39
SEED TESTING INTERNATIONAL APRIL 2026
• RULES DEVELOPMENT
Inter-Laboratory Reproducibility of Water Activity 
Measurements
Fiona R. Hay1, Jayanthi Nadarajan2, Deborah de Souza Vidigal3, Laura Bowden4, Ruud Nabben5, Marco Porceddu6,  
Gianluigi Bacchetta6, Maria Enrica Boi6, Siem Reijnders7, Céline Herbert8 and Paul Keijzer9
1Department of Agroecology, Aarhus University, Slagelse, Denmark
2New Zealand Institute for Bioeconomy Science Ltd, Food Industry Science Centre, Fitzherbert Science Centre, Palmerston North, New Zealand
3Bejo Zaden BV, Warmenhuizen, The Netherlands
4SASA, Edinburgh, UK
5BASF Nunhems, Nunhem, The Netherlands
6Sardinian Germplasm Bank (BG-SAR), Department of Life and Environmental Sciences, University of Cagliari, Italy
7Pedak Meettechniek, Heythuysen, The Netherlands
8GEVES, Beaucouzé cedex, France
9Rijk Zwaan BV, De Lier, The Netherlands
THIS ARTICLE REPORTS THE RESULTS 
OF A SPECIAL PROJECT carried out by 
the ISTA Moisture and Storage Committees. 
Samples of three vegetable seeds equilibrated 
to four different levels of moisture were tested 
for water activity by six different laboratories. It 
was found that the results were both repeatable 
within a laboratory and reproducible between 
laboratories. These results are an important 
step towards introducing measurement of water 
activity into the International Rules for Seed 
Testing (ISTA Rules). A survey of ISTA accredited, 
genebank and research laboratories was also 
conducted as part of this Special Project. Water 
activity is already being measured by some 
laboratories, because of the advantages of this 
approach compared with moisture content 
determination. Wider adoption is limited 
because this method is not yet in the ISTA Rules.
Introduction
An ISTA Special Project carried out by the ISTA 
Moisture and Storage Committees, with the 
collaboration of ISTA accredited laboratories 
and genebank/research laboratories, aimed to 
evaluate the reproducibility of water activity 
measurements across different facilities and 
using different instruments, and to better 
understand demand for the introduction of water 
activity into the ISTA Rules or as an official ISTA 
method. Water activity (αw) is measured using a 
water activity meter and is usually expressed on 
a scale of 0–1, though it may also be converted 
to ‘equilibrium relative humidity’ (eRH) by 
multiplying by 100 to give a percentage scale 
(Hay et al., 2022, 2023). This method is widely 
used as an alternative to moisture content, in 
the food industry, in particular. It is considered 
more insightful than moisture content since it 
provides an indication of the availability of water 
for chemical reactions to take place, especially 
during seed development, germination 
and storage. Furthermore, it can be applied 
universally since the type and relative rate of 
reactions along the αw scale will be consistent for 
seeds of different species. Measuring αw also has 
the advantage of being non-destructive, relatively 
quick and simple, typically requiring from 7 to 30 
min per sample. This Special Project was divided 
into two parts, where part one looked at the 
water activity measurements for three selected 
vegetable seeds (i.e. cabbage, carrot and onion), 
while part two involved a survey on the use of αw 
in seed laboratories.
Materials and Methods 
Water Activity Measurements
Seeds of cabbage, carrot and onion were 
provided by Bejo Zaden BV, The Netherlands. 
For each species, seeds were divided into four 
lots. Three of these lots were equilibrated to 
different moisture levels by placing them at 45, 
65 or 70% relative humidity at 20 °C. The fourth 
lot of each species was used as provided (αw 
0.30–0.35), hermetically sealed in a Mabfoil 
trilaminate-aluminium foil packet. Following 
equilibration, each lot was subdivided into 18 
samples of approximately 5 g and vacuum sealed 
in Mabfoil trilaminate-aluminium foil packets. 
These packets were then sent to each of the six 
laboratories: GEVES (France), the Sardinian 
Germplasm Bank (Italy), SASA (UK), and BASF 
Nunhems, Pedak Meettechniek and Rijk Zwaan 
(The Netherlands). Thus, each laboratory received 
36 packets (three species ´ four moisture levels 
´ three replicates). Packets were randomly 
numbered to avoid any systematic error.
Each laboratory was asked to measure the 
αw of the seeds at 20 °C and to measure the 
sample from each packet twice. They were then 
requested to determine the moisture content 

View this content as a flipbook by clicking here.