Abstract: The effect of seed ageing of 14 Bangladeshi rice cultivars was investigated to aid the identification of rice genotypes tolerant of low temperature during germination. This would facilitate breeding cultivars suitable for direct wet-seeding in the cooler Boro season in Bangladesh. The present study was carried out at the University of Aberdeen, UK during 1999. The results of the experiment on temperature gradient plate at a range of constant temperatures (13.7-37.3oC) revealed a number of cultivars (BR1, KS and KG) were to be of lower physiological quality than the rest. It was therefore, necessary to confirm whether the reason for their relatively poor performance was physiological deterioration. Seed survival curves of all cultivars at 24% moisture content (mc) and 45oC for up to 96 h tested at 21oC showed a clear separation in germination after 48 h ageing. Cultivars BR1, KS and KG were identified as the lowest quality seed lots with 0, 35 and 17% germination, respectively. Cultivar samples had different Ki (initial seed quality) after probit transformation with a range 79.30% (e.g. cv. KG) to 99.36% (e.g. cv. BR29), but surprisingly, had different slopes. The steepest slope was found for cv. BR11 (-0.046) and shallowest of that was for cv. BR24 (-0.017). The rates of germination of the faster germinating cultivars (8 cultivars, around 0.30 seed d-1) declined more rapidly and at 72 h ageing the rates of germination of all cultivars were closer. Cultivars KS and KG had the least rates of germination (around 0.15 seed d-1). Only when the lower quality cultivars (BR1, KS and KG) were included, were significant relationships found between measures of physiological age (48 h ageing germination, Ki and viability period) and final germination at lower temperature. The results of the study suggested that seed quality as well as genotype might be responsible for reducing final germination of cultivars. The present study also revealed that germination of seed lots of 14 rice cultivars in low temperature was influenced more by genotype than seed quality.
Introduction
There are four criteria which normally determine seed quality: physical purity, seed health, germination capacity and seed vigour. The germination capacity and seed vigour are the most important physiological properties of the seed which determine the response to stress conditions (e.g. low and high temperatures). When seed lots with high germination percentage are sown, large differences in emergence have been observed (Ellis and Roberts, 1980a). These differences in field emergence of seed lots with high germination capacities have been attributed to differences in seed vigour (Matthews, 1980).
The major cause of differences in seed germination and vigour is seed ageing or physiological deterioration (Powell, 1988; Naylor and Gurmu, 1990; Powell and Matthews, 1992; Matthews, 1994). Seeds may deteriorate both on the mother plant and in storage and the rate of deterioration is highly dependent upon the environmental conditions, particularly temperature and relative humidity (RH). Thus, the prevailing hot and humid climates in the rice growing regions of the world including Bangladesh also favour rapid deterioration of conventional seeds, result in lower percentage germination.
When this declination in germination percentage of seeds is plotted against storage period, it follows a sigmoidal survival curve which represents a negative cumulative normal distribution (Ellis and Roberts, 1980b). The viability of a seed lot is affected by a given period of ageing and this viability is also dependent upon the initial position of the seed lot on this survival curve. When the survival curve is converted into probits and then plotted this against time the results is a straight line with a negative slope. The intercept on the y-axis is termed Ki, a constant for the seed lot and is probit percentage germination at zero time which is an indicator of initial seed quality and the slope is 1σ-1, where σ is the standard deviation of the distribution of seed death in time. The intercept Ki, is specific to the seed lot and its value is dependent on the pre-storage environment, genotype and the interaction between these two factors. Storage conditions affect only the value of σ and not Ki. Therefore, the value of Ki can be determined using any set of storage condition; and so can be carried out quite quickly by storing seeds under poor conditions where viability is lost fairly rapidly and sampling several times (Ellis and Roberts, 1980b).
Thus, for any seed lot stored under constant conditions:
V = Ki – pσ-1
Where, V is the probit percentage viability, p is the storage period (days). If the period of seed deterioration required for viability to become zero, the storage period or the survival period (p), can be determined as:
| 0 | = | Ki – p¯σ1 | |
| pσ-1 | = | Ki | |
| p | = | Ki x σ |
As relative differences in longevity between seed lots are maintained in all storage environments, differences in survival period can be used as a method for ranking the vigour of different seed lots. However, a little was known about the seed quality of Bangladesh Rice Research Institute (BRRI)- released rice cultivars regarding low temperature germination. The present study was, therefore, carried out to assess the physiological quality of the seed lots of 14 rice cultivars using ageing test (Controlled Deterioration) so as to separate the influence of seed quality and genotype on low temperature germination which would facilitate cultivar selection for winter sown Boro rice in Bangladesh.
