40
SEED TESTING INTERNATIONAL   www.seedtest.org
RULES DEVELOPMENT •
of the seeds according to the ISTA Rules (ISTA, 
2024). This was done as a check in case of any 
anomalous αw results and are not reported.
Survey
A survey was also prepared to better understand 
how widely αw is already being used in seed 
laboratories, including ISTA accredited 
laboratories, genebank and research laboratories. 
The survey questionnaire covered different areas 
such as type of laboratory and their current 
method of moisture content testing; if they would 
be interested in measuring αw; if they are already 
measuring αw; and if they would like to see αw 
measurement included as a method in the ISTA 
Rules. Links to the questionnaire were broadcast 
on the ISTA social media channels.
Results
Water Activity Measurements
Two brands of water activity instruments, 
Rotronic (www.processsensing.com) and 
Novasina (www.novasina.ch), were used by  
the six laboratories (Fig. 1). Two laboratories  
(B and F) used Novasina instruments with built-
in temperature control; their measurements 
were made exactly at the prescribed temperature 
of 20 °C. The mean recorded temperature was 
slightly higher for the other laboratories (highest 
mean temperature reported by laboratory D for 
the carrot measurements), but this source of 
variation did not have a systematic effect across 
the different moisture levels and crops, and hence 
is not considered significant in this data set.
For cabbage, the maximum absolute difference 
between the value reported by a laboratory and 
the mean was 0.026; for carrot, it was 0.039; 
and for onion, it was 0.067. There were some 
systematic errors. For example, laboratories B 
and F generally reported water activity results 
that were lower than the mean; laboratory D 
tended to report higher water activity at the two 
lowest moisture levels, and lower water activity 
at the two highest moisture levels. Nonetheless, 
overall the repeatability within a laboratory was 
generally very high with low standard deviations 
for any given moisture level, crop and laboratory. 
Larger standard deviations were, however, 
observed for laboratory F, which may reflect a less 
efficient loading process. Reproducibility across 
laboratories was also acceptable.
The measurement times recorded by the 
laboratories ranged from <5 to 45 min.
Survey
There was a total of 31 responses to the survey, 
primarily (20 responses) from people working 
in a seed testing laboratory, seven of which 
were ISTA accredited for moisture content 
determination. Eleven of the 31 respondents said 
that they were already measuring water activity, 
six of these were from seed testing laboratories, 
three from seed (gene) banks and two from 
university or other research laboratories. These 
measurements were on seeds from diverse 
species: cereals, legumes, trees and shrubs, 
and native or wild species. Some of these user 
respondents commented that they find it easy 
and efficient to assess seed moisture status using 
this method, with a clear advantage that it is 
non-destructive, so seeds are not wasted. Some 
of those not already measuring water activity 
were concerned that it might be expensive to 
introduce, but the main limitation was that there 
is no demand for water activity measurement.
Conclusions
The results of this study suggest that, provided 
the water activity meter is within calibration 
and temperature is controlled, water activity 
measurements are highly repeatable (within a 
laboratory and on the same instrument). Water 
activity measurements are also reproducible 
across laboratories using different instruments. 
Seed water activity is already being applied in 
some seed laboratories. Many seed (gene) banks 
measure water activity to confirm that seeds 
have dried enough to be packed for storage. Some 
seed companies are also using water activity 
internally during seed processing and research. 
However, since ISTA has not recommended its 
use, companies face challenges in adopting 
this method more broadly, particularly at the 
seed production stages. The results of this study 
represent a significant step towards establishing 
water activity measurement as an official ISTA 
method.
Additional validation rounds could consider 
other types of seeds and different moisture 
levels. Also, the requirement for cutting larger 
seeds or seeds with physical dormancy could be 
included. Further work has already explored the 
measurement of αw for coated seeds (Vijayan et 
al., 2025), in which it was concluded that, in the 
case of film-coated seeds, measurements of αw 
provide an accurate indication of the αw of the 
seeds but the same cannot be said for pelleted 
seeds.
Acknowledgements
We gratefully acknowledge Jean-Louis Laffont 
and Kirk Remund (ISTA Statistics Committee) for 
analysing the data.
References
1. Hay, F.R., Rezaei, S. and Buitink, J. (2022). 
Seed moisture isotherms, sorption models, and 
longevity. Frontiers in Plant Science, 13, 891913.
2. Hay, F.R., Rezaei, S., Wolkis, D. and McGill, 
C. (2023). Determination and control of seed 
moisture. Seed Science and Technology, 51, 
267–285.
3. ISTA (2024). International Rules for Seed 
Testing. International Seed Testing Association, 
Wallisellen, Switzerland.
4. Vijayan, S.S., Rezaei, S. and Hay, F.R. (2025). 
Water activity-moisture content relations of 
primed, film-coated and/or pelleted vegetable 
seeds. Seed Science and Technology, 53, 197–210.
Figure 1. Water activity (αw ) of seeds of three vegetable crops equilibrated to four different moisture 
levels and tested by six different laboratories (A–F). Within each species, the laboratories are ordered 
according to the recorded measurement temperature. Values are the mean ± s.d. of measurements 
made for three replicate samples. Horizontal lines represent the means for the crop at each 
respective moisture level.

View this content as a flipbook by clicking here.