Soil Sci. Plant Nutr., 61 ( I ) , 101 - 108,2005 101

Genotypic Variation in Shoot Cadmium Concentration in Rice and Soybean in Soils with Different Levels

of Cadmium Contamination

Satoru Ishikawa, Noriharu Ae**, Megumi Sugiyama, Masaharu Murakami, and Tomohito Arao*

Department of Environmental Chemistry, Heavy Metal Group, Soil Biochemistry Unit; *Chemical Analysis Research Center. Environmental Chemicals Analysis Laboratory, National Institute for Agro-Environmental Sciences,

Tsukuba, 305-8604 Japan; and **Department of Biological and Environmental Sciences, Faculty of Agriculture, Kobe University, Nada-ku, Kobe, 657-8501 Japan

Received July 15,2004, accepted in revised form November 26, 2004

A pot experiment was conducted to investigate whether the shoot cadmium (Cd) concentra- tion in 11 rice and 10 soybean cultivars varied among 4 soils with different levels of Cd con- tamination. Significant differences in shoot Cd concentration were found among rice or soybean cultivars grown in the 4 soils. The ranking of the rice cultivars for the shoot Cd concentration varied considerably among the soils. On the other hand, the soybean culti- vars were ranked similarly in terms of shoot Cd concentration in the 4 soils. Significant and positive correlations were found between the Cd and Zn concentrations and between the Cd and Mn concentrations in the shoot of rice cultivars, when they were grown in 2 soils with relatively moderate levels of Cd contamination. The shoot Cd concentration in the soybean cultivars, however, was not correlated with the concentrations determined for any of the metals (Zn, Mn, Cu, and Fe) across the 4 soils. Significant and positive correlations between the concentrations of Cd in younger shoots and mature seeds were detected among the soybean cultivars in 2 soils used, unlike among the rice cultivars, indicating that it may be diftlcult to evaluate the genotypic variation in seed Cd concentration using rela- tively younger shoots in the case of rice. These results revealed that genotypic differences in shoot Cd concentration in rice or soybean are variable or invariable among soils, respec- tively.

Key Words: cadmium, Cd-polluted soils, heavy metals, rice cultivars, soybean cultivars.

The international standard of cadmium (Cd) concen- tration in foods has been discussed by the Codex Ali- mentarius Commission of FA0 /WHO due to the growing concern about the safety of foods and human health. Among the staple foods, since rice and soybean are the major sources of dietary intake of Cd for the Jap- anese people, the development of effective techniques for reducing the Cd levels of rice and soybean is urgent- ly required.

Several alleviating techniques, such as soil dressing, soil amendment, and soil submergence have been applied in Cd-polluted paddy fields to reduce the Cd levels of rice grains in Japan. However, an alternative approach is needed because these methods have draw- backs in terms of effectiveness and economic value. Selecting or breeding crop cultivars with low levels of Cd-accumulation (hereafter referred to as “low-Cd- accumulating cultivars”) has been proposed as one of

the promising approaches to minimize Cd contamina- tion. The uptake and accumulation of Cd vary consider- ably among plant species and also among cultivars within a species (Arao and Ae 2003; Arao et al. 2003). Most of the studies reported on the screening of low-Cd cultivars were carried out using a water culture system, by the addition of Cd to non-polluted soils, or by using one soil series. The selection and evaluation of low-Cd cultivars in one soil or water culture for breeding pur- poses may not be practical, because the uptake of Cd in plants is considerably affected by various soil factors, e.g., pH, CEC, clay content, total carbon content, con- centrations of heavy metals including Cd, etc. (Li et al. 1995). In addition, Cd in soils occurs in complicated forms because of its association with a number of physi- cochemical forms that in turn influence its availability.

It has been suggested that the uptake of Cd by roots and its subsequent translocation to shoots could be con-

102 S. ISHIKAWA et al.

siderably affected by other metals coexisting in Cd-pol- luted soils, such as Zn, Mn, Cu, etc. The synergistic or antagonistic interactions between Cd and other metals were reported in rice (Liu et al. 2003), wheat (Zhang et al. 2002), maize (Florijn and Van Beusichem 1993), let- tuce (Crews and Davies 1985), etc. The ratio of Cd to other heavy metals in soils depends on the source of Cd pollution, and the uptake and translocation of Cd may also vary depending on the metal ratio in different soils with complex contaminants.

