Christina Walters1*, N. Kameswara Rao2 and Xiaorong Hu3
1 United States Department of Agriculture, Agricultural Research Service, National Seed Storage Laboratory, 1111 South Mason Street, Fort Collins, CO, USAAbstract2Genetic Resources and Enhancement Program, International Crops Research Institute for the Semi-Arid Tropics, Patancheru, Andhra Pradesh 502 324, India
3Institute of Crop Germplasm Resources, Chinese Academy of Agricultural Science, 30 Bai Shi Qiao Road, Beijing, 100081, PR China
*Correspondence
This experiment was undertaken to gain consensus among laboratories regarding the best methods to process seeds for storage and to detect the progress of deterioration of stored seeds. A worldwide experiment, involving laboratories in the USA, India and China, was initiated in June 1995 using lettuce seeds as a model system. Seeds with water contents ranging from 0-0.10 g H2O/g dw (g/g) were stored at -18°C (control), 20, 35 and 50°C. The level of deterioration of seeds was measured periodically by assaying changes in the growth rate of radicles, percentage germination and time required to germinate. Aging kinetics were similar among the participating laboratories. While there was less variability in the percentage germination assay, significant changes in the time to germination and radicle length were detected before changes in percentage germination were found. Significant changes in seed growth parameters were not detected within the 800-1000 days of the experiment in seeds containing less than 0.07 g / g that were stored at 20°C, and so the presence of a critical water content at that temperature was not indicated. Significant changes were not noted in seeds containing 0.020-0.035 g/g (China and USA) or those having water contents as low as 0.013 g / g (India) that were stored at 35°C, and so an optimum water content for storage at 35°C is potentially indicated between 0.020 and 0.035 g/g. Deterioration was observed in the 50°C treatment for seeds at all water contents, but aging was slowest in seeds with water contents less than 0.018 g / g, suggesting that a critical water content exists at or below this water content. Relative humidities corresponding to maximum longevities were determined by constructing water sorption isotherms at the storage temperatures. The relative humidities corresponding to the optimum water content at 35°C and the critical water content at 50°C coincided, being between 10 and 19% at 35°C and £15% at 50°C. To achieve the optimum moisture levels at 35 and 50°C, seeds were equilibrated at 22°C to either 5.5-12% or 1% RH, respectively. The experiment confirms that lettuce seeds stored at 50°C must be dried to very low water contents to achieve their maximum shelf life; however, this extreme drying is not necessary if seeds are to be stored at 35°C, and may even be counter-productive to longevity.
Keywords: germplasm conservation, lettuce, Lactuca saliva, relative humidity, seed aging, seed storage, temperature, water content, water sorption isotherm.
Introduction
Storage of seeds in ex situ germplasm banks is essential to the preservation of biological diversity. In order to preserve the genetic integrity of stored samples, seed germinability must be maintained for several decades. For many seed species, this can be accomplished by adjusting the water content of seeds and by reducing the storage temperature.
The internationally recommended standards are that seeds be dried to 5±2% water prior to storage (FAO/IPGRI, 1994). However, experiments using a number of different species have shown that longevity may be improved if seeds are dried to moisture contents less than the recommended mean value (e.g. Ellis et al., 1988, 1989, 1990, 1995, 1996; Vertucci and Roos, 1990). It has thus been recognized that the range of water contents recommended by IPGRI and the Food & Agriculture Organization are guidelines, and the chemical composition of the seed must be considered to achieve the appropriate moisture content for storage. Differences in lipid composition can be taken into account if seeds are equilibrated to a specific RH rather than moisture content (Ellis et al., 1989; Roberts and Ellis, 1989; Vertucci and Roos, 1990). Several procedures have been proposed: equilibrate seeds to 15% RH and 15°C (www.rbgkew.org.uk/seedbank/); equilibrate seeds to 10% RH and 20°C (Ellis et al., 1989,1990); or equilibrate seeds to a moisture level that gives 20-25% RH at the storage temperature (Vertucci and Roos, 1990,1993).
