2023). International Rules for Seed 
Testing. International Seed Testing Association, 
Wallisellen, Switzerland.
4. Miles, S.R. (1963). Handbook of tolerances 
and measures of precision for seed testing. 
Proceedings of the International Seed Testing 
Association, 28(3), 168 pp.
5. Powell, A.A. (2022). Seed vigour in the 21st 
century. Seed Science and Technology, 50 
(1 Suppl.), 45–73. doi.org/10.15258/
sst.2022.50.1.s.04
Table 6. Reproducibility estimates for seven lots of Allium cepa using a linear mixed model
Table 7. Repeatability estimates for three lots of Allium cepa using a linear mixed model
Table 8. Reproducibility estimates for three lots of Allium cepa using a linear mixed model
Excluded 
samples %
72
40
2.09
1.78
𝑝𝑝̅… 
𝑓𝑓) 
𝑓𝑓*+,-. 
78
8.50
6.78
6.15
1.04
11.30
𝑝𝑝̅… 
𝜎𝜎$"#$ 
𝜎𝜎$"%&Γ—"#$ 
𝑆𝑆( 
𝑓𝑓( 
𝑆𝑆) 
Excluded 
samples %
81
30
2.14
1.71
𝑝𝑝̅… 
𝑓𝑓) 
𝑓𝑓*+,-. 
r = 0.716*
y = 33.5 + 0.626x
Appendix 1: Relationship between radicle emergence and field emergence in Allium cepa 
(unpublished data)
r = 0.716*
y = 33.5 + 0.626x
References
1. Demir, I., Kuzucu, C.O., Ermis, S., Memis, 
N. and Kadioglu, N. (2022). Estimation of 
seed longevity in onion seed lots by a vigour 
test of radicle emergence test in artificial 
ageing conditions. Horticulturae, 8, 1063. doi.
org/10.3390/horticulturae8111063
2. Ermis, S., KarslΔ±oglu, M., Ozden, E. and Demir, 
I. (2015). Use of a single radicle emergence 
count as a vigour test in prediction of seedling 
emergence potential of leek seed lots. Seed 
Science and Technology, 43, 308–312. doi.
org/10.15258/sst.2015.43.2.16

View this content as a flipbook by clicking here.