The main objectives of the present study were to investigate the genotypic differences in shoot Cd con- centration and the interactions between Cd and other heavy metals in terms of shoot concentration in rice and soybean plants grown in soils with various levels of Cd contamination. We also considered the possibility of evaluating the genotypic variation in seed Cd concentra- tion using relatively younger shoots.

MATERIALS AND METHODS

Characteristics of Cd-polluted soils. In 2002, samples of Cd-polluted soils were taken from four loca- tions in Japan and used in pot experiments in the same year, as described below. The soil types consisted of 2 Fluvisols (A and B, Table 1) and 2 Andosols (C and D, Table 1). Heavy metal contamination was derived from effluents and waste ores of mine stations for A, emission from metal-smelters for B and D, and both sources for C. Heavy metal contamination in soils was evaluated based on the concentrations of 0.1 M HCI-extractable Cd, Zn, Cu, and Mn. The fractionation of Cd in the 4 soils was performed according to the method of Sada- mot0 et al. (1994). The Cd, Zn, Mn, and Cu concentra- tions in the soils were determined by inductively coupled plasma optical emission spectroscopy (ICP- OES, Vista-Pro, Varian, Australia). The characteristics of the 4 soils are shown in Table 1 .

Plant materials and growth conditions. Eleven rice (Oryza sativa L.) cultivars with 3 different geno- types [Japonica (cvs. Nipponbare, Akitakomachi, Aki- hikari, and Mizuhatamochi), Indica (cvs. Kasalath, LAC 23, Habataki, HU-LO-TAO, and PEH-KU-TAO-TU), and Japonica-Indica crossing (cvs. IR-8 and Milyang 23)], and 10 soybean (Glycine max (L.) Merr.) cultivars (Sakukei 4, Enrei, Kiyomidori, Fukuyutaka, Suzunone, Tachinagaha, Tachiyutaka, Tohoku 128, Suzuyutaka, and Harosoy) were selected in the present study. It has been reported that these rice and soybean cultivars dis- played a large genotypic variation in seed Cd concentra- tion (Arao and Ae 2003; Arao et al. 2003). These cultivars were grown in a greenhouse in plastic pots with three replications for each cultivar. The pots were

filled with 500 g of each of the Cd-polluted soils. Fertil- izers were applied at the following rates: 0.1 g of N, 0.1 g of P,O,, and 0.1 g of K,O per pot for rice and 0.02 g (N), 0.15 g (P,O,), and 0.1 g (K,O) for soybean in the form of ammonium sulfate, single superphos- phate, and potassium sulfate, respectively. The pH of the soil solution was adjusted to 6.0 by the application of lime to the A and B soils prior to soybean cultivation, but it was not adjusted for rice cultivation. The seeds of the rice and soybean cultivars were directly sown in soils, and tap water was supplied when required. The rice or soybean seedlings were thinned to 3 or 1 per pot after emergence, respectively. Rice was grown under upland conditions to reveal the maximum capacity of Cd uptake among cultivars. Shoots were harvested at 40d after sowing.

Determination of Cd and other heavy metals in shoots. Plant samples were dried to a constant weight and ground to a fine powder using a stainless steel grinder (P-14, Fritsch, Germany). The powder (0.5 g) was digested with 10 ml of a HN0,-HC10,-H,SO, (5 : 1 : 1, v / v) mixed solution using the digest system (K-437, BUCHI, Germany). The Cd, Zn, Mn, Cu, and Fe concentrations of the samples were determined by ICP-OES.

RESULTS

Genotypic differences in shoot Cd concentra- tion in rice and soybean plants grown in soils with various levels of Cd pollution

A summary of the shoot dry weight, Cd concentra- tion, and Cd content of 11 rice and 10 soybean cultivars grown in the four soils is presented in Table 2. The aver- age value of the shoot dry weights of the rice and soy- bean cultivars was lowest in the D soil; chlorotic symptoms appeared in several cultivars due to the high heavy metal concentration in the soil (Table 1 ) . The average value of the shoot Cd concentration was higher in the rice cultivars than in the soybean cultivars across the 4 soils. Similar results were observed for the shoot Cd content. In the rice cultivars, the average value of the Cd concentration was the highest in the D soil, followed by the C, B, and A soils. However, the average value of the shoot Cd content in rice was the highest in the C soil, followed by the D, B, and A soils, due to the lowest value of the shoot dry weight in the D soil. In the soy- bean cultivars, the average value of the Cd concentration and Cd content in shoots was the highest in the D soil, followed by the B, C, and A soils. A wide range of shoot Cd concentrations was found among the rice or soybean cultivars, irrespective of the extent of Cd contamination in the soils. In order to determine whether the ranking of

'Igb

le 1

. C

hem

ical

char

acte

ristic

s of

met

al-c

onta

min

ated

soils

in a

pot

exp

erim

ent.