Experiments have also shown that there are limits to the beneficial effects of drying, such that drying below a critical moisture content will not improve longevity (e.g. Ellis et al., 1988, 1989, 1990) and may even have detrimental effects on seed survival in storage (Vertucci and Roos, 1990; Vetucci et al., 1994). Knowing the value of this critical moisture level is important as insufficient drying results in less than maximum longevity, and over-drying expends energy unnecessarily, may reduce vigour because of longer drying periods, and may even have an adverse effect on aging rates. It is well established that the critical moisture level varies with species and even with cultivars of the same species (Ellis et al., 1988, 1989, 1990; Vertucci and Roos, 1990). Determining the critical moisture level may be further complicated as theoretical considerations suggest that critical moisture contents are also a function of temperature (Vertucci and Roos, 1993; Vertucci et al., 1994).
This experiment was designed to elucidate the critical moisture level for lettuce seeds stored at three temperatures, and to determine the consequences of drying to below the critical level. The long-term nature of seed-storage experiments makes it difficult to obtain answers quickly to critical seed-storage questions. Hence we need tools to detect slight changes in seed viability and vigour. Different growth assays were compared in terms of the experimental variability (within and between laboratories) and the sensitivity for detecting change.
Materials and methods
Lettuce seed (Lactuca saliva cv. Black Seed Simpson, 1994 harvest, Gurney's Seed & Nursery Co., Yankton, South Dakota) were processed at the NSSL, Fort Collins, Colorado, USA, beginning in April 1995. The moisture contents of the seeds were manipulated by storing them for 5-10 days at 22°C and RH ranging from 0-75%. RH was controlled by saturated salt solutions as described by Vertucci and Roos (1993). The salts used were P2O5 (no water added, »1% RH), ZnCl2 (5.5% RH), KOH (8% RH), LiCl (13% RH), KC2H3O2 (23% RH), MgCl2 (33% RH), NaI (38% RH), K2CO3 (43% RH), Ca(NO3)2 (51% RH), NaBr (58% RH), NH4NO3 (62% RH) and NaCl (75% RH). In addition, seeds were dried to equilibrium over activated silica gel. Moisture contents ranging from 0.008 and 0.10 g / g were obtained. Aliquots of 1.5 g dw were then sealed into foil laminate bags, shipped, and stored at -18, 20, 35 and 50 °C at participating laboratories in the USA, China and India.
Water contents and growth parameters were evaluated in each laboratory at each sampling time. Water contents were measured by drying 0.4 g seeds at 103°C for 24 h (India, China, USA) or at 95°C for 5 days (USA). The two methods did not give significantly different water content determinations (data not shown). Prior to germination assays, seeds were humidified for 16-24 h on the laboratory bench in a sealed box containing water. Seeds were then placed on 0.4% agar (percentage germination assay, India only) or on wet paper towels (all other assays), and incubated at 20°C. The number of germinated seeds was evaluated every 6-12 h for 7 days in four replicate samples of 50-100 seeds each. Radicle growth was measured using a separate sample of 50 seeds, 120 h after planting. Total percentage germination was measured using four replicates of 50-100 seeds, 7 days after planting. The time to germination was calculated in two ways: as the midpoint of the steepest part of the germination time course (USA, China), and as the mean time of germination according to the equation
where T is the mean germination time and n is the number of seeds germinating at time t (India). A sample of seeds stored at - 18°C was included in each germination test. Assuming that deterioration does not occur at this storage temperature during the time course of the experiment, this treatment serves as a control. The rate of deterioration for each storage treatment was calculated from the slope of the least-squares line fitted through the control-corrected vigour parameters.
Water sorption isotherms give the relationship between temperature, moisture content and relative humidity. Isotherms were constructed at 5-50°C in 10°C intervals by storing seeds over saturated salt solutions (Vertucci and Roos, 1993). Fresh weights of seeds were measured periodically until a constant weight was achieved. Water contents were measured by drying seeds at 95°C for 5 days, and are expressed on a dry weight basis. Isotherm curves were drawn using van't Hoff analyses (Vertucci and Roos, 1993). Briefly, relative humidities corresponding to water contents ranging from 0-0.12 g / g in 0.005 g / g intervals were interpolated from water content-RH data measured at each temperature. In a van't Hoff isochore, the ln(RH/100) corresponding to a given water content is linearly related to 1/temperature (K), and the slope of the line is related to the enthalpy of water sorption at that water content. Water sorption isotherms can be calculated within the temperature range studied by a series of van't Hoff isochores. Isotherms can also be derived for temperature ranges where saturated salts do not give stable RH by extrapolating (with caution).