Materials and Methods
Cultivars: Seeds of fourteen indica rice cultivars were produced in Bangladesh between March and December 1997 (Table 1). The well dried and pure seed samples were brought to the United Kingdom (UK) from BRRI, Bangladesh and were stored in the Department of Agriculture and Forestry, University of Aberdeen, UK in sealed aluminum foil packets at 4oC and were used to carry out the present experiment, during 1999.
Seed moisture content (mc) measurement: The moisture content of the seed lot of each cultivar was determined by the Oven Dry method (ISTA, 1999) and expressed on a wet weight basis.
| Table 1: | List of rice cultivars used in the study with initial seed moisture content (mc) and initial final germination at 21oC |
| * Traditional cultivars | |
Five replicate samples of 25 seeds in each replication were ground to a fine texture using an electrical grinding mill and weighed to four decimal places before and after oven drying at 130oC for 1 h. Seed moisture content was then determined as a percentage of seed wet weight.
Ageing tests: Seed samples were screened under a low magnification binocular microscope and broken seeds were removed. Initial seed moisture content was determined as previously described.
The weight of the seed at the required moisture content was calculated by using the following formula (ISTA, 1995).
Where
| A | = | initial seed moisture content (%) wet weight basis, |
| B | = | required seed moisture content which is 24%, |
| W1 | = | initial seed weight (g), |
| W2 | = | seed weight (g) at the required moisture content. |
Seed moisture content was raised to 24% by placing a weighed sample of seeds of each cultivar on a filter paper moistened with distilled water in a 90 mm plastic petridish (with lid) to allow imbibition. Seeds were regularly removed from the filter paper and quickly surface-dried with a paper tissue and weight was taken until required weight was reached.
When the seed samples reached the required weights they were placed inside individual aluminum foil packets and heat-sealed. The packets were kept in the refrigerator overnight at a temperature of 5°C for moisture to equilibrate through out the seeds. The foil packets were then placed in a water bath at 45±0.5°C for periods of 0, 24, 48, 72 or 96 h. The packets were taken out from water bath after the relevant deterioration time and washed with cool running tap water for 2 min. Seeds were removed from packets, placed in a petridish and put in a germination room maintained at 21±1°C. Weights of samples were taken regularly until the samples had dried back to their initial weight indicating that the samples had returned to their initial moisture content.
Germination test: A standard germination test using petri dishes was carried out using 4 replicates of 25 seeds for each period of deterioration for each cultivar with a control of 4 replicates of 25 undeteriorated seeds. The test was conducted in a germination room at 21±1oC in dark. Germination counting was done at 24 h interval and seeds were considered as germinated when radicles were extended more than 2 mm. Number of normal and abnormal seedlings were recorded at final counting (14 days).
Data analyses: Analyses of variance were carried out on all data collected by using MSTAT statistical programme. The correlations and regressions were studied by using graphical routines in Microsoft Excel (Version 5). The data were examined by probit analysis (Ellis and Roberts, 1980b) and Ki, σ and p were calculated for each cultivar. The mean germination time (MGT) was calculated as
MGT = (Σnd)N-1
where
| n | = | number of germinated seeds on each day, |
| d | = | number of days from beginning of the test, |
| N | = | total number of germinated seeds and the rate of germination was determined as 1(MGT)-1. |
Results
In the experiment on temperature gradient plate conducted at a range of constant temperatures, seed lots of a number of rice cultivars (BR1, KS and KG) were thought to be of lower physiological quality than the rest because germination at the optimum temperature was close to 80%. These cultivars (BR1, KS and KG) were also consistently lower in their final germination and rate of germination at the lower temperatures of 13.7 and 15.8oC (Ali, 2001).
| Table 2: | Percentage germination at 21oC of fourteen rice cultivars as affected by five levels of seed ageing at 24% mc and 45°C |
| Number (s) in parentheses are the percentages of germinating seeds that gave abnormal seedlings | |
It was therefore, necessary to confirm whether the reason for their relatively poor performance was physiological deterioration. Seeds of 14 cultivars (Table 1) were aged for 0, 24, 48, 72 and 96 h at 24% mc and 45°C and subsequently germinated at 21oC. Distinct differences in germination between cultivars began to be seen at and after 48 hours ageing (Table 2). Three assessments of physiological quality related to position on the seed survival curve (Fig. 1) were determined: germination after 48 h ageing, Ki and viability period. Germination after 48 h ageing identified cvs. BR1, KS and KG as the lowest quality seed lots (Table 2). These were also the seeds with a high number of abnormal seedlings (Table 2). The initial viability (Ki) also identified the four poorest cultivars as KS (80.92%), KG (79.30%), BR24 (91.65%) and BR26 (93.43%). Ki could not be calculated for cv. BR1. Three cultivars (BR11, BR28 and BR29) had Ki’s above 99% (Table 3).
