全 文 ：遗 传 学 报 Acta Genetica Sinica, November 2006, 33 (11)：1007–1013 ISSN 0379-4172
Analysis of Genotype × Environment Interaction Effects for
Starch Pasting Viscosity Characteristics in Indica Rice
BAO Jin-Song①, SHEN Sheng-Quan, XIA Ying-Wu
Institute of Nuclear Agricultural Sciences, Zhejiang University, Hangzhou 310029, China

Abstract: Milled rice (Oryza sativa L.) is composed of approximately 90% starch. The properties of starch have considerable ef-
fects on cooked rice palatability and consumer acceptability. Starch pasting viscosity parameters serve as important indices in the
estimation of eating, cooking, and processing qualities of rice. In the present study, four cytoplasmic male-sterile (CMS) lines and
eight restorer (R) lines have been used in an incomplete diallel cross to analyze seed effects, cytoplasmic effects, maternal gene
effects, and their genotype × environment (GE) effects on the following starch pasting viscosity parameters: breakdown (BD), con-
sistency (CS), and setback (SB). The results demonstrated that the total main genetic variances (VG) accounted for over 64% of the
total genetic variance (VG + VGE) for the three traits, indicating that these traits were mainly controlled by the main genetic effects in
addition to the GE interaction effects. The estimated total narrow-sense heritability were 67.8%, 79.5%, and 79.5% for BD, CS, and
SB, respectively. The general heritability (h2G) accounted for over 75% of the total heritability (h2G + h2GE), indicating that early
selection would be effective for those traits and the selection efficiencies were relatively stable in different environments.
Key words: rice; starch; RVA; pasting viscosity; genotype × environment effect

Received: 2005-12-12; Accepted: 2006-05-17
This work was supported by the National Natural Science Foundation of China (No. 30300227).
① Corresponding author. E-mail: jsbao@zju.edu.cn
In rice (Oryza sativa L.) quality improvement
programs, amylose content is widely recognized as
one of the most important determinants of eating,
cooking, and processing qualities. However, it is also
agreed that there are secondary differences among
varieties that have similar amylose contents, one of
which is starch pasting viscosity[1].
Starch pasting viscosity has long been used in
rice quality breeding to develop rice varieties with
desirable eating, cooking, and processing properties,
in various rice breeding programs focusing on the
quality improvement[2,3]. In the United States of
America, it has been established that the intermediate
amylose content is a good indicator of an intermedi-
ate amylograph viscosity profile, when evaluating
crosses of the southern United States of America
long-grain rice with low amylograph viscosity profile
and high amylose content long grain rice cultivars[4].
In Australia, rice breeders have used Rapid Visco
Analyser (RVA) to test pasting viscosity at all stages
from F4 grains onward to estimate the cooking quality
and class[5]. From the traces, the peak viscosity,
breakdown after peak, setback, and viscosity on
cooling to 50℃ are measured. Although these meas-
urements are useful, the major use of RVA traces is
direct comparison with class standard varieties. In
China, starch pasting viscosity characteristics among
different rice have been well characterized in the past
few years[2,6―9]. The parameters, such as breakdown
(BD), consistency (CS), and setback (SB) viscosity,
are more useful than peak viscosity (PV), hot paste
viscosity (HPV), and cool paste viscosity (CPV)[6],
and their relationships with eating quality are well
established[6,8,9].
Only a few genetic studies have been reported
for the starch pasting viscosity parameters[4,10-12]. The
PV, HPV, and CPV may be controlled by a single lo-
cus with additive effects[4,12]. Using combining
1008 遗传学报 Acta Genetica Sinica Vol.33 No.11 2006

