SOYBEAN SEED QUALITY AS AFFECTED BY TIME OF PLANTING IN THE DRY ZONE OF SRI LANKA

I Eight soybean (Clycine max (L.) Merr.) cultivars were planted during the first week of each calender month in the dry zone of Sri Lanka. The soybean crops established from May and June plantings produced yield and size of seed significantly greater than the other plantings. These seeds, which matured under more favourable weather conditions with the combination of daily mean temperature (28O~) and daily maximum relative humidity (75%). were of the highest quality as indicated by their high level of viability and vigour. In contrast, the crop of January and February plantings gave seeds of lowest quality and at the time of their maturation the daily mean temperature was remarkably high (30 31°c). The predicted model explaining the effect of weather factors on seed quality reveals that daily mean temperature and daily maximum relative humidity are the significant attribu.tes t o seed quality of soybean. Weather is probably the preharvest factor that has the greatest influence on seed quality in many crops, including cereals and legumes. Marked weather injuries in wheat seeds have been attributed to results from temperature1 or preharvest rains.' l Exposure to dampness6 or hot weather9 during maturation could lead to the production of low quality seeds in soybean. There are distinct advantages iq producing soybean seeds under favourable weather conditions and such conditions are rather uncommon in the subtropics and tropics.l Under favourable conditions, seed set and recovery is usually optimum; there is low incidence and severity of pests and diseases and germination and seed vigour are quite high. Hence, seed quality is usually good if seeds are produced under favourable conditions. In soybean growing areas of Sri Lanka, the prevailing practice has been to plant soybean in the early part of May and November which are the beginning of dry and wet seasons, respectively. However, it could be possible to extendthe Part of Ph.D. thesis submitted by senior author, to the Postgradvate Institute bf Agriculture of the University of Peradediya, Sri Lanka. 2 V. Arulnandhy and Y.D.A. Senanayake planting time within these growing seasons. This is important for various reasons which include unavailability of seed material and labour at the peak time of seasons and crop failure due to uncertain weather. Therefore, an investigation was undertaken to determine the most appropriate time of planting for quality seed production in soybean in the dry zone of Sri Lanka. This investigation also attempts to critically analyse the major weather factors that are responsible for seed quality of soybean. 2. Experimental The experiment consisted of eight soybean cultivars (Bossier, Improved Pelican, M 534, Pb-1, PM 78-13, PM 78-25, W S L 216, 30120-49-3) arranged in a randomized complete block design with four replications. The block was 27.5m x 1.6m in size and consisted of eight plots. Eachplot had four 3 m rows spaced at 40 cm, having plants at 5 cm apart. The seeds of these cultivars were planzed on the fifth day of each month for 12 months (May 1984 to April 1985) at the Agricultural Research Station, Maha Illuppallama, which is situated in the dry zone of Sri Lanka Maha llluppallama, being at a latitude of 8O5'N, could be classified as a humid tropical area. The experiment was well managed by providing the optimum conditions for growth and development. Seeds were harvested from two centre rows of each plot after 3 3% months, at development stage R8 at which 95% of the pods were matures8 karvested seeds were sun-dried to a seed moisture content of 10% and stored under cold storage at 10°C and 45% relative humidity until assayed for viability and vigour. The seed yield and the weight of 100 seeds were also recorded. 'From seed lots produced by each planting date, 25 seeds from each cultivar and each replicate were drawn at random for testing of seed quality. These seeds were mixed with Captan 80% WP (N[(Trichloromethyl) thio] 4-cyclohexane 1, 2 dicarboximide) at 3 g/l kg. of seeds and planted in moist sterilized sand contained in plastic boxes, Seeds were germinated under illumination at 8 h a day, using flourescent lamps of 750-1000 lux placed 1 m above the sand boxes, and at a constant temperature of 30 * 1°C, the known approximate optimum for soybean seed Another 25 Captan-treated seeds of each.sample were planted in the field where conditions were favourable for germination. Seedlings that emerged were counted five and eight days after planting in both germination tests. Seedlings were carefully uprooted eight days after planting, examined and classified in accordance with the criteria of the Association of Official Seed Analysts.2 Seedling length and dry weight ye*? deterpined using five seedlings taken at random from each sample. Seedling length was measured from the point of attachment of the cotyledons to the Seed Quality Af fec ted by Planting Time root tip. The same seedlings were