0.1 M

HC1

-ext

ract

able

hea

vy m

etal

s (m

g kg

-l so

il)

Cd

zn

Mn

cu

Frac

tiona

tion

of C

d in

soi

ls (m

g kg

-' so

il)

Exch

ange

able

In

orga

nic

Org

anic

O

cclu

ded

into

si

tes

site

s si

tes

sesq

uiox

ides

So

il So

il ty

pes

PH(H

,O)

A

Fluv

isol

s 5.

5 0.

3 9.

1 17

.1

5.1

0.02

0.

1 0.

1 0.

2 B

Fl

uvis

ols

5.3

1.7

46.0

49

.2

5.2

0.7

0.6

0.4

0.4

C

And

osol

s 6.

1 2.9

22

.3 18

2 6.

6 0.5

1.

1 1.5

0.5

D

-A

ndos

ols

6.3

10.4

42

2 10

6 20

.2

1.1

4.5

3.4

2.6

Tab

le 2

. Su

mm

ary

of th

e sh

oot dry w

eigh

t, C

d co

ncen

tratio

n, and

Cd c

onte

nt in

11

rice

and

10 s

oybe

an c

ultiv

ars g

row

n in

4 C

d-co

ntam

inat

ed so

ils.

A s

oil

B so

il C

soi

l D

soi

l R

ice

soyb

ean

Ric

e so

ybea

n R

ice

soyb

ean

Ric

e so

ybea

n

Dry

wt.

(gpo

t-')

Mea

n 4.

16

6.25

3.

43

6.79

3.

46

7.82

2.

02

4.98

R

ange

3.

13-4

.96

4.82

-7.1

4 2.

28-4

.86

5.89

-8.0

0 1.

92-4

.77

6.39

-9.0

6 S.

D.

0.54

1.0

3 0.

87

0.86

0.

88

1.03

0.62

1.

10

Cd

conc

. (m

g kg-')

Mea

n 4.

66

1.52

11.6

3.7

5 15

.7 1.

91

24.7

6.

34

Ran

ge

2.09

-6.4

0 0.

94-2

.76

6.49

- 17

.0

1.81

-7.3

0 10

.3-2

0.3

0.92

-3.4

8 14

.6-3

2.2

3.91

- 13

.9 S.

D.

1.19

0.63

3.

71

2.1

1 3.

34

0.97

4.

97

3.30

Cd

con

tent

(kg

pot-'

) M

ean

19.5

9.44

41

.4

25.4

55

.8 15

.0 50

.7

30.5

S.D

. 6.

70

3.93

20

.2

14.5

20.7

7.

70

19.7

11.6

3.46

-6.6

7 0.

73-3

.01

Ran

ge

10.3

-3 1

.7 6.

2-15

.4

15.5

-7 1

.7 10

.5-4

8.5

19.8

-82.

7 16

.3-7

5.7

14.5

-47.

7 6.

5-28

.0

8 ¶

Y

0 8 z.

v1

a-

0 2 2 5'

v1

0

Y 8

9

104 S. ISHIKAWA et al.

%ble 3. Correlation (r) between Cd concentration in the shoots of 11 rice cultivars and 10 soybean cultivars grown in four soils.