ability analysis, Jin et al. [11] found that RVA parame-
ters are mainly controlled by additive effects. Using a
more complex quantitative genetic model that could
dissect main genetic effects, cytoplasmic effects, and
maternal plant effects on the endosperm traits, Bao
and Xia[2] found that the genetic effects of the seed
(endosperm), cytoplasm, and maternal plant control
the viscosity parameters of indica rice. Analysis of
quantitative trait locus (QTL) confirmed that the wx
locus located on chromosome 6 controls all the vis-
cosity parameters except for PV[13―15]. However, no
QTLs for pasting viscosity were detected at the wx
locus when the genetic population was derived from
varieties with similar amylose content (IR64 and
Azucena), although different (CT)n microsatellite for
the parents was detected[16].
The effects of environment, in particular, the
ambient temperature during rice ripening, on the
starch pasting viscosity are significant[17,18]. However,
few studies have been concerned about the genotype
by environment effects on the starch pasting viscosity
parameters. Wan et al.[19] analyzed the stability for
the viscosity parameters across different environ-
ments and found that some rice varieties are more
stable than the others. Bao et al.[18] analyzed the
genotype by environment effects on eight rice culti-
vars in different locations and seasons, and found that
the seasonal effects mainly affect PV and HPV,
whereas genotypic effects mainly affect CPV, BD,
and SB. A better understanding of the impact of the
environment and genetic factors on the control of rice
starch pasting viscosity parameters would help
breeders select suitable rice with the desired starch
property. Rice grains represent the next generation of
the maternal plant, so the rice quality traits may be
controlled by genes of the triploid endosperm, cyto-
plasm, and maternal plant. In the present article, the
genetic model proposed by Zhu[20] and Shi et al.[21]
for quantitative traits of the endosperm in cereal crops,
has been used to dissect genetic and GE interaction
effects of the seed, cytoplasm, and maternal plant for
rice pasting viscosity characteristics, and to predict
their genetic components of variance and heritability.
1 Materials and Methods
1. 1 Plant materials and experimental design
Four cytoplasmic male-sterile (CMS) lines (P1 =
Zhenshan 97 A, P2 = Xieqingzao A, P3 = II32 A, and
P4 = Longtefu A), and eight restorer lines (P5 = 371,
P6 = Minghui 63, P7 = Zhehui No. 3, P8 = Wenhui
No. 4, P9 = Milyang 46, P10 = Ce 48-2, P11 = Ce
64-7, and P12 = IR36) were used in this experiment.
All possible single crosses were made with female
CMS lines and male R lines in an incomplete diallel
cross (4 × 8) in Hangzhou in September. The seeds of
the parents and F1s were further sown in November in
the Hainan Province and in next May in Hangzhou,
respectively. A single plant per hill was transplanted
to the paddy field at a spacing of 20 cm × 20 cm. The
experimental design was a completely randomized
block with two replications. There were 36 plants in
each plot. Female plants (CMS) were planted close to
the plots so that they could be crossed with the male
plants during the flowering period to obtain the F1
seeds by artificial pollination. The seeds of maternal
plants, F1s on the maternal plants and F2s on the F1
plants were harvested at maturity from 10 plants in
the middle part of the plot. Rice samples were first
milled to white rice using a Satake Rice Machine
(Satake Corporation, Japan), and then ground to flour
on a Cyclone Sample Mill (UDY Corporation, Fort
Collins, Colorado, USA) by a 100-mesh sieve.
1. 2 RVA pasting viscosity profiles
The RVA paste viscosity was determined on a
Rapid Visco Analyser (RVA, Model 3D, Newport
Scientific, Warriewood, Australia). According to Bao
and Xia[10], three original parameters of the pasting
viscosity can be obtained from the pasting curve:
peak viscosity (PV, first peak viscosity after gelatini-
zation), hot paste viscosity (HPV, paste viscosity
during the 95℃ holding period), and cool paste vis-
cosity (CPV, paste viscosity at the end of the test). In
addition, breakdown (BD = PV-HPV), setback (SB =
CPV-PV), and consistency (CS = CPV-HPV) can be
derived from the original three parameters (Fig. 1).
BAO Jin-Song et al.: Analysis of Genotype × Environment Interaction Effects for Starch Pasting Viscosity Characteristics in Indica Rice 1009
Breakdown refers to the maximum drop in viscosity
during cooking with reference to the peak viscosity,
whereas setback represents the rise or fall in the vis-
cosity at the end of the cooling cycle also with refer-
ence to the peak viscosity. The consistency represents
the rise in viscosity at the end of the cooling cycle
following the cooking of the suspension. All the vis-
cosity parameters were measured in Rapid Visco
Units (RVU).

Fig. 1 An illustration of Rapid Visco-Analyser (RVA)
pasting profile for rice starch
BD: breakdown; CS: consistency; SB: setback.