used to obtain the dry weight gravimetrically after drying in an oven at 60°C for 48 h. Percentage germination was used as a measure of seed viability. Reported germination values include only normal seedlings counted after eight days. The seedling length and dry weight were the parameters assigned to estimate seed vigour. In addition, the emergence rate was computed as a modification of the method suggested by ~ a ~ u i r e " using h e formula given below: of seedlings number of seedlings on 5th day + emerged on 8th day after planting / 1 3 The data obtained on viability and vigour of seed lots harvested at different times were processed and subjected to the standard analysis of variance. Gemination percentages were transformed to arcsine degree scale before statistical analysis. Means were compired using Duncan's Multiple Range ~ e s t . ~ Correlations were carried out to see whether seed yield could be considered as a guide to predict seed quality. Regressions were computed as appropriate. 3. Results and Discussion Results indicate a considerable variation in yield, size and quality of seeds produced from soybean crops that were established from different planting dates. The mean germination of seed lots of eight cultivars in sand was over 91.8% for the plantings of May, June and July. These values were significantly higher than % germination of seed lots from other planthgs, except for the August and November plantings (Tabl< 1). Likewise, the field germination of seeds produced from the plantings of May, June and July was over 89.8% and significantly higher than the other plantings, with the exception of August planting (Table 1). The % germhation of January and February plantings was lowest in sand germination test9while field germination test showed lowest germination in February and March plantings. These results showed that seeds produced from field plantings made during the period May to July were generally of high viability. Emergence rates higher than 3.0 per day were noticed in the plantings made from May to August and also in November in both sand and field germination and these values did not differ significantly from each other (Table 1). Lowest values for emergence rate were obtained for the plantings of January, February and March. VArulnandhy and Y.D.A. Senanayake V) C t 7 m ~ S m m 8 5: a t ; , > u n 2 a 0 u . 3 0) z n a Z b Seed Quality Affected by Planting Time Tne highest mean seedling length was recorded for May and August plmtings in sand germination. In field germination, maximum length was observed from July planting; however, it did not differ statistically from May ,and June plantings(Tab1e 1). The highest mean seedling dry weight wasfor the November plantirig in sand germination which however, did not show any significant difference from plantings made in May and June (Table 1'). With respect to seedling dry weight under field conditions, May and June plantings showed the highest weight. The above data generally suggest that the seeds produced fr0.m plantings made during the period May to July and also in November have good seed vigour. TrbJe 2. Mean seed yield and size of eight soybean cultbars harvested from 12 different p lantings in one year, Planting time Mean Mean secd yield seed size Year Date iKglha) (@I00 seeds) 1984 May 5 4842.5 a 15.3, a June 5 July 5 , Augu'st 5 September 5 October 5 November 5 December 5 . January 5 February 5 March 5 April 5 Mean followed by same letters in each column do nor differ s@nifichtly at P = 0.05 based on Duncan's Multiple Range Test 6 V.aAmlnandhy and Y.D.A. Senanayake The highest seed yield was also obtained from May and June plantings which had the highest see.d size as well (Table 2). Seed yield and seed size of these plantings were found to be significantly greater than those of other plantings, including November planting which is the regular planting time during the wet season. Considering the collective data on seed yield, seed size, seed viability and vigour t ~ g ~ t h e r , the best time of planting for seed production appears to be the period during -May to June in the dry zone of Sri~ a n k i High quality seeds that were produced from May and June plantings matured under more fav~urable weather conditions with, the combination of daily mean temperature of about 2g°C and daily maximum relative humidity of 75-78% (Table 3). In contrast, seeds of plantings in January and February which w>re of lowest quality (Table l) , matured when the daily mean temperature was highest (30-31°C). Several workers have pointed out earlier that adverse weather conditions during seed maturation could cause severe seed uality roblems which reduce seed viability ! and vigour in s ~ ~ b e a n . ~ * ~ * ~ , " Gregg reported that dry hot weather and inadequate soil moisture during seed maturation resulted in green seeds of soybean which were of low quality with respect to germability. Table 3. ' Mean temperature and maximum relative humidity during seed maturation of the soybean cultivars p

