Plant Prod. Sci.

7 (2) : 129 ź 137 (2004)

Drought Tolerance Characteristics of Brazilian Soybean Cultivars

ź Evaluation and characterization of drought tolerance of
various Brazilian soybean cultivars in the field ź
Tetsuji Oya1, Alexandre Lima Nepomuceno2, Norman Neumaier2, José Renato Bouças Farias2,
Satoshi Tobita1 and Osamu Ito1

(1Japan International Research Center for Agricultural Sciences, Tsukuba 305-8686, Japan; 2National Soybean Research
Center of Brazilian Agricultural Research Corporation, Caixa Postal 231, 86001-970, Londrina-PR, Brazil)

Abstract : Drought is one of the major constraints for soybean production in Brazil. Seed yield of ten Brazilian
soybean cultivars sheltered from rain (drought stress) for one month after the first flowering was examined over
two growing seasons in the field in Londrina, Brazil. The drought tolerance on the basis of seed yield varied
with the cultivar, and the yield ranking among cultivars was nearly the same across two years. In cultivars with
higher drought tolerance, crop growth rate (CGR) during the drought stress period was higher than in other
cultivars. They also maintained a larger leaf area during the stress period. Although reproductive development
was retarded by the drought stress, it tended to be retarded less in drought-tolerant cultivars. The information
obtained in this research may be useful for breeding drought-tolerant cultivars or selecting diverse germplasms of
soybean cultivars.

Key words : Brazil, Crop growth rate, Cultivar difference, Drought tolerance, Harvest index, Rain shelter, Seed
yield, Soybean.

ǽThe commercial production of soybean in Brazil that exhibited less reduction in yield component such
started in the 1960s and exceeded 40 million tons as 100-seed weight and the number of seeds under
in 2002. Brazil is now the second greatest producer water stress conditions than other cultivars among
of soybean in the world next to the United States of four American cultivars. A soybean genotype originally
America (FAO STAT). Average yield of soybean in introduced from Japan, PI416937, also showed less
Brazil is slightly lower than 3 tons per hectare that yield reduction with slow wilting leaves under a
can be considered highly productive and similar to drought condition (Sloane et al., 1990). Hida et al.
that in other countries in North and South America (1995) found cultivar differences in drought tolerance
(FAO STAT). Stable and high productivity of soybean based on yield in Japanese and American cultivars,
in Brazil is needed not only for the Brazilian economy and Neumaier et al. (1995; 1997) also found cultivar
but also for the world economy (Kokubun and Fujisaki, differences in drought tolerance in Brazilian cultivars.
1997). Although Sammons (1978) found cultivar differences
ǽAlthough the total annual precipitation in Brazil is in drought response at the seedling stage, those
sufficient for soybean cultivation, water deficiency is cultivars were unable to be consistently categorized by
caused by a dry spell of more than a few weeks without the several measured variables such as water potentials,
rainfalls in the middle of growth period in a particular leaf development, and leaf photosynthesis. Moreover,
location. the differences in drought response in the seedling
ǽSoybean has been reported to have a wide variation stage of the cultivars were not entirely consistent with
in drought tolerance. In this study, drought tolerance the drought response of the cultivars based on yield
was defi ned as high yield under water defi cit. A data (Mederski and Jeffers, 1973). Therefore, some
cultivar is considered drought tolerant when the indices at a specific growth stage should be established
yield was significantly higher than the other cultivars to relate with drought tolerance on a yield basis.
in a drought environment but not in a non-drought ǽDrought stress during the vegetative growth stage is
environment and also when the yield difference among compensated to some extent by the subsequent rainfall
the cultivars is significant in both drought and non- during the reproductive growth stage. However, stress
drought environments and the yield level was ranked during the reproductive growth stage tends to reduce
higher under drought environments (Sneller and yield directly (Doss et al., 1974; Sionit and Kramer,
Dombek, 1997). Brown et al. (1985) found a cultivar 1977; Hirasawa et al., 1994; Saitoh et al., 1999). The

Received 2 June 2003. Accepted 25 October 2003. Corresponding author : S. Tobita (bita1mon@jircas.affrc.go.jp, fax
+81-29-838-6355).
 130 Plant Production Science Vol.7, 2004

Tableǽ1.ǽGenetic backgrounds and characteristics of the ten Brazilian soybean

cultivars (Embrapa Soja, 2002).

