Previous Page Table of Contents Next Page

Ultra-dry technology: perspective from the National Seed Storage Laboratory, USA - Walters, C.

Christina Walters

United States Department of Agriculture, Agricultural Research Service, National Seed Storage Laboratory, 1111 South Mason Street, Fort Collins, CO, USA

About 10 years ago, one of the seed technologists from the Seed Storage and Viability Unit came to me with a problem. She had been trying to prepare sunflower seeds for storage. Standard procedures at the NSSL at the time were to dry seeds at 5°C and 10% RH until the water content was about 5%. But the sunflower seeds would not comply; it had been about 3 months and the water content of the seeds was still less than 3%. That water content was not within the standards used at the NSSL (i.e. seeds must be adjusted to water contents of 5±1%). What should she do? To meet the standards, the seeds would have to be exposed to a relative humidity of 75 or 80%; yet clearly that would be foolish since we know that pathogenic fungi flourish at these high relative humidities. The problem was not in the sunflower seeds, it was in how we defined our seed storage standards.

In the fall of 1987, I gave a seminar at the University of Wyoming on how water affects physiological activities in seeds and how there are critical water contents that change the nature of chemical reactions. Dr Bill Smith of the Botany Department asked me why all of my measurements were expressed in terms of water content; why not use terms with established relevance in plant water relations - such as water potential. My blithe response was that water content was easier to measure.

Around the same time, the National Seed Storage Laboratory was struggling for funds for sorely needed additions and renovations. We were nearing capacity (about 275 000 accessions) and were expecting a total collection of over 1.5 million accessions by the year 2050. Our storage vaults had originally been designed for use at 5°C, but were retrofitted for storage at - 18°C about 12 years previously. The energy used to maintain the retrofitted freezer units was enormous and the maintenance was astronomical. In the early days of the NSSL (the 1960s and 1970s) seeds were dried with heat, but that was shown to be damaging to them, so a far gentler, albeit slower, drying protocol was introduced, using a dryer held at 5°C, 10% RH. Even in Fort Collins, Colorado, with an ambient yearly RH of 30%, it was difficult to maintain 10% RH at 5°C and constant repairs of the seed dryers were required. Clearly, changes were needed and a comprehensive statement of why seeds are handled in a particular way was necessary to legitimize the expense.

These three stories were the background to the paper by Vertucci and Roos (1990), Theoretical basis for protocols of seed storage', which apparently launched a decade of controversy. In the paper, we resigned ourselves to extrapolation, since the data we needed to determine optimum storage conditions for seeds simply did not exist. In this case, extrapolation was not such an unforgivable transgression as the necessary data would take a couple of hundred years to collect. The basis of the extrapolation was thermodynamic principles, mostly because they had been worked out years before, on a myriad of systems - polymer and colloidal surfaces, proteins, membranes and foods. We felt that the conclusions arrived at in the paper could be supported by these thermodynamic principles as well as the existing data on seed aging kinetics. What is more, we felt that the conclusions were rather self-evident, as they resulted in no real changes in how seeds should be prepared for long-term storage: drying protocols based on relative humidity did not substantially change the IBPGR (1976) standards of 5±2% water, they just specified which seeds should be dried to 3% and which tolerate higher water contents. Essentially, the paper argued towards switching from water content-based to RH-based drying protocols, and was very similar in scope to a paper published just months before by Roberts and Ellis (1989). We were surprised, therefore, by the attention that the first Theoretical basis' paper received (Ellis et al., 1991; Smith, 1992).

Initially, there were two scientific issues that were considered controversial in our paper. The first was an elaboration of an earlier discovery that removal of what was then called 'strongly bound water' damaged plant tissues (Vertucci and Leopold, 1987b). The second disagreement dealt with the water content at which this occurred. All of our studies showed that water was 'strongly bound' to tissues at RH as high as 20-30% (Vertucci and Leopold, 1984, 1986, 1987a; Vertucci et al., 1985; Leopold and Vertucci, 1986, 1989; Vertucci, 1990; Vertucci and Roos, 1990) and our aging experiments demonstrated the existence of a water content optimum within this relative humidity range (Vertucci and Roos, 1990). The coincidence was sufficiently compelling to warn of the dangers of over-drying seeds. Subsequent reports of more rapid deterioration in very dry seeds and pollen (Carpenter and Boucher, 1992; Vertucci et al., 1994; Ellis et al., 1995; Dickie and Smith, 1995; Buitink et al., 1998; Chai et al., 1998; Hu et al., 1998; Kong and Zhang, 1998; Shen and Qi, 1998) clearly demonstrate the potential risk of over-drying. The questions remain, what is the water content at which this occurs, and what is the corresponding relative humidity?