Results and discussion
This experiment was designed to determine aging rates of seeds stored at various water contents and temperatures. Here we report preliminary results after 800-1000 days' storage. Deterioration was evaluated by changes in percentage germination, radicle growth after 120 h incubation, and time required for seeds to germinate.
Assuming no changes occur in samples stored at -18°C during the time course of the experiment, this treatment serves as a control and a way to evaluate the experimental variability among and between laboratories at different sampling times (Fig. 1). Germination percentages for the control treatment were very reproducible, with average and standard deviation values of 0.95+0.02, 0.97±0.02 and 0.97±0.03 for laboratories in India, China and USA, respectively. Time to germination of control samples was also fairly constant within laboratories, averaging 30.5±1.8 h (India, mean germination time) or 21.2±0.4 or 21.7±0.4h (China and USA, respectively, time to 0.5 maximum germination). Within laboratories, variability in the radicle length assay was problematic, with standard deviations of control treatments ranging from 9-16% of the mean radicle length; however, average radicle lengths of the control treatment were similar among laboratories (average and standard deviation of radicle length measurements were 40±6, 44±4 and 45±7 in India, China and USA laboratories, respectively).
Deterioration in stored samples was evaluated by reductions in germination percentage and radicle growth or increases in the time for germination. A representative time course is given in Fig. 2, which shows the deterioration of lettuce seeds containing 0.033 g / g water stored at 50°C. In this treatment, seeds begin to lose the ability to germinate at about 300-100 days (Fig. 2A). Statistically significant differences in the percentage germination were noticed at different times in different laboratories (after 196 and 248 days in India and USA, respectively, and not at all within the 392 days of the experiment in China). A clear increase in the time to germination was measured at 224 days' storage (India and China) and 245 days' storage (USA) (Fig 2B). Reductions in radicle growth were often measured before changes in percentage germination were noticed (at 108, 112 and 252 days for USA, India and China, respectively), but additional sampling was necessary to confirm a consistent change (Fig. 2C). These results are consistent with the established idea that losses of seed vigour (measured by germination time and radicle growth in these experiments) occur before losses in seed viability (measured by germination percentage in these experiments) (e.g. Justice and Bass, 1978), and indicate that vigour assays are more sensitive indicators of seed aging. Because of the large noise-to-signal ratio in radicle length measurements, this assay is only useful if done repetitively.
Figure 1. Germination assays of lettuce seeds stored at 0.054 g/g and -18°C for 800-100 days. Data are presented for the participating laboratories in China (circles), India (diamonds) and USA (squares). This treatment is considered the control treatment, and demonstrates the experimental variability intrinsic to percentage germination (A), time to germination (B) and radicle growth after 120 h incubation at 20°C (C).
Figure 2. The change in percentage germination (A), time to germination (B) and radicle growth of lettuce seeds stored at 50°C and 0.033 g/g (C). Seeds were prepared in the USA by equilibrating them at room temperature over a saturated solution of KOH (8% RH) and then stored at participating laboratories in China (circles), India (diamonds) and USA (squares). While the deterioration rates observed are unique to this treatment, the general features of this data set are representative of the all the other treatments.
The rate of deterioration was calculated from the slope of linear regressions of percentage germination, time to germination and radicle length time courses. This treatment is most useful in quantifying the trend towards reduced radicle growth. It is also useful in quantifying the decrease in percentage germination and increase in time to germination during the earlier stages of aging, before these parameters change abruptly (i.e. about 300 days in Fig. 2). Aging rates calculated from experiments performed in different labs were similar (Figs 3-5).
For 20°C storage, a significant trend toward reduced percentage germination and radicle growth of germinated seeds was observed for seeds containing ³0.09 g/g (Fig. 3A) and ³0.075 g/g (Fig. 3C). Significant trends for increased time to germination were also observed for seeds containing ³0.075 g/g (Fig. 3B). Because deterioration was not detected within the 800-1000 days of the experiment, it is impossible, based on these data, to ascertain whether a critical water content for storage exists at 20°C.