1. 3 Genetic model
The genetic model used in this experiment,
which was proposed by Zhu[20], can analyze the ge-
netic effects and genotype by environment interaction
effects of quantitative traits of endosperm in cereal
crops [21]. Genetic components of seed additive vari-
ance (VA), seed dominance variance (VD), cytoplasmic
variance (VC), maternal additive variance (VAm), ma-
ternal dominance variance (VDm), seed additive inter-
action variance (VAE), seed dominance interaction
variance (VDE), cytoplasmic interaction variance
(VCE), maternal additive interaction variance (VAmE),
maternal dominance interaction variance (VDmE),
phenotypic variance (Vp), and residual variance (Ve)
were estimated by the MINQUE(0/1) method with
the Jackknife procedure [20-22]. Estimation of vari-
ances were further used for calculating seed (en-
dosperm) heritability he2 = VA/Vp, cytoplasmic
heritability hc2 = VC/Vp, maternal heritability hm2 =
VAm/VP, seed interaction heritability heE2 = VAE/Vp,
cytoplasmic interaction heritability hcE2 = VCE/Vp,
and maternal interaction heritability hmE2 = VAmE/VP.
All data were analyzed by C programs (supported by
Dr. Jun Zhu at Zhejiang University) running on an
IBM PC computer.
2 Results and Discussion
2. 1 Phenotypic performance of the parents and
their progenies
Differences in means and ranges of generations
for the three RVA starch pasting viscosity parameters
(BD, CS, and SB) were significant (Table 1). The
variation ranges of CMS lines were smaller than
those of the restorer lines in both environments. The
major reason was that more male parents with wider
diversity were involved in the study. All the means of
F1 and F2 were between the means of parents, sug-
gesting that the heterosis in F1 seeds for the three
traits, if existed, was minor. The three pasting viscos-
ity traits varied in different environments, indicating
that genotype effects as well as GE interaction effects
could affect the variations of the RVA profiles. Fig. 2
shows an example of RVA profiles of Longtefu (P4),
371 (P9), and their F1 and F2 generations. The peak
viscosity, hot paste viscosity, and cool paste viscosity
in HZ were much higher than those in HN (Fig. 2).
However, differences in BD, CS, and SB were not as
high as the differences in PV, HPV, and CPV (Table
1), because some of the differences were set off by
minus calculations.
2. 2 Estimation of seed, cytoplasmic, and mater-
nal genetic variances
Table 2 summarizes the estimated variances of
starch pasting viscosity profiles of rice. The total
main genetic variances (VG) accounted for over 64%
of the total genetic variance (VG + VGE) for the three
traits, indicating that the main genetic effects were
the chief cause for the variations in these traits.
However, genotype × environment interaction effects
also affected them.
In the genetic main variances (VG), no cytoplasmic
effects were detected for BD, but all three traits were

1010 遗传学报 Acta Genetica Sinica Vol.33 No.11 2006

Table 1 Phenotypic means (range) of generations for starch pasting viscosity parameters, breakdown (BD), consistency
(CS), and setback (SB) in Hainan (HN) and Hangzhou (HZ), respectively
Generation Environment BD (RVU) CS (RVU) SB (RVU)
Female HN 88.6 ± 6.4 158.2 ± 12.8 69.6 ± 17.9
(77.6-94.6) (143.8-177.1) (51.6-99.5)
HZ 62.9 ± 11.9 143.1 ± 20.4 80.3 ± 21.7
(43.3-81.0) (126.3-176.3) (49.7-109.7)
Male HN 111.9 ± 30.6 97.7 ± 33.4 -14.1 ± 61.2
(73.1-164.8) (59.3-156.1) (-102.9-61.0)
HZ 101.7 ± 26.6 86.0 ± 32.3 -15.7 ± 57.1
(63.1-140.9) (50.1-138.8) (-90.0-57.8)
F1 HN 81.5 ± 15.0 147.8 ± 16.9 65.8 ± 22.0
(54.9-107.0) (119.0-174.8) (30.0-98.8)
HZ 68.6 ± 10.9 128.7 ± 13.2 60.1 ± 15.2
(48.5-90.3) (107.4-160.3) (29.6-90.9)
F2 HN 94.4 ± 18.7 123.9 ± 13.0 29.5 ± 18.8
(56.7-133.8) (95.4-151.9) (-28.1-85.3)
HZ 81.8 ± 16.4 112.3 ± 14.3 29.9 ± 25.6
(19.5-108.3) (87.1-140.0) (−15.2-74.5)