Days from emergence to physiological maturity are -115 d in early (E), 116-125
d in semi-early (S), and 126-137 d in medium (M) cultivars. *Flowering dates of
individuals in each cultivar with different maturities differed 3-4 days at most. **R :
Resistant, M : Moderately susceptible.

drought stress at the pod fi lling stage seems to be 2.ǽPlant cultivation and drought treatments
most serious (Fukui, 1965; Doss et al., 1974; Sionit and ǽStandard cultural practices in Brazil were employed
Kramer, 1977; Saitoh et al., 1999). including soil fertilization with N, P and K of 0-28-20
ǽIt is difficult to predict exactly when the plant will chemical fertilizer at 250 kg ha-1 and inoculation of
encounter drought stress during its cultivation period. the seed with Bradyrhizobium japonicum at sowing.
Preventive measures include the change of planting Herbicides and pesticides were used when necessary.
dates and/or adoption of the cultivars with different The soil at this location is classified as kaolinitic, clayey
maturities, to avoid encountering a severe drought (very fine) thermic typic Haplorthox.
at the reproductive growth stage. However, once a ǽThe ten cultivars were sown on 11 and 22 November
drought occurs unexpectedly, soybean plants cannot and harvested on 24 to 27 March and 21 March in
escape from it. Therefore, cultivars with high drought the season of 1999/2000 and 2000/2001, respectively.
tolerance are strongly desired. There were four replicates of each cultivar that
ǽIn this study, drought tolerance was evaluated by the consisted of eight rows spaced 0.5 m apart and 4 m
yield in drought environments relative to that in non- long (6 mą4 m). There were approximately 20 plants
drought environments (relative yield). The drought m-1 of row, giving a population of 40 plants m-2.
environment was simulated by sheltering the plants ǽThe plants were irrigated (IR plots) manually when
from rain in the fi eld. The objectives of this study soil water potential, as measured with tensiometers,
were to investigate whether there were differences in decreased to −0.05 MPa at a depth of 0.30 m. No
drought tolerance among soybean cultivars currently water was applied to the non-irrigated (NI) rain-fed
cultivated in Brazil and to examine the physiological plots. A split-plot design was employed, where the
characteristics causing the differences in drought irrigation regime was assigned to the main plot and
tolerance in terms of yield and growth characteristics. cultivar to the sub-plot. There were also plots of 6 mą
3 m, where the plants were artificially drought stressed
Materials and Methods
by sheltering them from rain (RS plots) with three
1.ǽLocation and cultivars replicates adjacent to the other two plot areas. The
ǽExperiments were conducted in two growing seasons plants were kept from rain for one month after the
of 1999/2000 and 2000/2001 at Soybean Research first flowering (from R1 to R5 stage, according to Fehr
Center of Brazilian Agricultural Research Corporation et al., 1971). Soil moisture contents were determined
ǽ(Embrapa Soja) (23º11'37"S, 51º11'03"W, elevation by collecting soil samples from top soil layer (10-20
630 m) in Londrina, Paraná, Brazil. cm in 1999/2000, 0-20 cm in 2000/2001) in each
ǽTen determinate soybean cultivars bred by Embrapa plot. The soil samples were dried for the estimation of
Soja (BR-16, BR-37, Embrapa 48, Embrapa 59, BRS gravimetric water content. Data were averaged from
132, BRS 133, BRS 134, BRS 183, BRS 184, BRS four soil samples for the IR plot and five samples for
185) were used as the materials (Table 1). They are the RS plot on each sampling day.
all adapted to and recommended for the southern ǽIt can be considered that yield in the IR plot
regions in Brazil, especially in Paraná State (Embrapa represents the potential yield, and the yield in the NI
Soja, 2002). Among them, BR-16 is considered to be a plot represents the yield under a rain-fed condition,
cultivar with less drought tolerance according to the and the yield in the RS plot represents the yield under
previous studies by Embrapa Soja (Nepomuceno et al., drought stress during the early reproductive stage.
1994; Farias et al., 1995; Neumaier et al., 1995; 1997).
 Oya et al. źź Drought Tolerance Characteristics of Brazilian Soybean Cultivars 131