Deterioration was more rapid for seeds stored at 35°C and significant trends were observed for all water contents except between 0.020 and 0.035 g/g (Fig. 4). Results reported from the laboratory in China showed a detrimental effect of extreme drying (drying to water contents £0.02 g/g) in the percentage germination and radicle length assays. Results from the USA showed detrimental effects of drying to water contents £0.023 g / g in all assays. Deterioration was not detected at extremely low water contents in the experiments in India, and no detrimental effects of low water contents were reported from this laboratory. The collective results suggest the possibility that an optimum water content for storage exists at 35°C between 0.02 and 0.035 g/g (Fig. 4). Longer storage times are required to substantiate this conclusion.
Seeds deteriorated quickly when stored at 50°C (Fig. 5, notice ordinate scale is 10 × greater than scale in Figs 3 and 4). Significant deterioration was observed within 400 days in all water content treatments using the radicle length and time to germination assays. Generally, aging rates were progressively less as seeds were dried to progressively lower water contents (Fig. 5B, C). Only slight reductions in percentage germination were observed within 400 days for seeds stored with water contents less than 0.037, 0.023 and 0.018 g/g water according to data from the laboratory in China (Fig. 2), India and USA, respectively. In China and India, all samples stored at 50°C had been assayed by 392 days (15 samples), and further testing was not possible. Sampling continued in the USA, and after 1000 days, germination of seeds stored with 0.018 and 0.016 g/g was 29 and 0%, respectively. The collective results confirm earlier reports of a clear benefit of drying to extremely low water contents when seeds are stored at elevated temperatures. If a critical water content exists for lettuce seeds stored at 50°C, the value is £0.018 g / g (Fig. 5).
Figure 3. Rate of deterioration of lettuce seeds stored at 20°C and different water contents. Rates are calculated from the slopes of control-corrected deterioration time courses similar to those given in Fig. 2. Symbols represent data sets from each of the participating laboratories as described in Fig. 2.
The results from this experiment do not give a clear indication of critical water contents for seed storage, either because deterioration was not detected within the reporting period (800-1000 days for 20 and 35°C) or because samples were not dried sufficiently to see the presumed limit of the beneficial effects of drying (50°C). Based on information available to date, an optimum water content between 0.02 and 0.035 g / g is suggested for lettuce seeds stored at 35°C, and a critical water content near 0.018 g / g is suggested for seeds stored at 50°C. A critical water content of 0.026 g / g for lettuce stored at 65°C (Ellis et al., 1989) and an optimum water content between 0.02 and 0.04 g / g (Ellis et al., 1995) or at 0.047 g / g (Vertucci and Roos, 1990) for seeds stored at 35°C have been reported previously. The optimum water content of 0.047 g / g is considerably higher than the range of values predicted here (0.023-0.035 g / g) for lettuce, and this can be partially explained by the difference in lipid content in the two seed lots: 37% for seeds used by Vertucci and Roos (1990) and 44% for the seeds used in this experiment (unpublished data). An optimum water content of 0.035 g / g was calculated for peanut seeds containing 45% lipid (Vertucci and Roos, 1990), which is closer to the value suggested here for this particular lot of lettuce seed. A critical water content of 0.026 g / g determined at 65°C (Ellis et al., 1989) is also high compared to the value reported here of £0.018 g / g for 50°C, and this difference can also be rationalized by the higher lipid content of the seeds used in this experiment.
Figure 4. Rate of deterioration of lettuce seeds stored at 35°C and different water contents. Rates are calculated from the slopes of control-corrected deterioration time courses similar to those given in Fig. 2. Symbols represent data sets from each of the participating laboratories as described in Fig. 2.
Figure 5. Rate of deterioration of lettuce seeds stored at 50°C and different water contents. Rates are calculated from the slopes of control-corrected deterioration time courses similar to those given in Fig. 2. Symbols represent data sets from each of the participating laboratories as described in Fig. 2.