A cursory examination of the data cited above and those produced by Ellis and colleagues (Ellis et al., 1988, 1989,1990,1991,1993,1996) suggests that the incidence of optimum water contents is so inconsistent that the above questions cannot be addressed now. We believe that the apparently conflicting results can be explained on the basis of temperature differences among storage experiments and experimental design. Basically, we believe that optimum water contents for storage change with storage temperature: they are very low at high temperatures (where most of the experimental work has been done, e.g. Ellis et al., 1988, 1989, 1990) and they increase as the temperature for storage decreases (where fewer studies have been attempted). In other words, we believe that the reason why Ellis and colleagues concluded that 'ultra-dry storage is unlikely to be hazardous' (Ellis et al., 1989) is that they performed their experiments at 65°C and this high temperature masked any deleterious effect that over-drying may have caused. This point of view has been argued in several publications (Vertucci and Roos, 1991, 1993a,b; Vertucci et al., 1994; Buitink et al., 1998; reviewed by Walters, 1998) and the interested reader is referred to these works for a more detailed explanation. In studies at lower temperatures, deleterious effects of over-drying could not possibly be observed because only one water content was studied (Ellis et al., 1993) or the lowest water content studied was near the predicted optimum (Ellis et al., 1991,1996).

Let's be honest: no-one really knows the value of the optimum or critical water content for storage at 25°C or lower. This is because the experiments were initiated only recently at 20-25°C and are almost nonexistent at lower temperatures. Until the data are available, we have no alternative but to resort to extrapolation. What is the basis of that extrapolation? Apparently no one disputes that 'seed water potentials and hence air equilibrium RH values are more likely to yield a unifying principle than seed moisture contents' (Smith, 1992). Assuming this, we will use a constant water potential or relative humidity as the basis of extrapolation (e.g. Vertucci and Roos, 1993a). Therefore, we need to understand the relationship between water content, relative humidity and temperature in seeds. This relationship is described by water sorption isotherms.

I believe that the resolution of this entire controversy rests on the shape of isotherms at low relative humidities and how this changes with temperature. An excellent primer on water sorption isotherms is given by Priestley (1986) and the uninitiated reader is referred to chapter 1 of that publication for the definition of 'zone 1'. My laboratory has produced several studies of isotherms that roughly conform to isotherm properties of other biological and organic materials (Vertucci and Leopold, 1984, 1986, 1987a; Vertucci et al., 1985; Vertucci and Roos, 1990, 1993a; Walters-Vertucci et al., 1996; Buitink et al., 1998). We show that the water content-RH relationship in zone 1 changes with temperature. There is a good reason for this. Water adsorption is an exothermic reaction and desorption is an endothermic reaction. Exothermic and endothermic reactions are temperature-dependent. The energy produced during adsorption (or consumed during desorption) can be calculated in a variety of ways (see Appendices in Walters, 1998 for derivations). Smith (1992) argues that this energy approximates to zero. Ellis and colleagues indirectly argue a sorption energy of about zero in their statement that the inflection point on the isotherms 'is comparatively unaffected by temperature' (Ellis et al. 1989). By assuming that there is no energy associated with the sorption-desorption reaction in zone 1, these researchers are, in essence, claiming that there is no water binding. But zone 1 has traditionally been identified as the region with strong water binding (e.g. Brunauer et al., 1938; D'Arcy and Watt, 1970). The isotherms that I have produced and the sorption energies that I have calculated are consistent with 60 years of sorption theory. Isotherms constructed at different temperatures that justify the assumption of zero binding energy in zone 1 are conspicuously missing.

The shape of the isotherms at different temperatures is important because it describes the water content-RH relationship during storage. If we assume that there is a critical RH in zone 1 that defines maximum longevity and if the water content-RH relationship in zone 1 is independent of temperature (Ellis et al., 1989, 1990), then the value of the critical water content will be independent of temperature. In this case, critical water contents for low storage temperatures can be determined directly from experiments conducted at extremely high temperatures such as 65°C. If, on the other hand, we assume that there is a critical RH that defines maximum longevity, and we know that water adsorption in zone 1 is an exothermic process that is, by definition, temperature-dependent (Vertucci and Roos, 1993a,b), then we must deduce that the critical water content increases as storage temperature decreases. In other words, one cannot assume constant critical water contents with temperature. Each time Professor Ellis and I try to come to consensus on the ultra-dry debate, we get stuck on this issue.