Weather is probably the preharvest factor that has the greatest influence on seed quality in many crops, including cereals and legumes.Marked weather injuries in wheat seeds have been attributed to results from temperature1 or preharvest rains.' l Exposure to dampness6 or hot weather9 during maturation could lead to the production of low quality seeds in soybean.
There are distinct advantages iq producing soybean seeds under favourable weather conditions and such conditions are rather uncommon in the subtropics and tropics.lUnder favourable conditions, seed set and recovery is usually optimum; there is low incidence and severity of pests and diseases and germination and seed vigour are quite high.Hence, seed quality is usually good if seeds are produced under favourable conditions.In soybean growing areas of Sri Lanka, the prevailing practice has been t o plant soybean in the early part of May and November which are the beginning of dry and wet seasons, respectively.However, it could be possible to extend-the planting time within these growing seasons.This is important for various reasons which include unavailability of seed material and labour at the peak time of seasons and crop failure due t o uncertain weather.Therefore, an investigation was undertaken to determine the most appropriate time of planting for quality seed production in soybean in the dry zone of Sri Lanka.This investigation also attempts to critically analyse the major weather factors that are responsible for seed quality of soybean.

Experimental
The experiment consisted of eight soybean cultivars (Bossier, Improved Pelican, M 534, Pb-1, PM 78-13, PM 78-25, W S L 216, 30120-49-3) arranged in a randomized complete block design with four replications.The block was 27.5m x 1.6m in size and consisted of eight plots.Eachplot had four 3 m rows spaced at 40 cm, having plants at 5 cm apart.The seeds of these cultivars were planzed on the fifth day of each month for 12 months (May 1984 to April 1985) at the Agricultural Research Station, Maha Illuppallama, which is situated in the dry zone of Sri Lanka Maha llluppallama, being at a latitude of 8O5'N, could be classified as a humid tropical area.The experiment was well managed by providing the optimum conditions for growth and development.Seeds were harvested from two centre rows of each plot after 3 -3% months, at development stage R8 at which 95% of the pods were matures8 karvested seeds were sun-dried t o a seed moisture content of 10% and stored under cold storage at 10°C and 45% relative humidity until assayed for viability and vigour.The seed yield and the weight of 100 seeds were also recorded.
'From seed lots produced by each planting date, 25 seeds from each cultivar and each replicate were drawn at random for testing of seed quality.These seeds were mixed with Captan 80% WP (N[(Trichloromethyl) thio] -4-cyclohexane 1, 2 dicarboximide) at 3 g/l kg. of seeds and planted in moist sterilized sand contained in plastic boxes, Seeds were germinated under illumination at 8 h a day, using flourescent lamps of 750-1000 lux placed 1 m above the sand boxes, and at a constant temperature of 30 * 1°C, the known approximate optimum for soybean seed Another 25 Captan-treated seeds of each.samplewere planted in the field where conditions were favourable for germination.
Seedlings that emerged were counted five and eight days after planting in both germination tests.Seedlings were carefully uprooted eight days after planting, examined and classified in accordance with the criteria of the Association of Official Seed Analysts.2Seedling length and dry weight ye*? deterpined using five seedlings taken at random from each sample.Seedling length was measured from the point of attachment of the cotyledons to the root tip.The same seedlings were used to obtain the dry weight gravimetrically after drying in an oven at 60°C for 48 h.
Percentage germination was used as a measure of seed viability.Reported germination values include only normal seedlings counted after eight days.The seedling length and dry weight were the parameters assigned to estimate seed vigour.In addition, the emergence rate was computed as a modification of the method suggested by ~a ~u i r e " using h e formula given below: of seedlings number of seedlings on 5th day + emerged on 8th day after planting / 1 3 The data obtained on viability and vigour of seed lots harvested at different times were processed and subjected to the standard analysis of variance.Gemination percentages were transformed t o arcsine degree scale before statistical analysis.Means were compired using Duncan's Multiple Range ~e s t .~ Correlations were carried out to see whether seed yield could be considered as a guide to predict seed quality.Regressions were computed as appropriate.

