Xiaorong Hu*, Yunlan Zhang, Chenglian Hu, Mei Tao and Shuping ChenAbstract*CorrespondenceInstitute of Crop Germplasm Resources, Chinese Academy of Agricultural Sciences, 30 Bai Shi Qiao Road, Beijing 100081, China
The water content at which seeds are stored is one of the most important factors determining storage life. Procedures for achieving the proper water content have not been fully described. The purpose of this study was to compare the efficiency of two methods commonly used in seed genebanks: vacuum freeze drying and drying over desiccants. Efficiency was evaluated in terms of the water contents that could be obtained, the speed at which seeds could be dried, the incidence of damage to seeds during drying and the cost. Seeds of eight crop species [barley (Hordeum vulgare), rice (Oryza sativa, Japonica type), oat (Avena sativa), kenaf (Hibiscus cannabinus), sesame (Sesamum indicum), rape (Brassica campestris), amaranth (Amaranthus tricolor) and rice bean (Phaseolus calcaratus)] were dried from ambient conditions of about 20°C and 80% RH using a freeze-drier or silica gel as a desiccant. Water content of different species at the onset of the experiment and after freeze drying or drying over silica gel was an inverse function of the lipid content. Drying to low water contents had no measurable effect on germination percentage. Drying rates using silica gel or a freeze-drier were comparable, but lower water contents could be obtained using silica gel; also, silica gel was readily available and the daily rental fees made the cost of freeze drying prohibitive, thus drying using silica gel was the preferred method.
Keywords: seed moisture content, freeze drying, silica gel, seed drying, lipid, barley, oat, rice, amaranth, rape, rice bean, sesame.
Introduction
Conservation of germplasm using seed genebanks requires techniques that will prolong the life span of seeds. Adjusting the water content of seeds is an important factor in maintaining viability during long-term storage (FAO/IPGRI, 1994). While much research has focused on the water content to which seeds should be dried (e.g. Ellis et al., 1989,1990; Vertucci and Roos, 1990; FAO/IPGRI, 1994; see also other papers in this supplementary issue of Seed Science Research), very little information exists on how seeds should be dried. Conventional methods of seed drying often do not reduce the water content of seeds sufficiently (Justice and Bass, 1978). Genebank operators need methods for drying seeds quickly and inexpensively to low water contents without causing damage.
Many genebanks and researchers dry seeds in desiccators using desiccants such as silica gel (e.g. Ellis et al., 1989) or CaO (Cheng et al., 1991). Extremely low water contents could be achieved with no apparent detrimental effects. Woodstock and colleagues introduced the idea of freeze drying seeds to achieve the low water contents necessary for long life spans. Seeds of parsley, onion and pepper were dried in 4 days from 9.9, 9.0 and 7.4% water to 3.8, 3.3 and 2.5% water, respectively, with no measurable reduction in germination percentage (Woodstock et al., 1976). Seeds freeze dried for 1 or 2 days had higher germination percentages after 12 months at 40°C than seeds that were not dried (Woodstock et al., 1976). Freeze drying for 1 day dramatically slowed aging rates of onion seeds stored at 21 °C for 9 years (Woodstock et al., 1983). In a previous study using Brassica spp., the benefits of drying seeds were clear but there was no particular advantage of freeze drying over drying with desiccants (Cheng et al., 1991).
In this research, two methods of drying seeds were compared, using a freeze-drier and using silica gel, to determine the more efficient method. Efficiency was evaluated in terms of the rate of drying, the lowest water content achieved, the relative cost and relative damage to seeds.