Fig.ǽ1.ǽMeteorological data during the experimental periods. In 1999/2000, soybean cultivars were
planted on 11 Nov., flowered on 3 Jan., and were harvested on 24 to 27 Mar. In 2000/2001, they were
planted on 22 Nov., flowered on 8 Jan., and were harvested on 21 Mar. Precipitation in NI and RS
plots are shown by upper and lower lines, respectively.

3.ǽMeasurements ǽŽRelative yieldžwas calculated by dividing mean

ǽFive plants were periodically (every one or two weeks yield of each cultivar in the RS plot by mean yield in
in 1999/2000, only four times during reproductive the IR plot (potential yield). Similarly relative values
growth stage in 2000/2001) harvested from each for yield components were calculated.
replicate to determine the developmental stage, leaf Results
area development, and plant dry weight. Leaves were
detached and the leaf area was measured using a leaf 1.ǽMeteorological data and soil water content
area meter (LI-3100, Li-Cor, Lincoln, NE, USA). All ǽIn each growing season, meteorological data were
plant materials were then dried at 60ºC for at least 48 recorded daily adjacent to the experimental site. Fig.
h for the estimation of plant dry weight per unit land 1 shows precipitation, solar radiation and mean air
area. Crop growth rates in RS plots were calculated as temperature. Total amount of rainfall received by
the slope of linear regression of dry mass during the soybean plants in NI plots in 1999/2000 (471.0 mm)
period under the shelter from rain in 1999/2000. High was almost a half of that in 2000/2001 (741.8 mm),
regression coefficients (more than 0.75) were obtained but the plants in RS plots received almost the same
except Embrapa 48 and Embrapa 59 (less than 0.50). amount of rainfall during the season in both years
ǽAt maturity, all plants in a 1 mą2 m section of each (354.3 and 335.0 mm in 1999/2000 and 2000/2001,
plot were harvested by hand and machine-threshed. respectively). Amount of rainfall before the flowering
The seed yield was expressed on the basis of 130g time also differed between 1999/2000 (121.7 mm) and
moisture kg-1. Then yield analysis was carried out. 2000/2001 (256.8 mm).
132 Plant Production Science Vol.7, 2004

Fig.ǽ2.ǽChanges in soil moisture content in three

experimental plots. Soil samples were collected Fig.ǽ3.ǽSeed yield in RS plots and relative seed yield (seed
from soil layer of 10-20 cm (1999/2000) or 0-20 yield in RS plots relative to that in IR plots) in the ten
cm (2000/2001). After drying them, soil moisture Brazilian soybean cultivars in two growing seasons. 1 :
contents were calculated. RS plots were sheltered BR-16; 2 : BR-37; 3 : Embrapa 48; 4 : Embrapa 59; 5 : BRS
from rain from 53 and 55 days after planting (shown 132; 6 : BRS 133; 7 : BRS 134; 8 : BRS 183; 9 : BRS 184; 10 :
by arrows) in 1999/2000 and 2000/2001, respectively. BRS 185. *Significant at 0.05 level.
Data are presented as means and SDs of four
replicates.

Tableǽ2.ǽSeed yield of the ten Brazilian soybean cultivars under three

different water availabilities.

