Seed Viability Equation Research Articles

Derivation of seed viability constants (KE and CW) and prediction of longevity of Solanum aethiopicum seeds

The seed viability equation uses species viability constants to predict seed longevity under different storage conditions. There are already established universal values of constants that account for the effect of temperature on seed longevity (CH and CQ) which can be applied to different crop species. However, constants accounting for the effect of seed moisture (CW) and genotype (KE) are species-specific. This study aimed to derive viability constants and predict seed longevity for African eggplant (Solanum aethiopicum) under genebank conditions. The seeds were subjected to experimental storage at 40°C and different seed moisture contents (6, 8, 10 and 12%) for 120 days to derive survival curves and hence, CW and KE. CW and KE values for S. aethiopicum were 3.32 and 7.4978, respectively. The validation of the derived constants show that they can be used to predict seed longevity of other seed lots within the same species. The genebank curators and other commercial seed technologists may be able to accurately predict longevity of African eggplant seeds under cold dry storage conditions using the improved viability equation and the constants established in this research.

Seed ageing, survival and the improved seed viability equation; forty years on

Seed Science and Technology storage environment, and the implications for seed ageing and seed quality, including likely seedling emergence. Developments in seed science, and anhydrous biology more widely, associated with that well-cited paper are summarised here.

Royal Botanic Gardens, Kew’s Seed Information Database (SID): A compilation of taxon-based biological seed characteristics or traits

The Seed Information Database (SID) is one of the Royal Botanic Gardens (RBG), Kew's strategic databases in the public domain (http://data.kew.org/sid). SID is a compilation of taxon-based information on a range of biological seed characteristics or traits for about 52,564 seed plant taxa. Assembling and mobilizing such data involved challenges in establishing data standards for specific seed traits and taxonomy and integrating data sets from different sources. Data are derived from seed collections stored at the RBG Kew’s Millennium Seed Bank (MSB) and from other published and unpublished sources. Seed traits data available in SID includes (number of records in brackets): storage behaviour (24,926), seed weights (86,287) seed dispersal (5,377), germination (54, 453), oil content (4,526), protein content (3,603), morphology (599), morphology images (1,741), halophytes (1,546) and viability constants (72). Database can be searched for any, or all of these categories by clade, order, family, genus, or species. SID also offers various online tools useful for ex situ seed conservation and storage (e.g. seed viability equation to predict the proportion of seeds in a population that are viable after any period of storage in a wide range of environments). SID data has been used by a wider community (general public, researches, conservationists, policymakers, etc.) in a range of disciplines including biodiversity conservation, food security, climate change, agriculture, and seed industry across the globe for large-scale analysis, modelling, predicting, or describing plant seed traits.

Longevity of 285 seed lots of wheat in hermetic storage compared with independent estimates from the seed viability equation

The longevity in hermetic storage at 40°C with 14.1-15.7% moisture content of 285 seed lots of wheat (Triticum aestivum L., cv. Tybalt), harvested at or before maturity from control or modified environments under rain shelter or after simulated rainfall in three consecutive years (2010-12), was compared with independent estimates from the seed viability equation and previously-published viability constant values for wheat. Around half of the observations provided good agreement with these independent estimates of longevity, but some of the seed lots harvested close to maturity in 2012, or before the end of the seed-filling phase in 2011, survived longer in storage than prior estimates.

Effect of simulated flooding during rice seed development and maturation on subsequent seed quality

AbstractThe resilience of seed quality in rice (Oryza sativaL.) to flooding was investigated. Pot-grown plants of thejaponicacv. Gleva, theindicacv. IR64, and the introgressed line IR64-Sub1 were submerged in water, to simulate flooding, for 3‒5 days at different stages of seed development and maturation. Mean seed weight, pre-harvest sprouting, ability to germinate, and subsequent longevity in air-dry storage were assessed. Whereas seed quality in both IR64 and IR64-Sub1 was resilient to submergence, in Gleva the longer the duration of submergence and the later in development when plants were submerged the greater the pre-harvest sprouting. Thousand seed dry weight was reduced more by submergence in Gleva than IR64 or IR64-Sub1. At harvest maturity, few pre-harvest sprouted seeds were able to germinate upon rehydration after desiccation to 11‒12% moisture content. Seed longevity of the non-sprouted seed fraction in air-dry hermetic storage (40°C, 15% moisture content) was not affected greatly by submergence, but longevity of thejaponicarice was less than that of theindicarices due to the former's steeper seed survival curves. Longevity of the twoindicarices was predicted well by the seed viability equation and previously published estimates of viability constants for rice. The greater dormancy of IR64 and IR64-Sub1, compared with Gleva, enhanced resilience to pre-harvest sprouting and reduced thousand seed dry weight from plant submergence. There was little or no effect of plant submergence on subsequent air-dry storage longevity of non-sprouted seeds in any genotype.