Surprisingly the slopes of all the lines were different despite the seeds being aged under the same conditions (24% mc and 45oC). The steepest slope was found for cv. BR11 (-0.046) followed by cvs. BR5 and BR28 (both -0.040), the shallowest slopes were seen for cvs. BR24 (-0.017) and BR33 (-0.018) (Table 3). These differences in slope combined with differences in Ki led to a wide range in calculated viability period (p) from as low as 24.2 h for cv. KG to as high as 117.3 for cv. BR33 (Table 3).
Rate of germination: The rate of germination decreased significantly and linearly (R2 = 0.9961, P <0.001) with increasing ageing level (up to 72 h ageing) in all cultivars (pooled data) (Fig. 2). Initial rates of germination were found to be higher in cvs. BR11, BR14, BR23, BR28, BR29, BR31, BR32 and BR33 (around 0.30 seed d-1) (Fig. 3, Table 4) and lower initial rates were found in cvs.
| Table 3: | Seed vigour parameters from probit analysis of controlled deterioration for 13 rice cultivars. The average viability period, p is based on the survival of seeds incubated at 24% moisture content and 45°C |
| Notes: | (1) The initial seed quality is the percentage detransformed from the probits Ki |
| (2) The viability period is calculated using Ki (probits) | |
| Table 4: | Intercept, slope and R2 values of the fitted linear curves of germination rate fitted against ageing for different rice cultivars |
BR5, BR24 and BR26 (around 0.20 seed d-1) and the least were seen for cvs. KS (0.15 seed d-1) and KG (0.17 seed d-1) (Fig. 3, Table 4). Up to 24 h ageing, the differences in rates between germination of these three groups of cultivars were maintained but as ageing progressed, the rates of germination of the faster germinating cultivars declined more rapidly and at 72 h ageing the rates of germination of all cultivars were closer.
Assessments of physiological quality in relation to low temperature germination: The three assessments of physiological quality (germination at 48 h ageing, Ki and viability period) were examined for their relationship to final germination at the lower temperatures of 13.7oC and 15.8oC. The relationships were examined both including the clearly identified cultivars of low physiological quality (cvs. BR1, KS and KG) and excluding them.
When the low quality cultivars were included, final germination at 13.7 and 15.8oC was significantly and positively related to germination at 48 h ageing (Fig. 4a, c).
| Fig. 1: | The seed survival curve showing the pattern of decline in viability (% germination) of individual seeds in a population of orthodox seeds under constant temperature and moisture content (Priestley, 1986) |
The cultivars BR1, KS and KG were lower in their low temperature germination and in the assessment of their positions on the survival curve as indicated by germination at 48 h ageing. When the lower cultivars were excluded the significant relationships were lost (Fig. 4b, d) although there was a range of germination after ageing from 56% (cv. BR11) to 91% (cv. BR29) (Table 2). When Ki (initial germination) was used as an assessment of physiological quality, a significant relationship between Ki and germination was found only at 15.8oC when cvs. KS and KG were included (Fig. 5c). This was also the case when seed viability period (p) was related to germination.
Discussion
The results of the experiment on temperature gradient plate demonstrated that cultivars differed significantly in their response to constant temperature.
| Fig. 2: | Effect of rapid ageing at 24% mc and 45oC on germination rate (pooled data) of 13 rice cultivars (fitted linear regression) at 21oC |
| Fig. 3: | The effect of ageing at 24% mc and 45oC on germination rate at 21oC of 13 rice cultivars |
The cultivars that showed a particularly low rate of germination and final germination (BR1, KG and KS) at low temperature had a relatively low level of germination at higher temperatures suggesting low physiological quality (Ali, 2001). The present experiment was, therefore, conducted to confirm this and to ascertain how much physiological quality and not genotype was the explanation for differences in temperature response.
Controlled Deterioration (CD), a standard vigour testing method (ISTA, 1995) was carried out for the assessment of seed physiological age.
| Fig. 4: | Relationship between final germination at 21oC after 48 h ageing (24% mc and 45oC) and final germination at 13.7oC and 15.8oC, (a) and © including all cultivars, (b) and (d) excluding cvs. BR1, KS and KG |
According to the vigour concept described by Delouche (1969) there could be differences between seed lots in the degree of deterioration.
| Fig. 5: | Relationship between initial seed vigour (Ki) and final germination at 13.7oC and 15.8oC, (a) and © including all cultivars, (b) and (d) excluding cvs. BR1, KS and KG. Ki was calculated for each lot from the survival curve determined during ageing at 24% mc and 45oC followed by germination test at 21oC |
Germination after additional standard amounts of ageing would therefore reflect the relative positions of the seed lots in the deterioration process before the ageing was imposed. Therefore, CD was designed to force seeds out from phase I of seed survival curve where they are similar in final germination (Fig. 1) into phase II where they can be discriminated. In this test, seed lots with low final germination percentage after deterioration or ageing are said to be a low vigour with poor field emergence and storage potential.