IR : Irrigated plot, NI : Non-irrigated plot, RS : Plot sheltered from rain (drought

stress) during one month after the first flowering. Values in parentheses in NI
and RS plots are relative yield (%) to IR plots.
 Oya et al. źź Drought Tolerance Characteristics of Brazilian Soybean Cultivars 133

Tableǽ3.ǽHarvest index of the ten Brazilian yield compared with BR-16 in 1999/2000, although a
soybean cultivars in IR and RS plots. significant difference was not found in 2000/2001. On
the basis of the yield, BRS 183 was considered to have
higher drought tolerance than BR-16, Embrapa 59 and
BRS 134.
ǽThe yield ranking among cultivars in RS plots was
stable in the two years (Fig. 3). The coeffi cient of
correlation between the yields of individual cultivars
in 1999/2000 and 2000/2001 was 0.64 in RS plots
and that between the relative yield (yield in RS plots
relative to that in IR plots) in the two years was 0.69
(both signifi cant at 5% levels). Cultivars with high
MeanĂSE of three replicates. potential yields (yield in IR plots) like BRS 134 and
Embrapa 59 had low relative yields (26% and 36%
for BRS 134, and 22% and 42% for Embrapa 59,
ǽFig. 2 shows the changes in the soil moisture content in 1999/2000 and 2000/2001, respectively) (Table
in three experimental plots. Soil moisture content in 2). On the contrary, cultivar with higher relative
RS plots declined already on seven days after the start yield like BRS 183 (34% and 78% in 1999/2000 and
of sheltering from rain in both years as compared with 2000/2001, respectively) had not so high potential
the other two plots. No difference in soil moisture yield, but had high yields in RS plots. Low yield of
content existed between IR and NI plots. BR-16 in RS plots was due to low potential yield rather
than low relative yield (30% and 48% in 1999/2000
2.ǽSeed yield and 2000/2001, respectively). From the viewpoint of
ǽIrrigation had no significant effects on seed yields relative yield also, BRS 183 was considered to have
in either year between IR and NI plots (Table 2). high drought tolerance and Embrapa 59 and BRS
However, seed yield in RS plots were only 22-34% and 134 were considered to have low drought tolerance.
36-78% of potential yield (IR plots) in 1999/2000 and Among the cultivars BRS 183, BRS 134, and Embrapa
2000/2001, respectively. In RS plots, BRS 183 showed 59, difference in potential yield, relative yield, and RS
significantly higher seed yield compared with BR-16, yield were obvious, however, a negative association
Embrapa 59, and BRS 134 in 2000/2001, although a observed among these cultivars was not found in all
signifi cant difference was not found in 1999/2000. cultivars.
In IR plots, BRS 134 showed significantly higher seed ǽDrought stress reduced harvest index (HI) in most

Tableǽ4.ǽYield components of the ten Brazilian soybean cultivars in IR and RS plots.

Values in RS plots relative to that in IR plots (%) are presented in the parentheses.
 134 Plant Production Science Vol.7, 2004

Tableǽ5.ǽReproductive development of the ten

Brazilian soybean cultivars in IR and RS plots
(1999/2000).

Fig.ǽ4.ǽDry matter accumulation of the ten Brazilian soybean

cultivars in RS plots in 1999/2000. At each sampling,
five plants were harvested for each plot, and the data are
presented as the means of three replicates.

Reproductive development stages (R1-R8) are shown as respectively (Table 4). Drought stress in RS plots
means of two replicates for each cultivar.
reduced the yield mainly through the reduction of
pod number per unit land area. It reduced the seed
of the cultivars in 1999/2000, but not in 2000/2001 number per pod in 1999/2000 and the seed size in
(Table 3). HI in RS plots was lower than 0.30 in 2000/2001, but increased the seed size in 1999/2000
1999/2000 in all cultivars, but was higher than 0.40 in and did not affect the seed number per pod in
all cultivars even in RS plots in 2000/2001. No cultivar- 2000/2001.
specific responses were found. ǽThe rate of reduction in pod number per unit land
area by drought stress varied with the cultivar in both
3.ǽYield components years, but no significant difference among cultivars
ǽAveraged relative values (values in RS plots relative was found in reduction rate. Only BRS 183, a cultivar
to that in IR plots) of each yield component in the ten with high drought tolerance, showed relatively low
cultivars were, 36 and 61% (number of pods per unit reduction rate in pod number in 2000/2001.
land area), 75 and 106% (seed number per pod), 109
and 87% (seed size) in 1999/2000 and 2000/2001,