One-step fitting of seed viability constants for two Australian plant species, Eucalyptus erythrocorys (Myrtaceae) and Xanthorrhoea preissii (Xanthorrhoeacea)

Long-term ex-situ seed storage under controlled conditions in gene banks has become an important tool for conserving threatened Australian plants; however, there is scant information about the seed longevity of most species. The aim of the present study was to determine whether the seed longevity of two contrasting Australian species could be modelled using the seed viability equation, and whether the universal temperature constants are applicable to these species. Seeds of Eucalyptus erythrocorys F.Muell. (Myrtaceae) and Xanthorrhoea preissii Endl. (Xanthorrhoeaceae) were aged at moisture contents ranging from 3.9 to 15.7% and temperatures between –20 and 60°C. Survival data were fitted to the seed viability equation in one step and the species constants for each species determined. Both E. erythrocorys and X. preissii seeds exhibited orthodox seed storage behaviour whose longevity could be modelled using the seed viability equation. The viability constants were KE = 8.81, CW = 4.97, CH = 0.0412 and CQ = 0.000379 for E. erythrocorys and KE = 8.77, CW = 5.29, CH = 0.0382 and CQ = 0.000473 for X. preissii. The universal temperature constants could not be used without a significant increase in error. The storage behaviour of these two Australian species is in keeping with that of orthodox species from around the world. Predictions are that E. erythrocorys will be long-lived under gene bank conditions, whereas X. preissii would be moderately long-lived. Current long-term gene bank storage conditions appear suitable for storage of these species; however, recommendations for short-term storage need to be re-evaluated.

Derivation of Constants (KE, CW) for the Viability Equation for Pepper Seeds and the Subsequent Test of Its Applicability

The effect of moisture on seed longevity during experimental storage was investigated in pepper (Capsicum annuum) cultivar Demre. Seeds were stored hermetically at 7.0%, 8.6%, 9.5%, 10.5%, and 12.1% moisture content (MC; percent fresh weight basis) and 35 °C for up to 306 d. Viability (normal germination) was assessed periodically and the seed viability equation moisture constants, KE and CW, estimated by regression of log σ (the sd of distribution of seed death in time) with log MC. KE and CW values were found to be 7.767 and 4.670, respectively. The newly found moisture constants were combined with the temperature constants that had already been proposed as universal for all orthodox seeds (CH = 0.0329, CQ = 0.000478) and their validity was tested by predicting the viability loss in another pepper cultivar, Carliston. The Ki of the Carliston cultivar was determined by controlled deterioration at 40 °C and 14% MC. The observed viability loss between 30 and 270 d at 25 °C and 10% seed MC was closely related to that predicted by the equation R2 = 0.982 (P < 0.001).

Primed Lettuce Seeds Exhibit Increased Sensitivity to Moisture Content During Controlled Deterioration