Fig. 2 RVA profiles of the starch of two parents (P4 =
Longtefu; P9 = Miyang 46), and their F1 and F2 seeds in
Hainan (HN, top) and Hangzhou (HZ, bottom)
controlled by seed direct effects (i.e., endosperm ef-
fects) and maternal effects. The seed direct variances
(including VA and VD) accounted for over 53% of the
total main genetic variances. The second largest
variance component came from the maternal plant
effects (VAm + VDm), which accounted for 47%, 23%,
and 30% for BD, CS, and SB, respectively.
For the variant components of genotype × envi-
ronment interaction effects, no cytoplasm by envi-
ronment effects (VCE) were detected for the three
traits, indicating that the cytoplasmic effects were
stable among different environments. The seed direct
interaction effects (VAE + VDE) were the major com-
ponents of variances for BD and SB, accounting for
over 55% of the total genetic interaction variances
(VGE). This indicated that the environment easily af-
fected the expression of seed endosperm genes for
both traits. The maternal interaction variances ac-
counted for 74% for CS, indicating that the expres-
sion of the maternal genes for CS was easily influ-
enced by the environment.
Although the residual variances (Ve) were sig-
nificant for the three traits, they accounted for a very
small proportion of the total variance (VG+ VGE).

BAO Jin-Song et al.: Analysis of Genotype × Environment Interaction Effects for Starch Pasting Viscosity Characteristics in Indica Rice 1011


Table 2 Estimation of genetic variances of amylograph
viscosity traits in indica rice (BD: breakdown; CS: consis-
tency; SB: setback)
Variance BD CS SB
VG 1181.92 1124.39 2716.36
VA 516.45** 572.05** 1292.72**
VD 106.92** 81.36** 184.84**
VC 0 211.50** 422.22**
VAm 433.75** 198.88** 659.66**
VDm 124.80** 60.60** 156.92**
VGE 654.27 268.93 794.27
VAE 316.74** 0.00 435.01**
VDE 130.67** 70.65** 0.00
VCE 0.00 0.00 0.00
VAmE 0.00 134.81** 0.00
VDmE 206.86** 63.47** 359.26**
Ve 32.37** 11.31** 21.74**
VP 1868.56** 1404.63** 3532.37**
VG/(VG + VGE) 0.644 0.807 0.774
**: significant at 1% level; VG = main genetic variance; VA
=seed additive variance; VD = seed dominance variance; VC =
cytoplasmic variance; VAm = maternal additive variance, VDm =
maternal dominance variance; VGE = genotype by environment
interaction variance; VAE = seed additive interaction variance;
VDE = seed dominance interaction variance; VCE = cytoplasmic
interaction variance VAmE = maternal additive interaction vari-
ance; VDmE = maternal dominance interaction variance; Ve =
residual variance; VP = phenotypic variance.

2. 3 Estimation of components for general herita-
bility and interaction heritability
The estimated total narrow-sense heritabilities
were 67.8%, 79.5%, and 79.5% for BD, CS, and SB,
respectively. Our previous study indicated that the
heritability of the traits might be overestimated if
genotype by environment interactions existed. As a
result, the heritability estimated in the present study
was a little lesser than the previous reports because of
the dissection of genotype by environment interaction
variances. The general heritability (h2G) accounted for
75%, 87.9%, and 84.5% of the total heritability (h2G +
h2GE) for BD, CS, and SB, respectively, indicating
that early selection would be effective for those traits
that were controlled by the additive gene (including
cytoplasmic effects), and the selection efficiencies
were stable in different environments (Table 3).

Table 3 Estimation of heritability of amylograph viscosity
traits in indica rice
Heritability BD CS SB
hG2 0.509 0.699 0.672
hGe2 0.276** 0.407** 0.366**
hGc2 0.000 0.151** 0.120**
hGm2 0.232** 0.142** 0.187**
hGE2 0.170 0.096 0.123**
hGeE2 0.170** 0.000 0.123
hGcE2 0.000 0.000 0.000
hGmE2 0.000 0.096** 0.000
hG2 + hGE2 0.678 0.795 0.795
hG2/( hG2 + hGE2) 0.750 0.879 0.845
**: significant at 1% level; hG2 = general heritability; hGe2 =
seed general heritability; hGc2 = cytoplasmic general heritabil-
ity, hGm2 = maternal general heritability; hGE2 = Interaction
heritability; hGeE2 = seed interaction heritability; hGcE2 = cyto-
plasmic interaction heritability, hGmE2 = maternal interaction
heritability.