Fig.ǽ6.ǽThe relationship between crop growth rate during

the period sheltered from rain (drought stress) and seed
Fig.ǽ5.ǽLeaf area development of the ten Brazilian soybean yield in the ten Brazilian soybean cultivars in RS plots
cultivars in RS plots in 1999/2000. At each sampling, in 1999/2000. 1 : BR-16; 2 : BR-37; 3 : Embrapa 48; 4 :
five plants were harvested for each plot, and the data are Embrapa 59; 5 : BRS 132; 6 : BRS 133; 7 : BRS 134; 8 : BRS
presented as the means of three replicates. 183; 9 : BRS 184; 10 : BRS 185. *Significant at 0.05 level.
 Oya et al. źź Drought Tolerance Characteristics of Brazilian Soybean Cultivars 135

ǽRelative seed size (values in the RS plot relative to emphasized the usefulness of the yield-basis selection
that in IR plot) in 1999/2000 was 90, 94, and 97% (less for drought tolerance under irrigated conditions. In
than 100%) in Embrapa 48, BRS 134, and BRS 184, this study, the nearly same yield ranking in the RS plots
respectively, but was 126, 123, and 122% (higher than in the two years suggested that the effect of drought
120%) in BRS 132, BR 37, and BRS 183, respectively for one month after the fi rst fl owering was large
(Table 4). Relative values in these two groups were enough to be detected by the yield-basis analysis. For
significantly different. In 2000/2001, no significant breeding and analysis of the physiological mechanisms
difference in the relative seed size among cultivars was of drought tolerance, relative yield (yield in RS plots
observed. relative to that in IR plots) may be a more important
ǽRelative seed number per pod was 73 and 87% index. High correlation coefficients between yield in
in BRS 134 and Embrapa 59, respectively, and one RS plots and relative yield were observed in both two
in Embrapa 48 was 137% in 2000/2001 (Table 4). years (r=0.89 and 0.98 in 1999/2000 and 2000/2001,
Relative values in these two groups were significantly respectively), whereas only low correlation coefficients
different. In 1999/2000, no significant difference in between yield in RS plots and yield in IR plots were
the relative values among cultivars was observed. obser ved (r=0.02 and −0.01 in 1999/2000 and
ǽHigh drought tolerance of BRS 183 resulted from 2000/2001, respectively). This indicated that the
small reduction in pod number and seed size under yield in RS plots related more to relative yield than to
drought condition. On the other hand, low drought potential yield in this experiment.
tolerance of Embrapa 59 resulted from a large ǽBRS 183 was considered to be drought-tolerant
reduction in seed size, and that of BRS 134 from a because relative yield was higher than the other
large reduction in seed number per pod and seed size. cultivars. On the other hand, BR-16, Embrapa 59,
and BRS 134 were considered to be drought-sensitive
4. Growth and development because the relative yield was lower than other
ǽReproductive development of soybean plants was cultivars. Moreover, BR 134 was considered to be
retarded in RS plots, but the developmental stage at less tolerant than BR-16 because its yield in IR plot
the end of the drought-stress period varied with the was signifi cantly higher than that of BR-16. This is