The Ellis-Roberts seed viability equation is used to predict seed survival after storage at specified temperatures and moisture contents. Seed priming, which can break dormancy and accelerate germination, can also reduce seed storage life. Because primed seeds were not used in developing the Ellis-Roberts equation, the reciprocal nature of specific seed moisture content (MC, fresh weight basis) and temperatures that applies to nonprimed lettuce (Lactuca sativa L.) seeds may not apply to primed seeds. To determine how priming affects lettuce seeds in relation to the viability equation, an experiment was conducted using two cultivars, ‘Big Ben’ and ‘Parris Island Cos’. Seeds primed in polyethylene glycol 8000 (–1.45 MPa, 24 h at 15 °C) and nonprimed seeds were first adjusted to 6% and 9% moisture contents and then stored at 48 and 38 °C for up to 30 days, respectively. These storage conditions (6% MC and 48 °C; 9% MC and 38 °C) were predicted by the viability equation to result in equal longevities. Subsequent viability assays at 20 °C revealed that nonprimed seeds in both storage environments exhibited similar losses in viability over time, thus validating the Ellis-Roberts equation and the use of these conditions to apply different but equal aging stress. Primed seeds of both cultivars deteriorated faster than nonprimed seeds as expected. However, primed seeds did exhibit different rates of deterioration between the storage environments. Primed seeds stored at 9% MC and 38 °C deteriorated faster than primed seeds stored at 6% MC and 48 °C. The rate of decline in probit viability percentage was three times greater in primed ‘Big Ben’ seeds stored at 9% MC and 38 °C than for those stored at 6% MC and 48 °C (–1.34 versus –0.26 probits per day, respectively). ‘Parris Island Cos’ seeds stored at 9% MC and 38 °C had twice the rate of deterioration that those stored at 6% MC and 48 °C (–1.19 and –0.49 probits per day, respectively). The results indicate that primed lettuce seeds were more sensitive to the adverse effects of higher seed MC than were nonprimed seeds during storage at elevated temperatures.

Temperature sensitivity of the low-moisture-content limit to negative seed longevity--moisture content relationships in hermetic storage.

BACKGROUND AND AIMS The negative logarithmic relationship between orthodox seed longevity and moisture content in hermetic storage is subject to a low-moisture-content limit (m(c)), but is m(c) affected by temperature? Red clover (Trifolium pratense) and alfalfa (Medicago sativa) seeds were stored hermetically at 12 moisture contents (2-15 %) and five temperatures (-20, 30, 40, 50 and 65 degrees C) for up to 14.5 years, and loss in viability was estimated. Viability did not change during 14.5 years hermetic storage at -20 degrees C with moisture contents from 2.2 to 14.9 % for red clover, or 2.0 to 12.0 % for alfalfa. Negative logarithmic relationships between longevity and moisture contents >m(c) were detected at 30-65 degrees C, with discontinuities at low moisture contents; m(c) varied between 4.0 and 5.4 % (red clover) or 4.2 and 5.5 % (alfalfa), depending upon storage temperature. Within the ranges investigated, a reduction in moisture content below m(c) at any one temperature had no effect on longevity. Estimates of m(c) were greater the cooler the temperature, the relationship (P < 0.01) being curvilinear. Above m(c), the estimates of C(H) and C(Q) (i.e. the temperature term of the seed viability equation) did not differ (P > 0.10) between species, whereas those of K(E) and C(W) did (P < 0.001). The low-moisture-content limit to negative logarithmic relationships between seed longevity and moisture content in hermetic storage increased the cooler the storage temperature, by approx. 1.5 % over 35 degrees C (4.0-4.2 % at 65 degrees C to 5.4-5.5 % at 30-40 degrees C) in these species. Further reduction in moisture content was not damaging. The variation in m(c) implies greater sensitivity of longevity to temperature above, compared with below, m(c). This was confirmed (P < 0.005).

Seed survival in Chilean Nothofagus in response to desiccation and storage

Nothofagus alpina,N. obliqua,N. glauca,N. leonii,N. dombeyiandN. pumilioseeds exhibited consistent, albeit slight, sensitivity to extreme desiccation, but nevertheless maintained viability at low moisture contents and cool temperatures (–10° to –20°C) over 2 years.Nothofagus alpina,N. obliqua,N. glauca,N. leoniiandN. dombeyiconformed to the seed viability equation of Ellis and Roberts; sensitivity of longevity to temperature was quantitatively similar to that of crop seeds, sensitivity to moisture was somewhat less, and a low-moisture-content limit to the equation was detected at 4.8% moisture content in hermetic storage at 65 °C, and possibly similar moisture contents at 30–40°C. These five species show orthodox seed storage behaviour. Therefore,ex-situconservation of theseNothofagusspecies in seed banks is possible, but the quality of seed lots collected requires attention. Seed storage behaviour was not defined inN. pumilio: initial seed quality was poor and loss of viability was detected over 2 years at 0°, –10° and –20°C at 2.7% moisture content, but not at 5.2%. The results confirm that the economy of nature in seed storage physiology extends to forest tree seeds, but the repeated observation of reduced sensitivity of longevity to moisture in forest tree seeds requires further investigation.