Materials and methods
Seeds from barley (Hordeum vulgare cv. Baomai), rice (Oryza sativa, Japonica type, cv. Modao 110), oat (Avena sativa cv. Huabe 2), amaranth (Amaranthus tricolor cv. Maguo 1), kenaf (Hibiscus cannabinus cv. Xianghong 2), rape (Brassica campestris cv. Jianxia 1), rice bean (Phaseolus calcaratus cv. Jihong 2) and sesame (Sesamum indicum cv. Xiongtong 1) were harvested in 1991 by the Chinese Academy of Agricultural Sciences, Beijing (barley, rice and oats), the Jiang Xi Academy of Agricultural Sciences, Jiang Xi Province (amaranth, kenaf, rape, sesame) and the He Bei Academy of Agricultural Sciences, He Bei Province (rice bean). The initial moisture contents of seeds ranged from a high of 10.8% for barley to a low of 4.4% for sesame, and reflected ambient conditions of Beijing of about 20°C and 80% RH. Initial germination rates of seeds ranged from 86% for kenaf and barley to 99% for rice.
Seeds of barley, rice, amaranth, kenaf, rape, rice bean and sesame were dried by freeze drying. Seeds were placed in a freezer at - 47°C for 3 h and then placed in a TFD-500-5 freeze-drier (Tokara Seisakusho Co. Ltd, Tokyo). Samples were spread out in layers 1 cm thick on stainless-steel shelves in a tray-drying accessory. Seed samples were removed after 1, 2, 4 and 8 days (barley only) and evaluated for water content and germination. Seeds of barley, oat and kenaf were also dried with silica gel in desiccators (6 L capacity) at 25°C. About 2 kg of silica gel were placed at the bottom of the desiccator and covered by a wire mesh. A 200 g sample of seeds was folded into nylon netting and placed on top of the wire mesh in the desiccator. Small rotating fans were installed in a similar set of desiccators to test whether drying rates could be increased with air currents. Silica gel was regenerated regularly by heating at 103°C. Seeds were removed periodically for up to 167 days and tested for water content and germination percentage.
The water content of seeds was measured gravimetrically according to International Seed Testing Association rules (ISTA, 1985). Dry weights were determined by heating ground seeds of rice, oat, barley and rice bean at 130-133°C for 2 h, or whole seeds of sesame, rape, amaranth and kenaf at 103-105°C for 17 h. Two replicates of 5 g each were used for moisture content measurements, and values are expressed on a wet weight basis.
The viability of seeds before and after drying was evaluated using standard germination procedures (ISTA, 1985). Except for oats, which were planted in vermiculite, seeds were germinated between damp paper towels at 20°C (barley, kenaf, rape and oat), 25°C (rice bean, sesame, and amaranth), or 28°C (rice) for 7-10 days. To avoid damage during imbibition, large-seeded crops (barley, oat, rice, kenaf and rice bean) were placed at 40-60% RH for 7 days and then at 80-90% RH for 2 additional days. Seeds of amaranth, rape and sesame were prehydrated by placing them over water (100% RH) for 24 h. Each germination assay consisted of four replicates of 100 seeds (50 seeds for rice bean).
The lipid content of seeds was determined using the Shuo Shi extraction method (Huang and Cheng, 1982). Pre-weighed samples of ground seeds were packaged in oil-free filter paper and submerged in petroleum ether. After 16 h soaking, the solvent layer was drained and the remaining sediment was heated at 80°C for 8 h. Lipid content was calculated from the weight difference between the original sample and the soaked-then-dried sample. Lipid content was determined from two replicates of about 5 g each.
Figure 1. Drying time courses for barley (top), kenaf (middle) and rice and oat (bottom) seeds that were freeze dried (open squares) or dried over silica gel with (solid stars) and without (solid diamonds) a circulating fan in the desiccator. In the bottom panel, freeze-drying data are for rice, silica gel-drying data are for oats.
Results
Both freeze drying and drying over silica gel were effective in reducing the water content of seeds. Freeze drying appeared to remove water more quickly, at least initially, compared to drying in desiccators over silica gel, but the addition of a rotating fan in the desiccators eliminated any perceivable difference in drying rates (Fig. T). After 2-4 days, seeds in the freeze-drier reached minimum water content, but seeds in desiccators continued to dry for several more months. The fans were particularly useful in reducing the time required to dry seeds to an equilibrium weight. Without fans, constant water contents were achieved after 167, 130 and 110 days for barley, oat and kenaf, respectively; with fans, equilibrium water contents were achieved in 112, 60 and 60 days (time courses in Fig. 1 shown for 20 days only).