cultivar from R2.5 to R5 (Table 5). The retardation of consistent with the comparatively higher relative yield
development somewhat reduced seed yield (compare of BR-16 than that of BRS 134.
Table 5 with Fig. 3). ǽThe seed yield significantly correlated with CGRs
ǽBRS 183 accumulated more dry matter by the end during the early reproductive growth stage (Table
of the drought-stress period (85 d after planting), 4). This is presumably because soybean has a certain
compared with the other cultivars (Fig. 4). Dry period during which vegetative and reproductive
weight at the end of the drought-stress period was g r o w t h p r o g r e s s s i m u l t a n e o u s l y. D u r i n g t h i s
low in BR-16, Embrapa 59, and BRS 134. The dry period the leaf area becomes maximum and dry
weight of Embrapa 59 was kept low till the end of the matter production is most active. Progress of the
experiment. developmental stage during this period was retarded
ǽThe leaf area of BRS 183 was larger than that of by drought stress in RS plots (Table 5). Spaeth et al.
other cultivars at the end of the drought-stress period (1984) reported that harvest index (HI) in a cultivar
(85 d), and that of Embrapa 59 was smaller throughout was stable unless plants received drought stress in the
the experimental period (Fig. 5). Maximum leaf area late reproductive growth stage. In 1999/2000 little
was achieved on 85 d in BRS 183, but on 69 d in BR-16 precipitation with large yield reduction from potential
and BRS 134, which might have lowered their dry yield, HI was decreased in most cultivars. However, HI
weight at the end of the drought-stress period (85 d). did not decrease in most cultivars in 2000/2001. Thus,
ǽSignificant correlations were observed between crop the cultivar that maintained its dry matter production
growth rates (CGRs) during the drought stress period higher during the drought-stress period exhibited
(early reproductive growth period) and seed yields in higher seed yield in this study. Maintenance of a high
1999/2000 (Fig. 6), suggesting the importance of this CGR during drought-stress period seems to be a key
period for yield determination. for high seed yield under drought conditions, that is,
Discussion high tolerance to drought. Since there was a significant
correlation (r=0.63, significant at 0.05 level) between
ǽIn this study, yield ranking of the ten soybean dry matter accumulation before flowering and seed
cultivars was fairly stable in IR and RS plots across two yield in RS plot in 1999/2000, the amount of vegetative
years (r=0.90 and 0.64 in IR and RS plots, respectively). growth is considered to affect the subsequent growth
Sneller and Dombek (1997) reported that the rank during the drought-stress period presumably through
correlation of yield between the sets of two years was development of root systems. Comparing two yearsż
not signifi cant under conditions without irrigation results, cumulative precipitation in 1999/2000 was
but significant under irrigated conditions. Thus, they half of that in 2000/2001 (Fig. 1). The shortage
 136 Plant Production Science Vol.7, 2004