Seed survival of four tropical tree species in response to environment

The response of seed survival to storage duration and environment (temperature and moisture content) in the four tropical tree species: Cedrela odorata L., Ceiba pentandra (L.) Gaertn., Dalbergia spruceana Benth. and Tabebuia alba (Cham.) Sandwith. from Amazonia conformed to the seed viability equation of Ellis and Roberts. Estimates of the seed viability constants to calculate seed longevity in these species are provided.

Storage of West African yam (Dioscorea spp.) seeds: modelling seed survival under controlled storage environments

Survival data was collected from seeds of two yam species, D. dumetorum and D. togoensis stored at moisture contents ranging from 3% to 25% under temperatures ranging from 5°C to 55°C for 400 days (13 months). The data was subjected to probit analysis to estimate seed longevity under each of the seed storage treatment combinations. The relationship between seed longevity and storage environment was evaluated by a seed viability equation whose constants were determined by linear regression modelling. Seed viability constants estimated for D. dumetorum were K E = 5.859, C W = -3.060, C H = -0.0007, C Q = -0.0007 and for D. togoensis were K E = 4.505, C W = -1.646, C H = -0.020, C Q = -0.00011. R-values of the data fit to the equation were approximately 0.92 for D. dumetorum seeds and 0.87 for D. togoensis seeds. The estimated constants are recommended for monitoring the viability of seed lots of the species in gene banks operating controlled storage facilities within the range of the experimental storage conditions.

Application of the Normal Distribution for Testing the Potential of the Controlled Deterioration Test

The idea of vigor testing by the controlled deterioration test is usually illustrated by nonlinear courses of germination loss during the aging period. The aim of this analysis was to estimate the discrimination power of the controlled deterioration test by applying the assumptions about the nonlinear course made for establishing the Ellis and Roberts seed viability equation. The calculations showed that during the aging period, the differences in germination between seed lots increase according to the assumed normal distribution, while the sampling error of the germination test increases according to the binomial distribution. A gain of discrimination power was achieved when initial germination was higher than 50% and the duration of the aging period was optimized to achieve an average germination after the aging period of 50%. Experimental errors of the aging tests impair the discrimination power, whereas deviations from the assumptions made for the seed viability equation show the potential to improve the discrimination power.

Irrigation and Seed Quality Development in Rapid-cycling Brassica: Soluble Carbohydrates and Heat-stable Proteins

Abstract Changes in soluble carbohydrates and heat-stable proteins have been examined in relation to the acquisition of desiccation tolerance and/or potential seed longevity during seed development in rapid-cycling brassica [ Brassica campestris (rapa) L.]. Rates of seed development were moderated by different irrigation regimes. At the early stages, glucose, fructose and sucrose predominated. The raffinose series oligosaccharides accumulated during seed maturation, and occurred earliest in seeds from plants irrigated only until 16 days after pollination. Stachyose content correlated positively, and monosaccharide content correlated negatively, with the ability of seeds to tolerate rapid desiccation and with their potential longevity (the constant K i of the seed viability equation). Similarly, the ratio of oligosaccharide:total sugars provided strong positive correlations with ability to tolerate desiccation and with potential longevity. Most of the heat-stable proteins selected for study accumulated comparatively late, i.e. during maturation drying. The imposition of water stress induced earlier accumulation of heat-stable proteins. The ability to tolerate desiccation was correlated with the content of selected heat-stable proteins, but potential longevity provided stronger correlations. The content of a 58 kDa heat-stable protein provided the strongest positive correlation with potential longevity. A simple multiple regression model of the relations between potential longevity and both the oligosaccharide:total sugar ratio and the 58 kDa heat-stable protein content was developed for all three plant irrigation regimes to show the combined effect of certain sugars and proteins on seed quality. The model suggests that these sugars and proteins are equally likely to be required for seed quality development, and that initially the sugars tend to accumulate at a greater rate than the proteins, but that during maturation drying the heat-stable proteins accumulate at the greater rate.