For the general heritability components, the seed
direct heritability (hGe2) was the major component for
all of the three traits, which accounted for 54.2%,
58.2%, and 54.5% of the total general heritability
(hG2). Because the inheritance of cytoplasmic effects
was derived from the maternal plant, the seed general
heritability was larger than the total maternal general
heritability and cytoplasmic heritability for all the
three traits, suggesting that selection based on a sin-
gle seed would be more effective than on a single
plant. This agreed with our previous study[10]. How-
ever, it should be mentioned that the viscosity of a
single seed was unable to be tested by RVA, and the
selection of these traits could be delayed in later gen-
erations. In our rice breeding practice, we began to
analyze the RVA profile with the F4 seeds (on the F3
plant). Blakeney[5] also reported that they had used
the RVA to test pasting viscosity at all stages from F4
grains onward to estimate cooking quality and class.
For the interaction heritability, because signify-


1012 遗传学报 Acta Genetica Sinica Vol.33 No.11 2006


cant V

[7] Shu Q Y, Bao J S, Wu D X, Xia Y W. Application of the
Rapid Visco Analyser in the improvement of rice eating
quality. In: Shu Q Y, Xia Y W ,eds. Breeding and Appli-
cation of the Quality Early Season Rice in the Middle
and Lower Reaches of Yangtse River. Zhejiang Univer-
sity Press, 1999, 125-131.
AE were detected for BD and SB, the seed direct
interaction heritability was 17% for BD and 12.3%
for SB. For CS, it was only a small amount (9.6%)
but significant maternal interaction heritability was
detected.
In conclusion, even though the RVA profiles of
rice starch were highly affected by different environ-
ments, only a small amount of interaction variances
were detected, so the RVA profiles were mainly con-
trolled by the genetic main effects. The general
heritabilities (h2G) were 50.9%, 69.9%, and 67.2% for
BD, CS, and SB, respectively, which was higher than
the interaction heritability (hGE2). This indicated that
early selection would be effective for those traits.
Early selection was more efficient based on a single
seed method than that of a single plant method.
However, selection of viscosity parameters tested by
RVA would begin from F4 grains onwards, because
RVA was unable to test the viscosity of a single seed.
References:
[1] Juliano B O. Criteria and tests for rice grain qualities. In:
Jaliano B O, eds. Rice Chemistry and Technology .The
American Association of Cereal Chemists, Ins. St. Paul,
Minnesota, USA. 1985, 443-524.
[2] Zhu M S, Tang S Z, Gu M H. Progresses in the study on
the assessing, genetic and breeding of the rice starch
RVA profile in rice eating quality. Chinese Agric Sci Bull,
2005, 21(8): 59-64 (in Chinese with an English ab-
stract).
[3] Bao J S, Shen S Q, Xie J K. A study on the improvement
of eating quality of early indica rice with the assistance
of RVA. Mol Plant Breed, 2004, 2(1): 49-53.
[4] Gravois K A, Webb B D. Inheritance of long grain rice
amylograph viscosity characteristics. Euphytica, 1997,
97(1): 25-29.
[5] Blakeney A B. Rice. In: Henry R J, Kettlewell P S , eds.
Cereal Grain Quality. London: Chapman & Hall, 1996,
55-76.
[6] Shu Q Y, Wu D X, Xia Y W, Gao M W, McClung A.
Relationship between RVA profile character and eating
quality in Oryza sativa L. Scientia Agricultura Sinica,
1998, 31(3): 25-29 (in Chinese with an English ab-
stract).
[8] Hu P S, Zhai H Q, Tang S Q, Wan J M. Rapid evaluation
of rice cooking and palatability quality by RVA profile.
Acta Agronomica Sinica, 2004, 30(6): 519-524 (in Chi-
nese with an English abstract).
[9] Sui J M, Li X, Yan S, Yan C J, Zhang R, Tang S Z, Lu J
F, Chen Z X, Gu M H. Studies on the rice RVA profile
characteristics and its correlation with the quality. Scien-
tia Agricultura Sinica, 2005, 38(4): 657-663 (in Chinese
with an English abstract).
[10] Bao J S, Xia Y W. Genetic control of paste viscosity
characteristics in indica rice (Oryza sativa L.). Theor
Appl Genet, 1999, 98 (6/7): 1120-1124.
[11] Jin Z X, Jiang W Z, Chin Y, Koh H. Analysis on the
combining ability of taste meter value and starch RVA
properties in indica rice. Acta Agronomica Sinica, 2004,
30(12): 1210-1214 (in Chinese with an English abstract).
[12] Li X, Zhang R, Sui J M, Liang G H, Shen X P, Yan C J,
Gu S L, Gu M H. Performance and inheritance of rice
starch viscosity (RVA profile) characteristics. Chinese
Journal of Rice Science, 2004, 18(5): 384-390 (in Chi-
nese with an English abstract).
[13] Bao J S, Zheng X W, Xia Y W, He P, Shu Q Y, Lu X,
Chen Y, Zhu L H. QTL mapping for the paste viscosity
characteristics in rice (Oryza sativa L.). Theor Appl
Genet, 2000, 100(2): 280-284.
[14] Bao J S, Corke H, He P, Zhu L H. Analysis of quantita-
tive trait loci for starch properties of rice based on an
RIL population. Acta Botanica Sinica, 2003, 45(8): 986-
994.
[15] Shen S Q, Zhuang J Y, Shu Q Y, Bao J S, Wu D X, Xia Y
W. Analysis of QTLs with main, epistasis and G × E
interaction effects of starch paste viscosity in rice. Acta
Agronomica Sinica, 2005, 31(10): 1289-1294.
[16] Bao J S, Wu Y R, Hu B, Wu P, Cui H R, Shu Q Y. QTL
for rice grain quality based on a DH population derived
from parents with similar apparent amylose content.
Euphytica, 2002, 128(3): 317-324.
[17] Bao J S, Xu J J, Wu W Q, Shu Q Y, Xia Y W. The
changes of eating and cooking quality of indica early
rice in different cropping seasons. Journal of Zhejiang
University (Agric & Life Sci), 2002, 26(1): 103-106 (in
Chinese with an English abstract).
[18] Bao J S, Kong X L, Xie J K, Xu L J. Analysis of geno-
typic and environmental effects on rice starch. 1. Ap-
parent amylose content, pasting viscosity, and gel texture.
BAO Jin-Song et al.: Analysis of Genotype × Environment Interaction Effects for Starch Pasting Viscosity Characteristics in Indica Rice 1013