of soil moisture might restrict the dr y matter ǽSoybean has a characteristic to remobilize carbon
production of soybean before the flowering period in and nitrogen from leaves and stems to seeds. Westgate
1999/2000, causing stronger effect of drought stress et al. (1989) observed that steady seed growth was
after flowering. Severe stress might have masked the maintained even when photosynthetic rate was
difference in drought tolerance among cultivars. decreased by drought stress. Leaves, stems and
ǽPI416937 has been known as a drought-tolerant pericarps reduced their carbohydrate content during
genotype. Its characteristics related to drought the stressed period, suggesting that the remobilization
tolerance are high capability of osmotic adjustment of carbon would support the seed growth. The
and development of abundant fi ne adventitious reduction in seed size may be inescapable when
roots in a shallow soil layer (Sloane et al., 1990; drought continues longer. However, the cultivar with
Hudak and Patterson, 1996). However, a more less reduction in seed size can be considered superior
important characteristic may be its vigorous growth in remobilization capacity. Thus, it is speculated that
at the vegetative stage (Hudak and Patterson, 1995). both the vigor in vegetative growth and remobilization
Hirasawa et al. (1994) showed that the soybean plants capacity may be important physiological characteristics
which experienced drought before fl owering had for drought tolerance in soybean.
higher seed yields under a drought condition after ǽWe found a wide variation in drought tolerance
flowering because they had already developed a larger in ten released-cultivars in Brazil across two years. At
root system before flowering. These reports clearly least higher growth rate under a drought condition
indicate that growth analysis of the aboveground during early reproductive stage was associated with
of plants alone may be insuffi cient to explain higher yield, that is, higher drought tolerance in
drought tolerance. To adapt to short day length at soybean. The present information may be useful for
low latitudes, many Brazilian cultivars have a long- breeding drought-tolerant cultivars or selecting diverse
juvenile characteristic that prevents soybeans from too germplasms of soybean cultivars. It should be noted
early flowering without sufficient vegetative growth that for effective selection, it would be necessary to
(Hatwig and Kiihl, 1979). Although this characteristic find any characteristics exhibited also in the earlier
is considered favorable for drought tolerance, the growth stage in soybean plants.
relationship is still unclear.
Acknowledgements
ǽThe effects of drought stress on yield components
may vary with the growth stage at the time of exposure ǽThe authors wish to thank Nelson Delattre, Roseli
to the stress (Fukui, 1965; Saitoh et al., 1999; Desclaux Dagmar Rossi Cardoso, Dirceu Antonio de Oliveira,
et al., 2000). Drought stress reduces the number of Everaldo Correia do Carmo, Mário Luiz Maxwell
pods and seed size when applied from flower initiation Zaparoli, Glaucia Maxwell Zaparoli, Júlio Cesar
till fl owering period, the number of pods when Pedroso, Alexandre Carvalho, Rafael, Scalloni, Jean,
applied in flowering period, the number of seeds per Marcio, and Marcos for their assistance in the field
pod when applied from late flowering till early pod experiments and measurements of the plant materials.
filling period, and the seed size when applied in the
References
late pod filling period. Hida et al. (1995) formulated
three sets of soybean cultivars, each of which had a Brown, E.A., Caviness, C.E. and Brown D.A. 1985. Response of
pair of cultivars with similar phenology but different selected soybean cultivars to soil moisture deficit. Agron. J. 77 :
response to drought. Cultivar differences in response 274-278.
to drought were found in seed size but not in pod Desclaux, D., Huynh, T.-T. and Roumet, P. 2000. Identification
number in all cases. The cultivars with little reduction of soybean plant characteristics that indicate the timing of
drought stress. Crop Sci. 40 : 716-722.
in yield in response to drought stress before flowering
Doss, B.D., Pearson, R.W. and Rogers, H.T. 1974. Effect of soil
showed little reduction in seed size. Also in this study,
water stress at various growth stages on soybean yield. Agron.
even though pod number was greatly reduced, cultivar
J. 66 : 297-299.
differences in response to drought stress were larger Embrapa soja 2002. Cultivares de soja 2001/2002. Londrina, PR,
in the other yield components. The reduction in the Brazil*.
number of pods and seed size by the drought stress Farias, J.R.B., Neumaier, N. and Nepomuceno, A.L. 1995. Efeito
from fl ower initiation till fl owering period appears da variação diária da umidade do solo sobre a fotossíntese
through the reduction of dry matter production e resistência estomática da soja. In Sociedade Brasileira
(Fukui, 1965). Therefore, drought tolerance based Agrometeorologia, 9 ed., Anais. Congresso Brasileiro de
on yield may refl ect dry matter production under Agrometeorologia, 9. Campina Grande, PB, Brazil. 68-70*.
drought and the response of seed size to drought. In Fehr, W.R., Caviness, C.E, Burmood, D.T. and Pennigton, J.S.
this study, BRS 183 showed smaller reduction in yield 1971. Stage of development descriptions for soybeans [Glycine
as well as dry matter production during the sheltered max (L.) Merrill]. Crop Sci. 11 : 929-931.
(drought-stress) period with less reduction in seed size Fukui, J. 1965. Physiological and ecological studies of soybean
compared to the other cultivars. in relation to soil moisture condition. Journal of the Central
 Oya et al. źź Drought Tolerance Characteristics of Brazilian Soybean Cultivars 137