(Rice cultivars)

B soil C soil D soil

y = 0.1 89 t 2 .45~ r = 0.787 ** A C soil y = 8.29 + 1 . 5 8 ~ r = 0.565 A Dsoil y=26.9-0.464x r=-0.111

,6 35 - '6 z * m

30 Asoil 0.787** 0.565 -0.1 1 1 B soil 0.594 0.406 C soil 0.337

(Soybean cultivars)

B soil C soil D soil

A soil 0.965 *** 0.952*** 0.941 *** B soil 0.994*** 0.926*** C soil 0.910***

**. *** Significant at the 0.01 and 0.001 probability levels, respectively.

the shoot Cd concentrations in the rice and soybean cul- tivars varied among the four soils, the correlations of shoot Cd concentration between soils are shown in Table 3. In addition, the shoot Cd concentration of the cultivars grown in the soils with a relatively high level of Cd pollution (B, C, and D soils) versus the soil with a low level of Cd pollution (A soil) is plotted in Fig. 1. No correlations were observed in the shoot Cd concentra- tion of the rice cultivars among the soils, except between the A and B soils (Fig. I and Table 3). Mean- while, a significant conelation in the shoot Cd concen- tration of the soybean cultivars was found among the four soils (Fig. 1 and Table 3). The shoot Cd concentra- tions of the soybean cultivars were categorized into two extreme groups; three cultivars of soybean (Tohoku 128, Suzuyutaka, and Harsoy) showed a higher shoot Cd concentration than the other cultivars across the four soils.

Interaction between Cd and four heavy metals in shoots

The correlations between the concentrations of Cd and Zn, Cd and Mn (Fig. 2). Cd and Cu, and Cd and Fe (data not shown) in the shoots of the rice and soybean cultivars were observed for the four soils. The rice shoots accumulated high concentrations of not only Cd but also of other metals, compared to soybean. A remarkably high Zn concentration in the shoots of the rice and soybean cultivars was found in the D soil due to its high Zn concentration. However, the shoot Mn con- centration in the rice and soybean cultivars was the low- est in the D soil compared to the other soils, irrespective of the relatively high level of extractable Mn in the D soil, presumably due to the inhibition of Mn uptake by highly bio-available Cd and Zn. Significant and positive correlations were found between the Cd and Zn concen- trations and between the Cd and Mn concentrations for

A A

A A A

2 3 4 5 6 7 Cd concentration in shoots of plants grown in soil

with a low level of Cd pollution (A) (mg kg-')

y=-1.14+3.21x r=0.965"' y = -0.296 + 1 . 4 5 ~ r = 0.952"'

.c 0

1 1.5 2 2.5 3 s - 0.5

Cd concentration in shoots of plants grown in soil with a low level of Cd pollution (A) (mg kg-')

Fig. 1. Genotypic variation in shoot Cd concentration (rng kg-' dry weight) in rice and soybean plants grown in 4 soils with different Cd contamination levels. Three soybean cultivars (Tohoku 128, Suzuyutaka, and Harosoy) with a high shoot Cd concentration across the 4 soils showed consanguinity (refer to Arao et al. 2003). The straight or dotted regression lines indicate significant or non-significant differences, respectively. **.***Sig- nificant at the 0.01 and 0.001 probability levels, respectively.

the rice cultivars cultivated in the B and C soils. The correlations between the Cd and Cu concentrations and between the Cd and Fe concentrations were not signifi- cant in the rice cultivars across the soils. On the other hand, the shoot Cd concentration in the soybean culti- vars was not correlated with that determined for any of the metals.

Genotypic Variation in Shoot Cd in Rice and Soybean

500

- pi, In 0

In C

C 0 '= 2oo E

Y - 2 300- .-

- c 100

15 0

105

0 A soil (r4.476) Soybean B soil ( r d.451)

n C soil ( r d.289) A Dsoil(rd.150)

A AA A

A A A A

- . % 0-

0. -

A " " ~ ' ~ ' ~ ' " " '

L - -n- C soil fr d.762")

F

0 5 lOo0 800 t A A I

0 ~ ~ " " ' ~ ' ~ " " ~ ~

0 5 10 15 20 25 30 3! Cd concentration in shoots (mg kg")

250

E

8 150-

- u)

5,

5 L

C .- 'ij 100- - c a

00

0 0 0

a

0

0

A A A

"0 5 10 15 20 25 30 35 "0 2 4 6 8 10 12 14 16

Cd concentration in shoots (mg kg") Cd concentration in shoots (mg kg-')

Fig. 2. Correlations between the concentrations of Cd and other metals (Zn and Mn) in the shoots of rice and soybean culti- vars. The straight or dotted regression lines indicate significant or non-significant differences, respectively. ****Significant at the 0.05 and 0.01 probability levels, respectively.