Results and Discussion
Results indicate a considerable variation in yield, size and quality of seeds produced from soybean crops that were established from different planting dates.The mean germination of seed lots of eight cultivars in sand was over 91.8% for the plantings of May, June and July.These values were significantly higher than % germination of seed lots from other planthgs, except for the August and November plantings (Tabl< 1).Likewise, the field germination of seeds produced from the plantings of May, June and July was over 89.8% and significantly higher than the other plantings, with the exception of August planting (Table 1).The % germhation of January and February plantings was lowest in sand germination test9while field germination test showed lowest germination in February and March plantings.These results showed that seeds produced from field plantings made during the period May to July were generally of high viability.
Emergence rates higher than 3.0 per day were noticed in the plantings made from May to August and also in November in both sand and field germination and these values did not differ significantly from each other (Table 1).Lowest values for emergence rate were obtained for the plantings of January, February and March.Tne highest mean seedling length was recorded for May and August plmtings in sand germination.In field germination, maximum length was observed from July planting; however, it did not differ statistically from May ,and June plantings(Tab1e 1).The highest mean seedling dry weight was-for the November plantirig in sand germination which however, did not show any significant difference from plantings made in May and June (Table 1

').
With respect to seedling dry weight under field conditions, May and June plantings showed the highest weight.The above data generally suggest that the seeds produced fr0.m plantings made during the period May to July and also in November have good seed vigour.
The highest seed yield was also obtained from May and June plantings which had the highest see.d size as well (Table 2).Seed yield and seed size of these plantings were found to be significantly greater than those of other plantings, including November planting which is the regular planting time during the wet season.Considering the collective data on seed yield, seed size, seed viability and vigour t ~g ~t h e r , the best time of planting for seed production appears to be the period during -May t o June in the dry zone of Sri-~a n k i High quality seeds that were produced from May and June plantings matured under more fav~urable weather conditions with, the combination of daily mean temperature of about 2g°C and daily maximum relative humidity of 75-78% (Table 3).In contrast, seeds of plantings in January and February which w>re of lowest quality (Table l), matured when the daily mean temperature was highest (30-31°C).Several workers have pointed out earlier that adverse weather conditions during seed maturation could cause severe seed uality roblems which reduce seed viability ! and vigour in s ~~b e a n .~* ~* ~, " Gregg reported that dry hot weather and inadequate soil moisture during seed maturation resulted in green seeds of soybean which were of low quality with respect to germability.Seed yield was significantly and positively correlated with germination, emergence rate, seedling length and dry weight (Table 4).The close association of seed yield with seed quality characteristics such as seed viability and vigour suggests that seed yield may be considered as a useful guide to predict seed quality.Table 4.
Correlation between seed yield and parameters of viability and vigour of seeds of different plantings.
-- A multiple linear regression model was used to estimate the effect of weather parameters that prevailed during the period from physiological to harvest maturity on seed quality (Y) in terms of seed viability (% germination in sand at 30°C and field) and vigour (emergence rate) (Table 5).The weather parameters included in the estimation were daily maximum relative humidity (XI), daily minimum relative humidity (X2), daily mean relative humidity (X3), daily mean temperature (X4), daily rainfall (X5) and daily sunshine hours (X6).From the available data, the predicted model by estimation is given as: The model indicates that daily mean temperature and daily maximum relative humidity had a significant influence on seed viability and vigour.The other parameters considered did not influence seed quality significantly (Table 5).The regression coefficient values predict that % germination in sand and field respectively decrease by 6.1626 and 7.1747 for each OC increase in daily mean temperature and 1,4968 and 1.0305 for each % increase in daily maximum relative humidity (Table 5).Emergence rate which measures seed vigour declines by 0.0442 and 0.2016 per day for each 0 % increase in daily maximum relative humidity and each C increase in daily mean temperature (Table 5).The data obtained reveal that daily mean temperature and daily maximum relative humidity during the period from physiological to harvest maturity ofTsoybean crop are the most important determinants of seed quality.

Conclusions
Collective data on yield, size, viability and vigour of seeds produced from different plantings suggest that the best time of for the production of maximum yield of high quality soybean seed in the dry zone of Sri Lanka would be during-the period of May to June when moderate daily mean temperature (2g°C) and daily maximum relative humidity (75%) prevail during seed maturation.In contrast, seeds from January and February plantings which mature under high daily mean temperature (30-31°c) would be of lower quality.The data also show that, of the weather factors, daily mean temperature and daily maximum relative humidity during seed maturation are the probable major determinants of soybean seed quality.
Significant a t P = 0.01

Table 5 .
Estimates of the coefficients of regressions for variables in the model for determinirig the effect of weather parameters, during the period from physiological to harvest maturity (seed maturation), on seed quality.