Agricultural Experiment Station 9 : 1-68**. Grande, PB, Brazil. 80-82*.

Hartwig, E.E. and Kiihl, R.A.S. 1979. Identifi cation and Neumaier, N., Farias, J.R.B. and Nepomuceno, A.L. 1997.
utilization of a delayed flowering character in soybeans for Diferenças no índice de tolerância à seca entre cultivares
short day conditions. Field Crops Res. 2 : 145-151. de soja. In UFU/DEAGO ed., Ata e Resumos. Reunião de
Hida, Y., Hirasawa, T. and Ishihara, K. 1995. Varietal differences Pesquisa de Soja da Região Central do Brasil, 18, Uberlândia,
in growth response of soybean to soil moisture depletion. Jpn. MG, Brazil. 274*.
J. Crop Sci. 64 : 565-572***. Saitoh, K., Mahmood, T. and Kuroda, T. 1999. Effect of moisture
Hirasawa, T., Tanaka, K., Miyamoto, D., Takei, M. and Ishihara, stress at different growth stages on flowering and pod set in
K. 1994. Effects of pre-flowering soil moisture deficits on dry determinate and indeterminate soybean cultivars. Jpn. J. Crop
matter production and ecophysiological characteristics in Sci. 68 : 537-544***.
soybean plants under drought conditions during grain filling. Sammons, D.J., Peters, D.B. and Hymowitz, T. 1978. Screening
Jpn. J. Crop Sci. 63 : 721-730. soybeans for drought resistance. I. Growth chamber
Hudak, C.M. and Patterson, R.P. 1995. Vegetative growth procedure. Crop Sci. 18 : 1050-1055.
analysis of a drought-resistant soybean plant introduction. Sionit, N. and Kramer, P.J. 1977. Effect of water stress during
Crop Sci. 35 : 464-471. different stages of growth of soybean. Agron. J. 69 : 274-278.
Hudak, C.M. and Patterson, R.P. 1996. Root distribution and Sloane, R.J., Patterson, R.P. and Carter, T.E., Jr. 1990. Field
soil moisture depletion pattern of a drought-resistant soybean drought tolerance of a soybean plant introduction. Crop Sci.
plant introduction. Agron. J. 88 : 478-485. 30 : 118-123.
Kokubun, M. and Fujisaki, K. 1997. Agriculture and soybean in Sneller, C.H. and Dombek, D. 1997. Use of irrigation in
South America. JIRCAS working Rep.No.7. Tsukuba, Japan**. selection for soybean yield potential under drought. Crop Sci.
Mederski, H.J. and Jeffers, D.L. 1973. Yield response of soybean 37 : 1141-1147.
varieties grown at two soil moisture stress levels. Agron. J. 65 : Spaeth, S.C., Randall, H.C., Sinclair, T.R. and Vendeland, J.S.
410-412. 1984. Stability of soybean harvest index. Agron.J. 76 : 482-486.
Nepomuceno, A.L., Farias, J.R.B. and Neumaier, N. 1994. Westgate, M.E., Schussler, J.R., Reicosky, D.C. and Brenner,
Efeitos da disponibilidade hídrica no solo sobre a cultura M.L. 1989. Effect of water deficits on seed development in
da soja. In EMBRAPA-CNPSo ed., Ata - Documentos 72. soybean. II. Conservation of seed growth rate. Plant Physiol.
Reunião de Pesquisa de Soja da Região Central do Brasil, 15. 91 : 980-985.
Londrina, PR, 1994. 42-43*.
Neumaier, N., Farias, J.R.B. and Nepomuceno, A.L. 1995. ǽ* In Portuguese.
Índice de tolerância à seca em quatro cultivares de soja. ǽ** In Japanese with English summary.
In Sociedade Brasileira de Agrometeorologia ed., Anais. *** In Japanese with English abstract.
Congresso Brasileiro de Agrometeorologia, 9, Campina