Irrigation and Seed Quality Development in Rapid-cycling Brassica: Seed Germination and Longevity

Abstract Irrigation of rapid-cycling brassica ( Brassica campestris [rapa] L.) plants either ended 16 or 24 days after pollination (DAP) or continued throughout the experiment (control). Seeds were harvested serially from these plants during their development and maturation. The earlier irrigation to the mother plant ended, the earlier mass maturity (end of seed-filling phase) occurred, the lower the final seed dry weight, and the more rapid the decline in seed moisture content. The onset of ability to germinate normally occurred as early as 12 DAP, when seeds were less than half-filled. The onset of ability to tolerate rapid enforced desiccation (to 10% moisture content) occurred at 16 DAP. Desiccation tolerance developed within most seeds in both populations about 5 d sooner in seeds harvested from plants in which irrigation was stopped at 16 DAP than in control plants, but maximum desiccation tolerance occurred at about 28 DAP in all treatments. Survival curves (percentage normal germination vs. period of storage) of seeds hermetically stored at 40 °C with 15% moisture content conformed to negative cumulative normal distributions, and provided a common estimate of the standard deviation of the frequency distribution of seed deaths in time for seed lots harvested at different times from the three environments (σ=4.7 d). Potential seed longevity (i.e. the constant K i of the seed viability equation) varied greatly with plant irrigation treatment and with duration from pollination. Maximum K i values were attained 44, 36 and 32 DAP (10, 6 and 7 d after mass maturity) for seeds from control plants, or plants irrigated until 24 or 16 DAP, respectively, and declined thereafter. Maximum potential longevity was greatest ( K i =4.61) for seeds from plants where irrigation stopped 16 DAP and least ( K i =3.88) for those from plants irrigated throughout the experiment. Thus the water status of the mother plant affected seed quality development substantially: terminal drought resulted in more rapid seed quality development and also greater maximum seed quality.

Seed Longevity of Rice Cultivars and Strategies for their Conservation in Genebanks

Changes in seed quality during ripening were studied in sixteen cultivars of rice, representing the three ecogeographic races of Oryza sativa, and one cultivar of O. glaberrima, grown during one dry season (Nov. –May) 1992 –1993 at Los Ba ños, Philippines. Mass maturity (defined as the end of seed filling period) among the cultivars was attained between 18.5 and 21.6d after anthesis (DAA). The seed moisture content at mass maturity varied between 24 and 40%. Germination ability of seeds in the early stages of development varied significantly, but as mass maturity approached, germination increased to the maximum and no significant differences were found among cultivars. The seeds were stored hermetically at 35 °C with 15±0.2% moisture content and the resultant seed survival data were analysed by probit analysis. Potential longevity (quantified by the value of seed lot constant K i of the seed viability equation) was greatest between 33 and 37 DAA, i.e. about 2 weeks after mass maturity. The stage during development at which seeds achieve maximum potential longevity is described by the term storage maturity. Lowland japonicacultivars, large seeded accessions (seed mass ≥40mg) and O. glaberrimahad shorter storage longevity ( σ, standard deviation of the frequency of seed deaths in time=1.47 weeks) while cultivars with purple pericarp survived longer than other cultivars ( σ=2.33 weeks). The initial germination of the japonicacultivars at storage maturity was high (99 –100%) and the estimates of maximum potential longevity ( K i ) which ranged between 3.3 (Shuang cheng nuo) and 4.4 (Minehikare) were close to those of the indicacultivars. This research suggests that seed production environment between Nov. and May at Los Ba ños is benign for the temperate japonicacultivars. The implications of these results on management of rice genetic resources are discussed.

The effect of temperature and CO2on seed quality development in wheat (Triticum aestivumL.)

Winter wheat (Triticum aestivum L.) cv. Hereward was grown in the field in two double-walled polyethylene-covered tunnels within each of which a temperature gradient was superimposed on diurnal and seasonal fluctuations in temperature. The mean temperature between anthesis and harvest maturity varied from 14.3 to 18.4°C among plots within these tunnels. The CO 2 concentration was controlled at different values in each tunnel ; seasonal mean concentrations were 380 and 684 μmol CO 2 mol -1 air. Crops were also grown outside the tunnels at ambient temperatures and CO 2 . Samples of seeds were harvested sequentially from each plot between anthesis and harvest maturity. Seed germination and seed survival during subsequent air-dry storage were determined for each sample. The onset of both ability to germinate and desiccation tolerance (ability to germinate after rapid desiccation to 10-15% moisture content and subsequent rehydration) coincided in all environments. Full germination capacity (>97%, determined at 10°C) was reached 4-18 d before the end of the seed-filling phase (mass maturity) in most cases. There was little or no decline in germination capacity during subsequent seed development and maturation. Differences in seed quality were evident, however, throughout seed development and maturation when seed survival curves during subsequent storage were compared. Potential longevity in air-dry storage (assessed by the value K i of the seed viability equation) improved consistently both before and after mass maturity. There was a significant positive relation between the rate of increase in potential longevity (dK i /dt) and temperature (the minimum temperature for seed quality development was 4.8°C), but neither CO 2 concentration nor production within the polyethylene tunnels affected this relation.