J Agric Food Chem, 2004, 52(19): 6010-6016.
[19] Wan X Y, Chen L M，Wang H L, Xiao Y H, Bi J C, Liu X,
Zhai H Q, Wan J M. Stability analysis for the RVA profile
properties of rice starch. Acta Agronomica Sinica, 2004,
30 (12): 1185-1190 (in Chinese with an English abstract).
[20] Zhu J. Analysis methods for seed models with genotype
× environment interactions. Chinese Journal of Genetics,
1996, 23(1): 56-68.
[21] Shi C H, Zhu J, Zang R C, Chen G L. Genetic and het-
erosis analysis for cooking quality traits of indica rice in
different environments. Theor Appl Genet, 1997, 95(1/2):
294-300.
[22] Miller R G. The Jackknife—a review. Biometrika, 1974,
61(1): 1-15.

籼米淀粉粘滞性的基因型与环境互作研究
包劲松，沈圣泉，夏英武
浙江大学农学院核农所，杭州 310029

摘 要: 水稻精米中大约含有 90%的淀粉，因此淀粉的特性对水稻的食味品质有很大的影响。淀粉粘滞性是预测稻米食用、
蒸煮和加工品质的重要指标。本研究利用 4 个细胞质雄性不育系和 8 个恢复系配置的不完全双列杂交组合来分析淀粉粘滞
性指标（崩解值、回复值和消减值）的胚乳、细胞质和母体基因效应及环境互作效应。结果表明：在崩解值、回复值和消
减值的遗传变异中，遗传主效应方差分量占了 64%以上，表明它们主要受遗传主效应控制，同时也受到基因型与环境互作
效应的影响。崩解值、回复值和消减值的总遗传率分别为 67.8%、79.5%、79.5%，而且普通遗传率占了总遗传率的 75%以
上，表明对这些性状的早世代选择有效，且在不同环境中选择效果相对稳定。
关键词: 水稻；淀粉粘滞性；RVA；基因型与环境互作
作者简介: 包劲松（1971-），男，浙江人，博士，研究方向：水稻遗传与生物技术。E-mail: jsbao@zju.edu.cn