I believe that to resolve the debate, we must have reliable isotherms constructed for RH between 0 and 25% for temperatures ranging from 0 to 65°C. I use saturated salt solutions to manipulate RH and then measure the equilibrium water content. This is a 'tried and true' method, but it is extremely cumbersome. Alternatively, the water contents of seeds can be adjusted and the equilibrium RH of the air above them measured. Instruments that relate relative humidity to electrical conductance (e.g. Novasina Humidat used by Ellis and colleagues) are usually calibrated with the salt solutions I use directly. Instruments that measure dew point temperature and use psychometric charts to determine RH are extremely accurate, and rumour has it that isotherms of pharmaceutical powders can be produced in an afternoon.

Relative humidity or water potential may not be the 'unifying principle' defining critical water contents. Perhaps a change in glassy structures or a critical viscosity are the important regulators of chemical reactions in seeds (reviewed by Walters, 1998). In these instances, one must also conclude that critical water contents increase with decreasing temperature.

The value of the critical RH or water potential or viscosity is not known for sure. Early data from my laboratory suggested critical relative humidities between 20 and 25%, but data collected more recently suggest that this value should be modified downwards, perhaps to 15-20% RH (Walters-Vertucci et al., 1996; Buitink et at., 1998; Walters et al., 1998). The existence of a critical RH and its value will be confirmed or refuted with future research.

The scientific issues discussed here pertain to germplasm facilities where seeds are stored at ambient temperatures, under refrigerated or freezer conditions and even in liquid nitrogen. The challenge is to determine how to maximize seed longevity using the resources available at each facility. The dearth of data make this a difficult task with inherent uncertainty. Based on the data available, I have concluded that seeds should probably not be stored at RH <15% and never at RH <10%. Most probably, this means that the water content that maximizes longevity varies with the storage temperature used.


Brunauer, S., Emmett, P. and Teller, E. (1938) Adsorption of gases in multimolecular layers. Journal of the American Chemistry Society 60, 309-320.

Buitink, J., Walters, C, Hoekstra, F.A. and Crane, J. (1998) Storage behavior of Typha latifolia pollen at low water contents: interpretation on the basis of water activity and glass concepts. Physiologia Plantarum 102 (in press).

Carpenter, W.J. and Boucher, J.F. (1992) Temperature requirements for the storage and germination of Delphinum × cultorum seed. HortScience 27, 989-992.

Chai, J., Ma, R., Li, L. and Du, Y. (1998) Optimum moisture contents of seeds stored at ambient temperatures. Seed Science Research 8, Supplement No. 1, 23-28.

D'Arcy, R.L. and Watt, I.C. (1970) Analysis of sorption isotherms of non-homogeneous sorbents. Transactions of Faraday Society 66, 1236-1245.

Dickie, J.B. and Smith, R.S. (1995) Observations on the survival of seeds of Agathis spp. stored at low moisture contents and temperatures. Seed Science Research 5, 5-14.

Ellis, R.H., Hong, T.D. and Roberts, E.H. (1988). A low-moisture-content limit to logarithmic relations between seed moisture content and longevity. Annals of Botany 61, 405-408.

Ellis, R.H., Hong, T.D. and Roberts, E.H. (1989) A comparison of the low-moisture-content limit to the logarithmic relation between seed moisture and longevity in twelve species. Annals of Botany 63, 601-611.

Ellis, R.H., Hong, T.D., Roberts, E.H. and Tao, K.L. (1990) Low-moisture-content limits to relations between seed longevity and moisture. Annals of Botany 65, 493-504.

Ellis, R.H., Hong, T.D. and Roberts, E.H. (1991) Seed moisture content, storage, viability and vigour (correspondence). Seed Science Research 1, 275-277.

Ellis, R.H., Hong, T.D., Martin, M.C., Perez Garcia, F. and Goméz-Campo, C. (1993) the long-term storage of seeds of seventeen crucifers at very low moisture contents. Plant Varieties and Seeds 6, 75-81.

Ellis, R.H., Hong, T.D. and Roberts, E.H. (1995) Survival and vigour of lettuce (Lactuca sativa L.) and sunflower (Helianthus annuus L.) seeds stored at low and very low moisture contents. Annals of Botany 76, 521-534.