Glassy State and Seed Storage Stability: A Viability Equation Analysis

Dry seeds exist generally in a glassy (or vitrified) state. The high viscosity of the glassy state would be expected to have a retarding effect on deteriorative reactions in the cytoplasm. Thus the glassy state may be considered to be a biophysical barrier for seed deterioration. The present study aims to test the hypothesis that seed storage stability is associated with the glassy state. With the equations derived from the seed viability equation, we have calculated the maximum temperature (Tmax) for long-term storage of corn, pea and soybeans. The Tmax for long-term seed storage is found to be in a good agreement with the glass transition temperature (Tg) in each instance, suggesting that seed deterioration would be accelerated when seeds are not in the glassy state. Experiments with soybeans given accelerated ageing show that the loss of glassy state is followed by a rapid decrease in seed viability. These observations provide indirect evidence that the glassy state may play a significant role in seed storage stability.

Desiccation Tolerance and Potential Longevity of Developing Seeds of Rice (Oryza sativa L.)

The longevity and desiccation tolerance of samples of seeds of a japonica rice (Oryza sativa L.) harvested serially during development from plants grown in two temperature regimes, viz 28/20 °C and 32/24 °C (12/12 h) were determined. Mass maturity (defined as the end of the seed-filling phase) occurred 19·7 and 18·3 d after 50% anthesis, respectively. Longevity (determined at 40 °C with 15% moisture content and quantified by the value of the constant Ki of the seed viability equation) improved during seed development and maturation until 17 and 14 d after mass maturity in the cooler and warmer regimes, respectively, but declined thereafter. Changes in Ki with time were similar in the two environments until mass maturity, but the increase in Ki values after mass maturity was much greater in the cooler regime. Tolerance of desiccation to low (4%) moisture contents improved until 22 and 14 d after mass maturity in the cooler and warmer regimes, respectively, when maturation drying had reduced seed moisture contents naturally to 24 and 32% moisture content, respectively. Further delays to seed harvest reduced desiccation tolerance, particularly in the warmer environment. Comparison among 15 samples of seeds harvested at different times in the two environments showed a strong correlation (r = 0·947, P < 0·01) between longevity (Ki) and desiccation tolerance (to 4% moisture content). Hence, it is suggested that the regulation of desiccation tolerance to low moisture contents and potential air-dry longevity during seed development and maturation determined here may have a common cause.

Seed Production Environment, Time of Harvest, and the Potential Longevity of Seeds of Three Cultivars of Rice (Oryza sativa L.)

Changes in seed quality (assessed by potential longevity, i.e. the value of the seed lot constant K1 of the seed viability equation) in three contrasting cultivars of rice (Oryza sativa L.) were monitored during seed development and maturation in two temperature regimes, viz 28/20°C and 32/24°C (12/12 h), provided by controlled environments. Mass maturity (defined as the end of the seed-filling phase) varied only between 18 and 20 d after 50% anthesis. In five of the six treatment combinations maximum potential longevity was not achieved until 12-19 d after mass maturity. In contrast, the maximum potential longevity of seeds of a japonica rice cultivar produced in the warmer regime was obtained in the first harvest after mass maturity. After mass maturity, the potential longevity of the japonica rice seed lots produced in the warmer environment was much less than that for the cooler environments. Maximum potential longevity was also consistently greater in the cooler than the warmer regime for the two indica cultivars, although the difference in K1 was small (0·3-0·5). The deleterious effect of increase in temperature on seed quality development was not detected until after mass maturity. Maximum potential longevity in the cooler regime was greatest in the glutinous indica (K1 = 3·9) and least in the japonica cultivar (K1 = 3·1). It is concluded that the japonica cultivar is not as well adapted to warm seed production regimes as the indica cultivars. Consequently, subject to confirmation, this research suggest that the seed production of japonica cultivars for long-term genetic conservation should be undertaken, whenever possible, in warm temperate environments.