Main

It has been shown that seeds of many species can benefit from being cooled and dried. Their longevity over a range of conditions can be described by an improved seed viability equation3, although there are limits to the range of conditions over which this equation applies. For long-term seed storage, it is the lower temperature and moisture-content limits that are of most concern. The lower temperature limit is probably about 20 °C (ref. 3), although experiments at high temperatures (necessary to reduce longevity at low moisture contents to only a year or so) show that the low moisture-content limit is that produced by equilibrating seed at 10-12% relative humidity at 20 °C before storage4. It is this protocol, combined with storage at ambient or cooler temperatures, that has been described as ‘ultra-dry’ seed storage5.

Although practical tests have indicated the benefits of ultra-dry storage6, 7, some controversy remains. It has been reported that ultra-dry storage may adversely affect seed vigour at certain temperatures8, although otr results have not identified this problem9.

The results of Zheng et al.1 support the practical application of ultra-dry seed storage as a low-cost, simple technology for gene banks. However, we question their suggestion1 that this treatment provides longevity equal to that of more conventional gene banks. In some species at least, ultra-dry seed storage at ambient temperatures results in more rapid loss of seed viability than storage at about 5% moisture content at -18°C (ref. 6).