Three assessments of physiological quality of seed of all cultivars were determined from the seed survival curves at 25% mc and 45oC: germination after 48 h ageing, initial seed quality (Ki) and viability period (p). All assessments reflected the position of the seed lots of the cultivars on the seed survival curve before the ageing treatments. The relationships between germination at low temperatures (13.7 and 15.8oC) and the three assessments of physiological quality were only positive and significant when all cultivars were included. When the lower germinating cultivars BR1, KG and KS were excluded, no significant relationships were seen. This confirmed the low physiological quality of the seed of these cultivars and suggested that seed quality differences were not the explanation for differences in low temperature germination in the other cultivars.
There were some differences in seed quality seen in cultivars other than BR1, KS and KG, but these differences did not seem to influence the level of germination at the lower temperatures. For example, cv. BR11 was identified as the lowest quality and also had a relatively low germination at 13.7oC. In contrast, cv. BR5 appeared to be better quality seed but exhibited a relatively low and slow germination at 13.7 and 15.8oC (Ali, 2001). Thus the implication might be that once cvs. BR1, KS and KG were excluded; differences in low temperature germination performance were primarily the result of genotypic differences. There were no significant relationships between seed quality as measured for example by germination after 48 h ageing and low temperature germination even though relatively small differences in physiological quality were seen (Fig. 4). This would have resulted from the effect of genotype having greater influence on low temperature germination than quality. For example, the germinations of cvs. BR23 and BR5 are lower than might be expected from the 48 h ageing germination (Fig. 4) suggesting they were genotypically less able to germinate at 13.7oC. In contrast, cvs. BR26 and BR31 germinated at 13.7 and 15.8oC (Fig. 4) to a higher level than have been expected from 48 h ageing germination suggesting a genotypically determined ability to germinate at lower temperatures.
The increases in percentage germination and decrease in abnormal seedlings (Table 2) which occurred mostly in intermediate vigour seeds after mild ageing (24 h) in this study have been interpreted as a demonstration of repair mechanism resulting in invigoration of seeds (Naylor, 1989). Some authors similarly explained invigoration of seeds as metabolic repair which enhanced germination and this has been observed in wheat (Lush et al., 1981), onion (Ward and Powell, 1983) and Brussels sprouts (Burgass and Powell, 1984). Naylor and Syversen (1988) showed enhanced germination in Italian ryegrass which was subjected to ageing at high moisture content and high temperature. They also noted that lots with intermediate proportions germinating or intermediate germination times showed most improvement. This could be because in a high vigour lot there is little scope for improvement while in a low vigour lot deterioration has proceeded too far. The explanations for the improvement achieved have focused on the metabolic repair of previously sustained deterioration (in aged seed) and germination advancement (in high vigour seed) (Powell et al., 2000). Thornton and Powell (1992) provided physiological evidence of repair by increased germination after controlled deterioration in cauliflower and in Brussels sprouts seed following aerated hydration.
Ageing consistently reduced the rate of germination of all cultivars (Fig. 3). Comparisons of the rate of germination of unaged seeds showed that seeds clearly identified as being of lower physiological quality (BR1, KS and KG) all had slower rates of germination at all temperatures relative to most other cultivars (Ali, 2001). One clear exception was cv. BR5 which germinated slowly but was one of the better seed lots in terms of physiological quality (Table 3). This suggests that cv. BR5 is a slow germinating genotype.
Since seed longevity is strongly affected by seed quality (Tang et al., 1999), the deterioration rate may also be influenced by initial seed quality and possibly by genotype. Thus, the differences in seed longevity among seed lots of rice cultivars may result from a difference in initial seed quality, seed deterioration rate, or both. The differences in the rate of deterioration in this study were not very large and would, therefore, not result in dramatically different estimates of seed longevity. Thus, the vigour ranking (Table 3) which was done on the basis of initial seed quality and rate of seed deterioration was supported by the findings in other parameters, such as final germination percentage (Table 2), germination rate (Fig. 3, Table 4) and percentage abnormal seedlings (Table 2).
The results of the present study have shown that the differences in germination after ageing is the indication of the differences in initial seed quality and seed vigour of the cultivars. The results also revealed that germination of seed lots of fourteen rice cultivars in low temperature was influenced more by genotype than seed quality. There is no doubt that reduced seed vigour decreases field emergence. The magnitude of the decrease, however, depends on the severity of the stresses encountered in the field. It is desirable that a vigour test ranks different seed lots in order of superiority. In this way, it gives some rational measure of the absolute standard, so that relative difference between seed lots can be assessed.
Acknowledgments
The authors wish to thank the Director General of the Bangladesh Rice Research Institute for providing financial support for the study.