Drying rates for other species were calculated from the initial slope of similar time-course data (time courses not presented). Initial (within 1-2 days) drying rates in the freeze-drier averaged 3.3% water per day and ranged from 1 to 6% water per day (kenaf and rape, respectively), while rates of drying over silica gel ranged from 0.5 to 2.5% water per day (Fig. 2). Except for barley, where the low calculated rate may be an artefact since water contents were not sampled before 5 days (Fig. 1, top), there appeared to be no difference in initial drying rates using a freeze-drier and using desiccators containing silica gel and a circulating fan (Figs 1 and 2). Drying rates were not related to size of seed (Table 1; Fig. 2A) or to lipid content of seed (Table 1; Fig. 2B), and were poorly correlated with initial water content of seed (Table 1; Fig. 2C).
While initial drying rates did not appear to be affected by drying method or seed properties, the water content to which seeds were dried was dependent on the drying method and the lipid content of the seed (Table 1). With the exception of amaranth, seeds containing >5% lipid were dried to about 6.5% using a freeze-drier and to about 3.5% using silica gel. Similarly, kenaf seeds containing 22% lipid were dried to 4% water with the freeze-drier and to about 1% over silica gel. The coefficients of the linear regressions between lipid content and initial water content, final water content after freeze drying and final water content using silica gel (-0.115, -0.105 and -0.096, respectively, Table 1) were similar. Differences in water content between seed species can be attributed to the known relationship between lipid content and water content in seeds at the same water activity (Vertucci and Roos, 1990).
Figure 2. The relationship between initial drying rates and seed mass (A), seed lipid content (B) and initial water content (C) for different species. Initial drying rates are given for all species studied and were calculated from the initial slopes of time-course data similar to those in Fig. 1. Drying rates were calculated for freeze-dried samples (open squares) and samples dried over silica gel in the presence (solid stars) and absence (solid diamonds) of a circulating fan. Seed mass and lipid contents are listed for individual species in Table 1. The abscissa in D is calculated from the difference between initial water contents (Table 1) and the linear regression of initial water content and lipid content (lower portion of Table 1), and indicates the difference from a common water activity. The regression line drawn in D has a slope of 1.31, y-intercept of 3.3 and r2 of 0.79.
Table 1. Characteristics of different seed species and water contents before and after drying
|
Crop |
Lipid content (%) |
Seed mass (g/1000 seeds) |
Initial water content (%) |
Final water content (%) |
|
|
Freeze-dry |
Silica gel |
||||
|
Rice bean |
4.4 |
132.6 |
10.5 |
6.5 |
not done |
|
Rice |
4.7 |
20.9 |
9.9 |
6 |
not done |
|
Amaranth |
5.1 |
0.7 |
10.4 |
3.8 |
not done |
|
Barley |
5.3 |
40.9 |
10.8 |
6.9 |
2.8 |
|
Oat |
8.9 |
19.8 |
10.3 |
not done |
2.2 |
|
Kenaf |
22.1 |
27.8 |
7.2 |
3.8 |
1.1 |
|
Rape |
33.6 |
2.7 |
9.4 |
1.9 |
not done |
|
Sesame |
49.3 |
2.3 |
4.4 |
1.6 |
not done |
|
Regression analysis with lipid content as the independent variable |
|||||
|
x-coefficient |
|
|
-0.115 |
-0.105 |
-0.096 |
|
constant |
|
|
11.02 |
6.23 |
3.20 |
|
r2 |
|
|
0.76 |
0.76 |
0.97 |
Although there appeared to be a slight reduction in germination percentage in some species when seeds were dried, there was not a consistent relationship or a significant difference in the germination percentage between the undried and driest seeds (Fig. 3).