The relative humidity corresponding to these critical water contents can be determined by water sorption isotherms. The isotherms produced for lettuce seeds used in this experiment have the same reverse-sigmoidal shape reported earlier for other seeds (Fig. 6). Van't Hoff analyses were used to draw the isotherm curves: r2 values for isochores drawn for water contents between 0.005 and 0.12 g / g were ³0.90. Heat of sorption was highest (between 0.82 and 0.41 kJ/mol water) at water contents of 0.035 g / g or less. Isotherms in Fig. 6 have lower water contents for a given RH than reported previously for lettuce seeds at 25 and 20°C (Vertucci and Roos, 1990; Ellis et al., 1995), and this can be attributed to the relatively high lipid content in the lot used here (37% versus 44% lipid).
The relative humidity corresponding to presumed critical water contents of 0.018 g / g at 50°C and 0.023-0.035 g / g at 35°C is about 15% and 10-19% RH, respectively (Fig. 6). These RH ranges are consistent with those determined earlier for Typha latifolia pollen (Buitink et al, 1998) but are less than the 19-27% range predicted earlier from theoretical considerations (Vertucci and Roos, 1990). An optimum moisture level for storage between 10 and 19% RH is also consistent with results of a 5-year storage experiment in which the percentage germination of lettuce seeds stored at 20°C was higher in seeds stored at 10% RH compared to 35% RH (Ellis et al., 1996). It is generally agreed that there is a limit to the beneficial effects of drying on longevity. It is also accepted that this limit occurs within a narrow range of relative humidities. The data presented here suggest that the critical RH for lettuce approximates to 15%.
Different drying protocols were required to achieve maximum longevity at 50 and 35°C. All seeds were dried on the laboratory bench at about 22°C. P2O5 (RH»1%) was the only desiccant that could dry seeds to water contents as low as 0.018 g / g at room temperature. On the other hand, the critical moisture range for storage at 35°C was achieved by drying seeds over silica gel or saturated solutions of ZnCl2 (RH=5.5%), KOH (RH=8%) and LiCl (RH=12%). Drying over P2O5 increased aging rates at 35°C according to data from China and the USA (Fig. 4). This result supports the idea that a single drying protocol does not provide optimum moisture levels for all storage temperatures (Vertucci and Roos, 1993).
Figure 6. Water sorption isotherm of lettuce seed at 5-50°C. Data points are moisture contents determined after equilibrating seeds over saturated salt solutions at 5°C (squares), 15°C (diamonds), 25°C (circles), 35°C (triangles) and 45°C (stars). Isotherm curves are calculated from van't Hoff analyses (see text). The dashed and dotted lines at 50 and 20°C are extrapolations and interpolations, respectively, of van't Hoff isochores.
Preliminary results from this experiment suggest that there is a critical or optimum relative humidity for storage of lettuce seeds at about 15%. Water sorption isotherms presented here clearly show that seeds must be equilibrated to different relative humidities to achieve the critical RH of 15% at the storage temperature. Accordingly, equilibrating seeds at 20°C and 10% RH or 15°C and 15% RH will achieve the supposed critical moisture level for lettuce seeds (15% RH) that are to be stored at about 30 and 15°C, respectively. However, these recommended protocols are likely to under-dry seeds intended for storage at elevated temperatures and unnecessarily over-dry seeds stored under refrigerated conditions. Drying protocols that consider the storage temperature are needed.
Acknowledgements
C. Walters acknowledges the contributions of J. Crane and L. Hill for preparing the seeds for the storage experiment, coordinating the shipping of these seeds to the various laboratories, and performing the germination assays in the USA. X. Hu acknowledges the contributions of Professor Shuping Chen for managing the seed storage experiment in China. Professor Mingde Zhou is acknowledged for her supervision. N. Kameswara Rao acknowledges the contributions of Mr D.V.S.S.R. Sastry in conducting the germination and vigour assays. Drs Eric Roos and Richard Ellis are acknowledged for important discussions during the design of the experiment. Special thanks are addressed to Dr Jan Engels for providing the intellectual force to make the study feasible. This work was partially funded by cooperative agreements between IPGRI and each of the participating laboratories.
References
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).
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. (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.
FAO/IPGRI (1994) Genebank Standards. Rome, Food and Agricultural Organization of the United Nations/ International Plant Genetic Resources Institute.
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.
Roberts, E.H. and Ellis, R.H. (1989) Water and seed survival. Annals of Botany 63, 39-52.
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. (1993) 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., 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.
US Department of Agriculture, 1998