Interaction between the concentration of Cd in the shoots and seeds of the rice and soy- bean cultivars

The data on the Cd concentration in the seeds were cited from Arao and Ae (2003) for the rice cultivars grown under upland conditions and Arao et al. (2003) for the soybean cultivars, which were cultivated in 1999. These cultivars were grown in the soils derived from the same polluted sites as the A and D soils used in the present experiment. Similarity in the ranking of rice cul- tivars for grain Cd concentration was observed across 2 years (1999 and 2000), when they were cultivated in pots or containers filled with the same Cd-polluted soils under upland conditions, although the grain Cd concen- tration in the second year had somewhat decreased (Arao and Ae 2003). Similar results were obtained for the soybean cultivars (data not shown). Therefore, for the purpose of this study, namely the evaluation of geno- typic variation in seed Cd concentration in younger

shoots, it was considered that it might be possible to compare the Cd concentration between younger shoots and mature seeds of the cultivars grown in different years. The Cd concentration in the shoots of the rice and soybean cultivars, shown in Fig. 3, was derived from Fig. 1 in the present study. No correlation between the concentration of Cd in the shoots and seeds was found among the rice cultivars in the A and D soils (Fig. 3). Among the rice cultivars tested, LAC 23 showed a unique characteristic, namely, a low grain Cd concentra- tion in spite of a high shoot Cd concentration in both soils. Meanwhile, significant and positive correlations were found between the concentration of Cd in the shoots and seeds among the soybean cultivars cultivated in each soil (r = 0.823** for A soil and r = 0.932*** for D soil, **p < 0.01; ***p < 0.001). Similarly, a sig- nificant correlation was found after the data in A and D soils (r = 0.907***) were combined.

106 S. ISHIKAWA et al.

U p - 1 A

A A A

A 0

. 3 A 0

A 0 A

' @ LAC23 (HU-LO-TAO I . I . I . I . 1 . I .

" 0 5 10 15 20 25 30 35 Cd concentration in shoots (mg kg-')

- y= -0.632 + 0.776~ r =0.907*"

-- 12 Soybean

F 10 Y

0 2 4 6 8 1 0 1 2 1 4 Cd concentration in shoots (mg kg-')

Fig. 3. Relationship between the concentration of Cd in shoots and seeds in rice and soybean cultivars. The data on seed Cd concentration were cited from Arao and Ae (2003) for the rice cultivars grown under upland conditions and Ardo et al. (2003) for the soybean cultivars. The data on shoot Cd concen- tration were derived from Fig. 1. The straight or dotted regres- sion lines indicate significant or non-significant differences, respectively. ***Significant at the 0.001 probability level.

DISCUSSION

In all the Cd-polluted soils tested, the average value of the Cd concentration and Cd content in the shoots was higher in the rice cultivars than in the soybean cultivars, irrespective of the lower value of the shoot dry weight in the rice cultivars (Table 2). We did not determine the root Cd accumulation due to the difficulty in collecting whole roots from soil and removing Cd in the soil parti- cles adhering to the root surface for Cd analysis. How- ever, a preliminary experiment in water culture showed that the root Cd accumulation also varied among the rice and soybean cultivars and that the root Cd uptake was

generally higher in the rice cultivars than in the soybean cultivars (data not shown). Therefore, it can be speculat- ed that whole plant Cd uptake would be higher in rice at a younger stage. Interesting characteristics of the Cd uptake were found among the rice and soybean cultivars when these were cultivated in the B and C soils. The average value of the shoot Cd content in the rice culti- vars was higher when the plants were grown in the C soil containing relatively high amounts of inorganic and organic Cd than in the B soil containing a relatively high amount of exchangeable Cd. When high Cd-accumulat- ing rice cultivars were grown in Cd-polluted fields under upland conditions, the Cd content at the organic sites in the rhizosphere soil markedly decreased after harvest (unpublished data). On the other hand, the average value of the shoot Cd content in the soybean cultivars was higher when the plants were grown in the B soil than in the C soil. These results suggest that the forms of bio- available Cd may differ between rice and soybean. Rice can probably mobilize and absorb Cd not only from the exchangeable sites but also from the inorganic or organ- ic sites, under upland conditions. In several reports, it was suggested that root exudates, such as organic acids or phytosiderophores, play an important role in the solu- bilization of metals, including Cd, Zn, Cu, etc. in soils (Mench and Fargues 1994; Cakmak et al. 1996). On the contrary, root exudates of the hyperaccumulator plant Thlaspi caerulescens did not enhance metal mobiliza- tion (Zhao et al. 2001). Detailed studies on the Cd-solu- bilizing mechanisms in rice are ongoing.