Ellis, R.H., Hong, T.D., Astley, D., Pinnegar, A.E. and Kraak, H.L. (1996) Survival of dry and ultra-dry seeds of carrot, groundnut, lettuce, oilseed rape, and onion during five years; hermetic storage at two temperatures. Seed Science and Technology 24, 347-358.

Hu, C., Zhang, Y., Tao, M., Hu, X. and Jiang, C. (1998) The effect of low water contents on seed longevity. Seed Science Research 8, Supplement No. 1, 35-39.

IBPGR (1976) Report of IBPGR Working Group on Engineering, Design and Cost Aspects of Long-term Seed Storage facilities. Rome, International Board for Plant Genetic Resources.

Kong, X.-H. and Zhang, H.-Y. (1998) The effect of ultra-dry methods and storage on vegetable seeds. Seed Science Research 8, Supplement No. 1, 41-45.

Leopold, A.C. and Vertucci, C.W. (1986) Physical attributes of desiccated seeds, pp 22-34 in Leopold, A.C. (Ed.) Membranes, Metabolism and Dry Organisms. Ithaca, NY/London, Comstock Publishing.

Leopold, A.C. and Vertucci, C.W. (1989) Moisture as a regulator of physiological reaction in seeds, pp 51-68 in Stanwood, P.C. and McDonald, M.B. (Eds) Seed Moisture (Special Publication No. 14). Madison, Wisconsin, Crop Science Society of America.

Priestley, D.A. (1986) Seed Aging. Ithaca, NY/London, Comstock Publishing.

Roberts, E.H. and Ellis, R.H. (1989) Water and seed survival. Annals of Botany 63, 39-52.

Shen, D. and Qi, X. (1998) Short- and long-term effects of ultra-drying on germination and growth of vegetable seeds. Seed Science Research 8, Supplement No. 1, 47-53.

Smith, R.D. (1992) Seed storage, temperature and relative humidity. Seed Science Research 2, 113-116.

Vertucci, C.W. (1990) Calorimetric studies of the state of water in seed tissues. Biophysical Journal 58, 1463-1471.

Vertucci, C.W. and Leopold, A.C. (1984) Bound water in soybean seed and its relation to respiration and imbibitional damage. Plant Physiology 75, 114-117.

Vertucci, C.W. and Leopold, A.C. (1986) Physiological activities associated with hydration level in seeds, pp 35-49 in Leopold, A.C. (Ed.) Membranes, Metabolism, and Dry Organisms. Ithaca, NY/London, Comstock Publishing.

Vertucci, C.W. and Leopold, A.C. (1987a) Water binding in legume seeds. Plant Physiology 85, 224-231.

Vertucci, C.W. and Leopold, A.C. (1987b) Relationship between water binding and desiccation tolerance in tissues. Plant Physiology 85, 232-238.

Vertucci, C.W. and Roos, E.E. (1990) Theoretical basis of protocols for seed storage. Plant Physiology 94, 1019-1023.

Vertucci, C.W. and Roos, E.E. (1991) Seed moisture content, storage, viability and vigour: response (correspondence). Seed Science Research 1, 277-279.

Vertucci, C.W. and Roos, E.E. (1993a) Theoretical basis of protocols for seed storage. II. The influence of temperature on optimal moisture levels. Seed Science Research 3, 201-213.

Vertucci, C.W. and Roos, E.E. (1993b) Seed storage, temperature and relative humidity: response (correspondence). Seed Science Research 3, 215-216.

Vertucci, C.W., Ellenson, J.L. and Leopold, A.C. (1985) Chlorophyll fluorescence characteristics associated with hydration level in pea cotyledons. Plant Physiology 79, 248-252.

Vertucci, C.W., Roos, E.E. and Crane, J. (1994) Theoretical basis of protocols for seed storage. III. Optimum moisture contents for pea seeds stored at different temperatures. Annals of Botany 74, 531-540.

Walters, C. (1998) Understanding the mechanisms and kinetics of seed aging. Seed Science Research 8, 223-244.

Walters, C., Kameswara Rao, N. and Hu, X. (1998) Optimizing seed water content to improve longevity in ex situ genebanks. Seed Science Research 8, Supplement No. 1, 15-22.

Walters-Vertucci, C., Crane, J. and Vance, N.C. (1996) Physiological aspects of Taxus brevifolia seeds in relation to seed storage characteristics. Physiologia Plantarum 98, 1-12.

US Department of Agriculture, 1998

Previous Page Top of Page Next Page