Discussion
In this study, we measured the drying rates, minimum water contents and germination percentages of seeds dried in a freeze-drier and over silica gel. Initial drying rates (within 1-2 days) of freeze-dried seed averaged 3.3% water per day, and rates of drying over silica gel ranged from 0.5 to 2.5% water per day (Fig. 2). Drying rates measured here for seeds held over silica gel may be greater than those experienced in other laboratories because of the large ratio of silica gel to seeds (10:1). If a circulating fan was used in the desiccator, drying rates could be accelerated even more, such that there were no major differences in initial drying rates between freeze drying and drying over silica gel (Figs 1 and 2). The size of the seed and its lipid or initial water content did not affect initial drying rates (Fig. 2A-C); however, the initial water activity of the seed (inferred from the residual error of the lipid content versus initial water content regression; Table 1) had a major influence on seed drying rates (Fig. 2D).
The initial and final water contents of seeds were related to their lipid content (Table 1). This relationship has been well established previously (e.g. Ellis et al., 1989, 1990; Vertucci and Roos, 1990) and implies that the seeds were at similar water activities both before and after drying. The fact that seeds could be dried to lower water contents using silica gel as a desiccant compared to using a freeze-drier (Table 1) suggests that a lower water activity could be achieved with silica gel.
Based on the above experiments, there were two disadvantages of using silica gel to dry seeds compared to freeze drying: the possibility of slightly slower drying rates; and the amount of labour required to regenerate the silica gel daily. Neither freeze drying nor silica gel drying had detectable effects on germination percentage (Fig. 3). Silica gel drying had distinct advantages over freeze drying in that lower water contents can be achieved using silica gel. Silica gel also appears to be less expensive to use. The cost of desiccators and an oven to regenerate silica gel is considerably less than the cost of a freeze-drier ($6000 to $15000). The power required to heat an oven to 100°C (about 800-2000 watts) is similar to the power required for the freeze-drier (about 600 watts) and vacuum pump (about 600 watts), but drying using the freeze-drier requires continual power for 2-4 days while regenerating silica gel requires power for only a fraction of that time. Daily rental fees of $20 per day for freeze drying make it prohibitively expensive to use routinely.
Figure 3. The effect of drying in a freeze-drier (A) and over silica gel (B) on the percentage germination of seeds. Species are as indicated. Error bars represent the standard deviation of the mean of four replicates.
The capacity to dry seeds to lower water contents and the difference in cost make drying with silica gel the more efficient drying method.
Acknowledgements
The authors greatly appreciate the assistance of Dr Christina Walters in editing this paper.
References
Cheng, H.-Y., Zheng, G.-H. and Tao K.-L. (1991) Physiological, biochemical and ultrastructural studies on ultradried seeds of some Brassica species. Acta Phytophysiologica Sinica 17, 273-284.
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.
FAO/IPGRI (1994) Genebank Standards. Rome, Food and Agriculture Organization of the United Nations/ International Plant Genetic Resources Institute.
Huang, X. and Cheng, L. (1982) Experiment Manual of Seed Physiology. Beijing, Chinese Agricultural Press.
ISTA (1985) International rules of seed testing. Seed Science and Technology 13, 299-355.
Justice, O.L. and Bass, L.N. (1978) Principles and Practices of Seed Storage. Agriculture Handbook No. 506. Washington, D.C., US Government Printing Office.
Vertucci, C.W. and Roos, E.E. (1990) Theoretical basis of protocols for seed storage. Plant Physiology 94, 1019-1023.
Woodstock, L.W., Simkin, J. and Schroeder, E. (1976) Freeze drying to improve seed storability. Seed Science and Technology 4, 301-311.
Woodstock, L.W., Maxon, S., Faul, K. and Bass, L. (1983) Use of freeze-drying and acetone impregnation with natural and synthetic antioxidants to improve storability of onion, pepper and parsley seeds. Journal of American Society for Horticultural Science 108, 692-696.
© CAB INTERNATIONAL, 1998