Genotypic variation in shoot Cd concentration was found among rice and soybean cultivars across the four different soils (Fig. 1). The soybean cultivars were ranked similarly in terms of shoot Cd concentration in the four soils (Fig. 1 and Table 3). This result suggests that the Cd uptake of the soybean cultivars may not be affected by soil environments, and that the genotypic variation in the shoot Cd accumulation may be deter- mined by the Cd influx into the roots and root-to-shoot translocation. The physiological mechanism underlying the Cd translocation to shoots or seeds in soybean was considerably involved in Cd retention in the roots (Arao et al. 2003). A similar mechanism was found to operate in maize inbred lines (Florijn and Van Beusichem 1993). Meanwhile, the ranking of the rice cultivars for the shoot Cd concentration varied considerably among the soils. The Cd forms in soils and Cd availability for crop uptake vary with several soil factors, e.g., pH, clay con- tent, organic matter content, soil chloride levels, etc. (Li et al. 1995). In addition, the forms of bio-available Cd may differ among the rice cultivars. More detailed stud- ies on the variable ranking of rice cultivars for the shoot Cd concentration among soils should be performed from the viewpoint of plant-soil interactions, including the

Genotypic Variation in Shoot Cd in Rice and Soybean 107

characterization of the relationships between the soil factors and Cd uptake. We suggest that it would be desirable to evaluate the selection of low-Cd-accumulat- ing rice cultivars under various soil conditions. The shoot Cd concentrations of the soybean cultivars were categorized into two extreme groups (Fig. 1). Similar results were obtained for more than 100 soybean culti- vars (unpublished data). Three cultivars with a high shoot Cd concentration showed consanguinity (Arao et al. 2003), indicating the high level of inheritance of the Cd concentration in soybean. The mode of inheritance of the Cd uptake in rice has not yet been elucidated.

It has been suggested that the uptake and root-to- shoot translocation of Cd could be considerably affected by the synergistic or antagonistic interactions of other heavy metals such as Zn, Mn, Cu, etc. (Zhang et al. 2002; Liu et al. 2003). In the rice cultivars grown in the B and C soils, the shoot Cd concentration was positively correlated with the Zn or Mn concentration (Fig. 2). A similar trend was observed in the rice cultivars grown in the A and D soils, though it was not statistically signifi- cant. Using 20 rice cultivars grown in artificially Cd- contaminated soil, Liu et al. (2003) reported that the interactions of Cd and several heavy metals were syner- gistic in terms of their uptake and translocation from root to shoot. Thus, the physiological mechanisms of the uptake of Cd and its translocation to shoots in rice may be partly associated with several metal ions that are chemically related. Meanwhile, it is likely that the con- centrations of Cd and other metals in the shoots of the soybean cultivars are controlled by independent mecha- nisms, because no correlations between the Cd concen- tration and that of any of the metals were observed across the four soils (Fig. 2). Using two ecotypes of the Cd/Zn hyperaccumulator Thlaspi caerulescens, Lombi et al. (2001) suggested the existence of a high-affinity Cd transporter unrelated to Zn uptake and translocation. Their reports indicated that the high-Cd group in the soybean cultivars may harbor a high-affinity Cd trans- porter, which is independent of Zn.

If the genotypic variation in seed Cd concentration could be evaluated in relatively younger shoots, the effort to develop cultivars with a low seed Cd concentra- tion would be reduced. We compared the concentrations of Cd in the seeds and shoots of the rice and soybean cultivars grown in Cd-polluted soils across different years (Fig. 3). This direct comparison was possible because of the constant ranking of cultivars for seed Cd concentration across 2 years (Arao and Ae 2003). The shoot Cd concentration in the soybean cultivars was highly correlated with the seed Cd concentration, irre- spective of the soil types (Fig. 3). This indicated that it could be possible to evaluate the genotypic variation in seed Cd concentration in relatively younger shoots,

although similar studies should be carried out using a large number of soybean cultivars. Meanwhile, it appears to be difficult to conduct such an evaluation for rice cultivars, because there was no correlation between them in two soils (Fig. 3). Our recent study showed that the Cd concentration in rice grains was not affected by the Cd uptake rate at the maximum tiller number stage, but that it could be closely related to the Cd uptake rate at the heading stage (unpublished data). Hart et al. (1998) reported that the high grain Cd content in durum wheat compared with bread wheat was not associated with Cd influx rates into roots or root-to-shoot translo- cation, but that it could be related to phloem-mediated Cd transport to the grain. Direct evidence for phloem- mediated Cd transport to the rice grain was demonstrat- ed by Tanaka et al. (2003). We found a unique rice culti- var (LAC 23) that prevents Cd translocation to the grain due to the high Cd retention in the shoots. We also con- firmed that although LAC 23 showed a high Cd concen- tration in the straw and husk, Cd concentration in the grain was low at harvest time (data not shown). These results suggest that the genotypic differences in rice grain Cd may be closely associated with the Cd uptake rate at the heading stage and/or its subsequent phloem- mediated Cd transport to the grain.

We conclude that genotypic differences in shoot Cd concentration in rice and soybean are variable and invariable among the soils, respectively. It may be possi- ble to efficiently select soybean cultivars with low grain Cd using relatively younger shoots grown in a single Cd-polluted soil. Other approaches for efficiently select- ing rice cultivars with low grain Cd should be investigat- ed.

REFERENCES

Arao T and Ae N 2003: Genotypic variations in cadmium levels of rice grain. Soil Sci. Plant Nutr., 49, 413-419

Arao T, Ae N, Sugiyama M, and Takahashi M 2003: Genotypic differences in cadmium uptake and distribution in soybeans. Plant Soil, 251,241-253

Cakmak I, Sari N, Marschner H, Ekiz H, Kalayci M, Yilmaz A, and Braun HJ 1996: Phytosiderophore release in bread and durum wheat genotypes differing in zinc efficiency. Plant Soil, 180, 183-189

Crews HM and Davies BE 1985: Heavy metal uptake from con- taminated soils by six varieties of lettuce (Lactuca sutiva L.). J . Agr-ic. Sci., 105, 59 1-595

Florijn PJ and Van Beusichem ML 1993: Uptake and distribu- tion of cadmium in maize inbred lines. P lunf Soil, 150, 25- 32

Hart JJ, Welch RM, Norvell WA, Sullivan LA, and Kochian LV 1998: Characterization of cadmium binding, uptake, and translocation in intact seedlings of bread and durum wheat cultivars. Plant Physiol., 116, 1413-1420

108 S. ISHIKAWA et al.

Li Y-M, Chaney RL, Schneiter AA, and Miller JF 1995: Geno- typic variation in kernel cadmium concentration in sun- flower germplasm under varying soil conditions. Crop Sci.,

Liu J, Li K, Xu J, Liang J, Lu X, Yang J, and Zhu Q 2003: Inter- action of Cd and five mineral nutrients for uptake and accu- mulation in different rice cultivars and genotypes. Field Crops Res., 83, 271-281

Lombi E, Zhao FJ, McGrath SP, Young SD, and Sacchi GA 2001: Physiological evidence for a high-affinity cadmium transporter highly expressed in a Thlaspi carrulescens ecotype. New Phytol., 149, 53-60

Mench MJ and Fargues S 1994: Metal uptake by iron-efficient and inefficient oats. Plant Soil, 165, 227-233

35, 137-141

Sadamoto H, Iimura K, Honna T, and Yamamoto S 1994: Exam- ination of fractionation of heavy metals in soils. Jpn. J. Soil Sci. Plant Nutr., 66, 645-653 (in Japanese)

Tanaka K, Fujimaki S, Fujiwara T, Yoneyama T, and Hayashi H 2003: Cadmium concentrations in the phloem sap of rice plant (Oryza sativa L.) treated with a nutrient solution con- taining cadmium. Soil Sci. Plant Nutr., 49, 31 1-3 13

Zhang G, Fukami M, and Sekimoto H 2002: Influence of cad- mium on mineral concentrations and yield components in wheat genotypes differing in Cd tolerance at seedling stage. Field Crops Res., 77, 93-98

Zhao FJ, Hamon, RE, and McLaughlin MJ 2001: Root exudates of the hyperaccumulator Thlaspi caerulescens do not enhance metal mobilization. New Phytol., 151, 61 3-620